text
stringlengths 66
377k
|
|---|
The geometric phase controls ultracold chemistry
The geometric phase is shown to control the outcome of an ultracold chemical reaction. The control is a direct consequence of the sign change on the interference term between two scattering pathways (direct and looping), which contribute to the reactive collision process in the presence of a conical intersection (point of degeneracy between two Born-Oppenheimer electronic potential energy surfaces). The unique properties of the ultracold energy regime lead to an effective quantization of the scattering phase shift enabling maximum constructive or destructive interference between the two pathways. By taking the O þ OH-H þ O 2 reaction as an illustrative example, it is shown that inclusion of the geometric phase modifies ultracold reaction rates by nearly two orders of magnitude. Interesting experimental control possibilities include the application of external electric and magnetic fields that might be used to exploit the geometric phase effect reported here and experimentally switch on or off the reactivity.
T he geometric phase (GP) effect in molecules refers to the sign change associated with the Born-Oppenheimer adiabatic electronic wave function (typically the ground state) when transported along a closed path encircling a conical intersection (CI), a point of degeneracy between two electronic potential energy surfaces (PESs). The sign change was noticed by Longuet-Higgins 1 and Herzberg-Longuet-Higgins 2 over 56 years ago in the context of molecular spectra. In order for the total Born-Oppenheimer molecular wave function to remain single valued, the electronic sign change implies that a corresponding sign change must also occur on the nuclear motion wave function. This sign change occurs even when the energy of the nuclear motion lies well below the energy of the CI and the nuclear motion is confined to just one adiabatic electronic PES. The implications of the sign change for chemical reactions occurring on a single Born-Oppenheimer electronic PES including the effects of identical nuclei was treated by Mead and Truhlar 3 and Mead [bib_ref] The molecular Aharonov-Bohm effect in bound states, Mead [/bib_ref] [bib_ref] Superposition of reactive and nonreactive scattering amplitudes in the presence of a..., Mead [/bib_ref] [bib_ref] The geometric phase in molecular systems, Mead [/bib_ref]. They showed that a consistent treatment of the sign change and identical nuclei leads to a generalized Schrödinger equation for the nuclear motion that contains a vector potential interaction analogous to that of a magnetic solenoid centered at the CI 3 . Berry 7 later showed that the appearance of a vector potential and its associated GP (also known as the Berry phase) is a general consequence of the adiabatic transport of a quantum state [bib_ref] Quantum Mechanics: Foundations and Applications, Bohm [/bib_ref]. Mead originally referred to this effect as the 'Molecular Aharonov-Bohm' (MAB) effect [bib_ref] The molecular Aharonov-Bohm effect in bound states, Mead [/bib_ref] , and showed that the interference term between the reactive and non-reactive contributions to the scattering cross-section for the fundamental H þ H 2 reaction changes sign when the GP or MAB effect is included [bib_ref] Superposition of reactive and nonreactive scattering amplitudes in the presence of a..., Mead [/bib_ref]. However, the experimental detection of Mead's predictions has been elusive to this day despite heroic experimental efforts [bib_ref] Hunt for geometric phase effects in H þ HD-HD(v', j') þ H, Jankunas [/bib_ref]. Numerous theoretical scattering calculations for the H þ H 2 system and its isotopic variants that include the GP have also been performed beginning with the work of Lepetit and Kuppermann [bib_ref] Numerical study of the geometric phase in the H þ H 2..., Lepetit [/bib_ref]. Other molecular systems have also been treated [bib_ref] Geometric phase effects in H þ O 2 scattering. I. Surface function..., Kendrick [/bib_ref] [bib_ref] Geometric phase effects in H þ O 2 scattering. II. Recombination resonances..., Kendrick [/bib_ref] [bib_ref] Geometric phase effects in the resonance spectrum, state-to-state transition probabilities and bound..., Kendrick [/bib_ref]. Surprisingly, for the lower-symmetry purely reactive isotopic reactions H þ D 2 -D þ HD and D þ H 2 -H þ HD, it was shown that the GP effects seen in both the partial integral and differential cross-sections all but cancel out when the partial cross-sections are summed over all values of total angular momentum J to obtain fully converged cross-sections [bib_ref] Geometric phase effects in the H þ D 2 -HD þ D..., Kendrick [/bib_ref] [bib_ref] Quantum reactive scattering calculations for the H þ D 2 -HD þ..., Kendrick [/bib_ref] [bib_ref] Quantum reactive scattering calculations for the D þ H 2 -HD þ..., Kendrick [/bib_ref] [bib_ref] Geometric phase effects in chemical reaction dynamics and molecular spectra, Kendrick [/bib_ref]. A similar cancellation also occurs in the H þ H 2 reaction for transitions that involve only reactive contributions to the cross-sections 17 . However, for the transitions that involve both reactive and non-reactive contributions, the cancellation does not occur and Mead's original prediction of GP effects for this case were confirmed [bib_ref] Geometric phase effects in chemical reaction dynamics and molecular spectra, Kendrick [/bib_ref]. The surprising cancellation of the GP effects for the purely reactive transitions with respect to the partial wave sum was later confirmed by Juanes-Marcos et al. [bib_ref] Geometric phase effects in the H þ H 2 reaction: quantum wave-packet..., Juanes-Marcos [/bib_ref]. Furthermore, by considering direct and looping contributions to the reactive scattering amplitude, it was shown that they scatter into distinctly different angular regions [bib_ref] Theoretical study of geometric phase effects in the hydrogen-exchange reaction, Juanes-Marcos [/bib_ref]. Consequently, very little interference occurs between the two pathways resulting in essentially no observable GP effects in both the differential and integral cross-sections [bib_ref] Theoretical study of geometric phase effects in the hydrogen-exchange reaction, Juanes-Marcos [/bib_ref]. At higher collision energies and larger values of total angular momentum J, some interference is observed but it results in rapid oscillations of the differential cross-section as a function of scattering angle [bib_ref] Theoretical study of geometric phase effects in the hydrogen-exchange reaction, Juanes-Marcos [/bib_ref]. These rapid oscillations average out and vanish in the integral cross-sections [bib_ref] Theoretical study of geometric phase effects in the hydrogen-exchange reaction, Juanes-Marcos [/bib_ref]. Therefore, the most promising experimental detection of GP effects to date has focused on the elusive small-amplitude interference oscillations in the differential cross-sections at high collision energies below the CI 9 . Theoretical calculations have also been performed for energies above the energy of the CI [bib_ref] Strong geometric-phase effects in the hydrogen-exchange reaction at high collision energies, Bouakline [/bib_ref] [bib_ref] Strong geometricphase effects in the hydrogen-exchange reaction at high collision energies: II...., Bouakline [/bib_ref]. These calculations show large bi-modal structures in the DCS because of the GP but no significant effects were seen in the integral cross-sections.
As discussed above, all previous experimental and theoretical studies of GP effects in chemical reactions have been performed at thermal energies. Remarkable experimental progress in recent years in cooling and trapping molecules at cold (o1 K) and ultracold (o1 mK) temperatures has opened up an entirely new energy regime [bib_ref] Cold and ultracold molecules: science, technology and applications, Carr [/bib_ref] [bib_ref] Quantum-state controlled chemical reactions of ultracold potassium-rubidium molecules, Ospelkaus [/bib_ref] [bib_ref] Ultracold molecules under control!, Quéméner [/bib_ref]. At ultracold collision energies chemistry often becomes quantum-enhanced due in part to the very large de Broglie wavelengths, quantum tunnelling and quantum threshold effects. In the ultracold limit, the rate coefficients obey the Bethe-Wigner threshold laws [bib_ref] Theory of disintegration of nuclei by neutrons, Bethe [/bib_ref] [bib_ref] On the behavior of cross sections near thresholds, Wigner [/bib_ref] and approach a finite non-zero value in the zero temperature limit. For many reactions, the ultracold rate coefficient can often be larger than its value at thermal energies. A natural question to ask is 'Can GP effects become enhanced at ultracold collision energies?'. At ultracold collision energies only a single partial wave (s-wave) contributes to the reaction so that the cancellation of GP effects over the partial wave sum cannot occur as it does at thermal energies and the scattering is isotropic. In contrast to the situation discussed above for H þ H 2 at thermal energies, the direct and looping contributions to the cross-sections scatter into the same angular region with the potential for significant interference. For reactive collisions in which the direct and looping contributions are of similar magnitude, the interference term (depending on its sign and magnitude) can effectively turn on or turn off the reaction entirely. Since the GP changes the sign of the interference term between the direct and looping contributions, it can effectively control the reaction.
We report here theoretical predictions of GP effects in ultracold chemical reactions. Specifically, large GP effects are predicted to occur in ultracold chemical reactions, for which: (a) the relevant Born-Oppenheimer adiabatic electronic PES exhibits a CI, (b) the direct and looping contributions to the scattering amplitude are of similar magnitude and (c) their phases are quantized (or nearly so). These three conditions are readily met in the experimentally
[formula] relevant O þ OH (v ¼ 0, j ¼ 0)-H þ O 2 (v 0 , j 0 ) [/formula]
reaction, which will serve as our prototypical system. For this reaction, it is shown that the GP can enhance or suppress the rotationally resolved reaction rates by nearly two orders of magnitude. The OH radical has been cooled and trapped using the buffer gas and Stark decelerator techniques [bib_ref] Deceleration and electrostatic trapping of OH radicals, Van De Meerakker [/bib_ref] [bib_ref] Phase space manipulation of cold free radical OH molecules, Bochinski [/bib_ref] [bib_ref] Magnetoelectrostatic trapping of ground state OH molecules, Sawyer [/bib_ref] , and successful experiments on evaporative cooling of OH have recently been carried out [bib_ref] Evaporative cooling of the dipolar hydroxyl radical, Stuhl [/bib_ref]. Numerous experimental and theoretical studies of collisions involving OH molecules in the cold and ultracold temperature regimes have been reported in recent years, and the O þ OH reaction continues to be the topic of considerable experimental and theoretical investigations [bib_ref] Kinetics of the radical-radical reaction, O( 3 Pj) þ OH(X 2 P..., Carty [/bib_ref] [bib_ref] Ultracold Rb-OH collisions and prospects for sympathetic cooling, Lara [/bib_ref] [bib_ref] Quantum dynamics of the O þ OH-H þ O 2 reaction at..., Quéméner [/bib_ref] [bib_ref] Formation of molecular oxygen in ultracold O þ OH collisions, Quéméner [/bib_ref] [bib_ref] Ultracold collisions and reactions of vibrationally excited OH radicals with oxygen atoms, Juanes-Marcos [/bib_ref] [bib_ref] Chemical reaction versus vibrational quenching in low energy collisions of vibrationally excited..., Pradhan [/bib_ref] [bib_ref] Accurate combined-hyperbolic-inverse-powerrepresentation of ab initio potential energy surface for the hydroperoxyl radical..., Varandas [/bib_ref] [bib_ref] Quantum dynamics study on the CHIPR potential energy surface for the hydroperoxyl..., Teixidor [/bib_ref].
# Results
The interference mechanism. A two-dimensional (2D) slice of the three-dimensional electronically adiabatic ground state 2 A 00 PES for the HO 2 molecule is plotted in [fig_ref] Figure 1 |: HO 2 potential energy surface [/fig_ref]. Most notable are the two deep symmetric wells (E À 26,000 K relative to the bottom of the H þ O 2 asymptote), which correspond to the bound HO 2 complex. Three CIs are also clearly visible: one occurs for T-shaped or C 2v geometries (that is, O-H-O) and the other two occur at collinear geometries (that is, H-O-O and O-O-H). In the present treatment, we consider GP effects due to the C 2v CI. The scattering methodology for including GP effects due to collinear CIs is under development [bib_ref] Geometric phase for collinear conical intersections. I. Geometric phase angle and vector..., Li [/bib_ref]. The reaction is exoergic (by B8,100 K) and proceeds along a barrierless reaction path through the deep potential well. The total available energy for the ultracold collision is that of the reactant OH (v ¼ 0, j ¼ 0) state, which is E9,400 K relative to the bottom of the asymptotic product H þ O 2 well. The minimum energy of the C 2v CI lies at E22,000 K and the maximum energy encountered along the minimum energy pathway (MEP) encircling the C 2v CI is E580 K (ref. 42). Thus, even though the ultracold collision of O þ OH initiates with a tiny kinetic energy (1 mK), the total energy (E9,400 K) is much larger than the maximum energy encountered along the MEP around the CI (E580 K). Thus, the CI can be easily encircled, and the direct and looping contributions to the scattering amplitude can have similar magnitudes (see [fig_ref] kFigure 2 |: cos [/fig_ref]. In general, the encirclement of a CI at ultracold collision energies does not require the reaction to be exoergic (although that is often a favourable situation). Endoergic reactions as well as reactions with an entrance barrier can still exhibit favourable encirclement of a CI at ultracold collision energies, provided the reactant state is in a suitably high vibrationally and/or rotationally excited state (and with total energy still well below the energy of the CI).
The total scattering amplitude can be decomposed as . The amplitudes f GP (GP) and f NGP (No GP) are the scattering amplitudes computed with and without including the GP, respectively. The magnitudes of f direct and f loop depend on the details of the scattering process. In general, the magnitudes can be similar for situations in which the maximum potential energy encountered along the MEP encircling the CI lies below the available total energy. The energetics depend on the particular reactant and product states involved and the details of the Born-Oppenheimer electronic PES for the particular molecule of interest. If the direct and looping contributions are equal in magnitude jf direct j ¼ jf loop j ¼ f and we write the complex scattering amplitudes
[formula] f AE ¼ 1= ffiffi ffi 2 p ðf direct AE f loop Þ where f þ ¼ f NGP and f À ¼ f GP (as f direct ¼ f expðiy direct Þ and f loop ¼ f exp iy loop À Á , then jf AE j 2 ¼ f 2 1 AE cos D ð Þwhere D ¼ y loop À y direct is the phase difference. [/formula]
Furthermore, if the phases are quantized so that the phase difference is a multiple of p (that is, D ¼ np where n is an integer), then the square modulus of the scattering amplitude that gives the angular distributions will be zero or 2f 2 depending on whether ±cosD is À 1 or þ 1, respectively. For example, if for a particular scattering process cosD ¼ 1, then jf NGP j 2 ¼ 2f 2 and jf GP j 2 ¼ 0. That is, the GP completely turns off the reaction. Conversely, if cosD ¼ À 1, then jf NGP j 2 ¼ 0 and jf GP j 2 ¼ 2f 2 so that the GP turns on the reaction in this case. In actual scattering situations, the phases are not exactly quantized and the scattering amplitudes are not exactly equal so that variations between the two extreme values (0 and 2f 2 ) occur. However, the possibility of realizing very large GP effects on the chemical reaction rate is clear. In situations where the magnitudes of the f direct and f loop and is plotted as a function of the xy location of the hydrogen (red sphere) relative to the centre of the O 2 bond. The prominent deep potential wells (dark blue region) correspond to the bound HO 2 molecule. Three CIs are readily visible: two are located at linear geometries x ¼ 0, y ¼ ± 3.5 (bohr) and a third is located at a T-shaped (C 2v ) geometry x ¼ 2.65, y ¼ 0 (bohr). The upper cones for each CI correspond to the excited electronic state (to aid visualization, a cut plane was used at 34,800 K so that the potential surfaces above this energy are not plotted). The contour lines are uniformly spaced at intervals of 1,800 K. The scattering amplitudes for each of these pathways are summed to obtain the total scattering amplitude for the reactive collision. The reactivity can be enhanced or suppressed if the two scattering amplitudes interfere constructively or destructively, respectively. The inset figures in a show the effective potentials along each pathway, which support a different number of bound states. (b) plots the theoretically derived phase shift d/p for swave (l ¼ o) scattering of a particle of mass m from a model spherical well potential with a depth of V o ¼ q 2 /(2m) and radius r o . Each curve corresponds to a different well depth separated by q ¼ 0.025 p/r o where k 02 À q 2 ¼ k 2 . The notable feature in this plot is that the phase shift becomes effectively quantized in the ultracold limit k-0.
are significantly different, the square modulus of the total scattering amplitude is given by the general expression jf AE j 2 ¼ ð1=2Þðjf direct j 2 þ jf loop j 2 AE 2jf direct jjf loop j cos DÞ. If the magnitude of one of the amplitudes is much larger than the other, then the relative magnitude of the interference term is small and jf AE j 2 % ð1=2Þjf direct j 2 or jf AE j 2 % ð1=2Þjf loop j 2 and there is little or no GP effect (that is, jf NGP j 2 % jf GP j 2 ).
We have shown that the ultracold reaction O þ OH-H þ O 2 meets the first two criteria for realizing a large GP effect, namely (a) the PES exhibits a CI and (b) the energetics are favoruable for encirclement of the CI. The third criterion requires the quantization (or nearly so) of the phase difference between the direct and looping scattering amplitudes. The phase quantization is a unique feature of the ultracold energy regime. [fig_ref] kFigure 2 |: cos [/fig_ref] summarizes the phase quantization mechanism. In the zero energy limit, the scattering is isotropic and a spherical well can be used as a simple model of the collision process where the effective depth and width of the well corresponds to different collision pathways (and different reactant and product states). [fig_ref] kFigure 2 |: cos [/fig_ref] plots the s-wave scattering phase shift d for a spherical potential well given by V(r) ¼ À V o for rrr o and V(r) ¼ 0 for r4r o as a function of collision energy and well depth. In the zero energy limit (that is, wave vector k-0), this figure shows that the scattering phase shift becomes effectively quantized. That is, the most likely value for the phase shift at zero energy is near a multiple of p regardless of the well depth. As the well depth is varied for collision energies near zero, the phase shift rapidly jumps from one value of np to another. The jumps are due to resonances that occur at d r ¼ (n þ 1/2)p as a continuum state drops into the potential well as the well depth is increased. In fact, the np phase shift indicates the number of bound states n supported by the well. As sketched qualitatively in [fig_ref] kFigure 2 |: cos [/fig_ref] , the direct and looping scattering amplitudes traverse different pathways along the PES encircling the CI as the reaction proceeds from reactants to products. The effective potential energy along each of these pathways can be modelled with a spherical potential well with an appropriate depth and width. The depth and width of each of the spherical potentials V direct eff and V looping eff are different for each pathway so that each one supports a different number of bound states n direct and n loop , respectively. In the ultracold limit, the scattering phase shift associated with each pathway approaches a different integral multiple of p: n direct p and n loop p, and the phase difference D that appears in the interference term (when the direct and looping scattering amplitudes are combined to construct the total amplitude) is also an integral multiple of p: D ¼ mp where m ¼ n loop À n direct . Furthermore, if the integer m is even (odd) then cos D ¼ þ 1( À 1) and f NGP is non-zero (zero) and f GP is zero (non-zero). Thus, in the ultracold regime, the effects of the GP can be large, essentially turning off or on the reactivity. In real molecular collisions, the phase quantization is not exact and the magnitudes of the direct and looping contributions are not identical so that the dynamic range is less than that described in the model above. However, the dynamic range of the GP effect can be dramatic in real ultracold reactions as we will now demonstrate for O þ OH where the reactivity is suppressed or enhanced by nearly two orders of magnitude.
Ultracold reaction rates for O þ OH-H þ O 2 . The effective Schrödinger equation for the nuclear motion is solved using an accurate time-independent quantum reactive scattering methodology that includes the GP (see Methods for more details) [bib_ref] Geometric phase effects in H þ O 2 scattering. I. Surface function..., Kendrick [/bib_ref] [bib_ref] Geometric phase effects in H þ O 2 scattering. II. Recombination resonances..., Kendrick [/bib_ref] [bib_ref] Quantum reactive scattering in three dimensions using hyperspherical (APH) coordinates, Pack [/bib_ref] [bib_ref] Hyperspherical surface functions for nonzero total angular momentum. I. Eckart singularities, Kendrick [/bib_ref] [bib_ref] Accurate quantum calculations on three-body collisions in recombination and collision-induced dissociation. I...., Parker [/bib_ref] [bib_ref] Ultracold chemistry with alkali-metal-rare-earth molecules, Makrides [/bib_ref]. [fig_ref] Figure 3 |: Geometric phase effects on a rotationally resolved reaction rate [/fig_ref]
## Plots the computed reaction rate for
[formula] O þ OH (v ¼ 0, j ¼ 0)-H þ O 2 (v 0 ¼ 2, j 0 ¼ 1) [/formula]
as a function of collision energy summed over total angular momentum J ¼ 0-3.
The red and black curves correspond to the calculations with (GP) and without (NGP) the GP, respectively. The ultracold energy regime where only s-wave scattering (that is, J ¼ 0) contributes to the reaction rate occurs at energies below 100 mK. Higher partial waves contribute for collision energies above 100 mK. The four values of total angular momentum J included in these calculations give converged reaction rates for collision energies up to B0.2 K (see [fig_ref] Figure 1 |: HO 2 potential energy surface [/fig_ref] , which plots the individual contributions from each value of J and the associated Supplementary Discussion). [fig_ref] Figure 3 |: Geometric phase effects on a rotationally resolved reaction rate [/fig_ref] shows that at ultracold collision energies (1 mK) the reaction rate that includes the GP (8.4 Â 10 À 13 cm 3 s À 1 ) is over 40 times larger than the reaction rate that does not include the GP (2 Â 10 À 14 cm 3 s À 1 ). As discussed above, the increased reaction rate for the GP calculations is due to the positive sign on the interference term between the direct and looping contributions to the scattering amplitude (see also [fig_ref] kFigure 2 |: cos [/fig_ref]. A positive (negative) sign on the interference term leads to maximum constructive (destructive) interference between the direct and looping contributions to the total scattering amplitude (see [fig_ref] kFigure 2 |: cos [/fig_ref]. This particular ultracold reaction corresponds to the idealized case discussed above where cosD ¼ À 1 and f NGP j j 2 ¼ 0 and f GP j j 2 ¼ 2f 2 . As the collision energy increases, the J ¼ 1 contribution to the reaction rate becomes significant leading to a rapid increase in the NGP reactivity (black curve) starting at B100 mK. The GP (red) and NGP (black) reaction rates become nearly identical in magnitude once the collision energy reaches 0.05 K where several values of non-zero J contribute. The increased reactivity for the NGP case and J ¼ 1 is due to the presence of additional hyperspherical angular basis functions with symmetric (even) exchange symmetry with respect to the identical oxygen nuclei. For J ¼ 0 the Wigner D-function is symmetric and only hyperspherical angular basis functions with antisymmetric (odd) exchange symmetry are physically allowed. For J40 the Wigner D-functions contain symmetric and antisymmetric components that require the multiplication by both even and odd hyperspherical angular basis functions [bib_ref] Geometric phase effects in the H þ D 2 -HD þ D..., Kendrick [/bib_ref] [bib_ref] Hyperspherical surface functions for nonzero total angular momentum. I. Eckart singularities, Kendrick [/bib_ref]. The presence of the additional even symmetry states alters the relative number of bound states (m) in the effective spherical well potentials along the direct and looping pathways (see [fig_ref] kFigure 2 |: cos [/fig_ref]. The change in m in turn alters the relative sign on the interference
[formula] O þ OH (v ¼ 0, j ¼ 0)-H þ O 2 (v 0 ¼ 2, j 0 ¼ 1) reaction [/formula]
rate is plotted as a function of collision energy summed over all values of total angular momentum between J ¼ 0 and 3. Owing to constructive interference between the direct and looping pathways (see [fig_ref] kFigure 2 |: cos [/fig_ref] term giving constructive instead of destructive interference for the NGP case and J ¼ 1 (see [fig_ref] kFigure 2 |: cos [/fig_ref] and the associated Supplementary Discussion). [fig_ref] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and... [/fig_ref] shows rovibrationally resolved and total reaction rates for the O þ OH (v ¼ 0, j ¼ 0) reaction. Reaction rates for j 0 ¼ 3 and 5 within the v 0 ¼ 2 vibrational level are plotted, respectively, in [fig_ref] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and... [/fig_ref] ,b. In contrast to the j 0 ¼ 1 result shown in [fig_ref] Figure 3 |: Geometric phase effects on a rotationally resolved reaction rate [/fig_ref] , the GP result (2 Â 10 À 14 cm 3 s À 1 ) for j 0 ¼ 3 at 1 mK is B60 times smaller than the corresponding NGP result (1.2 Â 10 À 12 cm 3 s À 1 ). On the other hand, the GP result (4.6 Â 10 À 13 cm 3 s À 1 ) for j 0 ¼ 5 [fig_ref] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and... [/fig_ref] is about nine times larger than the NGP result (5.2 Â 10 À 14 cm 3 s À 1 ) at 1 mK. These findings show that both the GP and NGP ultracold reaction rates can be either dramatically enhanced or suppressed depending on the particular product rotational state involved. As discussed above, different product (and reactant) states give rise to different effective spherical well potentials along the direct and looping pathways. Thus, the relative phase shift (mp) and the sign of the interference term between the direct and looping contributions to the scattering amplitude is sensitive to the states involved. Large differences between the GP and NGP ultracold reaction rates are also observed for j 0 45 and other vibrational states v 0 . For example, vibrationally resolved reaction rates for v 0 ¼ 1 and 2 summed over all open product rotational levels j 0 are plotted in [fig_ref] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and... [/fig_ref]. In both cases differences remain between the ultracold reaction rates computed with and without the GP even after summing over many values of j 0 . For v 0 ¼ 1 and 2 the ultracold reaction rate computed with the GP is 1.9 and 1.6 times smaller than the NGP rate, respectively. Similar results are also seen for v 0 ¼ 0 and 3 (not shown) where the ultracold reaction rate computed with the GP is 1.1 and 1.3 times smaller than the NGP rate, respectively. All of the vibrationally resolved ultracold reaction rates (summed over all j 0 ) are decreased when the GP is included. The total reaction rate summed over all open vibrational v 0 and rotational j 0 product states is plotted in [fig_ref] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and... [/fig_ref]. The total ultracold reaction rate computed with the GP is 1.4 times smaller than the NGP rate.
# Discussion
The answer to the previously posed question 'Can GP effects become enhanced at ultracold collision energies?' is a definite 'yes'. In fact, the ultracold reaction rate is often controlled by the GP. As demonstrated for the O þ OH-H þ O 2 reaction, the rotationally resolved reaction rates can be enhanced or suppressed by nearly two orders of magnitude when the GP is included in the theoretical calculations. The vibrationally resolved and total reaction rates can also be altered by the GP (by factors of nearly two). The enhanced effects of the GP reported here are general and originate from the unique features of the ultracold and cold energy regimes where only one or a few partial waves contribute to the scattering process. In this energy regime the scattering is essentially isotropic and results in an effective quantization of the relative phase shift between the direct and looping contributions to the total scattering amplitude (see [fig_ref] Figure 3 |: Geometric phase effects on a rotationally resolved reaction rate [/fig_ref] and the associated Supplementary Discussion). The phase shift quantization and isotropic scattering results in maximum constructive or destructive interference between the direct and looping pathways. Since the GP changes the sign of the interference term between the direct and looping pathways, it controls whether the interference is constructive or destructive. The effects of the GP on all of the O þ OH-H þ O 2 reaction rates can be explained using this simple interference mechanism (see [fig_ref] kFigure 2 |: cos [/fig_ref] and [fig_ref] kFigure 2 |: cos [/fig_ref]. The enhancement or suppression of the ultracold reactivity due to the GP is a general result independent of the details of the PES. Of
[formula] O þ OH (v ¼ 0, j ¼ 0)-H þ O 2 (v 0 ¼ 2, j 0 [/formula]
) reaction rates for j 0 ¼ 3 and 5 are plotted as a function of the collision energy in a,b, respectively. The rates are summed over all values of total angular momentum between J ¼ 0 and 3. The ultracold (1 mK) reaction rate for j 0 ¼ 3 computed without the geometric phase (black) is dramatically enhanced relative to the rate computed with the geometric phase (red). In contrast, for j 0 ¼ 5 it is the rate computed with the geometric phase (red) that is enhanced. The vibrationally resolved reaction rate is plotted in c for v 0 ¼ 1 and 2 summed over all open product rotational levels j 0 . For both vibrational states, the rates computed with the geometric phase (red for v 0 ¼ 1 and blue for v 0 ¼ 2) lie below the rates computed without the geometric phase (black for v 0 ¼ 1 and green for v 0 ¼ 2). The total reaction rate is plotted in d summed over all open product vibrational and rotational states of O 2 (v 0 , j 0 ). Effects of the geometric phase remain even after the sums that results in a decreased ultracold reaction rate (red) relative to the total rate computed without the geometric phase (black).
course which particular product states are enhanced or suppressed and the magnitude of the effect does depend on the PES. The GP must be included in the theoretical calculations of ultracold reaction rates for which the molecular Born-Oppenheimer electronic PES contains a CI and its encirclement is energetically favourable. Furthermore, the analysis reported here for the direct and looping contributions to the scattering amplitude also holds for reactive processes that involve interference between reactive and non-reactive (inelastic) pathways, such as H þ HD-H þ HD . Other experimentally relevant ultracold molecular systems are also known to contain CIs, such as Li þ Li 2 (ref. [bib_ref] Interactions and dynamics in Li þ Li 2 ultracold collisions, Cvitaš [/bib_ref]. An intriguing future application based on the results presented here might include using external electric and magnetic fields to tune through a zero energy resonance and alter the number of bound states (and hence the relative phase shift) for either the direct or looping pathways. The energy differences between the highest-lying bound states of HO 2 are very small (E5 K) and the highest bound state lies only 21 K below the dissociation threshold [bib_ref] Geometric phase effects in the resonance spectrum, state-to-state transition probabilities and bound..., Kendrick [/bib_ref]. The Stark energy shift for OH in a typical laboratory external DC electric field of 100 kV cm À 1 is E4 K, which is of comparable magnitude. Thus, it may be possible to use external fields to alter the relative number of bound states between the direct and looping pathways. This technique could provide experimentalists with a switch for turning the reactivity on or off.
# Methods
Computational approach. The three-body quantum reactive scattering problem for a given molecular Born-Oppenheimer electronic PES is solved using a threedimensional time-independent scattering method based on hyperspherical coordinates [bib_ref] Quantum reactive scattering in three dimensions using hyperspherical (APH) coordinates, Pack [/bib_ref]. The methodology has been generalized to include the GP [bib_ref] Geometric phase effects in H þ O 2 scattering. I. Surface function..., Kendrick [/bib_ref] and nonzero total angular momentum J (ref. [bib_ref] Hyperspherical surface functions for nonzero total angular momentum. I. Eckart singularities, Kendrick [/bib_ref] , and has been parallelized to run efficiently on large supercomputers [bib_ref] Geometric phase effects in the H þ D 2 -HD þ D..., Kendrick [/bib_ref] [bib_ref] Quantum reactive scattering calculations for the H þ D 2 -HD þ..., Kendrick [/bib_ref]. It has been applied to several reactive systems over a wide range of collision energies from ultracold to thermal. Details of the methodology are given elsewhere [bib_ref] Geometric phase effects in H þ O 2 scattering. I. Surface function..., Kendrick [/bib_ref] [bib_ref] Quantum reactive scattering in three dimensions using hyperspherical (APH) coordinates, Pack [/bib_ref] [bib_ref] Hyperspherical surface functions for nonzero total angular momentum. I. Eckart singularities, Kendrick [/bib_ref] [bib_ref] Accurate quantum calculations on three-body collisions in recombination and collision-induced dissociation. I...., Parker [/bib_ref] ; therefore, a brief summary will be presented here. The method is numerically exact (that is, no dynamical approximations are made) and well suited for accurately treating ultracold chemistry [bib_ref] Ultracold chemistry with alkali-metal-rare-earth molecules, Makrides [/bib_ref]. In the present work, the ab initio based DIMKP [bib_ref] Potential-energy surfaces for the low-lying 2 A'' and 2 A' states of..., Kendrick [/bib_ref] PES with an improved long-range interaction [bib_ref] Formation of molecular oxygen in ultracold O þ OH collisions, Quéméner [/bib_ref] is used for the lowest lying 2 A 00 electronic state of HO 2 . The DIMKP PES accurately treats the CIs and includes a long-range potential that was absent in the ab initio XXZLG PES for this system 52 . Recently, a CHIPR PES has been developed, which is based on the same ab initio data set as the XXZLG PES but includes a long-range potential [bib_ref] Accurate combined-hyperbolic-inverse-powerrepresentation of ab initio potential energy surface for the hydroperoxyl radical..., Varandas [/bib_ref] [bib_ref] Quantum dynamics study on the CHIPR potential energy surface for the hydroperoxyl..., Teixidor [/bib_ref]. We refer the reader to previous work for detailed comparisons of the scattering results using these PESs [bib_ref] Quantum dynamics of the O þ OH-H þ O 2 reaction at..., Quéméner [/bib_ref] [bib_ref] Formation of molecular oxygen in ultracold O þ OH collisions, Quéméner [/bib_ref] [bib_ref] Ultracold collisions and reactions of vibrationally excited OH radicals with oxygen atoms, Juanes-Marcos [/bib_ref] [bib_ref] Chemical reaction versus vibrational quenching in low energy collisions of vibrationally excited..., Pradhan [/bib_ref] [bib_ref] Accurate combined-hyperbolic-inverse-powerrepresentation of ab initio potential energy surface for the hydroperoxyl radical..., Varandas [/bib_ref] [bib_ref] Quantum dynamics study on the CHIPR potential energy surface for the hydroperoxyl..., Teixidor [/bib_ref]. Smith-Johnson symmetrized hyperspherical coordinates are used in the interaction region where the three atoms are in close proximity (that is, for small hyperradius r) [bib_ref] Quantum reactive scattering in three dimensions using hyperspherical (APH) coordinates, Pack [/bib_ref] [bib_ref] Hyperspherical surface functions for nonzero total angular momentum. I. Eckart singularities, Kendrick [/bib_ref]. For larger hyperradius where the reactant (O þ OH) and product (H þ O 2 ) channels become well defined, a properly symmetrized set of Fock-Delves hyperspherical coordinates are used (one for each channel) [bib_ref] Accurate quantum calculations on three-body collisions in recombination and collision-induced dissociation. I...., Parker [/bib_ref]. The hyperradius r is discretized into a set of sectors spanning the range from small to large r. The three-body Hamiltonian is diagonalized at each fixed value of r to obtain a set of 5D angular wave functions. The 5D angular solutions are independent of the collision energy so only have to be computed once for each value of total angular momentum J and inversion parity. The 5D angular solutions form the basis set for the coupled-channel equations in r and are used to compute a set of potential coupling matrices within each sector and the overlap matrices between the different 5D solutions at the boundaries of each sector. The coupledchannel equations are solved for a given collision energy using a log-derivative propagator method from small to large r. Finally, the asymptotic boundary conditions are applied at large r to compute the scattering S matrix from which the cross-sections and reaction rates can be computed [bib_ref] Quantum reactive scattering in three dimensions using hyperspherical (APH) coordinates, Pack [/bib_ref]. Detailed convergence studies, basis sets and other parameters used for the O þ OH reaction are reported elsewhere [bib_ref] Quantum dynamics of the O þ OH-H þ O 2 reaction at..., Quéméner [/bib_ref] [bib_ref] Formation of molecular oxygen in ultracold O þ OH collisions, Quéméner [/bib_ref] [bib_ref] Ultracold collisions and reactions of vibrationally excited OH radicals with oxygen atoms, Juanes-Marcos [/bib_ref] [bib_ref] Chemical reaction versus vibrational quenching in low energy collisions of vibrationally excited..., Pradhan [/bib_ref].
[fig] Figure 1 |: HO 2 potential energy surface. A 2D slice of the Born-Oppenheimer PES for the 2 A 00 ground electronic state of HO 2 is plotted in the product region for a fixed O 2 (blue spheres) separation of 2.28 bohr. The potential energy is relative to the bottom of the asymptotic H þ O 2 well [/fig]
[fig] kFigure 2 |: cos(kr o )sin(k′r o ) -k′ cos(k ′r o )sin(kr o ) k ′ cos(k′r o )cos(kr o ) + k sin(k′r o )sin(Interference between the direct and looping pathways. The direct and looping pathways around the C 2v CI for H-O 2 are sketched in a. [/fig]
[fig] Figure 3 |: Geometric phase effects on a rotationally resolved reaction rate. The rotationally resolved [/fig]
[fig] Figure 4 |: Geometric phase effects on other rotationally and vibrationally resolved reaction rates and the total rate. The rotationally resolved [/fig]
|
Nanomaterials-Based Novel Immune Strategies in Clinical Translation for Cancer Therapy
Citation: Wahab, S.; Ghazwani, M.; Hani, U.; Hakami, A.R.; Almehizia, A.A.; Ahmad, W.; Ahmad, M.Z.; Alam, P.; Annadurai, S.
# Introduction
Cancer evolves and spreads due to the inability of the host immune system (IS), which is one of the reasons to identify tumor antigens and mount successful protection against them. During the process of immune surveillance, the IS of the host is accountable for identifying and eliminating cancer cells in the body [bib_ref] Immune System: A Double-Edged Sword in Cancer, Lakshmi Narendra [/bib_ref] [bib_ref] Paradoxical Roles of the Immune System during Cancer Development, Visser [/bib_ref] [bib_ref] Cancer Antigens: Immune Recognition of Self and Altered Self, Houghton [/bib_ref]. Tumor cells evade immunosurveillance by dodging the anti-tumor mechanisms of the IS [bib_ref] Immunosuppressive Cells in Tumor Immune Escape and Metastasis, Liu [/bib_ref] [bib_ref] The Immune System and Cancer Evasion Strategies: Therapeutic Concepts, Muenst [/bib_ref]. The anti-cancer capabilities of immune cells that have infiltrated a tumor may be suppressed by signals generated by the tumor itself that are present in the tumor microenvironment. They are also altered to hasten the development and growth of tumors [bib_ref] The Tumor Microenvironment and Its Role in Promoting Tumor Growth, Whiteside [/bib_ref] [bib_ref] Immune Cells within the Tumor Microenvironment: Biological Functions and Roles in Cancer..., Lei [/bib_ref]. Therefore, the IS performs a contradictory function by contributing to developing an immunosuppressive microenvironment that promotes cancer advancement [bib_ref] Paradoxical Roles of the Immune System during Cancer Development, Visser [/bib_ref] [bib_ref] Immune Cell Promotion of Metastasis, Kitamura [/bib_ref]. Immunosuppression is the diminished capacity of the body to generate an immune response due to a lack of immune cells such as T and B lymphocytes [bib_ref] Immune System: A Double-Edged Sword in Cancer, Lakshmi Narendra [/bib_ref]. Chemotherapy is a standard treatment for various tumors; however, owing to a reduction in the number of immune cells such as T and B lymphocytes [bib_ref] The Immune System and Cancer Evasion Strategies: Therapeutic Concepts, Muenst [/bib_ref] and dendritic cells (DCs), this therapy often results in immunosuppression [bib_ref] Lack of Dendritic Cell Mobilization into the Peripheral Blood of Cancer Patients..., Ferrari [/bib_ref] [bib_ref] Subpopulations of Peripheral Blood Dendritic Cells during Chemotherapy of Ovarian Cancer, Wertel [/bib_ref]. Nevertheless, it has been shown that many chemotherapy agents affect immunity by preventing the bone marrow from producing blood cells, leading to a sharp drop in the body's total cell count and compromised defenses [bib_ref] Immunosuppressive Cells in Tumor Immune Escape and Metastasis, Liu [/bib_ref] [bib_ref] The Immune System and Cancer Evasion Strategies: Therapeutic Concepts, Muenst [/bib_ref] [bib_ref] A Noninflammatory mRNA Vaccine for Treatment of Experimental Autoimmune Encephalomyelitis, Krienke [/bib_ref].
Immunomodulation optimizes the immune response by inhibiting it, as in treating autoimmune illnesses, or stimulating it, as in cancer immunotherapy or vaccination. Immunotherapy, which uses the body's IS to fight cancer, has received much attention and is now a common way to treat cancer. However, immunology is only one of several medical subfields where the potential of nanotechnology has been investigated. The nanomaterials (NMs)-based transporter is a critical mechanism to accomplish these aims since nanocarriers may be customized to give a variety of treatments to modify the IS [bib_ref] A Noninflammatory mRNA Vaccine for Treatment of Experimental Autoimmune Encephalomyelitis, Krienke [/bib_ref] [bib_ref] Synthetic Nanoparticles for Vaccines and Immunotherapy, Irvine [/bib_ref]. The effectiveness of NMs depends on their direct involvement in the delivery of medicines and improved targeting of diseased immune cells and tissues. However, little attention has been paid to the impact of non-carriers on the control of the IS [bib_ref] Particle Shape Dependence of CD8+ T Cell Activation by Artificial Antigen Presenting..., Sunshine [/bib_ref] [bib_ref] Current Trends and Future Perspectives of Nanomedicine for the Management of Colon..., Wahab [/bib_ref]. To find a solution to this problem, researchers have been looking at a wide variety of nanomaterials that may either directly exploit the IS due to the composition of the nanomaterials themselves, or indirectly exploit the IS by acting as intact active carriers. This review examines the functioning of the IS in cancer and the relationship between NMs and the IS to locate and eliminate cancer cells actively or passively. The current evolutions in NMs and their applications in cancer treatment are presented, with a primary emphasis on explaining their mechanism of action to understand the interactions between nanomaterials and the IS.
## Cancer and the role of the immune system
Understanding carcinogenesis and the intricate relationship between the IS and the host has been the driving force behind many fundamental improvements in cancer treatment over the last several decades. The standard treatment for cancer includes surgery, radiotherapy, and cytotoxic drugs that simultaneously target many tumor cell types. The IS is the body's natural defense against pathogens and foreign materials. Considerable research has been conducted to either manipulate the IS to prevent allergy and autoimmune reactions or harness the IS to cure cancer and infectious illnesses [bib_ref] Autoimmunity: Antigen-Specific Immunotherapy, Wraith [/bib_ref] [bib_ref] Delivering Safer Immunotherapies for Cancer, Milling [/bib_ref]. These defensive systems affect a wide variety of organs, as well as cells, tissues, and chemical mediators of IS. They are based on the capacity to discern between "self" and "non-self" substances. The IS's two main building blocks, adaptive and innate immunity, coordinate the body's defenses against infections and transformed cells. As a result, the IS's actions and immune cells serve as a protective barrier against invasive diseases and cancer to restore and maintain homeostasis [bib_ref] An Overview of the Immune System, Parkin [/bib_ref]. The first line of defense of the IS is innate immunity, which utilizes cells and molecules already in place to protect against the disease within an hour of the first encounter with the pathogen. The IS's first line of defense comprises mucosal linings, skin, cilia, and many other physical barriers.
Non-specific innate immunity serves as the initial line of protection. Inflammation is triggered when pathogen-associated molecular patterns from bacteria and viruses are identified by pattern recognition receptors (PRRs) on immune cells. Nucleotide oligomerization domain-like receptors are cytoplasmic PRRs, whereas Toll-like receptors (TLRs) are cell membrane-bound PRRs. Neutrophils, dendritic cells (DC), natural killer cells, mast cells, and macrophages are a few of the cell types that make up innate immunity. During NMs' uptake by these immune cells, NMs may induce innate immunity by activating the inflammasome, a multiprotein complex that contains PRRs. In response to a trigger, a group of NLRP3 proteins, procaspase-1, and an adaptor protein involved in apoptosis, the speck-like protein, combine to create the NLRP3 inflammasome [bib_ref] The NLRP3 Inflammasome: Molecular Activation and Regulation to Therapeutics, Swanson [/bib_ref] [bib_ref] Inflammasomes: Mechanism of Action, Role in Disease, and Therapeutics, Guo [/bib_ref]. The pro-caspase-1 enzyme cleaves during this traditional activation method, generating active caspase-1. It breaks down pro-IL-18 and pro-IL-1β precursor cytokines to produce IL-18 and IL-1β, proinflammatory cytokines. The NLRP3 gene was shown to be sufficiently activated by potassium ion efflux, commonly known as the reduction of cytosolic potassium ions [bib_ref] K+ Efflux Is the Common Trigger of NLRP3 Inflammasome Activation by Bacterial..., Muñoz-Planillo [/bib_ref] [bib_ref] Cytokines as Targets for Immunomodulation, Wahab [/bib_ref].
Interactions between the IS and the tumor itself play a significant part in the susceptibility or invulnerability of cancer cells to the activity of anti-cancer drugs. On the other hand, the activation of the IS in healthy tissues after treatment with chemotherapy or radiation is connected to both immediate and long-lasting repercussions, including inflammation and fibrosis. Some immune responses may boost toxicity in normal tissue and lessen the effectiveness of anti-cancer treatments. Conversely, manipulating immune responses may increase anti-tumor therapy's efficacy and reduce the toxicity experienced by normal tissue [bib_ref] Modulation of the Immune System by Melatonin; Implications for Cancer Therapy, Moslehi [/bib_ref]. Chemotherapeutic and immunotherapeutic medications influence all tissues. Immunotherapy and chemotherapy are due to severe responses to the IS [bib_ref] Natural Product-Based Nanoformulations for Cancer Therapy: Opportunities and Challenges, Kashyap [/bib_ref] [bib_ref] From Chemotherapy to Biological Therapy: A Review of Novel Concepts to Reduce..., Schirrmacher [/bib_ref] [bib_ref] Targeting Immune Cells for Cancer Therapy, Gun [/bib_ref]. The activity of each kind of immune cell, which includes both immunosuppressive and inflammatory cells, can regulate the activity of other cell types [bib_ref] Cancer-Associated Secondary Lymphoedema, Rockson [/bib_ref]. The development of immunomodulatory medicines designed to restore the host's anti-tumor immune response has led to the identification of new therapeutic targets. Therefore, it is necessary to understand the complex interactions between the environment of the tumor, known cancer-causing agents, the IS, and traditional cancer treatments to make headway against the disease's refractory and recurrent complications and morbidity. This review highlights the complex milieu of cancer immune responses, pointing out the potential nanomaterialsmediated immunomodulation for cancer therapeutics.
Immune cells exhibit varied behavior governed by intricate interactions within the microenvironment of tumors. Although it is generally accepted that an immune response specific to the tumor inhibits cancer progression, it is now known that some forms of inflammation linked to tumors might have the opposite effect. Macrophages, neutrophils, mast cells, and natural killer cells are the cells associated with an adaptive immune response, i.e., T and B lymphocytes. Although it is generally believed that an immune response specific to the tumor suppresses cancer progression, it is evident that some forms of inflammation linked with tumors may potentially have the opposite effect. For example, the generation of immunosuppressive cytokines, an increase in cell proliferation, and resistance to apoptosis can all contribute to the growth of a tumor when TLRs 2,4, and 7/8 are stimulated. Contrarily, tumor inhibition can occur via many mechanisms when the stimulation of TLR 2,3,4,5,7/8, and 9 is coupled with chemo-or immunotherapy. In addition, the activation of TLRs on NK cells and APCs (such as DCs and macrophages) might lead to the production of cytotoxic T lymphocytes, which further suppresses the growth of tumors [bib_ref] Prognostic Value of Innate and Adaptive Immunity in Colorectal Cancer, Grizzi [/bib_ref]. The innate/adaptive immunity and the interaction of immune cells in a tumor with the mechanisms of inhibition and promotion have been shown in [fig_ref] Figure 1: The innate/adaptive immunity and interaction of immune cells in a tumor with... [/fig_ref].
## Nanomaterials for indirect and direct immunomodulation
Nanotechnology has been looked at in many different areas of medicine, including immunology, where its applications are helpful [bib_ref] Nanomaterials for the Delivery of Herbal Bioactive Compounds, Wahab [/bib_ref]. The IS plays a critical role in the body's defenses and has a dispersed nature; hence, using its strengths in therapy has long appealed. Immunomodulation is the process of optimizing the immune response, either by inhibiting it, as in treating autoimmune illnesses or by stimulating it, as in cancer immunotherapy or vaccination. To do this, a wide range of NMs have been studied to see if they can directly affect the IS through their make-up or indirectly through intact active carriers. In this section, NMs and drug delivery systems (DDS) that are utilized to influence the immune response are discussed. It outlines several nanoparticle structural kinds and discusses their makeup and interactions with the IS. In recent research, new nanoparticles and more conventional ones were used. The most advanced ones go beyond
## Nanomaterials for indirect and direct immunomodulation
Nanotechnology has been looked at in many different areas of medicine, including immunology, where its applications are helpful [bib_ref] Nanomaterials for the Delivery of Herbal Bioactive Compounds, Wahab [/bib_ref]. The IS plays a critical role in the body's defenses and has a dispersed nature; hence, using its strengths in therapy has long appealed. Immunomodulation is the process of optimizing the immune response, either by inhibiting it, as in treating autoimmune illnesses or by stimulating it, as in cancer immunotherapy or vaccination. To do this, a wide range of NMs have been studied to see if they can directly affect the IS through their make-up or indirectly through intact active carriers. In this section, NMs and drug delivery systems (DDS) that are utilized to influence the immune response are discussed. It outlines several nanoparticle structural kinds and discusses their makeup and interactions with the IS. In recent research, new nanoparticles and more conventional ones were used. The most advanced ones go beyond administering a single ingredient and go towards combination medication delivery systems and formulations sensitive to stimuli.
There is little room for debate about the fact that nanoparticles provide genuine and novel prospects in various sectors, including healthcare and materials research [bib_ref] Zinc Oxide Nanoparticle: An Effective Antibacterial Agent against Pathogenic Bacterial Isolates, Ahmad [/bib_ref]. These particles are tiny enough to access practically all body parts, including organelles and cells, which may develop novel nanomedicine strategies. The use of nanostructures as diagnostic instruments, biosensors, and carriers for the targeted administration of drugs [bib_ref] Nanotechnology for Drug Delivery: The Perfect Partnership, Farokhzad [/bib_ref] has a significant amount of untapped potential. In the broadest sense, NMs are structures with at least one dimension decreased to 1-100 nm. The European Medicine Agency and the Food and Drug Administration in the United States came up with this expanded range. NMs are distinguished from bulk materials of the same composition by having physicochemical qualities distinct from those of bulk materials, such as the huge surface area-to-mass ratio, ultra-small size, and high reactivity. Compared to their larger counterparts, NPs have more free functional groups accessible for reactions [bib_ref] Nanoparticles in Medicine: Therapeutic Applications and Developments, Zhang [/bib_ref]. Because of their subcellular size, the particles that are used may circulate freely throughout the body and traverse any biological barriers that may be present. This causes a buildup in tissues already prone to it or in specific cells [bib_ref] Nanotechnology for Drug Delivery: The Perfect Partnership, Farokhzad [/bib_ref]. These characteristics of NPs make passive targeting possible by using both their physical properties and the tissues' features, such as enhanced penetration and retention in malignancy or leaky vasculature. On the other hand, active targeting occurs due to NPs' surfaces being transformed in response to different markers produced on specific cells and interactions with their functionalized surfaces. Both approaches of targeted administration increase the amount of the drug at the intended site of action, while decreasing the concentration of that drug in other areas of the body. Consequently, the drug's effectiveness is enhanced, its adverse side effects are reduced, and the dose needed is decreased. People are more likely to take their medications exactly as directed, which may have therapeutic benefits [bib_ref] Nanomaterials for Direct and Indirect Immunomodulation: A Review of Applications, Kubackova [/bib_ref].
NMs may serve as modulators by directly interacting with the IS, altering the immunological response. In all other circumstances, a modulator may be carried by NPs that have drugs loaded onto them, with the payload being the only element intended to modify the IS's response. NPs only operate as intact delivery vehicles while modulating IS in a roundabout way. Monitoring immunological reactions to different stimuli is challenging due to the complexity and widespread nature of the IS. Everyone does not entirely understand the interactions between the IS and NMs. The biodistribution of nanoparticles in vivo is a significant challenge in nanomedicine that must be overcome if immunotherapy and vaccination are to be successful. It is vital to know the distribution of NMs in vivo to control the immune response using NMs successfully. Even though many studies detail NP research in vivo [bib_ref] Particle Size-Dependent and Surface Charge-Dependent Biodistribution of Gold Nanoparticles after Intravenous Administration, Hirn [/bib_ref] , there has only been a little research conducted on how polymeric NPs are distributed across the immune system cells. Much research using model nanocarriers was conducted to comprehend how NMs are distributed in the body. Studies employing proteins encapsulated in poly (lactic-co-glycolic acid) PLA NPs have shown the activation of cytotoxic CD8 + T cells and the development of significant anti-tumor action [bib_ref] Efficient Ex Vivo Induction of T Cells with Potent Anti-Tumor Activity by..., Rosalia [/bib_ref]. Researchers have shown a connection between the biodistribution of AuNPs and several immune system cell subtypes. These cell types include granulocytes, T cells, and DCs [bib_ref] In Vivo Immune Cell Distribution of Gold Nanoparticles in Naïve and Tumor..., Almeida [/bib_ref]. There have also been reports of polystyrene nanoparticles' immunological imprints and differential uptake by B cells, DCs, and macrophages [bib_ref] Differential Uptake of Nanoparticles and Microparticles by Pulmonary APC Subsets Induces Discrete..., Hardy [/bib_ref]. The direct connections between NPs and IS immunity were recently the subject of more thorough research [bib_ref] Nanoparticles and Innate Immunity: New Perspectives on Host Defence, Boraschi [/bib_ref]. When employing NMs as nanocarriers, the objective is to lessen the number of direct contacts between the IS and the carrier while indirectly influencing the IS due to the transported substance. To accomplish this goal, the surface will need to be modified appropriately. Several strategies have been developed to reduce the direct immunomodulatory impact, including altering the surface of NPs to increase their circulating half-life, actively targeting the formulation, or enhancing its uptake by the chosen set of cells [bib_ref] Modulation of Immune Response Using Engineered Nanoparticle Surfaces, Moyano [/bib_ref]. Nanomaterials for indirect and direct immunomodulation have been shown in. . Nanomaterials for indirect and direct immunomodulation. NMs serve as a carrier, and by releasing an immunosuppressive medication or antigen, the immune response is modulated indirectly. Interaction between NMs and IS is a form of direct immunomodulation.
The reticuloendothelial system is the primary organ responsible for phagocytic immune system cell recognition; therefore, NPs that are primarily hydrophobic and charged are promptly opsonized in circulation and may be recognized with greater ease. Conversely, uncharged hydrophilic NPs make for a far less appealing target [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref]. Particles of sizes ranging from 100 nm to 6 μm are the ones that set off the phagocytosis process [bib_ref] Threshold Size for Optimal Passive Pulmonary Targeting and Retention of Rigid Microparticles..., Kutscher [/bib_ref] [bib_ref] Nanoparticles in Cellular Drug Delivery, Faraji [/bib_ref] [bib_ref] Role of Particle Size in Phagocytosis of Polymeric Microspheres, Champion [/bib_ref]. The contact angle between the particle being targeted and the phagocytic cell is of critical significance concerning the form of targeted particles. The conditions most favorable for internalization include spherical or ellipsoid particles approaching at a 45° angle [bib_ref] Role of Target Geometry in Phagocytosis, Champion [/bib_ref]. It is also related to how the elasticity of particles affects their internalization. Particles with a higher elasticity have a greater chance of being distorted during the phagocytosis process; as a result, more stiff particles tend to concentrate in phagocytic cells [bib_ref] Impact of Particle Elasticity on Particle-Based Drug Delivery Systems, Anselmo [/bib_ref]. However, the ultimate interactions between NPs and phagocytic cells are complicated due to the NPs' charge, hydrophobicity, surface chemistry, shape, size, and elasticity [bib_ref] Impact of Particle Elasticity on Particle-Based Drug Delivery Systems, Anselmo [/bib_ref] [bib_ref] Harnessing Nanoparticles for Immune Modulation, Getts [/bib_ref] [bib_ref] Factors Affecting the Clearance and Biodistribution of Polymeric Nanoparticles, Pridgen [/bib_ref].
The half-life in the bloodstream is reduced by phagocytic clearance; however, surface changes may mitigate this effect. These NPs are called "stealth" NPs. Polyethylene glycol (PEG) makes a hydrophilic corona around NPs to protect them [bib_ref] Factors Affecting the Clearance and Biodistribution of Polymeric Nanoparticles, Pridgen [/bib_ref]. PEG surface density and molecular weight are two PEG corona parameters influencing the circulation time [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref]. These properties influence interactions with the IS, and administering the correct dose avoids the formation of PEG antibodies. A shorter circulatory half-life results from what is called the accelerated blood clearance phenomenon [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref] [bib_ref] Targeted Polymeric Therapeutic Nanoparticles: Design, Development and Clinical Translation, Kamaly [/bib_ref]. In addition, a recent study discovered that PEG antibodies modify the biodistribution of NPs in the mucosa [bib_ref] Anti-PEG Antibodies Alter the Mobility and Biodistribution of Densely PEGylated Nanoparticles in..., Henry [/bib_ref]. Therefore, phagocytic cells are an attractive potential target for the control of immunological responses. These phagocytic cells may be actively targeted to increase the distribution of the substance to phagocytic cells. Phosphatidylserine is an "eat me" signaling molecule used often. One of the components that make up the inner cell wall is called a. Nanomaterials for indirect and direct immunomodulation. NMs serve as a carrier, and by releasing an immunosuppressive medication or antigen, the immune response is modulated indirectly. Interaction between NMs and IS is a form of direct immunomodulation.
The reticuloendothelial system is the primary organ responsible for phagocytic immune system cell recognition; therefore, NPs that are primarily hydrophobic and charged are promptly opsonized in circulation and may be recognized with greater ease. Conversely, uncharged hydrophilic NPs make for a far less appealing target [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref]. Particles of sizes ranging from 100 nm to 6 µm are the ones that set off the phagocytosis process [bib_ref] Threshold Size for Optimal Passive Pulmonary Targeting and Retention of Rigid Microparticles..., Kutscher [/bib_ref] [bib_ref] Nanoparticles in Cellular Drug Delivery, Faraji [/bib_ref] [bib_ref] Role of Particle Size in Phagocytosis of Polymeric Microspheres, Champion [/bib_ref]. The contact angle between the particle being targeted and the phagocytic cell is of critical significance concerning the form of targeted particles. The conditions most favorable for internalization include spherical or ellipsoid particles approaching at a 45 - angle [bib_ref] Role of Target Geometry in Phagocytosis, Champion [/bib_ref]. It is also related to how the elasticity of particles affects their internalization. Particles with a higher elasticity have a greater chance of being distorted during the phagocytosis process; as a result, more stiff particles tend to concentrate in phagocytic cells [bib_ref] Impact of Particle Elasticity on Particle-Based Drug Delivery Systems, Anselmo [/bib_ref]. However, the ultimate interactions between NPs and phagocytic cells are complicated due to the NPs' charge, hydrophobicity, surface chemistry, shape, size, and elasticity [bib_ref] Impact of Particle Elasticity on Particle-Based Drug Delivery Systems, Anselmo [/bib_ref] [bib_ref] Harnessing Nanoparticles for Immune Modulation, Getts [/bib_ref] [bib_ref] Factors Affecting the Clearance and Biodistribution of Polymeric Nanoparticles, Pridgen [/bib_ref].
The half-life in the bloodstream is reduced by phagocytic clearance; however, surface changes may mitigate this effect. These NPs are called "stealth" NPs. Polyethylene glycol (PEG) makes a hydrophilic corona around NPs to protect them [bib_ref] Factors Affecting the Clearance and Biodistribution of Polymeric Nanoparticles, Pridgen [/bib_ref]. PEG surface density and molecular weight are two PEG corona parameters influencing the circulation time [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref]. These properties influence interactions with the IS, and administering the correct dose avoids the formation of PEG antibodies. A shorter circulatory half-life results from what is called the accelerated blood clearance phenomenon [bib_ref] PEGylation as a Strategy for Improving Nanoparticle-Based Drug and Gene Delivery, Suk [/bib_ref] [bib_ref] Targeted Polymeric Therapeutic Nanoparticles: Design, Development and Clinical Translation, Kamaly [/bib_ref]. In addition, a recent study discovered that PEG antibodies modify the biodistribution of NPs in the mucosa [bib_ref] Anti-PEG Antibodies Alter the Mobility and Biodistribution of Densely PEGylated Nanoparticles in..., Henry [/bib_ref]. Therefore, phagocytic cells are an attractive potential target for the control of immunological responses. These phagocytic cells may be actively targeted to increase the distribution of the substance to phagocytic cells. Phosphatidylserine is an "eat me" signaling molecule used often. One of the components that make up the inner cell wall is called a phospholipid. Phosphatidylserine is displayed on the surface of a cell as a signal for phagocytic cells to scavenge the injured cell if the cell membrane is broken or the cell is undergoing apoptosis [bib_ref] Eat Me, Bagalkot [/bib_ref]. The injection of these liposomes was shown to be effective in reducing inflammation in an in vivo model. The phospholipid of concern was employed as a component of the liposomal bilayer [bib_ref] Apoptotic Mimicry: Phosphatidylserine Liposomes Reduce Inflammation through Activation of Peroxisome Proliferator-Activated Receptors..., Ramos [/bib_ref]. Other nanomaterials, including polylactic acid nanoparticles (PLGA NPs), have their surfaces modified using phosphatidylserine [bib_ref] Nanoparticles Containing a Liver X Receptor Agonist Inhibit Inflammation and Atherosclerosis, Zhang [/bib_ref] and CNT [bib_ref] Phosphatidylserine Targets Single-Walled Carbon Nanotubes to Professional Phagocytes in Vitro and in..., Konduru [/bib_ref] , to target the compositions of macrophages. This was conducted to target the formulations. In addition to organic molecules, the attachment of tiny functional groups up to 500 daltons in size has been described. These functional groups include alcohols, carboxylic acids, primary amines, sulfhydryls, and anhydrides. Several distinct cell lines, including active and resting macrophages, were tested for the presence of magnetic nanoparticles with functionalized monocrystal surfaces. Even tiny functional groups effectively targeted various cell types and diverse physiological states in macrophages [bib_ref] Cell-Specific Targeting of Nanoparticles by Multivalent Attachment of Small Molecules, Weissleder [/bib_ref].
When a dead cell is merged with an antigen, an immunogenic cell death (ICD) process initiates the adaptive immunological response in the immune host. Dying tumor cells associated with damage-associated molecular patterns such as calreticulin exposure, ATP secretion, ANXA1, and type I interferons (IFNs) expression, and the release of non-histone nuclear protein high-mobility group box 1, altogether promote cell corpses and debris engulfment by antigen-presenting cells, resulting in dendritic cell maturation. Additionally, dendritic cell activation encourages CD4 + and CD8 + T cell priming, triggering cytotoxic T lymphocyte (CTL) and immunogenic T helper 1 (Th1) cell responses. These are the key steps in ICD-induced immune cell realization and immunosuppressive retaliation. Antigens and adjuvants are delivered via nanoparticle-based delivery systems that target lymph nodes that drain tumors [bib_ref] Nanoparticle-Mediated Immunogenic Cell Death Enables and Potentiates Cancer Immunotherapy, Duan [/bib_ref]. Recent research on NPs for ICD-inducer delivery into tumor cells aims to enhance the immunostimulatory effects, and, subsequently, cancer immunotherapy. The excitation of immunostimulatory cells, cytokines, and chemokines, aided by immunoinhibitory cells and cytokine suppression (IL-4, IL-6, and IL-10), revealed a potential oxaliplatin-mediated ICD in the nano-folox [bib_ref] Nano Codelivery of Oxaliplatin and Folinic Acid Achieves Synergistic Chemo-Immunotherapy with 5-Fluorouracil..., Guo [/bib_ref]. Studies have shown that ROS is crucial for ICD induction, suggesting that the combined effects of ROS-triggering strategies and NP-based ICD-induction treatments might improve cancer immunotherapy. A systemic delivery platform nanoscale coordination polymer core-shell particle aimed to transport both ROS-triggering agent dihydroartemisinin and chemotherapeutic agent oxaliplatin [bib_ref] Immunostimulatory Nanomedicines Synergize with Checkpoint Blockade Immunotherapy to Eradicate Colorectal Tumors, Duan [/bib_ref] [bib_ref] Cirsilineol Inhibits the Proliferation of Human Prostate Cancer Cells by Inducing Reactive..., Wahab [/bib_ref].
However, there are few kinetic investigations of polymeric NPs' cellular biodistribution in the IS. The results of research carried out by Yang and Luo showed that NPs (polystyrene yellow-green, 500 nm) are biodistributed in immunological organs, which implies that they might be helpful in the rations design of formulations. It is crucial for developing immunotherapies based on the targeted administration of NPs to comprehend the kinetics of biodistribution of polymeric NPs in the IS. Blood and bone marrow doublenegative cells, splenic dendritic cells (DCs), and monocytes were all evaluated for their phagocytic capacity. It would be helpful to grasp how NPs are distributed throughout the body in vivo to improve or modify immunity [bib_ref] Cellular Biodistribution of Polymeric Nanoparticles in the Immune System, Yang [/bib_ref].
The advent of nanotechnology has opened the possibility of modifying the immune response to either reduce undesirable immune overreactions or adjust interactions between the IS and incredibly deadly substances. In the domain of immunology, NPs are being used as vehicles for the delivery of drugs. In addition, transporters for immunoadjuvants or antigens are also included in the product. They are divided into groups according to the substance's chemical structure from the vast pool of natural materials already in existence. The non-viral, physiologically applicable drug delivery nanosystems are the focus of this review. In the field of life sciences, scientists are actively using a wide variety of nanomaterials for various applications. [fig_ref] Table 1: Different immunotherapies are based on nanomaterials [/fig_ref] provides an overview of several immunotherapies that use nanomaterials.
## Physicochemical properties of nanomaterials and their impact on the immune system
In biomedical research, nanomaterials consisting of polymers, lipids, and inorganic materials have several potential applications [bib_ref] Novel Drug Delivery Systems as an Emerging Platform for Stomach Cancer Therapy, Hani [/bib_ref] [bib_ref] Metal Nanomaterials: Immune Effects and Implications of Physicochemical Properties on Sensitization, Elicitation,..., Roach [/bib_ref] [bib_ref] Lipid Nanoparticle Formulations for Enhanced Co-Delivery of SiRNA and MRNA, Ball [/bib_ref] [bib_ref] Size-Dependent EPR Effect of Polymeric Nanoparticles on Tumor Targeting, Kang [/bib_ref]. The immunological homeostasis is interfered with by a wide range of NMs' physicochemical features, which increase the unpredictability of nanomedicine in vivo [bib_ref] Immunosuppressive and Anti-Inflammatory Properties of Engineered Nanomaterials, Ilinskaya [/bib_ref]. Nanocarriers, which can be changed to deliver different therapies to tune the IS, are a crucial way to reach these goals [bib_ref] A Noninflammatory mRNA Vaccine for Treatment of Experimental Autoimmune Encephalomyelitis, Krienke [/bib_ref] [bib_ref] Synthetic Nanoparticles for Vaccines and Immunotherapy, Irvine [/bib_ref]. Their effectiveness has been chiefly ascribed to the direct action of treatments that have been delivered and to improved targeting of immune cells or diseased tissues, with the impact of nanocarriers on immunological regulation receiving little attention [bib_ref] Particle Shape Dependence of CD8+ T Cell Activation by Artificial Antigen Presenting..., Sunshine [/bib_ref]. The dominant factors affecting the outcome of immune modulation are the shape, size, charge, rigidity, and surface chemical composition of NMs. These criteria are taken into consideration while reviewing the immunomodulatory abilities of NMs. Even though the results of immunological modulation are usually the combined impacts of several parameters, examining the effects of specific components may provide insights into how best to tune the physicochemical features of NMs for immunomodulation. The nm-induced immunological responses shown in [fig_ref] Table 2: Nanomaterial-induced immune responses concerning nanomaterials' physicochemical properties, such as surface, charge, rigidity,... [/fig_ref] indicate a wide range of nm physicochemical parameters, including shape, size, charge, rigidity, and surface composition. The ability of nanotechnology to modulate the IS opens the possibility of treating a wide variety of ailments [bib_ref] Dual-Targeting Nanoparticle Vaccine Elicits a Therapeutic Antibody Response against Chronic Hepatitis B, Wang [/bib_ref] [bib_ref] Applications of Nanotechnology for Immunology, Smith [/bib_ref]. However, the outcome is negatively impacted by the different NMs' ability to provoke uncontrolled immunological responses. The importance of natural molecules in immune regulation was further established by the phenomenal success of two mRNA COVID-19 vaccines developed by Moderna and Pfizer-BioNTech [bib_ref] Efficacy and Safety of the MRNA-1273 SARS-CoV-2 Vaccine, Baden [/bib_ref] [bib_ref] Safety and Efficacy of the BNT162b2 MRNA COVID-19 Vaccine, Polack [/bib_ref] [bib_ref] Use of Natural Compounds as a Potential Therapeutic Agent Against COVID-19, Wahab [/bib_ref]. This review discusses the immunomodulatory effects of NMs concerning these variables.
## Size-dependent immunomodulation of nms
NMs' size is a significant consideration for nanomedicines to modulate the IS [bib_ref] Size, Shape, and Protein Corona Determine Cellular Uptake and Removal Mechanisms of..., Ding [/bib_ref]. Their interstitial mobility and biodistribution in vivo are controlled by the nm size of the particles [bib_ref] The Effect of Nanoparticle Size, Shape, and Surface Chemistry on Biological Systems, Albanese [/bib_ref] [bib_ref] Glomerular Barrier Behaves as an Atomically Precise Bandpass Filter in a Sub-Nanometre..., Du [/bib_ref] [bib_ref] Exploiting Lymphatic Transport and Complement Activation in Nanoparticle Vaccines, Reddy [/bib_ref] [bib_ref] Influence of Size and Shape on the Anatomical Distribution of Endotoxin-Free Gold..., Talamini [/bib_ref]. The primary difficulty associated with their systemic administration is the elimination of NMs from the body during normal blood circulation [bib_ref] Engineering Long-Circulating Nanomaterial Delivery Systems, Fan [/bib_ref]. In a study, researchers examined the polymeric NPs to precisely target and control the co-delivery of medicine with various physicochemical properties to cancer cells. They co-delivered the docetaxel and cisplatin to prostate cancer cells with synergistic cytotoxicity. The results show that the kidney efficiently clears NMs smaller than 5 nm. When compared to NMs larger than this size range, those with a size between 50 and 100 nm often have a greater lifetime in circulation. In vitro toxicities showed that the targeted dual-drug combination NPs were better than NPs with a single drug or NPs that were not targeted [bib_ref] Engineering of Self-Assembled Nanoparticle Platform for Precisely Controlled Combination Drug Therapy, Kolishetti [/bib_ref]. PLLA-b-PEG polymersomes with a single crystal-like crystalline structure and an average diameter of around 200 nm have very long blood circulation times, with 47% of the injected NP still present in the blood 24 h after injection [bib_ref] Copolymer Crystalsomes with an Ultrathin Shell to Extend Blood Circulation Time, Qi [/bib_ref].
The health risks of AgNP are likely to increase with the increasing number of NPcontaining products and have shown an adverse reaction in various cell lines. The amount of exposure and a fixed time point measure the results of a toxicology test. The kinetics of NP uptake and the time-dependent intracellular concentration are not commonly considered. The initial line of defense against foreign invaders, such as NPs, is provided by macrophages. The macrophage response to NPs is crucial in determining whether the NPs are harmful. However, investigations on the uptake of nanometer-sized particles and macrophage-like cells are severely lacking. The research was conducted on which uptake rates were measured over 24 h for three different sizes of AgNPs (20, 50, and 75 nm) in a medium containing and without fetal calf serum. The non-toxic concentration of 10 ng Ag/mL for monocytic THP-1 cells was used for this study. This concentration represents a realistic exposure level for short-term exposures. The uptake of silver was more significant in a medium that did not include fetal calf serum, and the results demonstrated that the uptake increased with decreasing NP sizes, both in terms of the NP mass and the NP number. This study's findings indicate that the uptake rate of NPs by macrophages varies depending on the size of the NPs [bib_ref] Uptake of Silver Nanoparticles by Monocytic THP-1 Cells Depends on Particle Size..., Kettler [/bib_ref]. The nanoparticle shape influences antibody and cytokine production [bib_ref] From Immunotoxicity to Nanotherapy: The Effects of Nanomaterials on the Immune System, Smith [/bib_ref].
## Immunomodulation of nms by shape
Currently, many medicines have been authorized for use or are undergoing clinical studies, and it is anticipated that nanotechnology will soon be included in many commercial items [bib_ref] Toxic Potential of Materials at the Nanolevel, Nel [/bib_ref] [bib_ref] Immunological Properties of Engineered Nanomaterials. In Nanoscience and Technology: A Collection of..., Dobrovolskaia [/bib_ref]. The shape of the NMs is another critical element that plays a role in their immunomodulation and size [bib_ref] The Effect of Nanoparticle Size, Shape, and Surface Chemistry on Biological Systems, Albanese [/bib_ref] [bib_ref] Glomerular Barrier Behaves as an Atomically Precise Bandpass Filter in a Sub-Nanometre..., Du [/bib_ref] [bib_ref] Exploiting Lymphatic Transport and Complement Activation in Nanoparticle Vaccines, Reddy [/bib_ref]. The research was conducted to determine the precise impact of shape on the biodistribution of predetermined AuNPs after intravenous delivery in mice. They integrated quantitative data derived by inductively coupled plasma mass spectrometry with observational results from histochemistry. Researchers utilized healthy mice that could mount an immunological response since the bio-nano interaction involves the IS. It has been proven that the form of the nm has a role in determining whether they can avoid being removed by the reticuloendothelial system. In filter organs, the kinetics of AuNP buildup and excretion are significantly influenced by shape [bib_ref] Influence of Size and Shape on the Anatomical Distribution of Endotoxin-Free Gold..., Talamini [/bib_ref]. It is known that µm-sized flexible filaments are less absorbed by macrophages than spheres and short filomicelles, due to their expanded conformation under blood flow, lengthening the blood circulation of long filaments [bib_ref] Shape Effects of Filaments versus Spherical Particles in Flow and Drug Delivery, Geng [/bib_ref]. The varied ways NMs interact with cells are the root cause of the disparity in their biodistribution. The nm form partially governs the interactions. As a result, nm shape may be manipulated to control tissue-targeting and immunological regulation [bib_ref] The Biodistribution and Immuno-Responses of Differently Shaped Non-Modified Gold Particles in Zebrafish..., Van Pomeren [/bib_ref] [bib_ref] Shape-Dependent Cytotoxicity and Proinflammatory Response of Poly(3,4-Ethylenedioxythiophene) Nanomaterials, Oh [/bib_ref] [bib_ref] Influences of Nanocarrier Morphology on Therapeutic Immunomodulation, Frey [/bib_ref]. However, the ultimate interactions between NPs and phagocytic cells are complicated due to the NPs' charge, hydrophobicity, surface chemistry, shape, size, and elasticity [bib_ref] Impact of Particle Elasticity on Particle-Based Drug Delivery Systems, Anselmo [/bib_ref] [bib_ref] Harnessing Nanoparticles for Immune Modulation, Getts [/bib_ref] [bib_ref] Factors Affecting the Clearance and Biodistribution of Polymeric Nanoparticles, Pridgen [/bib_ref].
## Immunomodulation of nms by rigidity
Immune cells can detect and react to biophysical stimuli ranging from dynamic stresses to spatial characteristics throughout their formation, activation, differentiation, and expansion. These biophysical signals control the functions of immune cells, such as the release of leukocytes, the selection and activation of T cells, and the polarization of macrophages. In addition, integrins and focal adhesion complexes play a significant role in the contact between the cell and the matrix. Ion channels are another kind of mechanosensors. These channels gate soluble ions such as Na + , K + , and Ca 2+ [bib_ref] Mechanosensation of Cyclical Force by PIEZO1 Is Essential for Innate Immunity, Solis [/bib_ref]. The function of each cell is governed by the aggregate signals that are received from a variety of immunoreceptors. The expression and activity of immunoreceptors are contingent on the cell's current development stage and the surrounding environment [bib_ref] Mechanosensing through Immunoreceptors, Zhu [/bib_ref]. Recent research has shed light on the existence of mechanical force on several different immunoreceptor-ligand pairs, as well as the significant role that force plays in regulating the interaction and function of these couples. The pharmacokinetics of nm and the effectiveness of intracellular drug delivery are both affected by the mechanical forces created during interactions between NMs and cells [bib_ref] Size and Rigidity of Cylindrical Polymer Brushes Dictate Long Circulating Properties in..., Müllner [/bib_ref] [bib_ref] In Vivo Fate of Carbon Nanotubes with Different Physicochemical Properties for Gene..., Cifuentes-Rius [/bib_ref]. The effectiveness of cancer vaccines based on peptides is limited in people, despite the enormous promise these vaccines provide. Recent advancements have heralded a new age of personalized immunotherapy using patient-specific neoantigens in tumor exome sequencing. Yet, there is still a lack of a general strategy for inducing potent CD8α+ cytotoxic T-lymphocyte responses. In addition, vaccination with many epitopes led to broad-spectrum T cell responses, effectively preventing tumor development. When paired with treatment targeting anti-PD-1 and anti-CTLA-4 receptors, nanodiscs successfully eradicated both the MC-38 and B16F10 tumors. These results indicate a broad method for tailored nanomedicine and provide a novel, very effective approach to the treatment of cancer immunotherapy [bib_ref] Designer Vaccine Nanodiscs for Personalized Cancer Immunotherapy, Kuai [/bib_ref]. When developing delivery systems for vaccines and immunotherapies, it is essential to consider the relevance of biomechanics and rigidity. It is possible to alter the rigidity of NMs to direct them to the lymphoid organs and enhance their biodistribution. As a result, soft NMs could have an advantage in homing to lymph nodes, which would boost the administration of immunomodulators.
## Immunomodulation of nms by surface charge
Another physicochemical characteristic influencing nm fate is surface charge [bib_ref] Effects of Particle Size and Surface Modification on Cellular Uptake and Biodistribution..., Kulkarni [/bib_ref] [bib_ref] The Effect of Surface Charge on in Vivo Biodistribution of PEG-Oligocholic Acid..., Xiao [/bib_ref] [bib_ref] Effects of Nanoparticle Charge and Shape Anisotropy on Translocation through Cell Membranes, Nangia [/bib_ref]. Most DDS are colloidal and may have positive or neutral surface charges. Researchers attempted to control the surface charge of drug carriers through the change of surface chemistry or other techniques, including introducing positive or negative charges to the surface of DDS. The surface charge may be affected in several different ways. A neutral surface charge increases circulation time and inhibits the adsorption of plasma proteins to particle surfaces. On the other hand, a positive surface charge may facilitate improved contact with the cellular membrane and internalization; nevertheless, this may also have a toxic effect on the cells. A negative charge not only has a less harmful impact, but also decreases the cell's ability to adsorb particles [bib_ref] Enhanced Antigen Presentation and Immunostimulation of Dendritic Cells Using Acid-Degradable Cationic Nanoparticles, Kwon [/bib_ref] [bib_ref] Aerosol Delivery of Functionalized Gold Nanoparticles Target and Activate Dendritic Cells in..., Fytianos [/bib_ref]. It may be possible to create an appropriate carrier for DDS by imparting a positive charge on the surface of particles, combining with incorporating stealthy materials (such as PEG) or targeting agents [bib_ref] The Effect of Surface Charge on in Vivo Biodistribution of PEG-Oligocholic Acid..., Xiao [/bib_ref]. The charge that is present on the surface of the DDS has the potential to increase cellular uptake. This is particularly true if the charge is positive, as this may interact with the negative charge on the cell membrane to stimulate adsorption.
The contact between NMs and immune cells may be mediated by proteins adsorbed on their positively charged surfaces. There have also been contradictory reports about the influence of charge on nm internalization by immune cells [bib_ref] Differential Uptake of Functionalized Polystyrene Nanoparticles by Human Macrophages and a Monocytic..., Lunov [/bib_ref]. Aside from cell uptake, the nm surface charge influences nm biodistribution and immune cell activation in vivo [bib_ref] Liposome-Induced Complement Activation and Related Cardiopulmonary Distress in Pigs: Factors Promoting Reactogenicity..., Szebeni [/bib_ref] [bib_ref] Selective Organ Targeting (SORT) Nanoparticles for Tissue-Specific MRNA Delivery and CRISPR-Cas Gene..., Cheng [/bib_ref] [bib_ref] Delivery of Tacrolimus with Cationic Lipid-Assisted Nanoparticles for Ulcerative Colitis Therapy, Wang [/bib_ref] [bib_ref] Charge-Convertible Polymers for Improved Tumor Targeting and Enhanced Therapy, Li [/bib_ref]. Researchers intravenously injected mice with functionalized gold nanoparticles and employed quantitative imaging based on laser ablation inductively coupled plasma mass spectrometry to display the surface charge changes in the suborgan distributions of NPs in the liver, kidney, and spleen. The kidney images show that positively charged nanoparticles accumulate extensively in the glomeruli, which is the initial stage in the filtering process for the nephron. It suggests that the kidney may filter these nanoparticles at a different rate than the neutral or negatively charged nanoparticles that the kidney would filter. The red pulp of the spleen is where researchers observed a significant accumulation of nanoparticles with both positive and negative charges. However, unlike positively or negatively charged nanoparticles, uncharged particles build up more in the spleen's white pulp and peripheral zone. In addition, the likelihood of these uncharged nanoparticles being identified as connected with Kupffer cells in the liver is increased. Nanoparticles with a positive charge build up in the liver hepatocytes, while nanoparticles with a negative charge have a more widespread distribution throughout the liver. These observations, taken together, point to the possibility that neutral nanoparticles with cores measuring 2 nm may interact with the IS to a greater extent than charged nanoparticles. This finding highlights the importance of determining the suborgan distributions of nanomaterials for applications involving delivery and imaging [bib_ref] Surface Charge Controls the Suborgan Biodistributions of Gold Nanoparticles, Elci [/bib_ref]. The spleen is the body's largest secondary immune organ. It is vital to start immunological responses to blood-borne antigens and filter the blood of foreign substances and damaged or old red blood cells. The red and white pulp, the two primary compartments of the spleen, are responsible for performing these duties. The red and white pulps are quite distinct in their vascular organization, architecture, and cellular makeup. The charged AuNPs exhibited a substantially lower concentration in the spleen's white pulp and marginal zone than the neutral AuNPs [bib_ref] Normal Structure, Function, and Histology of the Spleen, Cesta [/bib_ref].
Furthermore, neutral AuNPs were found in Kupffer cells in the liver. Researchers have shown that DCs can be precisely and effectively targeted in vivo by giving RNA-lipoplexes based on well-known lipid carriers through an IV. This is done by adjusting the net charge of the particles in the best way possible without adding molecular ligands to the particles. RNA-LPX (RNA-lipoplexes) represent a universally applicable vaccine class for systemic DC targeting and synchronized induction of both highly potent adaptive as well as type-I-IFN-mediated innate immune mechanisms for cancer immunotherapy [bib_ref] Systemic RNA Delivery to Dendritic Cells Exploits Antiviral Defence for Cancer Immunotherapy, Kranz [/bib_ref]. Gene editing using CRISPR-Cas and protein replacement therapy based on messenger RNA promise to efficiently cure disease-causing mutations that may arise from various cell types. A method is known as selective organ targeting (SORT), in which several types of lipid nanoparticles are methodically built to specifically edit extrahepatic tissues by incorporating an additional SORT molecule. By changing the composition of the permanently charged lipids included in the lipid nanoparticles (LNPs), it was possible to target specific organs in mice after the systemic injection of mRNA-delivery LNPs [bib_ref] Selective Organ Targeting (SORT) Nanoparticles for Tissue-Specific MRNA Delivery and CRISPR-Cas Gene..., Cheng [/bib_ref]. These results suggested that NMs with neutral or low negative charges may be more efficient in targeting DCs and macrophages in the spleen than their positively charged counterparts. The ability of NMs to adsorb proteins is highly dependent on their surface charge. Further research to identify corona proteins that mediate APC targeting may aid in developing NMs for effective APC targeting. The physicochemical properties of nanomaterials and mechanisms of cell death in cancer induced by nanoparticles have been illustrated in [fig_ref] Figure 3: Physicochemical properties of nanomaterials and mechanisms of cell death in cancer induced... [/fig_ref]. Differently shaped nanocarriers have their spatial characteristics and some unique benefits and drawbacks. Although NP-based nanomedicine has significantly contributed to cancer immunotherapy, more research is needed to understand the disadvantages of NP-mediated immunogenicity, non-targeted cellular uptake and cytotoxicity, and the precise interactions between NPs and the IS. Furthermore, translating NP-based nanomedicine to clinical applications is difficult due to the scarcity of evidence on the immune system's role in tumor genesis and progression. Because of the intricacy of innate and adaptive immunity, determining the impact of one component's depletion or suppression on the entire immune network is difficult. Furthermore, different tumors have heterogeneous structures, making it difficult to predict how they respond to NP-based nanomedicine. Consequently, developing suitable NP platforms for distinct cancers is crucial and difficult [bib_ref] Advancement of Cancer Immunotherapy Using Nanoparticles-Based Nanomedicine, Gowd [/bib_ref]. Current toxicity studies for NPs are in their infancy due to their complex physiochemical characteristics and possible interactions with biological components already present in the body. Neither their ultimate structural shape nor their safety after therapeutic administration is known.
## Advantages and disadvantages of the different types of nanocarriers
Nanodrugs may enhance many pharmacological characteristics of traditional (or "free") drugs [bib_ref] Tumour-Specific Uptake of Anti-Cancer Drugs: The Future Is Here, Caraglia [/bib_ref]. NP systems' primary properties are the Zeta potential, particle size, and size distribution. NPs are distinguished from bulk materials (on the microscale) of the same composition by their ultra-small size, enormous surface area-to-mass ratio, and high Differently shaped nanocarriers have their spatial characteristics and some unique benefits and drawbacks. Although NP-based nanomedicine has significantly contributed to cancer immunotherapy, more research is needed to understand the disadvantages of NPmediated immunogenicity, non-targeted cellular uptake and cytotoxicity, and the precise interactions between NPs and the IS. Furthermore, translating NP-based nanomedicine to clinical applications is difficult due to the scarcity of evidence on the immune system's role in tumor genesis and progression. Because of the intricacy of innate and adaptive immunity, determining the impact of one component's depletion or suppression on the entire immune network is difficult. Furthermore, different tumors have heterogeneous structures, making it difficult to predict how they respond to NP-based nanomedicine. Consequently, developing suitable NP platforms for distinct cancers is crucial and difficult [bib_ref] Advancement of Cancer Immunotherapy Using Nanoparticles-Based Nanomedicine, Gowd [/bib_ref]. Current toxicity studies for NPs are in their infancy due to their complex physiochemical characteristics and possible interactions with biological components already present in the body. Neither their ultimate structural shape nor their safety after therapeutic administration is known.
## Advantages and disadvantages of the different types of nanocarriers
Nanodrugs may enhance many pharmacological characteristics of traditional (or "free") drugs [bib_ref] Tumour-Specific Uptake of Anti-Cancer Drugs: The Future Is Here, Caraglia [/bib_ref]. NP systems' primary properties are the Zeta potential, particle size, and size distribution. NPs are distinguished from bulk materials (on the microscale) of the same composition by their ultra-small size, enormous surface area-to-mass ratio, and high reactivity. NPs may encapsulate and transport medications that are not easily soluble when used as therapeutic carriers [bib_ref] Toxicological Considerations When Creating Nanoparticle-Based Drugs and Drug Delivery Systems, Sharma [/bib_ref]. These features are often linked to highly desired attributes (electrical, mechanical, and chemical) for specialized medicinal purposes. Still, they may also be the primary determinants defining their potentially harmful consequences on human health [bib_ref] Risk Assessment of Engineered Nanomaterials and Nanotechnologies-A Review, Savolainen [/bib_ref]. A comparison of various nanoparticles has been exhibited in . To better understand how nanostructures interact with biological systems, different international scientific societies have emphasized the significance of developing nanotoxicology, a key subdiscipline of nanotechnology. This discipline focuses on elucidating the connections between nanostructures' physical and chemical properties and the induction of toxic biological responses [bib_ref] Risk Assessment of Engineered Nanomaterials and Nanotechnologies-A Review, Savolainen [/bib_ref]. Nanodrugs can enter the body via main routes such as lung, subcutaneous, intraperitoneal, intravenous, and oral. The nanomaterials may be toxic at these phases through many mechanisms, such as inflammatory and pro-oxidant activities. The toxicity profiling of NMs has been a highly demanded research area worldwide in recent times. Natural NMs have been a part of the ecosystem for a long time and contain various processes that make them less toxic to living things. Research breakthroughs have shown some immediate hazardous impacts of nanosized particles in biological systems. Emerging NPs, such as viral NPs and nanozymes, should undergo thorough cytotoxicity experiments to determine safe dose levels and application procedures. The success of some nanoparticle-based medications, such as the COVID-19 mRNA vaccines, has generated interest among the public and scientific community regarding their potential application in the treatment of a variety of other diseases, including discussions about the development of a future cancer vaccine. A vaccination for cancer is different from one for an infectious illness. Cancer vaccines may need to use a variety of approaches to overcome treatment resistance. For example, a vaccination using nanoparticles must be designed differently for injection into the blood instead of the muscle. Although the area of nanomedicine has made significant strides in moving medications or diagnostics from the lab and into the clinic, there is still a long way to go. Learning from past successes and failures can help researchers develop breakthroughs that allow nanomedicine to live up to its promise. The benefits and drawbacks of the various nanomaterials have been listed in . . Benefits and drawbacks of the various kinds of nanocarriers.
## Types of nanocarriers drawbacks benefits
## Metallic nanoparticles
Particles' instability, impurity, biologically harmful, explosion, difficulty in synthesis, toxicity Biocompatible; strong plasma uptake; and uniformity in size, shape, and branch length. Tuned pharmacokinetics and biodistribution Dendrimers Low hydro solubility and high non-specific toxicity, poly(amidoamine) (PAMAM) dendrimers, and PPI dendrimers attributed toward toxic manifestations Water soluble and biocompatible, good PK behavior, flexibility in conjugation chemistry, and ability to encapsulate and deliver various bioactive agents [bib_ref] In Vitro/In Vivo Evaluation of Dexamethasone -PAMAM Dendrimer Complexes for Retinal Drug..., Yavuz [/bib_ref] [bib_ref] Dendrimers in Drug Delivery and Targeting: Drug-Dendrimer Interactions and Toxicity Issues, Madaan [/bib_ref] Polymeric micelles The low payload of drugs and less stability in an aqueous medium Biodegradable, self-assembling, and biocompatible. Potential targeting of functional modification, efficient carrier system for hydrophilic drugs, biodegradable
Carbon nanotubes Poorly soluble in water, not biodegradable, toxicity concerns, poor PK (pharmacokinetics) Ease of synthesis and conjugation of multiple bioactive agents, large surface area, ability to encapsulate and deliver various types of bioactive agents, protects entrapped drug and provides sustained releaseLiposomes Fewer stables, leakage, and fusion of encapsulated drug/molecules; high production cost; some may be allergic Targeted to specific cells or tissues, biocompatible, longer duration of circulation, high stability via encapsulation, high efficacy and therapeutic index of drug
# Nanomaterial-based immunotherapy
Cancer immunotherapy has received much attention recently due to its distinct features and consequences that no other cancer therapies can match. Immunotherapy for cancer may systematically target both primary and secondary tumor metastases. Nanotechnologies have opened up a new direction for research and development, allowing for the efficient delivery of drugs to specific areas of the body and the active targeting of particular cell populations, such as tumor cells or subsets of immune cells. NMs have been employed extensively in research on cancer immunotherapy because of their unique benefits. NMs' advancements in cancer immunotherapy, alone or in conjunction with other therapies, include dendritic cell (DC)-targeted delivery systems, self-adjuvants, combination therapy, and engineered APCs [bib_ref] Nanomaterial-Based Cancer Immunotherapy, Luo [/bib_ref]. In addition, to improve the response, co-delivery and targeting of nanocarriers are appealing options [bib_ref] Targeted PLGA Nano-but Not Microparticles Specifically Deliver Antigen to Human Dendritic Cells..., Cruz [/bib_ref] [bib_ref] Vaccination with OVA-Bound Nanoparticles Encapsulating IL-7 Inhibits the Growth of OVA-Expressing E.G7..., Toyota [/bib_ref] [bib_ref] Targeting the Tumor-Draining Lymph Node with Adjuvanted Nanoparticles Reshapes the Anti-Tumor Immune..., Thomas [/bib_ref] [bib_ref] Nanoparticle Conjugation of CpG Enhances Adjuvancy for Cellular Immunity and Memory Recall..., De Titta [/bib_ref]. Due to their outstanding physicochemical characteristics, which include size, shape, and surface features and result in preferable biological interactions, nanoparticles are a frequently utilized nanomedicine platform in cancer immunotherapy. Furthermore, cancer immunotherapy has excellent potential for using tailored nanostructural materials such as nano-emulsions, nanotubes, and NPs [bib_ref] Advancement of Cancer Immunotherapy Using Nanoparticles-Based Nanomedicine, Gowd [/bib_ref].
NPs' essential and physiochemical characteristics are influenced by many cancer treatment modalities, including chemotherapeutics, nucleic acid-based therapies, photothermal therapy, and photodynamic agents [bib_ref] Cancer Immunotherapy by Immune Checkpoint Blockade and Its Advanced Application Using Bio-Nanomaterials, Yadav [/bib_ref]. Antigen-presenting cells may take up nanoparticles to facilitate the cytosolic transport of encapsulated antigens and adjuvants [bib_ref] Advancement of Cancer Immunotherapy Using Nanoparticles-Based Nanomedicine, Gowd [/bib_ref]. They may be divided into numerous categories. The NP-based immunotherapy termed ARAC (Antigen Release Agent and Checkpoint Inhibitor) is designed to enhance the efficacy of PD-L1 inhibitors. PLK1(polo-like kinase 1) inhibition increases the expression of PD-L1 in cancer cells, reducing cytotoxic T lymphocytes' effectiveness. PLK1 inhibition and cancer immunosuppression support the use of the PD-L1 immune checkpoint blockade in conjunction with PLK1 inhibitors as a possible therapeutic approach [bib_ref] Development of a Nanoparticle-Based Immunotherapy Targeting PD-L1 and PLK1 for Lung Cancer..., Reda [/bib_ref]. Nanoparticles can modulate innate immune cells, including monocytes, NK cells, TAMs, neutrophils, DCs, and MDSCs. For instance, TAMs can act as antigen-presenting cells and produce different solubility to interact with other immune cells. In addition, they play a crucial role in cancer immunotherapy [bib_ref] Nanomaterials for Modulating Innate Immune Cells in Cancer Immunotherapy, Le [/bib_ref] [bib_ref] Nanotechnology-Based Products for Cancer Immunotherapy, Shams [/bib_ref]. Nanocarriers may enhance medications' pharmacokinetic and pharmacological characteristics by improving the drug's water solubility and stability in circulation. Additionally, they allow for tissue or cell-specific drug administration; reducing drug buildup in the kidneys, liver, and other non-targeted organs; and enhancing the therapeutic effectiveness and drug delivery of a medication cocktail [bib_ref] Cancer Drug Delivery in the Nano Era: An Overview and Perspectives (Review), Li [/bib_ref] [bib_ref] Therapeutic Advancements in the Management of Diabetes Mellitus with Special Reference to..., Chauhan [/bib_ref] [bib_ref] Effective Use of Nanocarriers as Drug Delivery Systems for the Treatment of..., Din [/bib_ref]. The conjugated therapeutic is shielded from deterioration by NPs. Furthermore, nanoparticles facilitate drug uptake by epithelial diffusion, enabling medicine concentration to reach optimum levels quickly. NPs alter medicines' pharmacokinetic and tissue distribution patterns in cancer cells and boost intracellular efflux [bib_ref] Cancer Drug Delivery in the Nano Era: An Overview and Perspectives (Review), Li [/bib_ref] [bib_ref] Therapeutic Advancements in the Management of Diabetes Mellitus with Special Reference to..., Chauhan [/bib_ref] [bib_ref] Effective Use of Nanocarriers as Drug Delivery Systems for the Treatment of..., Din [/bib_ref]. It has been shown that checkpoint inhibitors given by NPs increase the duration of the response rate of T cellbased immunotherapy [bib_ref] The Application of Nanotechnology in Immune Checkpoint Blockade for Cancer Treatment, Deng [/bib_ref] [bib_ref] Delivery Technologies for Cancer Immunotherapy, Riley [/bib_ref]. NPs can be fine-tuned and functionalized with specific moieties to promote their efficacy in targeting and delivering cargo materials to particular locations. Therefore, it has been concluded that NPs could be used as carriers in cancer immunotherapy.
## Future perspectives and conclusions
Nanomaterials' rapid advancement has given cancer immunotherapy a fresh perspective. Mainly NPs provide several advantages over traditional medication delivery methods. Although immunotherapy has shown promising outcomes in various therapeutic applications, there is still a considerable barrier preventing its widespread use in clinical settings. This barrier comprises a low patient response rate and restricted dosage toxicity. As a result, monotherapy still struggles to provide a positive response or prognosis in many individuals. Additionally, monoclonal antibodies' cost, preparation, and preservation make them unsuitable for widespread usage. To fix the problems with cancer immunotherapy and make it work better as a treatment, researchers have investigated nanomaterials-based combination therapy as an alternative therapeutic strategy to boost immune responses by controlling the many steps of the cancer immunity cycle. Stable, biocompatible nanomaterials may be modified with active targeting ligands or the EPR (enhanced permeability and retention) effect to increase drug accumulation at tumor locations. The development of nanomaterials for immunotherapy has sparked renewed optimism for improving this treatment's efficacy. A multipronged approach using nanomaterials, immunotherapy, and other medicines has emerged as the primary area of tumor therapy investigation.
The capacity of NPs to be tweaked and functionalized allows them to be constructed in various sizes, shapes, and capabilities to satisfy individual demands. In addition, the direct or indirect targeted delivery of NPs to tumor tissues exploits the tumor vasculature's hyperpermeability, bolstering cancer immunotherapy and reducing the toxic effects of anticancer drugs. Many in vitro and in vivo studies using NPs in cancer immunotherapy have shown positive results. These include considerable drug protection against degradation, prolonged and controlled intracellular delivery, and the avoidance of multidrug resistance in different kinds of NPs. In other trials, NPs have also proven crucial in combining treatment plans, including chemotherapy, phototherapy, and radiation. Furthermore, these nanoparticles can work with various immunotherapies to improve treatment outcomes by reprogramming the immunosuppressive tumor microenvironment and beginning systemic anti-tumor immune responses. However, the practical implementation of nanomaterialbased combination immunotherapies has a long way to go, even though some outcomes have been produced in the laboratory using these immunotherapies. For instance, in clinical studies, only radiotherapy and several combinations of chemotherapeutic medicines and immunotherapies are being investigated as potential treatments for cancer.
It is now widely accepted that nanomaterial-based immunotherapy has great promise for improving the efficacy of immunotherapy, and that a combination approach based on immunotherapy, nanomaterials, and other medicines is the primary area of tumor therapy research. This review demonstrates that various well-known and new polymeric and inorganic nanoparticles have been fabricated for immunotherapy and synergistic immunotherapy. By reprogramming the immunosuppressive TME and introducing systemic anti-tumor immune responses, nanoparticles may be considered ICD-inducing medications' modalities that synergize with various immunotherapies to improve treatment results. In clinical studies, only radiotherapy and several combinations of chemotherapeutic medicine and immunotherapies are being investigated as potential treatments for cancer. Nonetheless, a clinical trial assessment of the approach of combining PDT, PTT, or SDT (sonodynamic therapy) with immunotherapy has not been documented; therefore, these combination treatments are still in the early stages of study. Current tumor therapy has progressed from the lab to the clinic, but issues still need to be resolved, such as the low enrichment rate and more significant toxicity of nanomaterials. Cancer immunotherapy will soon be a breakthrough, even though nanomedicine-based immunotherapy and its combination treatment are still in their infancy. Chemists, biologists, and biochemists are all committed to this specialization.
[fig] Figure 1: The innate/adaptive immunity and interaction of immune cells in a tumor with the mechanisms of inhibition and promotion. [/fig]
[fig] Figure 2: Nanomaterials for indirect and direct immunomodulation. NMs serve as a carrier, and by releasing an immunosuppressive medication or antigen, the immune response is modulated indirectly. Interaction between NMs and IS is a form of direct immunomodulation. [/fig]
[fig] Figure 3: Physicochemical properties of nanomaterials and mechanisms of cell death in cancer induced by nanoparticles. [/fig]
[table] Table 1: Different immunotherapies are based on nanomaterials.Table 1. Cont. [/table]
[table] Table 2: Nanomaterial-induced immune responses concerning nanomaterials' physicochemical properties, such as surface, charge, rigidity, shape, and size. [/table]
|
Releasing the brakes: a case report of pulmonary arterial hypertension induced by immune checkpoint inhibitor therapy
Immune checkpoint inhibitors successfully treat various malignancies by inducing an immune response to tumor cells. However, their use has been associated with a variety of autoimmune disorders, such as diabetes, hepatitis, and pneumonitis. Pulmonary arterial hypertension due to checkpoint inhibitor use has not yet been described. We present a novel case of pulmonary arterial hypertension associated with systemic lupus erythematosus and Sjogren's syndrome overlap that was induced by therapy with the checkpoint inhibitor durvalumab.
A 44-year-old non-smoking African American woman with a history of stage 3 squamous cell lung cancer and recurrent pleural effusions was transferred to our hospital due to acute hypoxic respiratory failure. Two years prior, she was diagnosed with squamous cell lung cancer and was initially treated with cisplatin, etoposide, carboplatin, and paclitaxel. One year later, she was started on the programmed deathligand 1 (PD-L1) inhibitor durvalumab (10 mg/kg every two weeks for approximately eight months). Over this time, she developed new onset insulin-dependent diabetes, hypothyroidism, and adrenal insufficiency. These phenomena were attributed to autoimmunity induced by durvalumab, which was subsequently stopped.
Two months later, the patient presented with dyspnea, anasarca, bilateral exudative pleural effusions, and a pericardial effusion. Pleural fluid cytology was negative for recurrent lung cancer. She eventually was transferred to our center for further evaluation.
On presentation, the patient was severely volume overloaded and hypoxemic (requiring 5 L O 2 /min). She exhibited signs of tissue hypoperfusion including a venous lactate level of 3.9 mmol/L. B-type natriuretic peptide level was 628 pg/ ml, sodium was 126 mmol/L, and creatinine was 1.5 mg/dl. Chest computed tomography demonstrated mild diffuse ground-glass opacities with otherwise normal lung parenchyma as well as pulmonary artery enlargement and bilateral pleural effusions. In addition, right ventricular (RV) enlargement with reflux of contrast into the liver and a pericardial effusion were noted. Echocardiography revealed elevated RV systolic pressure, profound RV dysfunction, and a moderate pericardial effusion without hemodynamic compromise. The left atrium and ventricle (LV) were underfilled due to RV failure. There was no evidence of structural left heart disease.
After aggressive diuresis, right heart catheterization (RHC) was performed. This demonstrated severe precapillary pulmonary hypertension (PH) and RV failure. Infection, liver disease, congenital heart disease, pulmonary embolism, and illicit drugs were ruled out. The patient did not exhibit signs or symptoms of sleep-disordered breathing. There was no family history of PH. Hemoglobin was 7.3 mg/ dl, platelets were 124 per microliter, and haptoglobin was <6 mg/dl. Serologic workup revealed positive antinuclear antibody (ANA) of 1:1280 with a speckled pattern, positive SS-A antibody >8 U, and positive rheumatoid factor of 116 IU/ml. Rheumatology was consulted and a diagnosis of systemic lupus erythematosus (SLE)/Sjogren's syndrome overlap constellation was made based on sicca syndrome, recurrent pleural/pericardial effusions, hemolytic anemia, thrombocytopenia, positive ANA, positive SS-A, and high-titer positive rheumatoid factor. Treatment for SLE/ Sjogren's syndrome was initiated with prednisone and hydroxychloroquine for the duration of her hospitalization. Given the temporal relationship between symptom onset and durvalumab therapy without signs of connective tissue disease (CTD) or pulmonary arterial hypertension (PAH) prior to starting durvalumab, and given the lack of alternative causes for PAH development, the patient was diagnosed with PAH associated with SLE/Sjogren's syndrome overlap induced by PD-L1 inhibitor therapy.
The patient was treated with high-flow oxygen, dobutamine, norepinephrine, furosemide, intravenous treprostinil, and sildenafil for severe PAH with RV failure and hypoxemic respiratory failure. Unfortunately, she exhibited refractory RV failure with cardiogenic shock, and ultimately support was withdrawn.
# Discussion
To the best of our knowledge, this is the first report of checkpoint inhibitor-associated PAH. Programmed cell death protein 1 (PD-1) and its ligand PD-L1 limit activation of self-reactive T cells and protect against autoimmunity.Immune checkpoint inhibitors represent a new generation of cancer therapeutics that, through inhibition of immune checkpoints such as PD-1 or PD-L1, exhibit anti-tumor immunity by inducing removal of malignant cells. Their strong therapeutic potential has led to extensive and increasing clinical use. However, the associated increase in T-cell activation can cause excessive inflammation with development of immune-related adverse events such as pneumonitis, thyroiditis, neuropathy, and hepatitis.Furthermore, as checkpoint inhibitor use has increased, mounting evidence points towards the propensity of these drugs to lead to development of autoimmune conditions such as sicca syndrome, inflammatory myopathy, vasculitis, and lupus nephritis. These autoimmune phenomena are due to removal of PD-1-mediated inhibition of immune checkpoints that maintain self-tolerance and normally prevent autoimmunity.In patients with preexisting autoimmune disease, checkpoint inhibitor use has also been associated with worsening control or flare of the underlying autoimmune condition.In this patient, development of autoimmune conditions such as hypothyroidism and diabetes necessitated cessation of durvalumab. However, autoimmune phenomena can occur up to almost one year after checkpoint inhibitor use, suggesting that the patient's development of SLE/Sjogren's syndrome overlap was due to checkpoint inhibitor use as well.It is unclear whether she developed a CTD de novo or whether she had subclinical disease unmasked by durvalumab use. However, in either scenario, checkpoint inhibitor therapy led to the development of uncontrolled CTD.
Current evidence suggests that the PD-1/PD-L1 pathway may play a role in PAH development. PD-L1 expression is reduced in lungs of rats with experimental PAH. Regulatory T-cells (Tregs) protect from PAH development by increasing PD-L1 expression, and blocking PD-1/PD-L1 signaling abrogates Treg-mediated protection from experimental PAH.These data suggest that inhibition of PD-1/PD-L1 signaling in patients indeed may increase the risk of PAH development. PD-1/PD-L1 inhibition-induced PAH may be mediated by loss of protective Treg signaling and/or by activation of potentially pathogenic self-reactive T cells.While the patient presented with volume overload, LV disease and group 2 PH were ruled out by echocardiography and RHC. While ground-glass opacities are not typically thought to be associated with PAH, they have in fact been identified in >40% of PAH patients.They are typically seen in severe PAH when a dilated RV leads to LV compression with subsequent pulmonary venous congestion. We suspect this was the case in our patient. Another potential explanation for ground-glass opacity development would be pulmonary veno-occlusive disease, but the patient did not exhibit septal lines or lymphadenopathy, making this entity less likely. While ground-glass opacities could also result from checkpoint inhibitor-induced pneumonitis (indicating group 3 PH), lung parenchyma looked otherwise normal, and the severity of hemodynamic alterations was beyond those expected with group 3 PH.While multiple cardiovascular autoimmune phenomena have been associated with immune checkpoint inhibitor therapy, to the best of our knowledge, no invasively confirmed cases of pre-capillary PH due to immune checkpoint inhibitor therapy have been described, thus making our report both clinically relevant and novel. While there is one reported case of pulmonary arterial vasculopathy and RV dysfunction, 10 this patient had RV myocarditis and pulmonary vasculitis, thus presenting with an etiology and phenotype that are distinct from a typical PH phenotype with subsequent RV failure. In addition, no hemodynamic data were available in that case, making PH classification difficult. In another case report, also without invasive hemodynamics, PH was associated with left ventricular myocarditis, suggesting a post-capillary etiology of PH.This is the first report with a hemodynamically confirmed pre-capillary phenotype.
In summary, we report a novel case of PAH associated with SLE/Sjogren's syndrome overlap induced by checkpoint inhibitor therapy. While no clear diagnostic criteria exist for checkpoint inhibitor-induced autoimmune disease, the close temporal relationship between durvalumab use and CTD and PAH development and the absence of other potential causes for PAH make checkpoint inhibitorinduced CTD the most likely explanation for this patient's PAH. With the increasing use of checkpoint inhibitors and the well-known contribution of immune dysregulation and autoimmunity to PAH pathogenesis, these therapies may represent an increasingly prevalent novel risk factor for PAH development. |
Antiferromagnetic domain wall as spin wave polarizer and retarder
Lan et al. reports a theoretical study on the spin-wave (SW) propagation through an antiferromagnetic domain wall (DW), which is subject to the Dzyaloshinskii-Moriya interaction (DMI). The authors find that the in-plane and out-of-plane SW modes experience different SW gap and thus different SW scattering, which is induced by a combined effect of DMI and DW texture. As a result, when SW passes through DMI-DW, the SW polarization changes for an intermediate SW frequency and the SW phase changes for a higher frequency. These findings lead to a conclusion that an antiferromagnetic DW subject to DMI acts as a SW polarizer or a SW retarder, depending on the SW frequency. This conclusion is supported by SW theory and micromagnetic simulation, which have been well established in the SW community.Magnonics, which aims to utilize SWs as information carriers, is currently one of extensively studied subjects in spintronics/magnetism communities as it offers low-power computing due to negligible Joule heating and nonlinear computing due to its wave characteristics. As described in the introduction, most magnonic devices up to now have focused on the control of SW amplitude or phase, but largely ignored the possibility to use the SW polarization. As proven in spintronics and photonics, the control of relevant polarization is essential for various applications. In this respect, the idea proposed in this work is attractive and very stimulating. The manuscript is well written so that even nonspecialists may get the point of how the proposed idea works. Also the proposed idea seems to work for synthetic antiferromagnets which are easy to fabricate in lab so that I expect this work will immediately motivate experimental groups.On the overall, I find this work is very interesting and will attract considerable attention from magnonics/spintronics/magnetism communities. Thus I support the publication of this work in Nature Communications once the authors have a chance to consider the following minor comments.1. At the end of the 1st paragraph in page 4, it is written that "… in-plane (out-of-plane) mdoes x-polarized (y-polarized) modes". It seems that x and y may need to be interchanged.2. The authors mainly discuss the action of DW on SWs. It would be nice to discuss about the reaction: how the change in the SW polarization affects the DW? Is there any relevant magnonic torque enabling DW motion?Reviewer #2 (Remarks to the Author):The authors claim to use a domain wall as an "optical" element, a polarizer or a phase plate for magnons. This goal is achieved by introducing inhomogeneous DMI (Lifshitz invariant) into the system Hamiltonian and analysing magnon spectra and dispersion.Within the sine-Gordon model, which is usually used for FM and AF with uniaxial anisotropy, the domain wall does not scatter magnons (produces non-scattering potential for magnons). Any additional upgrading of the model (e.g., potential corresponding to tetragonal anisotropy) automatically gives rise to the scattering domain wall. If, in addition, this scattering is spin-dependent, then, the DW can have the properties of spin polarizer. Hence the idea is a clear one in principle, albeit not very innovative.
They start from LLG equations for each of magnetic sublattice and analyse excitations over the given equilibrium profile of the DW, analytically, by micromagnetic calc., and by what they claim to be a Green function technique.
There is however inconsistencies within their treatment. The authors consider the domain wall between two (equivalent? they do not explain) 180 domains. So, they implicitly assume that the ground state is homogeneous up/down long z axis. Moreover, as an equilibrium profile of the DW they use anzatz for AF without DMI.
The system with such a DMI is apt to formation of spiral and skyrmions. Homogenenous ground state (if any) can exist in a certain range of parameters which can be roughly estimated as follows. The system has two important lengthscales: the domain wall width $\sqrt{A/K}$ and a period of spiral $A/D$. Dimensionless parameter $AK/D^2$ separates two regimes. At $AK/D^2>>1$ the ground state is almost homogeneous, and DMI modifies the DW and can (I am not sure) modify spectra. However, two homogeneous (up/down) states are equivalent in the limit of vanishingly small DMI.
At $AK/D^2<<1$ the ground state is a spiral (or skyrmion). If there are two equivalent spiral states (I did not checked), then, the DW separates these states and its structure is different from the DW for 180 AF. In this case the author's approach is not valid at all.
The authors do not give any analysis of the parameters or validity of their model. However, at the end of the paper (Methods) give some numbers from which I deduce that they consider the most complicated case $AK/D^2=1$ in which the ground state is a spiral and spiral period is equal to the domain wall width. Analytical anzatz used for the DW profile is not valid.
The authors also consider DMI as correction (which is not justified in the text) and assume that DMI "acts" only within the DW (because DMI is related with space derivative).
In a homogenenous phase DMI modifies magnons spectrum, which should include linear in $k$ term. So, I cannot agree with the last statement in the paragraph below Eq.(3). Moreover, in a homogenenous state DMI modifies the profile of the DW, so, again, analytical anzatz used for the DW profile is approximate or not valid in the presence of DMI.
As a third comment, the authors consider DMI of the special type (inhomogeneous, Lifshitslike). The broken space-reversal symmetry allows not only inhomogenenous DMI, it can also allow homogenenous DMI. Homogenenous DMI means that the system is a weak AF, i.e., has nonzero magnetization in the initial state. This also means that magnon modes with different spins are nondegenerate from the beginning, DW also has nonzero magnetization and this can influence substantially on the polarising properties of the DW. I understand, that homogenenous DMI makes the model more complicated, but authors should explain explicitly why they disregard homogenenous DMI.
As a whole the idea is not, in my opinion, innovative enough (it has been around) to be considered for publication in Nature Communications. The many inconsistencies that are present in their analysis also brings the work to a quality that should be remedied. However, I would emphasized that even when these issues are addressed properly, I do not find the general claims (which some I question) of high enough interest for publication in Nature Communications.
## Reviewer #3 (remarks to the author):
This is an interesting, novel and generally correct manuscript coming from one of the theoretical leaders in the field.
The authors propose to use polarization of spin waves in a dielectric antiferromagnetic (AFM) waveguide as a new way for information coding in beyond-CMOS AFM-based information processing systems, and demonstrate, that in the presence of DMI, an AFM domain wall can work as a spin wave polarizer or as a waveplate.
The main text of the manuscript is clearly written and will be accessible to a wide audience of the Nature Communications.
However, in the opinion of this Referee, the abstract and the introduction to the manuscript contain several incorrect or/and ambiguous statements, and need to be rewritten.
1. In the first sentence of the abstract the authors write : "...spin wave, or magnon when quantized, can propagate in both conducting and insulating materials without Joule heating.
The text would become much clearer if the authors immediately mention that for spin wave Joule losses are simply replaced by losses associated with the Gilbert damping of propagating spin waves.
2. The authors, first, state that they like AFM materials for polarization manipulation, because, in contrast with ferromagnets (FM), the time-reversal symmetry is restored in AFM.
Then, they use AFM with DMI, where time-reversal symmetry is broken again by DMI.
The general relation between the polarization manipulation for signal processing and the timereversal symmetry should be explained in a correct and consistent way.
3. The authors write in the introduction:"...polarization is an intrinsic property of wave-like (quasi-)particles such as light, acoustic wave, and spin wave." It seems to me, that "wave-like (quasi-)particles" are not "light, acoustic wave, and spin wave", but photons, phonons and magnons. This is especially important , since the idea of AFM birefringence and rotation of the spin wave polarization in the course of propagation an AFM was explicitly introduced in [3].
In summary, this is and interesting and novel manuscript that can be published in the Nature Communication after a minor revision.
## Responses to referee reports on manuscript ncomms-17-01280-t
[formula] ---------------------------------------------------- Reply to remarks of the Referee #1 ---------------------------------------------------- [/formula]
Magnonics, which aims to utilize SWs as information carriers, is currently one of extensively studied subjects in spintronics/magnetism communities as it offers low-power computing due to negligible Joule heating and nonlinear computing due to its wave characteristics. As described in the introduction, most magnonic devices up to now have focused on the control of SW amplitude or phase, but largely ignored the possibility to use the SW polarization. As proven in spintronics and photonics, the control of relevant polarization is essential for various applications. In this respect, the idea proposed in this work is attractive and very stimulating. The manuscript is well written so that even nonspecialists may get the point of how the proposed idea works. Also the proposed idea seems to work for synthetic antiferromagnets which are easy to fabricate in lab so that I expect this work will immediately motivate experimental groups. On the overall, I find this work is very interesting and will attract considerable attention from magnonics/spintronics/magnetism communities. Thus I support the publication of this work in Nature Communications once the authors have a chance to consider the following minor comments.
Reply: We first thank the referee for the encouraging remarks regarding the quality and importance of the present work and also for his/her positive recommendation to Nature Communications.
At the end of the 1st paragraph in page 4, it is written that "… in-plane (out-of-plane) modes x-polarized (y-polarized) modes". It seems that x and y may need to be interchanged.
Reply: Thanks for the careful reading. We have corrected the text to "… in-plane (out-ofplane) modes y-polarized (x-polarized) modes …", and also the text above to " " and # coincide with and ".
The authors mainly discuss the action of DW on SWs. It would be nice to discuss about the reaction: how the change in the SW polarization affects the DW? Is there any relevant magnonic torque enabling DW motion?
Reply: We thank the referee for raising this interesting question. The answer is YES, the inverse effect also exists, i.e. SW-driven domain wall motion depends on the polarization. Actually, we discovered this inverse effect first. The present manuscript only discusses the effect of using 'rigid' domain wall to manipulate the SW polarizations. We will have a separate manuscript soon discussing the inverse effect by allowing domain wall to move.
[formula] ---------------------------------------------------- Reply to remarks of the Referee #2 ---------------------------------------------------- [/formula]
Within the sine-Gordon model, which is usually used for FM and AF with uniaxial anisotropy, the domain wall does not scatter magnons (produces non-scattering potential for magnons). Any additional upgrading of the model (e.g., potential corresponding to tetragonal anisotropy) automatically gives rise to the scattering domain wall. If, in addition, this scattering is spin-dependent, then, the DW can have the properties of spin polarizer. Hence the idea is a clear one in principle, albeit not very innovative.
Reply: We first thank the referee for his/her critical comments, which help us a lot in improving our manuscript for better presentation.
It is general that any type of polarizer needs some sort of polarization (spin)-dependent scattering. This 'principle' is of course a clear one, nothing to innovate in this respect. However, how to realize such spin-dependent scattering is highly non-trivial, especially when it needs to be simple and efficient. In our manuscript, we proposed that a simple domain wall in AFM nanowires with DMI can naturally realize this goal. This idea of using magnetic texture itself, instead of any artificial physical structure, as magnonic 'device' is quite innovative in our opinion.
They start from LLG equations for each of magnetic sublattice and analyse excitations over the given equilibrium profile of the DW, analytically, by micromagnetic calc., and by what they claim to be a Green function technique.
## Reply:
We employ the Green function technique to calculate the transmission coefficients in our system because our magnonic transport system is a close analogy to the electronic transport system, where Green's function is well developed and widely used. We have also checked that results from Green's function method agree well with that from the mode-matching method and the WKB method (not shown). We have modified our text about the Green function technique for better presentations in the revised manuscript. (Now extended and moved to the supplementary materials.)
There is however inconsistencies within their treatment. The authors consider the domain wall between two (equivalent? they do not explain) 180 domains. So, they implicitly assume that the ground state is homogeneous up/down long z axis. Moreover, as an equilibrium profile of the DW they use anzatz for AF without DMI.
## The system with such a dmi is apt to formation of spiral and skyrmions. homogenenous ground state (if any) can exist in a certain range of parameters which can be roughly estimated as follows. the system has two important lengthscales: the domain wall width $\sqrt{a/k}$ and a period of spiral $a/d$. dimensionless parameter $ak/d^2$ separates two regimes. at $ak/d^2>>1$ the ground state is almost homogeneous, and dmi modifies the dw and can (i am not sure) modify spectra. however, two homogeneous (up/down) states are equivalent in the limit of vanishingly small dmi. at $ak/d^2<<1$ the ground state is a spiral (or skyrmion). if there are two equivalent spiral states (i did not checked), then, the dw separates these states and its structure is different from the dw for 180 af. in this case the author's approach is not valid at all. the authors do not give any analysis of the parameters or validity of their model. however, at the end of the paper (methods) give some numbers from which i deduce that they consider the most complicated case $ak/d^2=1$ in which the ground state is a spiral and spiral period is equal to the domain wall width. analytical anzatz used for the dw profile is not valid.
Reply: We thank the referee to point out this extremely important point that we failed to mention in the previous manuscript. In the simplest model, when the DMI is stronger than a critical value ' , the ground state is indeed a spin spiral, which we have also confirmed, as shown in [fig_ref] Figure R1: Micromagnetic simulations of static magnetic profiles with different boundary conditions [/fig_ref] below. In order to achieve relative strong spin-dependent scattering effects, we adopt a relative large value of D (i.e. AK/D^2 ~ 1), for which a simple domain wall would not be stable. In order to stabilize a domain wall at this large D, we used a free (or pinned) boundary condition (rather than the DM boundary condition) at the two ends of the AFM nanowire. As shown in [fig_ref] Figure R1: Micromagnetic simulations of static magnetic profiles with different boundary conditions [/fig_ref] , a domain wall structure is stabilized for such relatively large D in the micromagnetic simulations. Note that in all simulations, before injecting spin waves, we always relax the magnetic texture to its stable state first. We would also like to emphasize that the presented polarizing and retarding effects in this manuscript exist for full range of DMI strength D. Smaller D values would result in similar polarizing and retarding effects, but less effective indeed. But this can be solved by chaining two or more domain walls together to enhance the effects, as shown in the supplementary materials.
As an example, we choose a smaller = = 1.4×10 ./ A < ' , for which a domain wall is always stable regardless of the boundary condition as shown in [fig_ref] Figure R1: Micromagnetic simulations of static magnetic profiles with different boundary conditions [/fig_ref]. [fig_ref] Figure R2: Theoretical calculations and micromagnetic simulations of the polarizing and retarding effects at... [/fig_ref] shows the polarization-dependent spin wave scattering behaviors across the domain wall remain for this smaller D. And as confirmed by micromagnetic simulations in [fig_ref] Figure R2: Theoretical calculations and micromagnetic simulations of the polarizing and retarding effects at... [/fig_ref] , a double-domain-wall structure can realize a very good spin wave polarizer at low frequency, and domain walls remain to function as spin wave retarders at high frequency.
We have now included separate supplementary materials discussing the effect of boundary condition on the stability of magnetic domain walls in AFM nanowire with DMI, and also the polarizing and retarding effects at smaller DMI strength D.
## The authors also consider dmi as correction (which is not justified in the text) and assume that dmi "acts" only within the dw (because dmi is related with space derivative). in a homogenenous phase dmi modifies magnons spectrum, which should include linear in $k$ term. so, i cannot agree with the last statement in the paragraph below eq. (3).
Reply: We do not consider DMI as correction. In fact, in our calculations DMI is present everywhere (except for boundary conditions). However, the effect of DMI only appears in the domain wall region where magnetization is varying in space.
As presented in Ref. and , DMI would modify the magnon spectrum, but only when the spin wave wavevector has finite component in the direction of magnetization, i.e. when - ≠ 0, there would be linear k term in the dispersion as mentioned by the referee. However, when the wavevector is perpendicular to the magnetization, the magnon dispersion is not modified by DMI. This is the situation investigated in this manuscript, where m is along and k is along the nanowire axis . Consequently, for the uniform domains in our structure, there is no linear k term in dispersion as indicated by Eq. .
Moreover, in a homogenenous state DMI modifies the profile of the DW, so, again, analytical anzatz used for the DW profile is approximate or not valid in the presence of DMI.
Reply: Whether the DMI would change the domain wall profile or not depends on direction of the easy-axis anisotropy, as shown in [fig_ref] Figure R3: Schematics of domain wall profiles with DMI [/fig_ref].
DMI would modify the domain wall profile when the easy-axis anisotropy (along x) is along the wire direction as the one in [fig_ref] Figure R3: Schematics of domain wall profiles with DMI [/fig_ref] , as also reported for an FM wire in Ref. . But for the one under study in this manuscript, where the easy-axis (along z) is perpendicular to the wire direction as in [fig_ref] Figure R3: Schematics of domain wall profiles with DMI [/fig_ref] , the profile remains unchanged, as reported in Ref. . We confirmed this by micromagnetic simulations (see supplementary materials) that the domain wall profile is the standard Walker profile, even in the presence of DMI.
We have modified the manuscript to highlight the role of DMI in the domain wall profile in our model. As a third comment, the authors consider DMI of the special type . The broken space-reversal symmetry allows not only inhomogenenous DMI, it can also allow homogenenous DMI. Homogenenous DMI means that the system is a weak AF, i.e., has nonzero magnetization in the initial state. This also means that magnon modes with different spins are nondegenerate from the beginning, DW also has nonzero magnetization and this can influence substantially on the polarising properties of the DW. I understand, that homogenenous DMI makes the model more complicated, but authors should explain explicitly why they disregard homogenenous DMI.
Reply: Indeed, we only considered the inhomogeneous (Lifshits-type) DMI. This assumption is exact in artificial synthetic antiferromagnet. In real antiferromagnet, the homogeneous DMI may exist, and causes a misalignment of two opposite magnetizations in two magnetic sublattices, which induces weak ferromagnetism. However, the misalignment and ferromagnetism induced by the homogenous DMI is extremely small since it directly competes with the dominating antiferromagnetic coupling between two magnetic sublattices in AFM. . Since ≫ , A , ≪ 1 thus the misalignment is extremely weak, and is expected to cause only negligible perturbations to the polarizing and retarding effects in our work.
Following the referee's suggestion, we have added a comment about why homogeneous DMI is omitted in our model in the revised manuscript. The main text of the manuscript is clearly written and will be accessible to a wide audience of the Nature Communications.
## Reply:
We thank the referee for the encouraging remarks regarding the quality and importance of the present work and also for his/her positive recommendation to Nature Communications.
In the first sentence of the abstract the authors write :"...spin wave, or magnon when quantized, can propagate in both conducting and insulating materials without Joule heating." The text would become much clearer if the authors immediately mention that for spin wave Joule losses are simply replaced by losses associated with the Gilbert damping of propagating spin waves.
## Reply:
To avoid any confusion about dissipation, we have modified the text in the abstract, and also the description in the introduction correspondingly.
The authors, first state that they like AFM materials for polarization manipulation, because, in contrast with ferromagnets (FM), the time-reversal symmetry is restored in AFM.Then, they use AFM with DMI, where time-reversal symmetry is broken again by DMI. The general relation between the polarization manipulation for signal processing and the time-reversal symmetry should be explained in a correct and consistent way.
## Reply:
We thank the referee for pointing out the inconsistencies in the discussions based on the time-reversal symmetry.
We removed the time-reversal discussion in the revised manuscript. We now attribute the polarization freedom in AFM to the structure of two opposite magnetic sublattices. This is in contrast to FM, which only possesses one type of magnetic lattice, thus the polarization freedom is absent. Since DMI does not change the basic structure of two magnetic sublattices in AFM, the polarization freedom still holds.
The authors write in the introduction:"...polarization is an intrinsic property of wave-like (quasi-)particles such as light, acoustic wave, and spin wave." It seems to me, that "wave-like (quasi-)particles" are not "light, acoustic wave, and spin wave", but photons, phonons and magnons.
Reply: Thanks for careful reading. We have modified the statement to "...polarization is an intrinsic property of wave-like (quasi-)particles such as photons, phonons and magnons."
The referencing of the previous work in the field is inadequate. The authors should mention at least some of the work of their competitors in the same field:
[fig] 4: The referencing of the previous work in the field is inadequate. The authors should mention at least some of the work of their competitors in the same field: 1.Gomonay, H. V. & Loktev, V. M. Spin transfer and current-induced switching in antiferromagnets. Physical Review B 81, 144427 (2010). 2. Gomonay, E. V. & Loktev, V. M. Spintronics of antiferromagnetic systems (review article). [/fig]
[fig] Figure R1: Micromagnetic simulations of static magnetic profiles with different boundary conditions. The DMI strength is (a) = 3.0×10 ./ A > ' (the value used in the main text), domain wall configuration is stable for free and pinned boundary condition, and spin spiral is stable for DM boundary condition. (b) = 1.4×10 ./ A < ' , domain wall configuration is always stable for all three types of boundary conditions. In both cases, the standard Walker profile is indicated by solid curves. [/fig]
[fig] Figure R2: Theoretical calculations and micromagnetic simulations of the polarizing and retarding effects at = 1.4×10 ./ A < ' . a. Theoretical calculations of the polarization-dependent scattering across antiferromagnetic domain walls at different DMI strengths D or across 2 consecutive domain walls. b. Micromagnetic simulation of the polarizing effect with 2 consecutive domain walls. c. Micromagnetic simulation of the retarding effect with 4 consecutive domain walls. [/fig]
[fig] Figure R3: Schematics of domain wall profiles with DMI. The easy-axis is (a) perpendicular to (b) along the wire. The magnetization profile is for < (red) and = (blue). [/fig]
[fig] -: Reply to remarks of the Referee #3 This is an interesting, novel and generally correct manuscript coming from one of the theoretical leaders in the field. The authors propose to use polarization of spin waves in a dielectric antiferromagnetic (AFM) waveguide as a new way for information coding in beyondCMOS AFMbased information processing systems, and demonstrate, that in the presence of DMI, an AFM domain wall can work as a spin wave polarizer or as a waveplate. [/fig]
|
The proteomics approach to find biomarkers in gastric cancer.
# Introduction
Gastric cancer is the most common cancer in Korea. It is an unfortunate reality for many Koreans. It has been a very serious disease and is naturally resistant to many anticancer drugs, so the results after surgery and chemotherapy have been disappointing. To reduce the mortality and improve the effectiveness of therapy, many studies have tried to find key biomarkers. Biomarkers are important molecular signposts of the biologic state of a cell at a specific condition. Active genes, their respective protein products, and other organic chemicals made by the cell create these signposts. Proteomic technologies are providing the tools needed to discover and identify disease associated-biomarkers.
After the completion of human genome project, the practice of medicine will be altered fundamentally, which includes the identification of genes (genomics) that are involved in the appearance, progression, and treatment of cancer, the answers to what those specific genes do and how they interact in communication networks (functional genomics), and the roles played by their protein products (proteomics) in molecular pathways.
There have been many studies including analysis of prognostic factors, development of chemotherapeutic agents, and the search for the early detecting molecules, but there are currently very few molecular markers that are clinically in use. Proteomics is one of the technologies that rapidly change our approach to cancer research. The conventional proteomics has been the high resolution 2-D gel electrophoresis (1) followed by computational image analysis and protein identification using mass spectrometry. The proteomic studies could lead to the molecular characterization of cellular events associated with cancer progression, signaling, and developmental stages [bib_ref] Multidrug resistance, Moscow [/bib_ref] , and these studies with clinical tumor samples have led to the defining of cancer-specific protein markers which will be the basis for developing new methods for early diagnosis and detection and clues to understand the molecular characterization of cancer progression [bib_ref] Preoperative treatment of rectal cancer with radiation, chemotherapy and hyperthermia: analysis of..., Rau [/bib_ref] [bib_ref] Heat-resistant variants of Chinese hamster fibroblasts altered in expression of heat shock..., Laszlo [/bib_ref] [bib_ref] Heat resistance in mammalian cells: lessons and challenges, Laszlo [/bib_ref].
In this study, we tried to find biomarkers in human gastric cancer tissue via proteomics as high throughput method.
# Materials and methods
# Materials
Eleven human stomach tissue samples were prepared from resection materials of cancer patients in Dankook University Jin-Woo Ryu*, Hyung-Jee Kim � , Young-Sun Lee, Na-Hye Myong � , Cheol-Hoh Hwang, Gae-Sung Lee ‖ , Heng-Cherl Yom �
## The proteomics approach to find biomarkers in gastric cancer
Gastric cancer is a very serious disease and is naturally resistant to many anticancer drugs. To reduce the mortality and improve the effectiveness of therapy, many studies have tried to find key biomarkers. Proteomic technologies are providing the tools needed to discover and identify disease-associating biomarkers. The proteomic study of gastric cancer establishes any specific events that lead to cancer, and it provides a direct way to define the true function of genes. Using two dimensional (2-D) electrophoresis of the stomach cancer tissue, we have gained about 1,500 spots in each gel, and 140 protein spots also were identified. Among the identified proteins, there were seven over-expressed proteins in stomach cancer tissue: NSP3, transgelin, prohibitin, heat shock protein (hsp) 27 and variant, protein disulfide isomerase A3, unnamed protein product and glucose regulated protein. There were also seven under-expressed proteins in stomach cancer: Apolipoprotein A-1, p20, nucleoside diphosphate isomerase A, alpha 1 antitrypsin, desmin, serum albumin and serotransferrin.
Hospital. Resections were examined by a pathologist and normal tissue samples were prepared from noncancerous regions. The margin tissue samples were prepared from noncancerous regions 5 cm apart from primary tumor. Their clinicopathologic characters were described in .
The power supply, the 2D electrophoresis system (PRO-TEAN IEF cell, PROTEAN II Xi cell), the Trans-Blot Cell, CHAPS, DTT, SDS, iodoacetamide, ampholyte, and IPG (Immobilized pH gradient) and strip were purchased from Bio-Rad (Richmond, CA, U.S.A.). DNase and the protein pI marker were from Amersham-Pharmacia (Bucks, U.K.). Acrylamide, N,N′ -methylene-bis-acrylamide, TEMED, ammonium persulfate, Trisma base, glycine, methanol, glacial acetic acid, glycerol, potassium chloride, urea, thiourea, phenylmethyl-sulfonylfluoride (PMSF), ethylenedinitrilotetra acetic acid disodium salt (EDTA), benzamidine, and brilliant blue R-250 were from Sigma-Aldrich (Missouri, U.S.A.).
## Sample preparation
Proteins were extracted from human stomach tissue as instructed by German Heart Center method (http://userpage. chemie.fu-berlin.de/~pleiss/tissue.html). Stomach tissue was washed in 50 mM Tris-HCl pH 7.1, 100 mM KCl, 60 mM EDTA, 5.8 mM benzamidine and 0.2 mM PMSF, and was lyophilized for 1hr. Samples of 30 mg were crushed by liquid nitrogen and were solubilized in 100 L buffer containing 7 M urea, 2 M thiourea, 4% CHAPS, 100 mM DTT, 25 mM tris-HCl pH 7.1, 50 mM KCl, and 0.2% Bio-Lyte 3/10 ampholyte. DNase was added to 210 U and was incubated at room temperature for 30 min. Insoluble materials were removed by centrifugation at 13,000 rpm for 20 min at 10℃. Protein concentration was determined using a commercial Bradford reagent (Bio-Rad) and the samples were stored at -70℃ until use.
## Two-dimensional electrophoresis
The first-dimensional gel separation was carried out with 17 cm pH 4-7 IPG strips following the manufacturer' s pro-tocol (Bio-Rad) with minor modifications. 1mg of the protein was loaded and Iso-electric focusing (IEF) was performed using PROTEAN IEF Cell for a total of 145 kVh at 20℃. After IEF, strips were equilibrated for 15 min in 6 M urea, 2% SDS, 0.05 M Tris-HCl pH 8.8, and 20% glycerol containing 2% DTT, and then equilibrated again for 15 min in the same buffer containing 2.5% iodoacetamide. Equilibrated IPG strips were transferred onto 13% uniform polyacrylamide gels and then were run in PROTEAN II xi cell tank at 30 mA per gel. The gels were visualized using the Coomassie brilliant blue R250 staining method.
After the staining, 2D gels were imaged using Powerlook 1,100 (UMAX, Fremont, CA, U.S.A.) and the images were analyzed using Melanie III (Genebio, Geneva, Switzerland).
## Protein identification
Proteins that were differentially pressed in the tumor tissue were identified by using the database from Gyeongsang National University (http://proteome.gsnu.ac.kr/index1.html) and by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) analysis
# Results
Two-dimensional electrophoresis of the stomach cancer tissue produced about 1,500 spots each. A electrophoresis feature is shown in . In this figure, all proteins that have been identified to date have been marked with arrows and accession number. Spots marked with circles and arrows are differentially expressed proteins in normal, margin, and cancer tissues [fig_ref] Figure 2: 2DE patterns showing the under-expression of desmoplakin I [/fig_ref]. In this study, we compared cancer tissue with normal tissue to obtain definite differential expressions of protein between them. A comparison of stomach cancer tissue with normal tissue in same patients showed the seven proteins, namely: NSP3, transgelin, prohibitin, heat shock protein (hsp) 27 and variant, protein disulfide isomerase A3, unnamed protein product, and glucose-regulated protein were over-expressed [fig_ref] Table 2: Over-expressed Proteins in Human Stomach Cancer Tissue [/fig_ref]. The seven proteins, namely: apolipoprotein A-1, p20, nucleoside diphosphate isomerase A, alpha 1 antitrypsin, desmin, serum albumin, and serotrans- ferrin were under-expressed [fig_ref] Table 3: Under-expressed Proteins in Human Stomach Cancer Tissue [/fig_ref].
# Discussion
Recent advances in genomics and proteomics hold great potential for diagnostic, prognostic, and therapeutic applications [bib_ref] Trends in biomarker research for cancer detection, Srinivas [/bib_ref]. Genes work at the protein level, and comparisons of transcripts and corresponding protein expressions have shown that mRNA and protein levels are not necessarily highly correlative. The proteomic study includes post-translational modifications such as acetylation, ubiquitination, phosphorylation, or glycosylation [bib_ref] Analysis of N-linked oligosaccharides released from glycoproteins separated by two-dimensional gel electrophoresis, Charlwood [/bib_ref]. Proteomic methods detect the functioning units of expressed genes (10) using protein fingerprint [bib_ref] Proteomics to study genes and genomes, Pandey [/bib_ref]. The proteomic results express both the intrinsic genetic effect on cell and the impact of its environment, so it is very valuable to determine biomarkers.
Today, it has become common to read that proteins are the main functional outputs of genes, and the proteomics will lead biology and medicine beyond genomics. The molecular events into neoplaplastic progression are complex and diverse, and they remain incompletely characterized. The identification, quantification, classification, and functional assignment of proteins will be essential to the full understanding of these molecular events. Such information will likely prove to be crucial in cancer prognosis, diagnosis, prevention and therapy with the ultimate goods being therapeutic target discovery, rational drug design and the identification of early detection surrogate biomarkers [bib_ref] Proteomics: a new approach to the study of disease, Chambers [/bib_ref] [bib_ref] Genomics and proteomics tools for the clinic, Rudert [/bib_ref].
In this study, we have evaluated our 2-D data from reference 2-D map (from Gyeongsang National University, Jinju, Korea). The over-and under-expressed proteins were confirmed by MALDI-TOF mass spectrometry (Yonsei Proteome Research Center, Seoul, Korea). There are useful URLS for other cancer, bladder, colorectal, breast, liver, and leukemia, but there is little data for stomach cancer [bib_ref] Cancer proteomics: from signaling networks to tumor markers, Simpson [/bib_ref]. We obtained seven over-expressed proteins (NSP3, transgelin, prohibitin, HSP27, protein disulfide isomerase A3, unnamed protein product and glucose regulated protein) in cancer. In thermoresistant human gastric cancer cell line study, stratifin, cytokeratin 7-9, calreticulin, glucose regulated protein 78, and heat shock proteins (hsp27, hsx70 variant, hsp 27 variant) were reported as over-expressed protein [bib_ref] Characterization of the differential protein expression associated with thermoresistance in human gastric..., Sinha [/bib_ref]. They declared that the development of thermoresistance in stomach cancer cell and their chemoresistance variants is a complex phenomenon that is not only associated with the expression of HSPs but of molecular chaperones at every subcellular level. In our study, the cytoplasmic proteins (transgelin, prohibitin, and HSP 27) were identified as over-expressed proteins which are molecular chaperones. Also, we obtained seven under-expressed proteins (apolipoprotein A-1, p20, nucleoside diphosphate isomerase A, alpha-1-antitrypsin, desmin, serum albumin, and serotransferrin) in human gastric cancer tissue. The extracellular proteins (serum albumin, serotransferrin, and alpha-1-antitrypsin) were identified as under-expressed proteins in cancer tissue. In thermoresistant cancer cell line study, eight proteins (14-3-3, aldehyde dehydrogenase 1, annexin 1, lipocortin 1 variant, initiation factor-5A, initiation factor-5A variant, reticulocalbin, and vimentin) were reported as under-expressed proteins [bib_ref] Characterization of the differential protein expression associated with thermoresistance in human gastric..., Sinha [/bib_ref].
We analyzed primary tumor biopsies and normal tissues in same patients. The primary tumor biopsies may contain several different cell types other than carcinoma cells, so the exact defining for major cellular changes during the conversion of normal to malignant stomach is limited. It is related to defects in a multiplicity of 2-D features. On the contrary cell line in culture, being free of any tissue components may have multiplicity, but may not represent their exact feature in vivo counter part. Clearly, more work and data are required to understand and clarify the exact role of proteins for progression of carcinoma.
In this study, we observed the over-and under-expressed proteins in human gastric cancer tissue. We hope that these results shall be correlated with clinical data in recent forthcoming day to define clinically useful biomarkers. The identification of functioning protein is essentially needed to provide more effective therapy for patients suffering from gastric carcinoma.
[fig] Figure 2: 2DE patterns showing the under-expression of desmoplakin I (indicated by arrows) and up-regulation of prohibitin (indicated by circles) in stomach cancer tissue samples (G-I) as compared to normal (A-C) and margin stomach tissue samples (D-F). [/fig]
[table] Table 2: Over-expressed Proteins in Human Stomach Cancer Tissue [/table]
[table] Table 3: Under-expressed Proteins in Human Stomach Cancer Tissue [/table]
|
Sickness Presenteeism as a Link between Long Working Hours and Employees’ Outcomes: Intrinsic and Extrinsic Motivators as Resources
# Introduction
Sickness presenteeism (presenteeism hereafter) refers to situations where people still turn up at their jobs despite ill health that should prompt rest and absence from work [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref]. The act of presenteeism due to ill health is the conceptualization we use in the present study. It is problematic for an individual, as it might generate a deterioration of health and well-being [bib_ref] Sickness presenteeism: Prevalence, attendance-pressure factors, and an outline of a model for..., Aronsson [/bib_ref] ; presenteeism also creates costs for organizations and wider society [bib_ref] Global patterns of workplace productivity for people with depression: Absenteeism and presenteeism..., Evans-Lacko [/bib_ref]. Thus, presenteeism is often viewed as a negative phenomenon that should be brought under control [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. Systematic literature reviews and meta-analyses have identified various antecedents of presenteeism and concluded that it is more susceptible to organizational demands (e.g., workload and downsizing) than to personal demands (e.g., financial needs) [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref] [bib_ref] Presenteeism in the workplace: A review and research agenda, Johns [/bib_ref] [bib_ref] Associations between health and combinations of sickness presence and absence, Gustafsson [/bib_ref]. Miraglia and Johns proposed a dual-path meta-analytic model explaining how job insecurity, personal financial difficulties, job control, job demands, collegial support, supervisor support, and optimism as key antecedents affecting job satisfaction and/or personal health, which then lead to sickness presenteeism or absenteeism [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref]. Their meta-analytic model, however, stops at the presenteeism decision, not including the consequences of the behavior. Other qualitative reviews mostly relied on evidence from cross-sectional studies to conclude that presenteeism was associated with negative outcomes for employees' well-being and organizational effectiveness [bib_ref] Associations between health and combinations of sickness presence and absence, Gustafsson [/bib_ref] [bib_ref] The consequences of sickness presenteeism on health and wellbeing over time: A..., Skagen [/bib_ref] [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref].
Researchers have argued that presenteeism-incurred productivity loss is over-estimated due to conceptual and measurement problems [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Presenteeism: A short history and a cautionary tale, Johns [/bib_ref]. Evidence linking presenteeism to Int. J. Environ. Res. Public Health 2022, [bib_ref] Cognitive activation theory of stress (CATS), Ursin [/bib_ref] , 2179 2 of 16 negative outcomes is equivocal in longitudinal studies [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. For example, although Demerouti, Le Blanc, Bakker, Schaufeli, and Hox found that presenteeism was related to burnout among Dutch nurses [bib_ref] Present but sick: A three-wave study on job demands, presenteeism and burnout, Demerouti [/bib_ref] , Lu, Peng, Lin, and Cooper found no lasting effects of presenteeism on mental health, physical health, or burnout for Chinese employees [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref]. Other studies found no evidence of long-lasting damaging effects of presenteeism on productivity and job performance [bib_ref] A cross-cultural examination of presenteeism and supervisory support, Lu [/bib_ref] [bib_ref] Supervisor support, role ambiguity and productivity associated with presenteeism: A longitudinal study, Zhou [/bib_ref]. The most recent study from Taiwan found mixed results: presenteeism had a damaging effect on employees' innovative performance 6 months later, but no such effect was found for burnout [bib_ref] The compensatory protective effects of social support at work in presenteeism during..., Chen [/bib_ref]. The latest comprehensive review thus concluded that consequences of presenteeism behavior for organizations and individuals are still poorly understood [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref].
One way to reconcile the inconsistencies is to look for potential mechanisms that might affect the link between presenteeism and its consequences [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. It is intriguing that factors reflecting positive motives such as job satisfaction, organizational commitment, and work engagement have been consistently linked to presenteeism [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref] [bib_ref] Dynamic associations among somatic complaints, human energy, and discretionary behaviors, Christian [/bib_ref] , the behavior that is often portrayed as aversive. However, motivational factors are under-studied in the presenteeism research [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. To address this void, Ma, Meltzer, Yang, and Liu differentiated between autonomous and controlled motivation for presenteeism [bib_ref] Motivation and presenteeism: The whys and whats, Ma [/bib_ref]. Cooper and Lu proposed a comprehensive theoretical framework to explain how intrinsic or extrinsic work value orientations motivate an individual to commit presenteeism behavior voluntarily or involuntarily, which then leads to good or bad outcomes pertaining to well-being and productivity [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. The central tenet of Cooper and Lu's motivational theory of excessive availability for work (EAW) is that the act of excessive working (e.g., presenteeism) per se is neutral; it is the configuration of personal, organizational, and cultural factors that determines outcomes for individuals and organizations [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. Taking the same motivational theoretical perspective, while Cooper and Lu's model views intrinsic and extrinsic work value orientations as motivational antecedents manifested as different types of driving forces (i.e., voluntarily or involuntarily) for committing presenteeism behavior, we in the present study attempted to unpack the mechanisms of work value orientations on the "presenteeismoutcomes" link. Specifically, we empirically tested the resource loss/gain dynamism postulated in the conservation of resource (COR) theory [bib_ref] Conservation of resources: A new attempt at conceptualizing stress, Hobfoll [/bib_ref] to examine whether resource gains (i.e., intrinsic/extrinsic drives to accomplish work) can protect an individual's wellbeing and task performance in the context of resource depletion (i.e., working when ill due to heavy workload) [bib_ref] Cognitive activation theory of stress (CATS), Ursin [/bib_ref].
For this purpose we drew on the core COR proposition of resource loss to theorize that working long hours due to a heavy workload can result in presenteeism behavior, which further depletes resources and leads to negative consequences for well-being and job performance over time. As an antidote, critical resource gains, such as the motivational drive for hard working (i.e., intrinsic and extrinsic work value orientations), may impact the final consequences for an individual [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. The ultimate goal was to arrive at a more comprehensive understanding of the chain of events, revealing how long work hours precipitate presenteeism behavior, which leads to deleterious consequences for well-being and productivity over time, as well as how personal motivational resources may alleviate such an impact.
## Theoretical framework and hypotheses development
## Resource loss cycles in the "long working hours-presenteeism-outcomes" linkage
Long working hours, presenteeism, and COR. The basic tenet of COR theory is that people strive to retain, protect, and build resources; moreover, the potential or actual loss of these valued resources threatens or is stressful for individuals [bib_ref] Conservation of resources: A new attempt at conceptualizing stress, Hobfoll [/bib_ref]. Resource loss is thus central to the stress and adaptation experience. Moreover, the loss of resources tends to lead to resource loss cycles that have increasing strength and speed. Working extended hours, often resulting from overwhelming workloads, precipitates the likelihood for working while sick to avoid the piling up of work tasks [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref]. Viewed from the COR perspective, sickness presenteeism thus represents a scenario for resource depletion [bib_ref] Burnout and Health, Ferreira [/bib_ref].
That is, when employees strive to accomplish work demands under suboptimal health states, the continuous resource depletion triggers the resource loss cycle.
Systematic reviews and the latest work on presenteeism have revealed that the majority of research on antecedents focused on quantitative, emotional, social, cognitive, and physical job demands [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref] [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref] [bib_ref] Work and health-related factors of presenteeism: A mediation analysis on the role..., Guidetti [/bib_ref] [bib_ref] A job demands-resources approach to presenteeism, Mcgregor [/bib_ref] [bib_ref] The impact of challenge and hindrance demands on burnout, work engagement, and..., Nair [/bib_ref] , while prolonged working hours were largely overlooked in this context [bib_ref] Long working hours and presenteeism in Asia, Lu [/bib_ref]. However, the heightened pressure to work longer hours has become a salient work condition worldwide as organizations strive to do more with less, compounded by the devastating disruptions and uncertainties caused by the COVID-19 pandemic [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref] [bib_ref] The compensatory protective effects of social support at work in presenteeism during..., Chen [/bib_ref] [bib_ref] The association between long working hours and health: A systematic review of..., Bannai [/bib_ref]. Extending previous studies, we thus focused on long working hours as a salient aspect of work conditions for Chinese employees in examining the resource loss cycle triggered by presenteeism.
Long working hours and employees' outcomes. In line with the COR theory, working for an extended period of time requires protracted, energy-draining efforts, which may overtax an individual's resources and impair well-being [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref]. We further argue that having to work long hours would deprive employees of the opportunity to recuperate and restore energy. Such continuous resource depletion situations exacerbate the risk for illness and precipitate the propensity of employees to work on days when they have suboptimal health to avoid the piling up of unfinished tasks [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. A longitudinal study in Taiwan found that working long hours was positively associated with employees' exhaustion and negatively associated with work engagement and job performance [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref]. We therefore propose that working long hours will damage employees' quality of work life (i.e., well-being and job performance).
Hypothesis 1a (H1a). Long working hours are negatively associated with well-being.
Hypothesis 1b (H1b). Long working hours are negatively associated with job performance.
The mediating role of presenteeism. The upswing in working hours is evidently more severe in some East Asian countries due to weaker employment legislation and regulations. The "long-working-hour norm" is sanctioned by the deeply rooted hardworking cultural values of Confucian ethics [bib_ref] Affective and normative motives to work overtime in Asian organizations: Four cultural..., Kang [/bib_ref]. These collectively held moral imperatives sustain the long-existing practice of prolonged work hours in countries such as Japan, Taiwan, Hong Kong, Singapore, and Korea [bib_ref] Long working hours and presenteeism in Asia, Lu [/bib_ref] [bib_ref] Karoshi (work to death)" in Japan, Kanai [/bib_ref] [bib_ref] Weekly work hours and stress complaints of workers in Korea, Park [/bib_ref]. In Taiwan there is no mandatory ceiling on working hours for most white-collar professionals (i.e., so-called "job responsibility system"), and working long hours is a widely used image-management tactic among office workers [bib_ref] Long working hours and presenteeism in Asia, Lu [/bib_ref]. One recent study found that white-collar workers, on average, worked over 46 h per week in Taiwan [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref]. Findings from the latest qualitative and crosscultural survey-based research corroborate the postulation that work norms and social doctrines affect presenteeism behavior [bib_ref] Presenteeism and Absenteeism at Work: An Analysis of Archetypes of Sickness Attendance..., Ruhle [/bib_ref] [bib_ref] Working sick and out of sorts: A cross-cultural approach on presenteeism climate,..., Ferreira [/bib_ref]. Recent cross-sectional studies in Australia [bib_ref] A job demands-resources approach to presenteeism, Mcgregor [/bib_ref] [bib_ref] The impact of challenge and hindrance demands on burnout, work engagement, and..., Nair [/bib_ref] also found that higher job demands were positively related to burnout, which then leads to presenteeism. We extended the existing studies to examine the lasting effects of long working hours as a job demand on well-being and job performance via presenteeism behavior.
While presenteeism may seem attractive for organizations at a first glance (avoiding productivity loss due to sickness absence), employers are now realizing that it may incur hidden productivity costs in addition to the more obvious wear and tear on employees' well-being. This is because when employees work in suboptimal health energy is sapped and attention diverted, such that productivity and performance quality (e.g., innovation) suffer [bib_ref] The compensatory protective effects of social support at work in presenteeism during..., Chen [/bib_ref] [bib_ref] A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests,..., Griffeth [/bib_ref]. In other words, heavy workloads tend to compel employees to commit presenteeism behavior, which over an extended period of time can lead to exhaustion, ill-being, and productivity damage in the ensuing resource loss cycle [bib_ref] Cognitive activation theory of stress (CATS), Ursin [/bib_ref]. Therefore, we hypothesized: Hypothesis 2a (H2a). Long working hours will be positively related to presenteeism, which in turn will lead to a decrease in well-being.
## Hypothesis 2b (h2b).
Long working hours will be positively related to presenteeism, which in turn will lead to a decrease in job performance.
## Work value orientations as resource gains in the "long working hours-presenteeism-outcomes" process
Existing cross-sectional studies have largely confirmed the damaging effects of presenteeism on well-being and productivity. However, the findings from the handful of longitudinal studies are inconsistent, suggesting that the consequences of presenteeism behavior may vary when assessed in an extended timeframe [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. One possibility for such heterogeneity is that some potent individual difference factors may act as buffers to weaken the noxious effects of presenteeism. We thus focused on work value orientations as motivational forces to investigate the moderating mechanisms in a longitudinal design in the integrated "long working hours-presenteeism-outcomes" process.
Work value orientations. Work value orientation refers to work-related reinforcement preferences or tendencies to value specific types of incentives in the work environment [bib_ref] Situational and dispositional determinants of managers' satisfaction, Glynn [/bib_ref]. People vary on the relative importance they attach to one versus another type of value, such that some people give greater weight to extrinsic values and others to intrinsic values [bib_ref] Facilitating optimal motivation and psychological well-being across life's domains, Deci [/bib_ref]. Organizational researchers have distinguished intrinsic from extrinsic work value orientations [bib_ref] Self-determination theory and work motivation, Gagne [/bib_ref] [bib_ref] Pay, intrinsic motivation, extrinsic motivation, performance, and creativity in the workplace: Revisiting..., Gerhart [/bib_ref] [bib_ref] Intrinsic and extrinsic work orientations as moderators of the effect of annual..., Malka [/bib_ref]. An intrinsic work value orientation is conceptualized as encompassing factors that are intrinsic to the job (e.g., autonomy, competence, interesting, and challenging). An extrinsic work value orientation is conceptualized as referring to financial and social rewards as well as security offered by a job. Numerous empirical studies and metaanalyses have revealed that an emphasis on intrinsic goals, relative to extrinsic goals, is associated with better well-being and task performance [bib_ref] Self-determination theory and work motivation, Gagne [/bib_ref] [bib_ref] Self-determination theory in work organizations: The state of a science, Deci [/bib_ref] [bib_ref] Do intrinsic and extrinsic motivation relate differently to employee outcomes?, Kuvaas [/bib_ref].
Work value orientations and the motivational mechanism of resource gains. Responding to Ruhle et al.'s observation that employees' positive motives are understudied yet important mechanisms for understanding presenteeism [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] , we contend that employees who have a high intrinsic work value orientation may consider presenteeism as a means to exercise autonomy, master competence, and receive satisfaction from interesting and challenging work [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. Research on presenteeism has found that in the challenging long working hours of the Chinese work setting, driven by intrinsic motivation, some employees voluntarily go to work when ill because they truly enjoy the job and want to show diligence and loyalty [bib_ref] Unhealthy and present: Motives and consequences of the act of presenteeism among..., Lu [/bib_ref]. Thus, when they commit presenteeism they will exert more effort to compensate for potential performance decrements due to illness [bib_ref] Compensatory control in the regulation of human performance under stress and high..., Hockey [/bib_ref]. Through such motivational self-regulation the detrimental effects of presenteeism on job performance and well-being may be alleviated.
On the contrary, extrinsically oriented people primarily focus on obtaining external indicators of worth, such as social approval and rewards. Two meta-analyses of organizational research have confirmed that providing financial incentives is associated with higher performance [bib_ref] The effects of incentives on workplace performance: A meta-analytic review of research..., Condly [/bib_ref] [bib_ref] Are financial incentives related to performance? A meta-analytic review of empirical research, Jenkins [/bib_ref]. A subsequent literature review further presented empirical evidence for the proposition that extrinsic incentives are as beneficial as intrinsic rewards in realizing high job performance [bib_ref] Pay, intrinsic motivation, extrinsic motivation, performance, and creativity in the workplace: Revisiting..., Gerhart [/bib_ref]. Research on presenteeism has also found that in the challenging long working hours of the Chinese work setting, driven by extrinsic motivation, some employees go to work when ill because they want to protect their income and job security, gain social approval from significant others (supervisors, colleagues, and customers), secure career prospects, and show team spirit [bib_ref] Unhealthy and present: Motives and consequences of the act of presenteeism among..., Lu [/bib_ref]. Through such motivational self-regulation the detrimental effects of presenteeism on job performance and well-being may be alleviated.
In the context of the present study, for Taiwanese employees working under the demanding conditions of long hours and through illness, resource depletion is high and continuous, amplifying the value of resource gains [bib_ref] Cognitive activation theory of stress (CATS), Ursin [/bib_ref]. Thus, intrinsic and extrinsic work value orientations may act as personal motivational resources of "drives within" for employees to construct meaning and purpose with which to protect well-being and persevere in maintaining performance standards. Accordingly, we proposed that work value orientations would moderate the indirect effect of long working hours on well-being and job performance via presenteeism behavior. We thus hypothesized:
Hypothesis 3a (H3a). Intrinsic work value orientation will moderate the indirect effects of long working hours on well-being and job performance via presenteeism, such that the effects will be weaker for employees with a higher intrinsic work value orientation.
Hypothesis 3b (H3b). Extrinsic work value orientation will moderate the indirect effects of long working hours on well-being and job performance via presenteeism, such that the effects will be weaker for employees with a higher extrinsic work value orientation. [fig_ref] Figure 1: The research framework [/fig_ref] depicts the research models presented above, one with an intrinsic work value orientation and the other with an extrinsic work value orientation as the moderator. may be alleviated.
In the context of the present study, for Taiwanese employees working under the demanding conditions of long hours and through illness, resource depletion is high and continuous, amplifying the value of resource gains [bib_ref] Cognitive activation theory of stress (CATS), Ursin [/bib_ref]. Thus, intrinsic and extrinsic work value orientations may act as personal motivational resources of "drives within" for employees to construct meaning and purpose with which to protect well-being and persevere in maintaining performance standards. Accordingly, we proposed that work value orientations would moderate the indirect effect of long working hours on well-being and job performance via presenteeism behavior. We thus hypothesized: Hypothesis 3a. Intrinsic work value orientation will moderate the indirect effects of long working hours on well-being and job performance via presenteeism, such that the effects will be weaker for employees with a higher intrinsic work value orientation.
Hypothesis 3b. Extrinsic work value orientation will moderate the indirect effects of long working hours on well-being and job performance via presenteeism, such that the effects will be weaker for employees with a higher extrinsic work value orientation. [fig_ref] Figure 1: The research framework [/fig_ref] depicts the research models presented above, one with an intrinsic work value orientation and the other with an extrinsic work value orientation as the moderator.
# Method
## Procedure and participants
All of our participants were white-collar full-time employees working in different organizations of diverse industries across Taiwan. We employed a two-wave study design in which all independent variables (working hours, presenteeism, and intrinsic as well as extrinsic work value orientations) were measured at time 1 (T1) and dependent variables (well-being and job performance) were measured twice with an interval of five months. In all of the analysis reported below well-being and job performance measured at time 2 (T2) were used as dependent variables, while those at T1 were treated as the baseline levels and controlled for. While there is a constant call for more longitudinal studies in organizational research, there is no consensus for the optimal temporal frame [bib_ref] The "quick start guide" for conducting and publishing longitudinal research, Ployhart [/bib_ref]. The Demerouti et al. study found a long-term effect (time frame of 1.5 years) of presenteeism
# Method
## Procedure and participants
All of our participants were white-collar full-time employees working in different organizations of diverse industries across Taiwan. We employed a two-wave study design in which all independent variables (working hours, presenteeism, and intrinsic as well as extrinsic work value orientations) were measured at time 1 (T1) and dependent variables (well-being and job performance) were measured twice with an interval of five months. In all of the analysis reported below well-being and job performance measured at time 2 (T2) were used as dependent variables, while those at T1 were treated as the baseline levels and controlled for. While there is a constant call for more longitudinal studies in organizational research, there is no consensus for the optimal temporal frame [bib_ref] The "quick start guide" for conducting and publishing longitudinal research, Ployhart [/bib_ref]. The Demerouti et al. study found a long-term effect (time frame of 1.5 years) of presenteeism on burnout for Dutch nurses [bib_ref] Present but sick: A three-wave study on job demands, presenteeism and burnout, Demerouti [/bib_ref] , and the two Taiwanese studies found a short-term effect (time frames of 2 and 3 months) of presenteeism on well-being and job satisfaction [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref] [bib_ref] Unhealthy and present: Motives and consequences of the act of presenteeism among..., Lu [/bib_ref]. However, neither of the Taiwanese studies found a significant relationship between presenteeism and job performance. We thus decided to test the effects of working hours and presenteeism on well-being and job performance in a medium-term timeframe (5 months), allowing sufficient time for presenteeism to incubate its effects on job performance and well-being.
The present study was approved by the Research Ethics Committee of the principal researcher's institute. A paper-pencil survey was carried out using convenient sampling to recruit participants through personal contacts of the principle researcher. Some participants were enrolled in executive education programs, and others were recruited through man-agers in various organizations. With the assistance of the contact persons we successfully collected data at two time points. At time 1 (T1) a cover letter accompanied the questionnaire, explaining the aim of our study and assuring confidentiality. The initial survey was completed by 474 persons (response rate: 80.75%). Five months later 291 persons completed the second survey (retention rate of 61.39%). Using respondents' self-generated "matching" codes, T1 and T2 data from 275 persons were combined. We examined the attrition bias by comparing the participants in the panel sample and the dropouts with regard to demographic characteristics and the mean scores of all variables (T1). We found no significant differences in any variables, indicating no serious attrition bias.
The sample was 36.3% male and 63.7% female, with a mean age of 36.44 (SD = 8.68, range = 23-63) and a mean job tenure of 6.86 years (SD = 6.79). Just over half of the sample (50.2%) were married. Most participants (96.50%) had a college education and a quarter of the respondents (24.70%) were managers. We asked participants to report the size of their organizations in three categories, namely SMEs employing under 250 people (35%), medium enterprises with between 251 and 1000 (27.5%) employees, and large companies employing over 1000 employees (37.50%). We also asked participants to identify the industries of their organizations and found that manufacturing (27.8%), high-tech (22.9%), and service (15.9%) were the top three.
## Measures
The structured questionnaire was written in Chinese, and all the standard measures have been used and validated with Chinese samples in previous studies (the Chinese validation reference is given for each scale below).
Working hours. Following the practice of a global survey of working hours (e.g., International Labour Organization, OECD) and the suggestion from a literature review on workaholism per working long hours, we assessed weekly work hours in the present study [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref] [bib_ref] Beyond workaholism: Towards a general model of heavy work investment, Snir [/bib_ref]. Participants reported their actual working hours in a typical week, including reported as well as nonreported overtime, and working on side jobs.
Presenteeism. As pointed out in the latest literature review, studies focusing on presenteeism as a behavior mainly draw upon unvalidated single items [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. We thus used the only non-single-item presenteeism behavior scale reviewed in Ruhle et al. [fig_ref] Table 2: Comparison of alternative factor structures for measurement validation [/fig_ref] , p. 349). The scale is developed and validated for Chinese populations by Lu et al. to measure the act of "sickness presenteeism" without evaluation or negative connotations (e.g., "Although you felt sick, you still forced yourself to go to work") [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref] [bib_ref] A cross-cultural examination of presenteeism and supervisory support, Lu [/bib_ref]. With a timeframe of the "past 6 months", four-point scales were used (1 = never, 4 = more than five times) to rate the behavioral frequency of presenteeism. This scale has shown high reliability and construct validity in Chinese samples [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref] [bib_ref] A cross-cultural examination of presenteeism and supervisory support, Lu [/bib_ref] [bib_ref] The compensatory protective effects of social support at work in presenteeism during..., Chen [/bib_ref] [bib_ref] Effects of leader-member exchange and workload on presenteeism, Wang [/bib_ref]. The internal consistency reliability of the scale was α = 0.88 in the present study.
Job performance. We used the four-item scale developed by Ang, Van Dyne, and Begley to assess job performance (e.g., "My supervisor is satisfied with the level of my performance") [bib_ref] The employment relationships of foreign workers versus local employees: A field study..., Ang [/bib_ref] [bib_ref] Relationships of work-family conflict coping strategies with adaptational consequences, Chang [/bib_ref]. Five-point rating scales were used (1 = disagree very much, 5 = agree very much). This scale has shown high reliability and validity for Chinese employees [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref]. (T1 α = 0.88, T2 α = 0.89.)
Well-being. As suggested by Ruhle et al. in the latest literature review, presenteeism behavior may be more related to general well-being than specific health consequences [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. One previous study measuring presenteeism behavior three months apart from self-rated symptoms failed to find a relationship [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref]. We therefore focused on subjective well-being [bib_ref] Subjective well-being: The science of happiness and a proposal for a national..., Diener [/bib_ref] , measured with the five-item version of the Chinese Happiness Inventory developed by Lu [bib_ref] Culture, self, and subjective well-being: Cultural psychological and social change perspectives, Lu [/bib_ref]. The scale has robust reliability and validity for Chinese populations [bib_ref] Chinese well-being, Lu [/bib_ref]. A sample item is "I am satisfied with most things in my life." Respondents rated each item on a four-point scale (0 = strongly disagree, 3 = strongly agree). A higher total score indicated a higher level of happiness. (T1 α = 0.89, T2 α = 0.89.)
Work value orientations. Intrinsic and extrinsic work value orientations were measured with the ten-item Chinese Work Value Scale [bib_ref] Work values and job adjustment of Taiwanese workers, Lu [/bib_ref]. Six items are for intrinsic work value orientation (e.g., "A sense of achievement", "Challenging work") and four are for extrinsic work value orientation (e.g., "Good income", "Job security"). Six-point rating scales were used (1 = disagree very much, 6 = agree very much). The internal consistency reliability was α = 0.87 for intrinsic work value orientation and α = 0.86 for extrinsic work value orientation in the present study.
In addition, information on sex (coded male = 0, female = 1), age, marital status (coded married = 1, not married = 0), education attainment (converted to years of education), tenure of current job (in years), and job position (coded managers = 1, employees = 0) were recorded. Finally, to detect self-report bias we assessed social desirability with 3 items (e.g., "There have been occasions when I took advantage of someone", negation indicates higher social desirability bias). Pearson correlation analysis showed that the social desirability score had no significant correlation with any of the main study variables. Similar patterns of results were obtained with or without social desirability as the control variable. For parsimony we report the results below, without controlling for the social desirability score.
# Results
## Correlations among variables
The bivariable correlations were computed, and the results are shown in [fig_ref] Table 1: Interrelations among research variables [/fig_ref]. Working hours positively correlated with presenteeism and negatively correlated with job performance in addition to well-being. The average working hours were 44.54 per week, much longer than the statutory cap of 40 h/week legislated in Taiwan. Notes: sex: 0 = female, 1 = male; marital status: 0 = not married, 1 = married; and job position: 0 = employees, 1 = managers. * p < 0.05, ** p < 0.01, and *** p < 0.001.
## Primary analyses
Measurement model. Prior to testing the structural relationships we conducted confirmatory factor analyses (CFAs) on T1 and T2 multi-item measures to examine the discriminant validity of our research variables. As shown in [fig_ref] Table 2: Comparison of alternative factor structures for measurement validation [/fig_ref] , the hypothesized three-factor model for T1 variables (i.e., intrinsic work value orientation, extrinsic work value orientation, and presenteeism) and the hypothesized two-factor model for T2 variables (i.e., well-being and job performance), respectively, had the best fit relative to the alternative models, indicating the unidimensionality of the measures. Additionally, all factor loadings for the measured variables on their latent factors were significant (presenteeism λ range = 0.68 to 0.76; intrinsic work value orientation λ range = 0.68 to 0.89; extrinsic work value orientation λ range = 0.77 to 0.87; well-being λ range = 0.84 to 0.92; and job performance λ range = 0.81 to 0.88). Furthermore, the composite reliability of the latent factors ranged from ρ = 0.73 to ρ = 0.92 (intrinsic work value orientation ρ = 0.73, extrinsic work value orientation ρ = 0.83, well-being ρ = 0.92, and job performance ρ = 0.89). The results indicate that our measures had adequate construct validity.
## Hypotheses testing
Testing main effects and mediating effects. We conducted a path analysis with Mplus (version 6) to test our hypotheses. All manifest variables were standardized prior to the analysis to put them on a common scale. Path estimates and conditional indirect effects were evaluated using bias-corrected confidence intervals based on 2000 bootstrap resamples with replacement [bib_ref] An index and test of linear moderated mediation, Hayes [/bib_ref].
Hypothesis one (a and b) proposed the negative relationships between long working hours and outcomes (well-being and job performance). The positive relationship between the intrinsic and extrinsic work orientations (r = 0.41, [fig_ref] Table 1: Interrelations among research variables [/fig_ref] might be due to the importance that employees attach to work values in general [bib_ref] On the relations among work value orientations, psychological need satisfaction and job..., Vansteenkiste [/bib_ref]. Thus, to examine the "net effect" of each work value orientation we controlled for the extrinsic orientation when analyzing the effect of intrinsic orientation, and vice versa.
As shown in [fig_ref] Table 3: Bootstrapping effects and 95% confidence intervals [/fig_ref] , working hours at T1 were negatively associated with well-being (β = −0.102, se = 0.05, and p < 0.01) and job performance (β = −0.134, se = 0.04, and p < 0.01) at T2, supporting hypotheses 1a and 1b. [fig_ref] Table 3: Bootstrapping effects and 95% confidence intervals [/fig_ref] also presents the mediating effects of presenteeism: (1) the indirect effect of working hours on well-being through presenteeism was −0.14 (p < 0.005; 95% CI (−0.1013, −0.0018)); (2) the indirect effect of working hours on job performance through presenteeism was -0.21 (p < 0.001; 95% CI (−0.1207, −0.0110)). The mediation effects of presenteeism on both the "working hours-well-being" and "working hours-performance" paths were significant. To further evaluate the effect size of the mediation effects we used Preacher and Kelley's recommended kappa-squared value (κ 2 ) [bib_ref] Effect size measures for mediation models: Quantitative strategies for communicating indirect effects, Preacher [/bib_ref]. We found that: (1) the effect size for the mediating effect of presenteeism on the relationship between working hours and well-being was medium in strength (κ 2 = 0.11); (b) the effect size for the mediating effect of presenteeism on the relationship between working hours and job performance was large in strength (κ 2 = 0.28) [bib_ref] Effect size measures for mediation models: Quantitative strategies for communicating indirect effects, Preacher [/bib_ref]. The above results provide full support for hypotheses 1a and 1b. Testing the latent moderated mediation effects. We followed Cheung and Lau's suggestion to conduct latent moderated structural equation modeling (LMS) in Mplus [bib_ref] Accuracy of parameter estimates and confidence intervals in moderated mediation models: A..., Cheung [/bib_ref]. We formulated four latent moderated mediation models (models three, four, five, and six in [fig_ref] Table 4: Model fit index and the index of moderated mediation models [/fig_ref] to examine the moderating effects of intrinsic/extrinsic work value orientations on the indirect effects of working hours via presenteeism on well-being and job performance, respectively (see [fig_ref] Figure 1: The research framework [/fig_ref]. These four hypothesized moderated mediation models were compared against two base models (models one and two), each with one focal dependent variable. Specifically, model one was the simple mediation model linking long working hours to well-being through presenteeism. Model two was the simple mediation model linking long working hours to job performance through presenteeism. Model three was the moderated mediation model with intrinsic work value orientation affecting the indirect effect of long working hours on well-being through presenteeism. Model four was the moderated mediation model with intrinsic work value orientation affecting the indirect effect of long working hours on job performance through presenteeism. Model five was the moderated mediation model with extrinsic work value orientation affecting the indirect effect of long working hours on well-being through presenteeism. Model six was the moderated mediation model with extrinsic work value orientation affecting the indirect effect of long working hours on job performance through presenteeism. In other words, in the latent moderated mediation models (model three to model six) we added the interactions among latent predictor variables (i.e., presenteeism × intrinsic/extrinsic work value orientations) to the mediation structural models (i.e., "working hours→presenteeism→well-being/job performance", model one and model two). As shown in [fig_ref] Table 4: Model fit index and the index of moderated mediation models [/fig_ref] , the moderated mediation models for well-being (model three and model five) had a better fit than the simple mediation model (base model one). Similarly, the moderated mediation models for job performance (model four and model six) had a better fit than the simple mediation model (base model two). These results indicated the improvement values for including the moderators in the overall structural models. [fig_ref] Table 4: Model fit index and the index of moderated mediation models [/fig_ref] also showed that only the two latent interaction terms, namely the presenteeism × intrinsic orientation (on job performance in model four, β = −0.138, p < 0.01) and presenteeism × extrinsic orientation (on well-being in model five, β = 0.162, p < 0.01), were significant. However, the other two latent interaction terms, namely the presenteeism × intrinsic orientation (on well-being in model three) and presenteeism × extrinsic orientation (on job performance in model six) were not statistically significant. Thus, the latent moderated mediation analysis results partially supported H3a and H3b. Specifically, intrinsic work value orientation moderated the indirect effect of long working hours on job performance via presenteeism, whereas extrinsic work value orientation moderated the indirect effect of long working hours on well-being via presenteeism.
To further probe the statistically significant conditional indirect (moderated mediation) effect we used estimates from the latent variable moderated mediation model to calculate the conditional indirect effect of working hours via presenteeism on well-being/job performance at various levels of intrinsic/extrinsic orientation. We estimated the indirect effects at low (−1 SD), medium (1 SD), and high (+1 SD) levels of the moderator. We explored the shape of the significant interactions by conducting a simple slope test. For job performance, the slope of presenteeism for employees with a high intrinsic work value orientation was non-significant (effect = 0.0112, bias-corrected bootstrap confidence interval, BCCI, = (−0.0014 to 0.0005)), whereas the slope of presenteeism for those with medium and low intrinsic orientations were significantly negative (medium: effect = −0.0028, BCCI = (−0.0252 to −0.0134); low: effect = −0.0143, BCCI = (−0.1231 to −0.0623)). This indicates that over a five-month period the relationship between presenteeism and job performance was negative for employees of medium and low levels of intrinsic work value orientation; presenteeism was unrelated to job performance when employees had a high level of intrinsic orientation (see [fig_ref] Figure 2: Interaction of presenteeism and intrinsic work value orientation [/fig_ref]. Thus, the results partially supported H3a, that intrinsic work value orientation moderated the indirect effect of long working hours on job performance via presenteeism, such that the effect became nonsignificant for employees with a higher intrinsic work value orientation. For well-being, the slope of presenteeism for employees of a low extrinsic work value orientation was significantly negative (effect = −0.0113, BCCI = (−0.0096 to −0.0029)), whereas the slope of presenteeism for those of a high extrinsic orientation was significantly positive (effect = 0.0088, BCCI = (0.0023 to 0.0036)). This indicates that over a fivemonth period the relationship between presenteeism and well-being was negative for employees of a low level of extrinsic work value orientation, but positive for those of a high level of extrinsic orientation (see [fig_ref] Figure 3: Interaction of presenteeism and extrinsic work value orientation [/fig_ref]. Thus, the results partially supported H3b, that extrinsic work value orientation moderated the indirect effect of long working hours on well-being via presenteeism, such that the effect was positive for employees with a higher intrinsic work value orientation. For well-being, the slope of presenteeism for employees of a low extrinsic work value orientation was significantly negative (effect = −0.0113, BCCI = (−0.0096 to −0.0029)), whereas the slope of presenteeism for those of a high extrinsic orientation was significantly positive (effect = 0.0088, BCCI = (0.0023 to 0.0036)). This indicates that over a five-month period the relationship between presenteeism and well-being was negative for employees of a low level of extrinsic work value orientation, but positive for those of a high level of extrinsic orientation (see [fig_ref] Figure 3: Interaction of presenteeism and extrinsic work value orientation [/fig_ref]. Thus, the results partially supported H3b, that extrinsic work value orientation moderated the indirect effect of long working hours on well-being via presenteeism, such that the effect was positive for employees with a higher intrinsic work value orientation.
cantly positive (effect = 0.0088, BCCI = (0.0023 to 0.0036)). This indicates that over a fivemonth period the relationship between presenteeism and well-being was negative for employees of a low level of extrinsic work value orientation, but positive for those of a high level of extrinsic orientation (see [fig_ref] Figure 3: Interaction of presenteeism and extrinsic work value orientation [/fig_ref]. Thus, the results partially supported H3b, that extrinsic work value orientation moderated the indirect effect of long working hours on well-being via presenteeism, such that the effect was positive for employees with a higher intrinsic work value orientation.
# Discussion
## Theoretical implications
As repeatedly found in reviews of the presenteeism literature, the bulk of existing empirical work in organizational research has focused on categorizing the determinants (or antecedents) of presenteeism, namely understanding the decision process of employees to be present or absent during sickness [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref] [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Presenteeism in the workplace: A review and research agenda, Johns [/bib_ref] [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref]. Only recently have some theoretical efforts been invested in clarifying and understanding the potential outcomes of presenteeism [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref] [bib_ref] Presenteeism: An introduction to a prevailing global phenomenon, Karanika-Murray [/bib_ref] [bib_ref] The health-performance framework of presenteeism: Towards understanding an adaptive behaviour, Karanika-Murray [/bib_ref]. Thus, the conclusion of an earlier review, that "the current body of scientific literature does not provide sufficient evidence to draw conclusions on the consequences of sickness presence" (p. 216), still largely holds [bib_ref] Sickness presence, Vingård [/bib_ref]. More importantly, we have little insight into the psychological mechanisms, contextual factors, and processes that link presenteeism behavior to its outcomes for individuals and organizations, regarding acclaimed well-being damage and productivity loss [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref] [bib_ref] The health-performance framework of presenteeism: Towards understanding an adaptive behaviour, Karanika-Murray [/bib_ref] [bib_ref] Presenteeism as a global phenomenon: Unraveling the psychosocial mechanisms from the perspective..., Cooper [/bib_ref].
To address this void in knowledge we in the present study examined the complete process of "antecedents-presenteeism-outcomes", incorporating motivational personal resources as the explanatory psychological mechanisms for affecting the outcomes five months later. More importantly, to answer the call for theoretical developments in presenteeism research [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Presenteeism: A review and research directions, Lohaus [/bib_ref] we pivoted our theoretical framework on the resource loss/gain dynamism of COR while adopting a motivational perspective on presenteeism [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref] [bib_ref] Motivation and presenteeism: The whys and whats, Ma [/bib_ref] [bib_ref] The health-performance framework of presenteeism: Towards understanding an adaptive behaviour, Karanika-Murray [/bib_ref]. Our results demonstrated that long working hours precipitated employees to work when sick, which led to impaired performance and depressed well-being, thus raising a red flag for the accumulating risk of presenteeism (in a medium-term timeframe). These findings corroborate the resource loss cycle posited in the COR theory by showing that enduring resource depletion caused by long work hours and sickness presenteeism can harm employees' job performance and well-being. However, intrinsic and extrinsic work value orientations can function as protective mechanisms, setting in motion resource gain cycles to counteract resource loss.
As Cooper and Lu purport in their motivational theory of excessive availability for work (EAW, e.g., sickness presenteeism), that "Good outcomes (work well-being, unharmed work performance) will be more likely among people who commit EAW as a voluntary act through freewill or through internalization" (proposition 4a, p. 9; italics added), our findings on work value orientations lend some support for this proposition. Employees who have a high endorsement of intrinsic work value orientation are more likely to commit presenteeism as a voluntary act through freewill, pursuing fun, interests, and challenges at work [bib_ref] Going to work ill: A meta-analysis of the correlates of presenteeism and..., Miraglia [/bib_ref] [bib_ref] Motivation and presenteeism: The whys and whats, Ma [/bib_ref] [bib_ref] Unhealthy and present: Motives and consequences of the act of presenteeism among..., Lu [/bib_ref]. Those who have a high endorsement of extrinsic work value orientation are more likely to commit presenteeism as a voluntary act through internalization, winning rewards and approval by serving one's duty, and securing career prospects through compliance with norms [bib_ref] A cross-cultural examination of presenteeism and supervisory support, Lu [/bib_ref] [bib_ref] Affective and normative motives to work overtime in Asian organizations: Four cultural..., Kang [/bib_ref].
Corroborating Cooper and Lu's proposition 4a, we have unraveled the motivational mechanisms of human agency by demonstrating the moderating effects of intrinsic and extrinsic work value orientations on the presenteeism-outcomes linkage [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. The buffering effects of work value orientation underlined the possibility that with the right configuration of personal, organizational, and cultural factors, presenteeism could have positive outcomes for employees and organizations [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. Our results highlighted the critical importance of motivational resources in understanding the functional aspect of presenteeism [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref] [bib_ref] Motivation and presenteeism: The whys and whats, Ma [/bib_ref] [bib_ref] The health-performance framework of presenteeism: Towards understanding an adaptive behaviour, Karanika-Murray [/bib_ref].
The protective mechanism of work value orientation can be explained by the satisfaction of basic psychological needs. Corroborating the existing research, individuals with a high intrinsic work value orientation cherish opportunities for intellectual fulfilment, creative self-expression, and the pleasure associated with task mastery on the job [bib_ref] Intrinsic motivation and extrinsic incentives jointly predict performance: A 40-year meta-analysis, Cerasoli [/bib_ref]. These employees view presenteeism as means to exercise autonomy and to master competence. Such positive motives thus act as personal resources attenuating the negative impact of presenteeism on job performance, as we predicted. Individuals with a high extrinsic work value orientation identify and internalize extrinsic values of monetary rewards, promotion, social recognition, realizing self-worth, and security [bib_ref] Self-determination theory and work motivation, Gagne [/bib_ref] [bib_ref] Pay, intrinsic motivation, extrinsic motivation, performance, and creativity in the workplace: Revisiting..., Gerhart [/bib_ref]. Such positive motives may enhance self-affirmation and uplift well-being [bib_ref] Chinese well-being, Lu [/bib_ref] , but may not necessarily help performance quality [bib_ref] Self-determination theory and work motivation, Gagne [/bib_ref]. Our finding that extrinsic work value orientation buffered the negative effects of presenteeism on well-being can also be explained by the effort-reward imbalance (ERI) model [bib_ref] Adverse health effects of high-effort/low-reward conditions, Siegrist [/bib_ref]. Research on the ERI has robustly found that rewards of esteem, security, pay, and promotion counteract work stress and strain by protecting well-being and health [bib_ref] Effort-reward imbalance at work and cardiovascular diseases, Siegrist [/bib_ref].
The intriguing protective effect of extrinsic work value orientation for employees' well-being during presenteeism pivots on the cultural context of presenteeism [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. Traditional Confucian values put a great emphasis on acquiring stability, security, wealth, and status for the prosperity and welfare of the family [bib_ref] Affective and normative motives to work overtime in Asian organizations: Four cultural..., Kang [/bib_ref] [bib_ref] Modernization, cultural change, and the persistence of traditional values, Inglehart [/bib_ref]. In contemporary Chinese societies, aspiring to realize extrinsic rewards offered by a job is thus not only meaningful and essential but also recognized as abiding by the cultural imperative. For employees in a collectivist Chinese society, committing presenteeism to protect their career image satisfies the basic psychological needs for relatedness and competence. Taken together, our findings added intrinsic and extrinsic work value orientations to the list of psychological mechanisms (e.g., self-efficacy, psychological detachment) that can ameliorate the problematic consequences of presenteeism behavior on both well-being and productivity [bib_ref] Presenteeism and health over time among Chinese employees: The moderating role of..., Lu [/bib_ref] [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref].
Last but not the least, our findings testified the resource loss/gain dynamism posited in the COR theory by showing that under intensive work situations (i.e., presenteeism triggered by long working hours) personally meaningful motivational resources, whether of an intrinsic or extrinsic source, are effective buffers, alleviating the accumulating deleterious effects of presenteeism on well-being and productivity.
## Managerial implications
What we have found in the present study has implications for both organizational strategies and managerial practices. As we have demonstrated, long working hours had a catalytic effect on sickness presenteeism and its subsequent damaging effects on employees' job performance and well-being. In our data the average working week was ca. 45 h, comparable to previous findings in Taiwan [bib_ref] Protecting job performance and well-being in the demanding work context: The moderating..., Lu [/bib_ref]. It is thus imperative that organizations endeavor to shorten actual work hours in order to stamp the origin of the "accumulative consequences of downstream health" [bib_ref] Presenteeism in the workplace: A review and research agenda, Johns [/bib_ref]. In addition, organizations and managers should reinforce the legitimacy of taking sick leave when needed and adjust task allocation or provide job replacement arrangements to lessen the felt pressure of employees to commit presenteeism. The appropriate use of sick leave and recuperation as a health-promoting strategy is well-documented in the recovery research [bib_ref] Job demands-resources theory: Taking stock and looking forward, Bakker [/bib_ref].
Moreover, organizations should create work environments that foster intrinsic work motivation, for example, equipping employees with autonomy and flexibility at work. Organizations should also provide attractive monetary or other extrinsic reinforcers (e.g., praise) that can enhance employees' perceptions of fairness and organizational support [bib_ref] Show them the money? The role of pay, managerial need support, and..., Olafsen [/bib_ref]. Such psychological states enhance employee well-being [bib_ref] Effects of reward on intrinsic motivation: Negative, neutral, and positive, Eisenberger [/bib_ref]. As shown in the present study and the ERI literature, extrinsic rewards help lift morale and spirits, which keep people going in demanding and precarious work conditions, at least in the short-run [bib_ref] Adverse health effects of high-effort/low-reward conditions, Siegrist [/bib_ref] [bib_ref] Effort-reward imbalance at work and cardiovascular diseases, Siegrist [/bib_ref].
## Limitations and directions for future research
Our study has some limitations. Firstly, we used self-report measures, which may increase the threat of common method variance (CMV) bias [bib_ref] Common method biases in behavioral research: A critical review of the literature..., Podsakoff [/bib_ref]. In an effort to minimize such bias we adopted a longitudinal design to separate the independent variables (working hours, presenteeism, and work value orientations) from dependent variables (job performance and well-being) in time. Furthermore, researchers have noted that CMV might enhance main effects, but that interaction effects were hardly attributable to such method bias [bib_ref] Common method biases in behavioral research: A critical review of the literature..., Podsakoff [/bib_ref]. Thus, it is unlikely that our evidence for a moderated mediation process is solely due to common method bias. However, future research may adopt a supervisoremployee dyadic design to crossvalidate our findings by obtaining supervisor-rated job performance data.
We extended the studies on presenteeism to an East Asian society (Taiwan); however, the relatively small and nonrandom sample we had may limit the generalization of our findings. Future studies should include more diverse samples from other Asian countries to establish the generalizability of our findings as well as those from Western studies. In addition, a lack of firm-level controls may be an omitted variable bias. Previous studies on presenteeism have highlighted the relevance of occupations, for instance in terms of the personal obligations teachers and care workers feel towards pupils and care recipients [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. We probed some of the organizational factors in the present study, such as the organizational climate of presenteeism and job replacement policy [bib_ref] Working sick and out of sorts: A cross-cultural approach on presenteeism climate,..., Ferreira [/bib_ref] ; these results were not reported in the tables as they were non-significant. Nonetheless, future research could target specific occupations, such as professionals and service workers, to better contextualize the study of presenteeism.
Finally, we measured the effects of presenteeism on well-being and performance after 5 months, but the time frame is nevertheless short for making strong statements about consequences. Additionally, a longer time frame may be useful for understanding if differences between intrinsic and extrinsic motivational drivers may have, in the long-run, different effects concerning well-being and productivity. Previous research on workaholism has found that workers who are driven only by external motivation can be exposed to a progressive exhaustion of resources. For example, perfectionism, conscientiousness, and self-efficacy were associated with an increase in workaholism only in contexts where the workload was very high or the overwork climate was strong [bib_ref] Does Workload Moderate the Association Between Perfectionism and Workaholism? A Longitudinal Study, Falco [/bib_ref] [bib_ref] Are workaholics born or made? Relations of workaholism with person characteristics and..., Mazzetti [/bib_ref]. Thus, the apparent benefit of extrinsic work value orientation for employees' well-being during presenteeism, as we found in the present study, requires more refined and nuance probing in future studies, ideally with a longer timeframe.
# Conclusions
As argued by Cooper and Lu in addition to Ruhle et al., sickness presenteeism as a work behavior should be viewed as neutral; it is through the interaction between a person (as a motivational/regulatory being) and the environment (with resources and constraints) that a process for actualizing both the potential costs and benefits of such work behavior is set in motion [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref] [bib_ref] Excessive availability for work: Good or bad? Charting underlying motivations and searching..., Cooper [/bib_ref]. In other words, the often negatively construed presenteeism behavior should be re-examined in contexts across multiple levels: individual, group, leader, organization, and overarching/social context [bib_ref] To work, or not to work, that is the question"-Recent trends and..., Ruhle [/bib_ref]. We have taken the pioneering step to investigate the outcomes of sickness presenteeism from the above contextual perspectives. Our findings make substantial contributions to the existing knowledge on presenteeism by highlighting the underlying psychological mechanisms (i.e., motivational resources) in the integrated process of the "working long hours-presenteeism-outcomes" linkage. Specifically, intrinsic work value orientation alleviated the negative effect of long working hours on job performance via presenteeism, and extrinsic work value orientation alleviated the negative effect of long working hours on well-being via presenteeism. Our moderated mediation models also highlighted the largely overlooked temporal issue. The lasting effects of overworking on individual outcomes (well-being and performance) provide a more holistic understanding of the prevailing phenomenon of the intensification of work in today's highly competitive, volatile, and uncertain business world.
[fig] Figure 1: The research framework. [/fig]
[fig] 17, Figure 2: SPIN = latent interaction of sickness presenteeism × intrinsic work value orientation. SPEX = latent interaction of sickness presenteeism × extrinsic work value orientation. β = standardized estimate; se = estimated standard error; p = p-value; BCLL = lower limit of bias-corrected bootstrap confidence interval; BCUL = upper limit of bias-corrected bootstrap confidence interval. Model specification: Base model one = simple mediation model for well-being: working hours→presenteeism→well-being. Base model two = simple mediation model for job performance: working hours→presenteeism→job performance. Model three = moderated mediation model for well-being with intrinsic work value orientation as moderator. Model four = moderated mediation model for job performance with intrinsic work value orientation as moderator. Model five = moderated mediation model for well-being with extrinsic work value orientation as moderator. Model six = moderated mediation model for job performance with extrinsic work value orientation as moderator.Int. J. Environ. Res. Public Health 2022, 19, x FOR PEER REVIEW 11 of Interaction of presenteeism and intrinsic work value orientation: moderation on the "presenteeism-job performance" path. [/fig]
[fig] Figure 2: Interaction of presenteeism and intrinsic work value orientation: moderation on the "presenteeism-job performance" path. [/fig]
[fig] Figure 3: Interaction of presenteeism and extrinsic work value orientation: moderation on the "presenteeism-well-being" path. [/fig]
[fig] Author: Contributions: Conceptualization, L.L.; methodology, L.L.; validation, C.L.C.; formal analysis, L.L.; resources, L.L. and C.L.C.; writing-original draft preparation, L.L.; writing-review and editing, L.L. and C.L.C.; funding acquisition, L.L. All authors have read and agreed to the published version of the manuscript. Funding: The empirical research was supported by a grant from the Ministry of Science and Technology, Taiwan, 108-2410-H-002-126-SS3. Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committee of National Taiwan University (protocol code 201905HS024 and date of approval 20190603). [/fig]
[table] Table 1: Interrelations among research variables (N = 275). [/table]
[table] Table 2: Comparison of alternative factor structures for measurement validation. [/table]
[table] Table 3: Bootstrapping effects and 95% confidence intervals (CI) for the mediation models. [/table]
[table] Table 4: Model fit index and the index of moderated mediation models. [/table]
|
Externalizing problems in childhood and adolescence predict subsequent educational achievement but for different genetic and environmental reasons
Background: Childhood behavior problems predict subsequent educational achievement; however, little research has examined the etiology of these links using a longitudinal twin design. Moreover, it is unknown whether genetic and environmental innovations provide incremental prediction for educational achievement from childhood to adolescence. Methods: We examined genetic and environmental influences on parental ratings of behavior problems across childhood (age 4) and adolescence (ages 12 and 16) as predictors of educational achievement at age 16 using a longitudinal classical twin design. Results: Shared-environmental influences on anxiety, conduct problems, and peer problems at age 4 predicted educational achievement at age 16. Genetic influences on the externalizing behaviors of conduct problems and hyperactivity at age 4 predicted educational achievement at age 16. Moreover, novel genetic and (to a lesser extent) nonshared-environmental influences acting on conduct problems and hyperactivity emerged at ages 12 and 16, adding to the genetic prediction from age 4. Conclusions: These findings demonstrate that genetic and shared-environmental factors underpinning behavior problems in early childhood predict educational achievement in midadolescence. These findings are consistent with the notion that earlychildhood behavior problems reflect the initiation of a life-course persistent trajectory with concomitant implications for social attainment. However, we also find evidence that genetic and nonshared-environment innovations acting on behavior problems have implications for subsequent educational achievement, consistent with recent work arguing that adolescence represents a sensitive period for socioaffective development.
# Introduction
The long-established importance of educational achievement for later-life success (e.g.has led to a significant body of work examining the antecedents of school achievement [bib_ref] Intelligence and educational achievement, Deary [/bib_ref] [bib_ref] Early conduct problems and later life opportunities, Fergusson [/bib_ref] [bib_ref] Strong genetic influence on a UK nationwide test of educational achievement at..., Shakeshaft [/bib_ref] [bib_ref] Mental health problems and educational attainment in adolescence: 9-year follow-up of the..., Veldman [/bib_ref]. One area of specific focus has been childhood and adolescent behavior problems. Several studies have reported that genetic and environmental factors underpin individual differences in both educational achievement [bib_ref] Heritability of educational achievement in 12-yearolds and the overlap with cognitive ability, Bartels [/bib_ref] [bib_ref] Strong genetic influence on a UK nationwide test of educational achievement at..., Shakeshaft [/bib_ref] and child and adolescent behavior problems [bib_ref] Genetics and developmental psychopathology: 2. The main effects of genes and environment..., Eaves [/bib_ref] [bib_ref] Heritable influences on behavioural problems from early childhood to mid-adolescence: Evidence for..., Lewis [/bib_ref]. Furthermore, these genetic and environmental factors have been found to be moderately to substantially overlapping: that is, some of the same genes and experiences affect both educational achievement and behavior problems [bib_ref] Moderating effects of personality on the genetic and environmental influences of school..., Hicks [/bib_ref] [bib_ref] Disruptive behavior and school grades: Genetic and environmental relations in 11-year-olds, Johnson [/bib_ref]. Less well understood is the extent to which genetic and environmental influences on childhood and adolescent behavior problems predict educational achievement at age 16 (the end of mandatory education in many countries).
In the current study, we used a longitudinal twin design to examine whether childhood and adolescent behavior problems share genetic and environmental influences with educational achievement, and how these influences relate over time. Specifically, we sought to estimate the extent to which genetic and environmental influences acting on behavior problems in early childhood, before formal schooling begins, can predict achievement in public examinations at age 16. Moreover, we examined whether novel genetic and environmental influences on behavior problems, emerging over the course of development [bib_ref] Heritable influences on behavioural problems from early childhood to mid-adolescence: Evidence for..., Lewis [/bib_ref] , would provide additional sources of prediction for educational achievement at age 16. Next we briefly introduce phenotypic and behavioral genetic investigations of educational achievement and behavior problems before moving to formal tests of the role of genetic and environmental influences on behavior problems and educational achievement over childhood and adolescence. problems, hyperactivity) will likely find it harder to pay attention in the classroom or to comply with school rules, and so it would be unsurprising to find inverse associations between externalizing behaviors and school success. A range of studies have examined whether behavior problems and educational achievement are inversely associated and have consistently confirmed this expectation. For example, in a large New Zealand birth cohort, conduct disorder at age 8 was found to predict leaving school at age 18 without educational qualifications [bib_ref] Early conduct problems and later life opportunities, Fergusson [/bib_ref]. Similarly, in a large Canadian community sample, hyperactivity-inattention and aggressiveness-opposition measured in kindergarten were found to predict noncompletion of high school [bib_ref] Mental health problems and educational attainment in adolescence: 9-year follow-up of the..., Veldman [/bib_ref]. A large number of other studies also provide support for the link between early-childhood externalizing behaviors and subsequent low educational achievement [bib_ref] Developmental cascades: Linking academic achievement and externalizing and internalizing symptoms over 20..., Masten [/bib_ref] [bib_ref] Mental health problems and educational attainment in adolescence: 9-year follow-up of the..., Veldman [/bib_ref].
Beyond externalizing behaviors, the links between educational achievement and behavior problems are more mixed. For example, one study noted that prosociality (often referred to as a behavioral strength: [bib_ref] The Strengths and Difficulties Questionnaire: A research note, Goodman [/bib_ref] at age 8 predicted educational achievement 5 years later in midadolescence [bib_ref] Prosocial foundations of children's academic achievement, Caprara [/bib_ref]. A similar observation in a 5-year longitudinal sample of Chinese school students followed from age 8 also found that prosocial competence predicted academic achievement in subsequent years. In contrast, other studies have failed to observe prosociality as a predictor of subsequent educational success (e.g. [bib_ref] Trajectories of prosocial behavior and physical aggression in middle childhood: Links to..., Kokko [/bib_ref]. In the domain of internalizing problems, similarly mixed results have been noted. For example, one study reported that higher levels of anxiety in the preschool years were predictive of higher school grades in early adolescence [bib_ref] Longitudinal effects of preschool behavioral styles on early adolescent school performance, Dilalla [/bib_ref] but other research has failed to observe such links .
Building on the well-replicated phenotypic links between externalizing behaviors and educational achievement have been genetically informative studies seeking to assess the relative roles of the genetic and environmental factors underpinning this association. In early childhood, the link between externalizing behavior and educational achievement has been reported to be mostly attributable to shared-environmental factors, although genetic factors have also been noted to play a role [bib_ref] Genetic and environmental influences on the co-occurrence of early academic achievement and..., Newsome [/bib_ref]. In midchildhood, this pattern appears to shift toward genetic factors accounting for the majority of the phenotypic links between externalizing behaviors and educational achievement. For example, in a large UK cohort [the Twins Early Development Study (TEDS): also used in the current study], hyperactivity and educational success at age 7 were found to share substantial genetic links, alongside more modest nonshared-environmental links. Similarly, results from the Minnesota Twin Family Study showed that at age 11 genetic influences on inattention and educational achievement were highly overlapping, although the genetic link between disruptive behavior and educational achievement, while statistically significant, was less pronounced [bib_ref] Disruptive behavior and school grades: Genetic and environmental relations in 11-year-olds, Johnson [/bib_ref]. In the same sample, achievement striving, self-control, and aggression at age 17 have been reported to be genetically related to educational success (also at age 17), alongside a modest link via nonshared-environmental influences [bib_ref] Moderating effects of personality on the genetic and environmental influences of school..., Hicks [/bib_ref]. Finally, work using the TEDS twin cohort reported that standardized UK high-school exam results at age 16 were heritable and genetically associated with many psychological traits including behavior problems, although associations between educational achievement and specific components of behavior problems were not detailed [bib_ref] The high heritability of educational achievement reflects many genetically influenced traits, not..., Krapohl [/bib_ref].
## The current study
These studies provide insights into common genetic and environmental influences underlying observed relationships between behavior problems and educational achievement. However, this literature is still in its infancy and a number of important questions remain unanswered. First, while childhood externalizing behaviors are phenotypically predictive of adolescent educational achievement [bib_ref] Early conduct problems and later life opportunities, Fergusson [/bib_ref] [bib_ref] Mental health problems and educational attainment in adolescence: 9-year follow-up of the..., Veldman [/bib_ref] , are these phenotypic links explained by genetic or environmental factors? Indeed, conduct problems show stable genetic and shared-environment influences from age 4 to age 16 [bib_ref] Heritable influences on behavioural problems from early childhood to mid-adolescence: Evidence for..., Lewis [/bib_ref]. As such, it is conceivable that either or both of these sources of variance might account for individual differences in their prediction of educational achievement at age 16.
Second, recent work has highlighted that childhood and adolescent externalizing behaviors are underpinned both by stable genetic and environmental influences, as noted above, and also by innovative genetic and environmental influences (i.e. effects that emerge across development: [bib_ref] Heritable influences on behavioural problems from early childhood to mid-adolescence: Evidence for..., Lewis [/bib_ref]. As such, do early-emerging (i.e. ≤age 4) and subsequent (i.e. >age 4) genetic and environmental factors independently relate to later educational achievement?
A range of perspectives have been indirectly informative on this issue. Perhaps most prominently, the developmental taxonomy proposed by [bib_ref] Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy, Moffitt [/bib_ref] stresses that antisocial behavior follows one of two main trajectories: life-course persistent or adolescent-limited. The former is argued to reflect disrupted neuropsychological functioning and temperament difficulties, which in turn negatively impact learning and interpersonal relations, and subsequently can serve to impair life outcomes. The latter trajectory is posited to reflect the extreme cases of the otherwise normative adolescent desire to attain status and a distinct personal identity. Here externalizing behaviors are argued to be simply the manifestation of these goals. Of importance, this subset of adolescents is believed to be relatively goal-directed in their externalizing behaviors: 'adolescence-limited delinquents are likely to engage in antisocial behavior in situations where such responses seem profitable to them, but they are also able to abandon antisocial behavior when prosocial styles are more rewarding' . As such, one would expect that genetic and environmental influences on childhood externalizing would predict educational achievement in adolescence, either as a result of a deleterious developmental cascade, or because the underpinning psychological characteristics of the behavior problems are stable over time and create issues for schooling in a more proximal fashion. In contrast, any genetic and environmental influences on externalizing that emerge in adolescence would be expected to contribute less to the prediction of educational achievement, despite the more proximal nature of these effects. However, recent observations suggest that adolescence is a sensitive period of development for a range of socioaffective processessuch as emotion regulation and impulse control [bib_ref] Risk taking in adolescence new perspectives from brain and behavioral science, Steinberg [/bib_ref]. These processes have well-established links to behavior problems [bib_ref] The relations of regulation and emotionality to children's externalizing and internalizing problem..., Eisenberg [/bib_ref] [bib_ref] Adolescents' emotion regulation in daily life: Links to depressive symptoms and problem..., Silk [/bib_ref] and so this sensitive period of socioaffective development may reflect important independent risk factors for subsequent life success in their own right [bib_ref] The developing social brain: Implications for education, Blakemore [/bib_ref]. As such, it is an open question whether early or later emerging genetic and environmental influences on behavior problems exert the greater impact on educational outcomes.
In addition to our core questions, here we also took the opportunity to examine how internalizing traits (i.e. anxiety, peer problems) and prosociality related to academic achievement, both phenotypically and via underlying genetic and environmental pathways. Establishing the presence and (where relevant) the etiology of such effects is important given the mixed results in these domains, as noted above.
To answer these questions, we used a large and population-representative sample of UK monozygotic (MZ) and dizygotic (DZ) twins who have been followed since birth as participants in the Twins' Early Development Study (TEDS; [bib_ref] Twins' Early Development Study (TEDS): A multivariate, longitudinal genetic investigation of language,..., Oliver [/bib_ref]. TEDS twins have been surveyed on a wide range of behavioral and cognitive characteristics throughout their life span. Here we used the Strengths and Difficulties Questionnaire (SDQ: [bib_ref] The Strengths and Difficulties Questionnaire: A research note, Goodman [/bib_ref] instrument to assess behavior problems with parental ratings obtained at each age in order to maintain a consistent mode of rating across each wave of measurement. We used SDQ scores at ages 4 and 12 for all SDQ scales, and also for age 16 for prosociality, conduct problems, and hyperactivity (SDQ anxiety and peer problems were not measured at age 16 and so were unavailable for analysis). Educational achievement was measured as the mean score of performance at age 16 in the standardized high-school completion exams taken in the United Kingdom: the General Certificate of Secondary Education (GCSE).
# Methods
## Participants
Participants were drawn from the TEDS, an ongoing longitudinal study following MZ and DZ twins born in England and Wales between 1994 and 1996 [bib_ref] Twins Early Development Study (TEDS): A genetically sensitive investigation of cognitive and..., Haworth [/bib_ref]. The TEDS sample is representative of the UK populationand the project received approval from the Institute of Psychiatry Ethics Committee (05/Q0706/228). Twin zygosity was determined using a parental rating measure of similarity and DNA genotyping [bib_ref] Infant zygosity can be assigned by parental report questionnaire data, Price [/bib_ref]. The number of complete twin pairs for each zygosity class across the three measurement points were as follows (also see [fig_ref] Table 1: Descriptive statistics for SDQ subscales and educational achievement © 2016 The Authors [/fig_ref] , for full details): MZ male pairs: n = 720-1,166; MZ female pairs: n = 1,028-1,350; DZ male pairs: n = 670-1,196; DZ female pairs: n = 886-1,247; and DZ opposite sex pairs: n = 1,513-2,352. About 52% of those individuals who were assessed at age 4 (for behavior problems) were reassessed at age 16.
## Measures
Strength and Difficulties Questionnaire. The SDQ is a short but reliable instrument (25 items: Goodman, 2001; Stone, Otten, Engels, Vermulst, & Janssens, 2010) for measuring psychosocial problems in children [bib_ref] The Strengths and Difficulties Questionnaire: A research note, Goodman [/bib_ref]. The SDQ consists of five scales measuring anxiety, conduct problems, hyperactivity-inattention, peer problems, and prosocial behavior. Higher scores indicate greater difficulties (i.e. for anxiety, conduct problems, hyperactivity, peer problems) or strengths (i.e. for prosociality). In the current study, we used parent-rated scores for the SDQ subscales acquired when the child was 4 and 12 years old. Scores for prosociality, conduct problems, and hyperactivity were also acquired by parental rating when the individual was 16. Cronbach's alpha was low for conduct problems (all ages: a range = .52-.57) and for peer problems (age 4: a = .47), although in line with previously reported values. Cronbach's alpha was broadly acceptable for the rest of the SDQ measures (anxiety: a range = .60-.68; hyperactivity: a range = .71-.77; peers age 12: a range = .64; prosociality: a range = .67-.73).
Educational achievement. General Certificate of Secondary Educations are graded from A* (the highest grade) to G (the lowest pass grade). We coded these grades from 11 (A*) to 4 (G): this scoring scheme reflects the fact that sub-GCSE levels of attainment represent National Curriculum levels 1, 2, and 3, and a G at GCSE was equivalent to Level 4 attainment. We constructed our measure of educational achievement as the mean score for the three required subjects: English (either the English language grade, or the mean of the English language grade and the English literature grade where both exams were taken), Science (the mean of all Science GCSEs taken), and Mathematics. environmental effects. Three sources of variance are typically estimated using data from MZ and DZ twins: additive genetic (A), sharedenvironment (C), and nonshared-environment effects (E). A reflects the aggregate impact of those genetic effects that sum up to influence a phenotype. C reflects the action of environmental factors shared by twins that serve to make them more similar on a particular phenotype. E reflects the action of environmental factors unique to individuals within a twin pair that serve to make them differ from each other on a particular phenotype. The presence of genetic effects on a given phenotype is typically inferred if the correlations between MZ twins are larger than the correlations for DZ twins. The presence of shared-environment effects is inferred if the correlations for DZ twins are larger than half the magnitude of the correlations for the MZ twins. Finally, nonshared-environmental effects are inferred if correlations for the MZ twins are less than unity. As such, this variance component also contains measurement error.
These correlational analyses were extended using formal model-fitting of variance-covariance matrices to the twin data. This approach allows parameter estimates in univariate models to be formally tested for significance as well as allowing multivariate modelsthe core focus of the current studyto be analyzed. The validity of inferences from twin analyses, as with all methods, rest on certain assumptions. First, that MZ and DZ twins are equally correlated with regard to environments of etiological importance for the trait under study (i.e. the equalenvironments assumption); second, the absence of parental assortative mating for the trait under study. Violations of the former will serve to (spuriously) increase estimates of heritability, whereas violation of the latter will serve to (spuriously) increase estimates of shared-environment effects.
In the current study, longitudinal analyses were central to our tests: We sought to estimate the extent to which genetic and environmental effects underlying SDQ measures at ages 4, 12, and 16 (for conduct problems, hyperactivity, and prosociality only) were associated with educational achievement at age 16. To perform this analysis, we used the Cholesky decomposition. The Cholesky decomposition specifies as many factors as there are variables for each source of variance, with each subsequent factor having one fewer pathway than the preceding factor (see . In other words, for additive genetic effects (A), the first latent factor loads on all of the n measured variables: The subsequent latent factors load on nÀ1, nÀ2. . .nÀi variables. In this way, each factor accounts for as much of the remaining variance as possible, until the last factor accounts for just the residual variance in the last measured variable. This is repeated for the shared-environment factors (C) and nonshared-environmental factors (E). This design makes it possible to estimate the extent to which earlyemerging genetic and environmental influences on an SDQ trait predict later educational achievement. Moreover, this design allows us to examine whether innovative genetic and environmental factorsthat is sources of variance independent of earlier genetic and environmental influencescontribute additional genetic and environmental prediction to educational achievement at age 16. Twin models were fitted using full-information maximum-likelihood in OpenMx 2.0 [bib_ref] OpenMx: An open source extended structural equation modeling framework, Boker [/bib_ref] running within R 3.2 (R Development Core Team, 2015).
# Results
Descriptive statistics for all of the study variables are detailed in full in [fig_ref] Table 1: Descriptive statistics for SDQ subscales and educational achievement © 2016 The Authors [/fig_ref]. Assumption testing using all twins indicated that means and variances could be equated across twin order, zygosity, and sex for most variables, with the small number of significant differences observed consistent with the large number of tests performed. Of note, however, was evidence for modest-to-moderate mean sex differences, particularly for hyperactivity and prosociality. Sex-limitation modeling (testing for quantitative and qualitative genetic and environmental differences across sex) largely indicated that genetic and environmental influences could be equated across sex, with the significant differences that were observed mostly being either small in magnitude or, again, consistent with the large number of tests performed. Following these observations, we pooled our sample across sex, but used sex-residualized variables for all twin analyses.
## Phenotypic analyses
We first examined whether age 4 SDQ traits predicted educational achievement at age 16. Correlational analyses showed significant negative links with anxiety (r = À.06, p < .001), conduct problems Graphical representation of the longitudinal Cholesky decomposition for Strengths and Difficulties Questionnaire (SDQ) traits and educational achievement. A = additive genetic influences; shared-and nonshared-environmental influences were also modeled, and took the same form as the A pathways (i.e. C 4 , C 12 , C 16 , E 4 , E 12 , and E 16 ), but are omitted here in the interests of visual clarity; SDQ-16 was only available for conduct, hyperactivity, and prosociality (r = À.19, p < .001), hyperactivity (r = À.23, p <.001), and peer problems (r = À.09, p < .001). No association between age 4 prosociality and educational achievement was observed (r = .02, p > .05). These associations remained significant when controlling for parental socioeconomic status (indexed by parental education level, occupation, and family income) and sex. The full set of intercorrelations is presented in . We next sought to test whether behavior problems at age 12 and 16 added to this prediction. As such we built a series of phenotypic Cholesky decomposition models (see [fig_ref] Figure 2: Phenotypic Cholesky decomposition modeling results for Strengths and Difficulties Questionnaire [/fig_ref] , which followed the same logic as detailed above for the twin analyses. These analyses indicated that all SDQ traits at age 4with the exception of prosocialitywere significant predictors of age 16 educational achievement (in line with the correlational analyses reported above). Of importance, we also observed that in all cases, SDQ traits provided incremental prediction at subsequent ages. The size of these effects ranged from moderate (hyperactivity age12 ? educational achievement b = À.29) to modest (prosociality age12 ? educational achievement b = .06; see [fig_ref] Figure 2: Phenotypic Cholesky decomposition modeling results for Strengths and Difficulties Questionnaire [/fig_ref]. Controlling for parental socioeconomic status and sex did not lead to any notable changes in the magnitude or significance of path estimates.
## Twin analyses
Educational achievement showed significant genetic (A = .55, p < .001), shared-environment (C = .35, p < .001), and nonshared-environment (E = .11, p < .001) effects (also see [bib_ref] Strong genetic influence on a UK nationwide test of educational achievement at..., Shakeshaft [/bib_ref]. The univariate twin analyses for the behavioral problems variables have been reported in other published work [bib_ref] Identical genetic influences underpin behavior problems in adolescence and basic traits of..., Lewis [/bib_ref] [bib_ref] Strong genetic influence on a UK nationwide test of educational achievement at..., Shakeshaft [/bib_ref] and so are not detailed in full here (but see . In brief, though, SDQ traits were all underpinned by moderate-to-large genetic and nonshared-environmental influences, with modest shared-environmental influences evident for SDQconduct problems and SDQ-prosociality.
We next turned to tests of genetic and environmental influences underpinning the phenotypic associations between SDQ traits and educational achievement. We built a series of Cholesky models with SDQ traits at age 4, age 12, and age 16 (conduct problems, hyperactivity, and prosociality only), and educational achievement score entered in chronological order from left to right (see or [fig_ref] Figure 3: Longitudinal additive genetic modeling results for Strengths and Difficulties Questionnaire [/fig_ref]. To test whether SDQ traits were genetically and environmentally associated with educational achievement, we examined each of the genetic paths shared between educational achievement and SDQ age 4, age 12, and age 16, respectively. These parameters correspond to A 4 , A 12 , and A 16 to educational achievement in . For conduct problems, these genetic paths were significant at all ages: conduct A4 -educational achievement: Dv 2 (1) = 17.88, p < .001; conduct A12 -educational achievement: Dv 2 (1) = 6.03, p = .01; conduct A16 -educational achievement: Dv 2 (1) = 20.67, p = <.001. Similar results were observed for hyperactivity: hyperactivity A4 -educational achievement: Dv 2 (1) = 98.85, p < .001; hyperactivity A12 -educational achievement: Dv 2 (1) = 27.22, p < .001; hyperactivity A16 -educational achievement: Dv 2 (1) = 22.59, p = <.001. No genetic associations were observed between the other three SDQ traits (at any age) and educational achievement: all Dv 2 (1) < 1.21, all p > .27.
Shared-and nonshared-environmental influences were examined using the same principles detailed above. For anxiety, conduct problems, and peer problems, shared-environmental influences were significantly associated with educational achievement, and these influences were all reflective of early-emerging shared-environment influences with broadly stable effects thereafter: anxiety C4 -educational achievement: Dv 2 (1) = 7.81, p = .005; conduct C4 -educational achievement: Dv 2 (1) = 14.56, p < .001; peer problems C4 -educational achievement: Dv 2 (1) = 19.91, p < .001. No further shared-environmental associations were observed between SDQ traits (at any age) and educational achievement. Nonshared-environmental influences were more nuanced. For conduct problems, these paths were significant at ages 12 (Dv 2 (1) = 4.80, p = .03) and 16 (Dv 2 (1) = 7.44, p = .006); and for hyperactivity, at ages 4 (Dv 2 (1) = 10.04, p = .002), 12 (Dv 2 (1) = 83.08, p < .001), and 16 (Dv 2 (1) = 95.30, p < .001). For prosociality, these overlapping influences were only apparent cross-sectionally at age 16 (Dv 2 (1) = 7.07, p = .008).
Finally, we examined the magnitude of the overlap between genetic and environmental influences on SDQ traits and educational achievement. Genetic influences on conduct problems that were present by age 4 accounted for 3.1% of the genetic effects underpinning educational achievement at age 16. Genetic influences on conduct problems that were present by ages 12 and 16 accounted for a further 1.5% and 5.3%, respectively. Genetic influences on hyperactivity that were present by age 4 accounted for 16% of the genetic effects underpinning educational achievement at age 16. Genetic influences on hyperactivity that were present by ages 12 and 16 accounted for a further 6.2% and 4.9%, respectively. Shared-environmental influences on peer problems fully overlapped (100%) with the shared-environment influences on educational achievement, and these overlapping influences were present from age 4. Shared-environmental influences on anxiety accounted for 25.7% of the shared-environment influences on educational achievement, and these overlapping influences were present from age 4. Nonsharedenvironmental influences on hyperactivity that were present by age 4 accounted for 1% of the nonshared-environmental effects underpinning educational achievement at age 16. Nonsharedenvironmental influences on hyperactivity that were present by ages 12 and 16 accounted for a further 5.8% and 5.8%, respectively. Conduct problems and prosociality at age 16 each showed nonsharedenvironmental influences that overlapped with educational achievement: <1% in both cases. Full model parameter estimates for additive genetic, shared-, and nonshared-environmental influences are detailed in [fig_ref] Figure 3: Longitudinal additive genetic modeling results for Strengths and Difficulties Questionnaire [/fig_ref].
## Subsidiary analyses
The above analyses independently addressed each of the behavioral problems and their respective links to educational achievement. Our results indicated that conduct problems and hyperactivity are both genetically linked with educational achievement. This observation gives rise to the question of whether the genetic contribution from conduct problems to educational achievement is specific to conduct problems, or overlaps with hyperactivity. Similarly, conduct problems, anxiety, and peer problems all showed shared-environment links with educational achievement. As such, is the shared-environmental contribution from conduct problems to educational achievement specific to conduct problems, or does it overlap with that of anxiety and peer problems? The Cholesky decomposition is ill-suited to address this issue as the general factor also necessarily includes specific variance to whichever variable is included first in the model. As such we used the independent pathways model (see . This model specifies both a general factor and specific factors for genetic, shared-, and nonshared-environment effects. Accordingly, if this model shows a good fit to the data (relative to the baseline Cholesky), it provides evidence that genetic and environmental covariance between the measured variables can be accounted for by the general factor. For these analyses, we focused specifically on age 4 behavioral problems as not all measures were available at age 16.
We first used this model to examine whether the genetic influences underpinning conduct problems and hyperactivity provide distinct or common genetic prediction for educational achievement. The independent pathways model provided a good fit to the data and was not appreciably different [bib_ref] Multimodel inference understanding AIC and BIC in model selection, Burnham [/bib_ref] to the Cholesky decomposition (AIC Cholesky = 15333.01 vs. AIC IPMod = 15335.40). We detail the parameter estimates of the independent pathways model in . As such, this analysis indicates that while conduct problems and hyperactivity (at age 4) are both genetic predictors of age 16 educational achievement, this prediction reflects a common etiology.
We next examined whether the shared-environment links between behavioral problemsspecifically, conduct problems, anxiety, and peer problems and educational achievement reflected processes specific to each behavioral problem or a more general etiology. The independent pathways model fitted substantially less well (AIC Cholesky = 21546.35 vs. AIC IP = 21561.16); however, this result may simply reflect the omission of a specific genetic effect common to conduct problems and educational achievement; that is the general genetic factor forces any link between age 4 conduct problems and educational achievement to also explain genetic influences on peer problems and anxiety. Indeed, a modified independent pathway model including this parameter provided a more parsimonious fit to the data than the Cholesky (AIC Cholesky = 21546.35 vs. AIC IPMod = 21539.22: see [fig_ref] Figure 2: Phenotypic Cholesky decomposition modeling results for Strengths and Difficulties Questionnaire [/fig_ref]. In aggregate, this set of analyses indicates that while conduct problems, anxiety, and peer problems (at age 4) are all shared-environment predictors of age 16 educational achievement, these environmental factors reflect generalized rather than specific sources of prediction.
# Discussion
The current study examined the association between child and adolescent behavior problems and educational achievement at age 16. At the phenotypic level, anxiety, conduct problems, hyperactivity, and peer problems (as rated by parents) at age 4 all predicted lower levels of educational achievement at age 16, although the magnitude of these predictions was modest for anxiety and peer problems. For each of these variables, incremental prediction for educational achievement was observed at the subsequent measurement points. Prosociality positively predicted educational achievement from age 12, with Panel E incremental prediction at age 16. These results support previous findings reporting early-childhood links from externalizing to school success [bib_ref] Early conduct problems and later life opportunities, Fergusson [/bib_ref] [bib_ref] Mental health problems and educational attainment in adolescence: 9-year follow-up of the..., Veldman [/bib_ref] , as well as help to clarify the role of internalizing behaviors and prosociality on educational achievement in light of mixed findings in the literature [bib_ref] Prosocial foundations of children's academic achievement, Caprara [/bib_ref] [bib_ref] Longitudinal effects of preschool behavioral styles on early adolescent school performance, Dilalla [/bib_ref] [bib_ref] Trajectories of prosocial behavior and physical aggression in middle childhood: Links to..., Kokko [/bib_ref]. The etiology (i.e. genetic and environmental underpinnings) of the association for the links between early-childhood behavior problems and later educational achievement was largely specific to each of the behavior problems. The link between earlychildhood conduct problems and later educational achievement was explained by genetic and sharedenvironmental factors. The link between earlychildhood hyperactivity and later educational achievement was also explained in part by genetic factors, but here nonshared-environmental factors accounted for the remainder of the association. This result is notable in light of the rarity of nonsharedenvironmental stability over time (e.g. [bib_ref] Do nonshared environmental influences persist over time? An examination of days and..., Burt [/bib_ref] [bib_ref] Nonshared environment: A theoretical, methodological, and quantitative review, Turkheimer [/bib_ref] ; although see [bib_ref] Does the environment have an enduring effect on ADHD? A longitudinal study..., Livingstone [/bib_ref]. The link between earlychildhood anxiety and peer problems and later educational achievement was explained by sharedenvironmental factors. Of note, the genetic influences linking conduct problems and hyperactivity with educational achievement were themselves substantially overlapping. Similarly, the shared-environmental contribution from conduct problems to educational achievement overlapped substantially with that of anxiety and peer problems. These findings, particularly the observations for the externalizing problems, are consistent with the notion that early-emerging behavior problems reflect long-standing challenges to life outcomes [bib_ref] Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy, Moffitt [/bib_ref] , here exemplified by the important life variable of educational achievement.
For both conduct problems and hyperactivity, we also found evidence for genetic and nonsharedenvironment innovations that emerged at ages 12 and 16 and provided incremental prediction for educational achievement. These results are consistent (particularly in the context of genetic innovation) with the notion of adolescence as a sensitive period of socioaffective development with implications in turn for educational outcomes [bib_ref] The developing social brain: Implications for education, Blakemore [/bib_ref]. In contrast, the associations from anxiety and peer problems to educational achievement did not show genetic or environmental innovations, instead being wholly accounted for by early-emerging sharedenvironment effects.
These findings raise some intriguing questions. First, what processes might explain the sharedenvironmental influences acting on age 4 conduct problems, anxiety, and peer problems, which in turn impact on later educational achievement? Some possible factors might include parental warmth/ support [bib_ref] On the links between attachment style, parental rearing behaviors, and internalizing and..., Roelofs [/bib_ref] , parental control [bib_ref] Parental psychological control: Revisiting a neglected construct, Barber [/bib_ref] , or family chaos [bib_ref] Chaotic homes and school achievement: A twin study, Hanscombe [/bib_ref]. Low levels of parental concern for the welfare and outcomes of the child might plausibly manifest in conduct problems, anxiety, and peer problems, and in turn impact educational outcomes, either directlyfor example through limited shared book reading or interactionor indirectlyfor example as a consequence of behavior problems impairing learning opportunities. Broader shared experiences beyond the homesuch as preschool quality or neighborhood-level deprivation [bib_ref] Neighborhood deprivation affects children's mental health: Environmental risks identified in a genetic..., Caspi [/bib_ref] [bib_ref] Area deprivation and child psychosocial problems, Reijneveld [/bib_ref] might similarly explain this pattern of sharedenvironmental effects.
Second, what processes might explain the genetic influences common between externalizing problems (i.e. conduct problems and hyperactivity) and educational achievement? One possibility is individual differences in executive functioning, particularly in the context of emotion and impulse management. Such mechanisms are likely to have direct effects on the expression of externalizing behaviors [bib_ref] Behavioral inhibition, sustained attention, and executive functions: Constructing a unifying theory of..., Barkley [/bib_ref] [bib_ref] Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy, Moffitt [/bib_ref] [bib_ref] A meta-analytic review of the relation between antisocial behavior and neuropsychological measures..., Morgan [/bib_ref] and may also impair educational development through failures to persevere when the workload becomes difficult, or indirectly as a consequence of exclusion from class activities due to poor behavior.
A number of limitations require discussion. First, the classical twin design is subject to a number of assumptions, such as the equal-environments assumption . Future studies that can capitalize on additional family structures in order to provide more assumption-free estimates would be valuable, although it is noteworthy that research testing whether violations of the equalenvironments assumption are apparent for psychopathology has found little evidence for this potential source of bias [bib_ref] A test of the equal-environment assumption in twin studies of psychiatric illness, Kendler [/bib_ref]. Second, with the current study design, we cannot draw inferences concerning the genetic and environmental mechanisms underpinning the observed links between early-childhood behavioral problems and later educational achievement. For instance, these traits may exert their influence on subsequent school success through the initiation of a deleterious developmental cascade (i.e. bad school behavior early on leads to poor skill development, with the associated knock-on effects for subsequent intellectual development) or because of stable influences that act more proximately. Third, although Cronbach's alpha was consistent with previous work (e.g., these values fell below conventional standards for conduct problems (all ages) and for peer problems (age 4). It is noteworthy, however, that some debate exists over whether modest Cronbach's alpha values signal need for concern. If one uses a broad content coverage and quickly administrable instrument with just a few items per scale, as is the case with the SDQ instrument, one should expect, and perhaps even desire, such 'modest' alphas (e.g. [bib_ref] Does item homogeneity indicate internal consistency or item redundancy in psychometric scales?, Boyle [/bib_ref].
In summary, in the current study, we have shown that genetic, shared-environmental, and (to a lesser extent) nonshared-environmental influences on behavior problems in early childhood are predictive of educational achievement in major public examinations at age 16, consistent with work emphasizing life-course persistence of behavior problems and the concomitant negative life outcomes. Of importance, we also observed that new genetic and nonshared-environmental influencesthat is, genetic influences on conduct problems and hyperactivity emerging during childhood and adolescent developmentwere also predictive of educational achievement at 16, consistent with the notion that adolescence represents a sensitive period for socioaffective development.
## Supporting information
Additional Supporting Information may be found in the online version of this article: . Independent pathway model for conduct problems (age 4), hyperactivity (age 4), and educational achievement (age 16). [fig_ref] Figure 2: Phenotypic Cholesky decomposition modeling results for Strengths and Difficulties Questionnaire [/fig_ref]. Modified independent pathway model for peer problems (age 4), anxiety (age 4), conduct problems (age 4), and educational achievement (age 16). [fig_ref] Table 1: Descriptive statistics for SDQ subscales and educational achievement © 2016 The Authors [/fig_ref]. Number of complete twin pairs across zygosity and sex for all study measures. . Phenotypic correlations between SDQ variables and educational achievement. . Cross-sectional twin analysis results for SDQ variables. . Correlations between SDQ variables and educational achievement for MZ twin pairs. . Correlations between SDQ variables and educational achievement for DZ twin pairs.
- Previous work shows childhood behavior problems predict subsequent educational achievement. - However, it is unclear whether these effects represent genetic or environmental factors. - To address this issue, we used a longitudinal classic twin design: behavior problems were assessed at ages 4, 12, and 16; educational achievement was assessed at age 16.
- Shared-environmental influences on anxiety, conduct problems, and peer problems at age 4 predicted educational achievement at age 16.
- Genetic influences on conduct problems and hyperactivity at age 4 predicted educational achievement at age 16.
- Novel genetic and nonshared-environmental influences acting on conduct problems and hyperactivity emerged at ages 12 and 16, adding to the genetic prediction from age 4.
[fig] Figure 2: Phenotypic Cholesky decomposition modeling results for Strengths and Difficulties Questionnaire (SDQ) traits and educational achievement. Bolded lines = p < .05; P = phenotypic effects; 4/12/16 = age 4/12/16; values are standardized path loadings. (A) Details the model for conduct problems and educational achievement; (B) details the model for hyperactivity and educational achievement; (C) details the model for prosociality and educational achievement; (D) details the model for anxiety and educational achievement; and (E) details the model for peer problems and educational achievement © 2016 The Authors. Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for Child and Adolescent Mental Health. [/fig]
[fig] Figure 3: Longitudinal additive genetic modeling results for Strengths and Difficulties Questionnaire (SDQ) traits and educational achievement. Bolded lines = p < .05; A = additive genetic effects; 4/12/16 = age 4/12/16; values are standardized path loadings. (A) Details the model for conduct problems and educational achievement; (B) details the model for hyperactivity and educational achievement; (C) details the model for prosociality and educational achievement; (D) details the model for anxiety and educational achievement; and (E) details the model for peer problems and educational achievement [/fig]
[fig] Figure 4: Longitudinal shared-environment modeling results for Strengths and Difficulties Questionnaire (SDQ) traits and educational achievement. Bolded lines = p < .05; C = shared-environment effects; 4/12/16 = age 4/12/16; values are standardized path loadings. (A) Details the model for conduct problems and educational achievement; (B) details the model for hyperactivity and educational achievement; (C) details the model for prosociality and educational achievement; (D) details the model for anxiety and educational achievement; and (E) details the model for peer problems and educational achievement © 2016 The Authors. Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for Child and Adolescent Mental Health. [/fig]
[fig] Figure 5: Longitudinal nonshared-environment modeling results for Strengths and Difficulties Questionnaire (SDQ) traits and educational achievement. Bolded lines = p < .05; E = nonshared-environment effects; 4/12/16 = age 4/12/16; values are standardized path loadings. (A) Details the model for conduct problems and educational achievement; (B) details the model for hyperactivity and educational achievement; (C) details the model for prosociality and educational achievement; (D) details the model for anxiety and educational achievement; and (E) details the model for peer problems and educational achievement [/fig]
[table] Table 1: Descriptive statistics for SDQ subscales and educational achievement © 2016 The Authors. Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for Child and Adolescent Mental Health. [/table]
|
The Accumulation and Transformation of Heavy Metals in Sediments of Liujiang River Basin in Southern China and Their Threatening on Water Security
# Introduction
Rivers are vital channels of water resource regulation and storage; however, with the development of industry and agriculture, they have gradually become an important channel for sewage discharging. Based on the previous investigations, the contamination of HMs with various degrees could be commonly found in rivers all over the world, indicating a severe situation of HMs pollution in rivers [bib_ref] Assessment of the effects of urban and industrial development on water and..., Gashi [/bib_ref] [bib_ref] Assessment of heavy metals bioavailability and toxicity toward Vibrio fischeri in sediment..., Rosado [/bib_ref] [bib_ref] Comprehensive risk assessment of heavy metals in lake sediment from public parks..., Yang [/bib_ref] [bib_ref] Heavy metals in surface sediments of the Jialu River, China: Their relations..., Fu [/bib_ref]. Therefore, the strict control of sewage discharging has been proposed to protect the ecological environment of rivers. The improvement of water quality in rivers actually confirmed the positive influences of wastewater regulation on degrading the environmental threat of HMs, but the HM pollution among aquatic biotas in rivers still suggested that the threat of HMs in rivers cannot be eliminated by only relying on the regulation of wastewater discharge [bib_ref] Analysis and health risk assessment of toxic and essential elements of the..., Miao [/bib_ref] [bib_ref] Health risk assessment of heavy metals in freshwater fish in the central..., Zhong [/bib_ref] [bib_ref] Metal concentrations in water, sediment and gudgeon (Gobio gobio) from a pollution..., Bervoets [/bib_ref]. In fact, a previous study indicated that the bio-enrichment of HMs among aquatic biotas is
# Materials and methods
## The description of study area and field sampling
The Liujiang River Basin is located in Guangxi Province, China with an area of 58,398 km 2 . The Liujiang River and Luoqingjiang River are the most important surface runoffs in this watershed. Liujiang River is the largest river in Liuzhou City and the second largest tributary of the Xijiang River in the Pearl River Basin with a total length of 272 km, which stems from the confluence of Longjiang and Duliujiang River in Fengshan Town, then flows through the main industrial areas, residential areas, and the city center of Liuzhou City. Therefore, the Liujiang River becomes the main waterway of Liuzhou City, in which HMs have been frequently discharged. The upper reaches of the Liujiang River include Duliujiang River, Xunjiang River, Rongjiang River and Longjiang River. Along these rivers, the mining industries are widely distributed, hence, the concentrations of HMs in these rivers are generally high. Luoqingjiang River, the tributary of Liujiang River, originates from Lingui County of Guilin City. Though the watershed of Luoqingjiang River includes mainly the rural areas without significant industrial operation, some industrial parks and part industrial areas in east suburban areas of Liuzhou City also belong to its watershed [bib_ref] Analysis and health risk assessment of toxic and essential elements of the..., Miao [/bib_ref]. The wide distribution of carbonate and karst landforms breed the intact karstic groundwater system in the Liuzhou River Basin, so that the continuous recharge of alkaline groundwater from karstic groundwater system to Liujiang River and Luoqingjiang River comes out, which also keeps the water constantly alkaline in these rivers.
The sample collection was conducted during May 12th to 18th, 2019, one week after two-week continuous rainfall to avoid the possible weakening of HMs conversion. A total of 24 sampling sites were selected along the Liujiang River and its tributary Luoqingjiang River and a total of 24 surface water and sediment samples were obtained [fig_ref] Figure 1: Sampling sites along the Liujiang River [/fig_ref]. The sediment samples were collected by using the grab sampler. The overlying water was simultaneously collected to detect the water chemistry (i.e., pH, redox potential (Eh), dissolved oxygen (DO), electrical conductivity (EC), total dissolved solids (TDS), and turbidity) and the content of dissolved HMs. A total of 5 mL nitric acid was added in water samples after being stored in 500 mL polypropylene plastic bottles, while the sediment samples were preserved in polythene self-sealing bags. All of these samples were transported at −20 - C until further processing.
## Sample preparation and analysis
The water samples were directly filtered via 0.22 µm syringe-driven filters for further testing.
For sediments, the sediment was freeze-dried to a constant weight at −80°C for 72 h, then sieved through a 0.15 mm nylon mesh. A total of 0.2 g of the sediment samples was digested by a solution of HNO 3 + HCl + HF (5:4:1 v/v) in a Teflon digestor at 140°C for 6 h. The residue obtained was diluted to make up the volume, then the samples were filtered with 0.22 µm syringe-driven filters for total contents of target HMs testing.
The distribution of HMs in different chemical forms (exchangeable (Fr1), carbonatebound (Fr2), reducible (Fr3), oxidizable (Fr4), and residual fractions (Fr5)) were obtained in accordance with the sequential extraction procedure (SEP), which is described in detail in a previous study [bib_ref] Determination and partitioning of metals in sediments along the Suez Canal by..., El-Azim [/bib_ref]. To obtain the percent recovery of the HMs using the adopted SEP, the total HMs concentrations were compared with the total concentrations of the HMs in the five SEP-derived fractions.
Cd, Cr, Cu, Pb, and Zn were analyzed by the inductively coupled plasma mass spectrometry method (ICP-MS, Thermo X series), while Hg and As were measured with the atomic fluorescence spectrometry method (AFS-920). Organic matter was tested by elemental analyzer (Vario EL-III, Elementar, German ), sediment particle size was tested by laser particle size analyzer (Mastersizer 2000, Malvern Instruments, UK).
## Sample preparation and analysis
The water samples were directly filtered via 0.22 μm syringe-driven filters for further testing.
For sediments, the sediment was freeze-dried to a constant weight at −80 ℃ for 72 h, then sieved through a 0.15 mm nylon mesh. A total of 0.2 g of the sediment samples was digested by a solution of HNO + HCl + HF (5:4:1 v/v) in a Teflon digestor at 140 ℃ for 6 h. The residue obtained was diluted to make up the volume, then the samples were filtered with 0.22 μm syringe-driven filters for total contents of target HMs testing.
The distribution of HMs in different chemical forms (exchangeable (Fr1), carbonatebound (Fr2), reducible (Fr3), oxidizable (Fr4), and residual fractions (Fr5)) were obtained in accordance with the sequential extraction procedure (SEP), which is described in detail in a previous study [bib_ref] Determination and partitioning of metals in sediments along the Suez Canal by..., El-Azim [/bib_ref]. To obtain the percent recovery of the HMs using the adopted SEP, the total HMs concentrations were compared with the total concentrations of the HMs in the five SEP-derived fractions.
Cd, Cr, Cu, Pb, and Zn were analyzed by the inductively coupled plasma mass spectrometry method (ICP-MS, Thermo X series), while Hg and As were measured with the atomic fluorescence spectrometry method (AFS-920). Organic matter was tested by elemental analyzer (Vario EL-III, Elementar, German ), sediment particle size was tested by laser particle size analyzer (Mastersizer 2000, Malvern Instruments, UK).
## Quality assurance and quality control
Standard and blank samples were randomly inserted in the test process of each batch of samples. The involved standard substances are: GBW 07401, CRM, GSS-5, SRM2976, which were purchased from the Chinese academy of measurement science. All measurements are minus the average of blank samples as the final sample test values, and a parallel sample was set for each sample, with the average value as the final result. The standard material was used in the sequential extraction process to ensure the accuracy of the extraction results, and the sample recovery ratio was calculated after the test. High sample recovery between 95-105% was reported and used towards QA/QC compliance, and for the detailed analytical limits of detection and quantitation (LOD, LOQ) and the operating
## Quality assurance and quality control
Standard and blank samples were randomly inserted in the test process of each batch of samples. The involved standard substances are: GBW 07401, CRM, GSS-5, SRM2976, which were purchased from the Chinese academy of measurement science. All measurements are minus the average of blank samples as the final sample test values, and a parallel sample was set for each sample, with the average value as the final result. The standard material was used in the sequential extraction process to ensure the accuracy of the extraction results, and the sample recovery ratio was calculated after the test. High sample recovery between 95-105% was reported and used towards QA/QC compliance, and for the detailed analytical limits of detection and quantitation (LOD, LOQ) and the operating parameters of analytical instrumentation refer to [bib_ref] Analysis and health risk assessment of toxic and essential elements of the..., Miao [/bib_ref] [bib_ref] Spatial distribution of heavy metals and their potential sources in the soil..., Miao [/bib_ref]. For all samples, the total concentrations in the five fractions of HMs were compared with the total HM concentrations, and the average percent recovery was 93.68% ± 5.52%.
# Statistical analysis
The correlations between the data were analyzed by SPSS 20, and p < 0.05 indicated significant difference in the data. Figures and tables were finished with Origin Pro 8 (OriginLab, USA) and Coreldraw X4 (Corel, Canada).
## Ecological risk assessment method
The potential ecological risk index (RI) is widely used to assess the ecological risks of toxic substances and pollutants to biotas [bib_ref] Spatial distribution of heavy metals and their potential sources in the soil..., Miao [/bib_ref] [bib_ref] An ecological risk index for aquatic pollution control-a sedimentological approach, Hakanson [/bib_ref] [bib_ref] Distribution and contamination assessment of heavy metals in surface sediments of the..., Liu [/bib_ref]. The calculation formulas of RI are shown below: where: C i , C i n and T i r , respectively, represent the heavy metal i concentration in sediments (µg/g), the relevant background values of soil in Guangxi Province, and the biological toxic factor of each heavy metal. The soil background values C i n and toxicity factors T i r of HMs can be found in [fig_ref] Table 1: The correlations of HMs in surface water and hydrochemical characteristics [/fig_ref]. The ratio of the measured concentration of the heavy metal in sediment (C i ) and the corresponding background value (C i n ) represents the single element contamination factor (C i f ). E i r is the monomial ecological risk factor of an individual heavy metal, while the sum of E i r is the total ecological risk index (RI). The grade of HM pollution and ecological risk assessment used in this study can be found in [fig_ref] Table 2: The correlations between HMs in sediments and environmental factors [/fig_ref].
[formula] C i f = C i C i n(1)E i r = T i r × C i f(2)RI = n ∑ i=1 E i r(3) [/formula]
# Results and discussion
## The variations of water chemistry in the surface water
The variations of water chemistry are given in [fig_ref] Figure 2: The properties of water and sediment in the Liujiang River Basin [/fig_ref] , the relative standard deviations (RSD) of pH and DO are all below 10%, indicating the indistinctive variation of these indexes. The pH generally shows weakly alkaline (with an average of 7.81), which conforms to the characteristics of rivers in karst regions [bib_ref] Effects of heavy metals speciations in sediments on their bioaccumulation in wild..., Miao [/bib_ref]. The faintly acid water was only found in upstream of Liujiang River (pH is 6.9 in S1), which suggested the perturbation of recharge from underground water in karst area. The water is considered to be oxygen-rich in the Liujiang River Basin with the ranges of DO from 6.22 to 8.43 mg/L. The relatively high RSD of Eh, EC, TDS, and turbidity suggest their larger spatial variations. The variation of Eh is 94.89-161.10 mV, while the variation of EC and TDS is 141.5-252.2 and 71.1-127.2, respectively. Turbidity is an index of suspended particulate matter in water, which ranges from 8.13-28.10 NTU, which is higher downstream of the Liujiang River Basin (Refer to [bib_ref] The Variation of Heavy Metals Bioavailability in Sediments of Liujiang River Basin,..., Hao [/bib_ref]. should be blamed on the variation of particulate sources and the sedimentary environment. The average value of OM in surface sediments is 0.72% [fig_ref] Figure 2: The properties of water and sediment in the Liujiang River Basin [/fig_ref] , which is relatively low, but significant space variation of OM can be found in this study, which is suggested by high RSD of OM (35.42%).
## The distribution of hms in surface water and sediments
## The distribution of hms in surface water
The concentrations of HMs in surface water are given in [fig_ref] Figure 3: HMs concentrations in the surface water [/fig_ref] and [fig_ref] Table 3: The correlations of HMs between surface water and sediments [/fig_ref]. The contents of Hg in water samples are all below the detection limit, while the contents of other HMs decrease with the order: Zn > Cr > As > Cu > Pb > Cd > Hg. HMs are all within the primary standard of national surface water quality released by State Environmental Protection Administration of China (GB 3838-2002), suggesting that the concentra-
## The distribution of particle size in surface sediments
Sediments are mainly clayey silt [fig_ref] Figure 1: Sampling sites along the Liujiang River [/fig_ref] , among them, silt (4-63 µm) is considered to be the dominant component, the mean proportion of which reach to 74.05%. Clay (<4 µm)
is the secondary component of sediments, which holds the ratio of 22.16%. The ratio of sand (63-125 µm) was found to be significantly low in sediments, which is only 3.79%. The grain size of sediments is relatively finer midstream and downstream of Liujiang River (S9 and S10, S13 and S14) and Luoqingjiang River (S23 and S24), which should be blamed on the variation of particulate sources and the sedimentary environment. The average value of OM in surface sediments is 0.72% [fig_ref] Figure 2: The properties of water and sediment in the Liujiang River Basin [/fig_ref] , which is relatively low, but significant space variation of OM can be found in this study, which is suggested by high RSD of OM (35.42%).
## The distribution of hms in surface water and sediments
## The distribution of hms in surface water
The concentrations of HMs in surface water are given in [fig_ref] Figure 3: HMs concentrations in the surface water [/fig_ref] and [fig_ref] Table 3: The correlations of HMs between surface water and sediments [/fig_ref]. The contents of Hg in water samples are all below the detection limit, while the contents of other HMs decrease with the order: Zn > Cr > As > Cu > Pb > Cd > Hg. HMs are all within the primary standard of national surface water quality released by State Environmental Protection Administration of China (GB 3838-2002), suggesting that the concentrations of HMs in surface water are generally not high and the strictly controlled sewage into the river in recent years has already exerted a positive effect in watershed of the Liujiang River. For Liujiang River, the contents of HMs in water decrease from upstream to downstream, the higher HMs concentrations mainly centralize in upstream (S1-S4), which should be attributed to the ongoing operation of prosperous mining in Longjiang River and Duliujiang River in the upper reach of Liujiang River [bib_ref] Spatial Distribution, Sources and Bioavailability of Heavy Metals in the Surface Sediments..., Lan [/bib_ref] [bib_ref] Seasonal varieties and influential factors of heavy metals in sediments of Taihu..., Wang [/bib_ref]. The concentrations of HMs gradually decrease with the river flowing downstream, but the lowest contents of HMs are observed only in midstream (S5-S12), where the main urban areas are located in. The HM concentrations slightly increase downstream of the Liujiang River (S13-S17), which should be blamed on some pollution industries moving to suburban areas. In addition, the HM concentrations in Luoqingjiang River are overall lower than that in the Liujiang River, which approach those in the midstream of the Liujiang River and are also related to the lower distribution of pollution industries.
## The distribution of hms in sediments
The distribution of HM contents in sediments is shown in [fig_ref] Figure 4: The distribution of HMs in surface sediments [/fig_ref]. The concentrations of HMs are in the decreasing order as follows: Zn > Cr > Pb > Cu > As > Cd > Hg. Among them, most HMs were between TEL (Threshold Effect Level) and PEL (Probable Effect Level) , suggesting a low risk exerting negative effects on the benthos, while Cu and Pb are usually lower than TEL. For the soil background of Guangxi (Table
## The distribution of hms in sediments
The distribution of HM contents in sediments is shown in [fig_ref] Figure 4: The distribution of HMs in surface sediments [/fig_ref]. The concentrations of HMs are in the decreasing order as follows: Zn > Cr > Pb > Cu > As > Cd > Hg. Among them, most HMs were between TEL (Threshold Effect Level) and PEL (Probable Effect Level) , suggesting a low risk exerting negative effects on the benthos, while Cu and Pb are usually lower than TEL. For the soil background of Guangxi , most HMs were close to or lower than their corresponding soil baseline, but Zn and Cd were significantly higher, which suggested a potential interference from human activities. The distribution of HMs in sediments is found to be similar with that in water, most HMs are higher in upstream of Liujiang River (S1 and S2). In particular, the concentration of Cd at S1 is 5.3 times higher than the average value. This further indicates that there is still a large amount of Cd remaining upstream of the Liujiang River in spite of the notorious Cd pollution event passing 8 years [bib_ref] Spatial Distribution, Sources and Bioavailability of Heavy Metals in the Surface Sediments..., Lan [/bib_ref]. The concentrations of HMs are relatively low in midstream of the Liujiang river (S3∼S9), while the concentrations rebound in the downstream (S10∼S17). The contents of HMs are overall lower in Luoqingjiang River (S18∼S24) than those in Liujiang River.
## The speciation of hms in sediments
The speciation of HMs in sediments is shown in [fig_ref] Figure 5: The chemical forms of HMs in surface sediments [/fig_ref]. HMs are mostly in the residual fraction, indicating that they are derived from crustal materials and less likely involve the biogeochemical circle of HMs due to the inertia of the minerals containing HMs [bib_ref] Accumulation and transformation of heavy metals in surface sediments from the Yangtze..., Liu [/bib_ref]. The ratio of As, Cr, and Hg in the residual fraction is extremely high (over 80%), which strongly indicates a natural source, while the proportions of Cu and Zn in the residual fraction are less than 50%, which suggests a certain amount of anthropogenic input. For Pb and Cd, their residual fractions are not the dominant fraction, and their dominant fractions are, respectively, reducible and carbonate-bound fractions, which expresses an intense exogenous input from human activities.
For non-residual fractions, the proportions of exchangeable HMs are extremely low, indicating that the exchangeable fraction is not the main speciation of HMs and also should be attributed to the low contents of HMs in water. Zn and Cd are primarily in the carbonate-bound fraction, which suggests they are primarily weakly reserved in sediments. Pb, Cu, and Cd are relatively high in reducible fractions, implying Fe/Mn oxyhydroxides play a vital role in stabilizing Pb, Cu and Cd. Due to the shortage of organic matter in sediments, organic matter fails to effectively compete with Fe/Mn oxy-hydroxides, and then directly lowers their combination with HMs, so that the proportions of HMs in the oxidizable fraction were commonly lower than 15%.
## The speciation of hms in sediments
The speciation of HMs in sediments is shown in [fig_ref] Figure 5: The chemical forms of HMs in surface sediments [/fig_ref]. HMs are mostly in the residual fraction, indicating that they are derived from crustal materials and less likely involve the biogeochemical circle of HMs due to the inertia of the minerals containing HMs [bib_ref] Accumulation and transformation of heavy metals in surface sediments from the Yangtze..., Liu [/bib_ref]. The ratio of As, Cr, and Hg in the residual fraction is extremely high (over 80%), which strongly indicates a natural source, while the proportions of Cu and Zn in the residual fraction are less than 50%, which suggests a certain amount of anthropogenic input. For Pb and Cd, their residual fractions are not the dominant fraction, and their dominant fractions are, respectively, reducible and carbonate-bound fractions, which expresses an intense exogenous input from human activities.
For non-residual fractions, the proportions of exchangeable HMs are extremely low, indicating that the exchangeable fraction is not the main speciation of HMs and also should be attributed to the low contents of HMs in water. Zn and Cd are primarily in the carbonatebound fraction, which suggests they are primarily weakly reserved in sediments. Pb, Cu, and Cd are relatively high in reducible fractions, implying Fe/Mn oxy-hydroxides play a vital role in stabilizing Pb, Cu and Cd. Due to the shortage of organic matter in sediments, organic matter fails to effectively compete with Fe/Mn oxy-hydroxides, and then directly lowers their combination with HMs, so that the proportions of HMs in the oxidizable fraction were commonly lower than 15%.
should be attributed to the low contents of HMs in water. Zn and Cd are primarily in the carbonate-bound fraction, which suggests they are primarily weakly reserved in sediments. Pb, Cu, and Cd are relatively high in reducible fractions, implying Fe/Mn oxyhydroxides play a vital role in stabilizing Pb, Cu and Cd. Due to the shortage of organic matter in sediments, organic matter fails to effectively compete with Fe/Mn oxy-hydroxides, and then directly lowers their combination with HMs, so that the proportions of HMs in the oxidizable fraction were commonly lower than 15%. The significant correlations of HMs between the total contents and various forms are found in this study , indicating that the elevation of HM concentration would have a vital impact on the speciation of HMs. For Cr, Cu, As and Hg, the correlations between total contents and residual fraction are found to be more significant, suggesting that the elevation of their residual fraction would be a priority with their total contents increasing, so that their increasing in total contents would be hard to affect their stability and bioavailability in sediments. For Zn, Pb, and Cd, the correlations were found to be more significant with the non-residual fraction (respectively, carbonate-bound, reducible and oxidizable fraction), instead of the residual fraction, suggesting that their non-residual fraction will be increasing in priority with their total content elevation, in other words, the elevations of these HMs are dominated by exogenous input. Given the high mobility and toxicity of HMs in the non-residual fraction [bib_ref] Accumulation and transformation of heavy metals in surface sediments from the Yangtze..., Liu [/bib_ref] [bib_ref] Changes in the total content and speciation patterns of metals in the..., Chen [/bib_ref] , it is, thus, essential to be vigilant on the excessive input of exogenous Zn, Pb, and Cd in the Liujiang River Basin.
## Accumulation mechanism of hms in surface sediments of liujiang river basin
## The existing forms of hms in surface water
The results of correlation analysis between hydrochemistry and HMs in water were shown in [fig_ref] Table 1: The correlations of HMs in surface water and hydrochemical characteristics [/fig_ref]. The correlations of HMs are all found to be non-significant with pH, Eh, indicating that pH and Eh are trivial in elevating the solubility of HMs in river. On the contrary, the correlations between HMs and other hydorchemical factors, i.e., EC, TDS, turbidity and DO, are found to be significant, which expresses the existing forms of HMs in water. Zn and Cr are significantly correlated with EC and TDS, indicating that Zn and Cr exist as dominant ions in water, which is obviously not consistent with the fact that HMs are just trace ions in water, and the excessive existence of Zn and Cr in water, thus, should be considered to be the result of anthropogenic emissions. Due to the higher bioavailability of dissolved HMs, the excessive existence of exogenous Zn and Cr in water must elevate their bio-enrichment among aquatic biotas, which was well expressed in our previous studies that Zn and Cr in wild fish are mostly beyond maximal residual limited [bib_ref] Analysis and health risk assessment of toxic and essential elements of the..., Miao [/bib_ref]. The significant correlation of Cu is found to be positive with turbidity, indicating that Cu is mainly in the form of particles, so that exogenous Cu may be discharged mainly in granular form. There is a significant positive correlation between Cd and DO, which indicates the dissolubility of Cd can be elevated with DO increasing. Despite the fact that exogenous emission of Pb is confirmed in this study, the correlations between Pb and hydrochemical parameters are not found, which maybe suggest the limited emission of Pb. Arsenic is determined mainly from natural sources in this study, so that As in water may be close to the background value. The significant correlation of As was only found to be negative with turbidity, indicating the suspended particles have a dilution effect on As in water.
## Impact factors of the accumulation of hms in surface sediments
The composition of particles size and organic matter have a vital impact on the accumulation of HMs in sediments [bib_ref] Seasonal varieties and influential factors of heavy metals in sediments of Taihu..., Wang [/bib_ref] [bib_ref] Influence of pH and OM concentration on Cu, Zn, Cd, and Al..., Gundersen [/bib_ref]. As shown in [fig_ref] Table 2: The correlations between HMs in sediments and environmental factors [/fig_ref] , the significant correlations between HMs and particle size are found to be negative. Fine particles have a strong binding capacity to HMs, which is not only related to their higher specific surface area, but also higher concentration of active components in fine particles, i.e., organic matter, carbonate, Fe/Mn oxides, etc., all of which can further facilitate the absorption of HMs in sediments [bib_ref] Spatial distribution of heavy metals and their potential sources in the soil..., Miao [/bib_ref]. The significant positive correlations between organic matter and HMs demonstrate that the increase of organic matter can significantly increase the concentrations of HMs in sediments, which is related to the combination of exogenous HMs and organic matter [bib_ref] Accumulation and transformation of heavy metals in surface sediments from the Yangtze..., Liu [/bib_ref]. However, the significant correlations of Cd, As, and Hg cannot be found with grain size and organic matter. For Cd, the effects of grain size and organic matter may be relieved by the large input of exogenous Cd. The effects of grain size and organic matter is non-significant on As and Hg, mainly due to their less exogenous input and relatively uniform distribution in sediments. The more significant correlations of HMs can be only found with Eh and turbidity rather than other hydrochemical factors, suggesting that Eh and suspended particles should be treated as the manipulators of HMs accumulation in rivers. It is well-known that higher Eh promotes the formation of Fe/Mn oxides in sediments, which will aggravate the aggregation of exogenous HMs [bib_ref] The Variation of Heavy Metals Bioavailability in Sediments of Liujiang River Basin,..., Hao [/bib_ref]. That explained the positive significant correlation between Zn, Pb, Cd and Eh. For turbidity, the significant correlation is found to be negative with Zn, Pb and Cu, indicating that these HMs eventually enter the sediments under the mediation of suspended particulate matter, regardless of the dissolved or particulate form they discharged. Given that Zn, Pb, Cd and Cu are all higher in reducible phase, we, thus, have a reason to believe that the regulation of hydrochemistry may be greater on HMs with higher content in the reducible phase. Therefore, the ecological threat resulting from higher concentration of HMs in the reducible phase should not be ignored. However, for Cr, there are no significant correlations with turbidity and Eh, indicating that exogenous Cr is not easy to be absorbed by suspended particles, which results in the limited input of exogenous Cr in sediments and the high proportion of Cr in the residual phase. Therefore, exogenous Cr may exist mainly in the water body and less in the sediments. pH fluctuation is an important factor that activates the HMs in sediments [bib_ref] Influence of pH and OM concentration on Cu, Zn, Cd, and Al..., Gundersen [/bib_ref] [bib_ref] Relationships Between Radionuclide Content and Textural Properties in Irish Sea Intertidal Sediments, Clifton [/bib_ref]. However, except for Cd, the significant correlations of pH cannot be found with HMs in this study, indicating that pH maybe a trivial factor in HM accumulation in river, while the significant correlation of Cd is negative with pH, which is obviously inconsistent with the traditional understanding that alkaline condition promotes the aggregation of HMs, but resulting from sewage discharging. In fact, the industries in this study area are dominated by electroplating, steel production and machinery manufacturing [bib_ref] Analysis and health risk assessment of toxic and essential elements of the..., Miao [/bib_ref] , the sewage from these industries is mainly acidic and the concentration of Cd in sewage could be superior. Therefore, the large emission of the sewage will elevate the input of exogenous HMs in sediment, while the large emission of the sewage will reduce the pH of water at the same time. The positive significant correlation between EC and Cd also confirms the input of exogenous Cd in sediments, the aggregation of Cd in sediment is found to be elevated with ions increasing, instead of diluting. This explains that why the decrease of pH does not destabilize the aggregation of Cd in sediments, but significantly increases the concentration of Cd in sediments. Therefore, it is reasonable to believe that Cd in sediments is dominated by exogenous input, so that the influence of pH fluctuation on the accumulation of Cd in sediments was reversed by sewage discharging. As and Hg are mainly from natural sources, with a high proportion of residual fraction, so that they are insensitive to changes in water conditions.
## Controlling factors on hm speciation in surface sediments
In order to further understand the specific mechanism of HM accumulation in sediments, the correlations analysis between different forms of HMs and various factors are shown in . In general, a large input of exogenous HMs can not only change the influences of hydrochemical fluctuations on the speciation of HMs in sediments, but also can reverse the influences of hydrochemical fluctuations on the speciation distribution of HMs in sediments [bib_ref] The Variation of Heavy Metals Bioavailability in Sediments of Liujiang River Basin,..., Hao [/bib_ref] , which is well-reflected in Cd. The increase of EC did not aggravate the competition of ions to reduce the content of Cd in the non-residual phase [bib_ref] A study of lead and cadmium speciation in some estuarine and coastal..., Chakraborty [/bib_ref] , but intensified the accumulation of Cd in all non-residual phases, which directly elevates the mobility and bioavailability of Cd. The significant degree of elevation disclosed by the correlation coefficients between EC and Cd in different forms, which decreased in the order: carbonate-bound fraction, reducible fraction, oxidizable fraction and exchangeable fraction, and this sequence is exactly consistent to their respective contents in different species of HMs , also suggesting the order of exogenous Cd that combined with active substances in sediments. Besides Cd, the negative correlations of pH were also found to be significant with Cu and Hg in the exchangeable phase and oxidizable phase, indicating the discharge of acidic wastewater can also increase the contents of these HMs in these forms, therefore, more strict controlling on wastewater discharging in Liujiang River Basin should be considered to be a critical approach that lowers the aggregation of Cd, Cu and Hg in non-residual forms.
According to the previous analysis in this study, the accumulation of Fe/Mn oxides is conducive to the aggregation of exogenous HMs in sediments [bib_ref] Changes in the total content and speciation patterns of metals in the..., Chen [/bib_ref] , which is also well-reflected in the speciation of HMs. The significant correlations between Eh and the speciation of most HMs indicate that the elevation of Eh can promote the accumulation of exogenous HMs in other non-residual fractions, but the higher significance indicates that the increase of Eh has a more important influence on the reducible phase and carbonatebound phase, which highlights the crucial role of Eh on the stability of HMs in sediments. There are significant positive correlations between the DO and Cr in carbonate-bound phase, reducible phase and Zn in oxidizable phase, indicating that the high DO content is beneficial to the accumulation of these forms. In fact, DO is the important index of photosynthesis [bib_ref] Determination of model and calculation parameters of oxygen production by photosynthesis of..., Li [/bib_ref] , the stronger photosynthesis will elevate the content of the dissolved oxygen and consume more carbon dioxide, eventually decreasing the acidity in river, which is conducive to the stabilization of HMs in the carbonate-bound phase. In addition, the high dissolved oxygen content is conducive to the accumulation of Fe/Mn oxides and promotes the aggregation of HMs in the reducible phase, while organic compounds that are synthesized by photosynthesis can contribute to the accumulation of HMs in the oxidizable phase [bib_ref] Effects of heavy metals speciations in sediments on their bioaccumulation in wild..., Miao [/bib_ref]. Therefore, although the correlation of dissolved oxygen is found to be non-significant with total concentrations of HMs, its impact on the accumulation of HMs in some forms exists. The correlations of turbidity are found to be negative with all forms of HMs, including residual fraction, indicating that the input of HMs mediated by suspended particles is not limited to the exogenous, but also the endogenous. For the physicochemical properties of sediments, coarse particles can significantly reduce the contents of various forms of HMs, while the impacts of organic matter are totally opposite. Regardless of HM forms, they are all preferred to aggregate in clay, instead of sand and silt.
## The conversion and accumulation of hms between sediments and water
The correlations of HMs between surface water and surface sediments are shown in [fig_ref] Table 3: The correlations of HMs between surface water and sediments [/fig_ref]. Most HMs are high in the residual fraction, therefore, the non-significant conversion of HMs between water and sediments should be blamed on the relatively limited exogenous input of most HMs, which is just confirmed by the fact that the significant correlation of most HMs in water is not found with those in sediments. However, the significant correlations of most HMs in some chemical forms are found to be correlated with HMs in water, suggesting some chemical forms play a vital role in regulating the conversion of HMs. The significant correlations between water and sediments are found to be Cd, As, Cu, Zn and Pb, but only the Cd is found to have the significant correlation in all non-residual fractions. The significant correlations of Cd between all non-residual fractions in sediment and surface water are found to be positive, which indicates that the large emission of wastewater elevated all non-residual fractions of Cd in sediments, however, the correlation coefficients of each non-residual fraction are relatively even, which suggested that the emission of Cd from wastewater was high enough to mitigate the preference of Cd combination with Fe/Mn oxide, organic matter and carbonate. Therefore, it is urgently needed to relieve the risk of Cd pollution in the Liujiang River Basin by reducing the emission of wastewater. Similar to Cd, the significant correlation of Zn between water and sediments is found to be positive with the reducible phase, indicating that the ions of Zn that discharged from wastewater will mainly aggravate the aggregation of Zn in the reducible phase with the impacts of the strong binding force between Fe/Mn oxides and Zn [bib_ref] Changes in the total content and speciation patterns of metals in the..., Chen [/bib_ref]. However, the correlations of Cu and Pb, which are different to Zn, are found to be negative in the oxidizable phase and in residual phase, respectively. For Cu, the significant correlation suggested that the oxidizable fraction in sediments should be treated as the main destination of exogenous Cu, therefore, the binding of exogenous Cu with organic matter and sulfide should be considered to be the main way to purify water. For Pb, the significant correlation expresses that the natural deposition of endogenous Pb has a significant dilution effect on the accumulation of exogenous Pb.
For Cr, the correlation of Cr in water was found to be non-significant with that in any forms and total content in sediment, which is obviously inconsistent with the existence of the exogenous Cr in water. Actually, the water is commonly found to be alkaline in this study, while the exogenous Cr was confirmed to be more likely to remain in water under alkaline conditions, that is the reason why the ratio of Cr in non-residual fraction (16%) fails to rise with the existence of the exogenous Cr in water. Despite the fact that the concentrations of Hg in water are all below detection limit, the extremely high ratio of Hg in residual phase similar to Cr also suggests that the accumulation of Hg is dominantly from natural deposition, while the aggregation of exogenous Hg in sediments from water may not be significant.
Although HMs in the non-residual fraction play a critical role in regulating the accumulation of exogenous HMs in the aquatic environment, HMs in the residual fraction can also affect the conversion of HMs between sediment and water, especially when microorganisms are involved. As was found to be the best instance in this study, which is dominant in the residual fraction, the correlation of As between water and sediments should have been nonsignificant, however, a significant correlation is observed. The significantly positive correlations of As in water with that in carbonate-bound and reducible phases suggest that carbonate and Fe/Mn oxides have an important role in the fixation of exogenous As. However, the significantly positive correlation of As in the residual phase indicates that the inert endogenous As can also impact water quality. In fact, the reductive dissolution of Fe/Mn oxide or minerals manipulated by microbes can even release the As contained in minerals [bib_ref] Stimulation of Fe(II) Oxidation, Biogenic Lepidocrocite Formation, and Arsenic Immobilization by Pseudogulbenkiania..., Xiu [/bib_ref] [bib_ref] Hot Topics and Trends in the Study of High Arsenic Groundwater, Jia [/bib_ref] , which indicates that As has a vital influence on water quality, regardless of that in non-residual or residual fractions.
## Potential ecological risk assessment of hms
The assessments of HM contamination in sediments are shown in [fig_ref] Figure 6: Risk assessments for the HMs in surface sediments [/fig_ref]. Except Cd, the CF values of HMs are almost all below 3, suggesting low-to-moderate pollution level. The mean CF of Hg, Cu and Cr are all below 1, expressing low pollution level, while the mean CF of Zn, Pb and As is within 1 to 3, indicating moderate pollution level. The CF of Cd is significantly higher and commonly beyond 3, but below 6, suggesting a considerable pollution level. The maximum E i r of Cd appears in the upper reach of Liujiang River (S1), reaching a very high risk, while S10 and S13∼S17 in the lower reach of Liujiang River present a high risk, and Luoqingjiang River shows a moderate risk overall. For Hg, the average of E i r is 56, presenting a moderate risk overall. The maximum value appears at S2 in the upper reach of Liujiang River, which reaches very high risk, while the rest shows low or moderate risk.
The average of RI reached 239, suggesting that the study area is generally with moderate risk. The RI values of S1 and S2 were high, which obviously exceed the limits of very high risk and considerable risk, respectively. The ecological risks of other sampling sites are close to or lower than the moderate risk. The RI of Luoqingjiang River is obviously lower, presenting low risk. In general, the potential ecological risk of HMs of surface sediments in the Liujiang River is higher than that in the Luoqingjiang River. The HM risk of the Liujiang River decreases gradually from upstream to downstream, but slightly rebounds downstream. Cd has a higher risk than other HMs. Therefore, Cd pollution of sediments in the Liujiang River Basin should be paid more attention.
# Conclusions
The obtained results suggested the concentrations of HMs in water are far below the primary standard of water quality, but in sediments, the contents of Cd and Zn are significantly higher than their corresponding baseline of soil. Cd and Pb are mainly in the carbonate-bound fraction and reducible fraction, respectively, while other HMs are mainly in residual fraction. The non-significant correlations suggested pH and Eh may be hard to influence HMs in water, while the significant correlations highlighted the regulations of Eh, organic matter and mean grain size on the accumulation of metals in sediments. The opposite correlations between EC, TDS, pH and Cd confirmed the emission of acid wastewater contributed to the accumulation of Cd in sediment. The conversion of metals between water and sediments was found to be significant only in the specific chemical form of Cd, As, Cu, Zn and Pb, suggesting the conversion of HMs in sediments should be
# Conclusions
The obtained results suggested the concentrations of HMs in water are far below the primary standard of water quality, but in sediments, the contents of Cd and Zn are significantly higher than their corresponding baseline of soil. Cd and Pb are mainly in the carbonate-bound fraction and reducible fraction, respectively, while other HMs are mainly in residual fraction. The non-significant correlations suggested pH and Eh may be hard to influence HMs in water, while the significant correlations highlighted the regulations of Eh, organic matter and mean grain size on the accumulation of metals in sediments. The opposite correlations between EC, TDS, pH and Cd confirmed the emission of acid wastewater contributed to the accumulation of Cd in sediment. The conversion of metals between water and sediments was found to be significant only in the specific chemical form of Cd, As, Cu, Zn and Pb, suggesting the conversion of HMs in sediments should be largely regulated by their specific chemical forms, so that cutting down the accumulation of HMs in specific forms should be considered as an effective approach to degrade their conversion. The very high risk disclosed by the higher values of E i r and RI are only found upstream, while the higher risk of Cd should be treated as a critical environmental threat.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph19031619/s1, [fig_ref] Table 1: The correlations of HMs in surface water and hydrochemical characteristics [/fig_ref] : Soil background values and toxicity factors of HMs; [fig_ref] Table 2: The correlations between HMs in sediments and environmental factors [/fig_ref] : Ecological risk assessment rating; [fig_ref] Figure 1: Sampling sites along the Liujiang River [/fig_ref] : The composition of grain size and mean particle size in surface sediments; [fig_ref] Table 3: The correlations of HMs between surface water and sediments [/fig_ref] : The concentrations of HMs (µg/L) in surface water and their respective thresholds in primary standard of environmental quality standards for surface water in China; : Sediment Quality Guidelines and baseline of HMs (mg/kg); : The concentrations of HMs (µg/L) in surface water and their respective thresholds in primary standard of environmental quality standards for surface water in China (SEPA, 2002); : The correlations between the total concentrations of HMs and their forms; : The speciations of heavy metal in surface sediments (%); : The correlations between chemical forms of HMs and environmental factors. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
# Data availability statement:
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.
[fig] Figure 1: Sampling sites along the Liujiang River. [/fig]
[fig] Figure 2: The properties of water and sediment in the Liujiang River Basin. [/fig]
[fig] of 17, Figure 3: Int. J. Environ. Res. Public Health 2022, 19, x FOR PEER REVIEW 7 HMs concentrations in the surface water. [/fig]
[fig] Figure 3: HMs concentrations in the surface water. [/fig]
[fig] Figure 4: The distribution of HMs in surface sediments. [/fig]
[fig] Figure 5: The chemical forms of HMs in surface sediments. [/fig]
[fig] Figure 6: Risk assessments for the HMs in surface sediments: (a) Box plots of contamination factors of the HMs; (b) E and RI of the HMs in different sampling positions. The dotted line and the solid line represent the grade of E And RI, respectively. [/fig]
[fig] Author: Contributions: Conceptualization, X.M. and H.Z.; data curation, Y.H. and M.S.; investigation, Y.H. and G.X.; visualization, Y.H.; writing-original draft preparation, Y.H. and X.M.; writingreview and editing, X.M. and H.Z. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research was funded by the Project funded by China Postdoctoral Science Foundation (2021M690935), Guangxi Natural Science Foundation (2021GXNSFBA220065, 2018GXNSFBA138051), China Geological Survey Project (DD20190825). Institutional Review Board Statement: Our research did not require ethical approval. [/fig]
[table] Table 1: The correlations of HMs in surface water and hydrochemical characteristics. [/table]
[table] Table 2: The correlations between HMs in sediments and environmental factors. * Correlation is significant at p < 0.05; ** Correlation is significant at p < 0.01. Correlation is significant in bold. [/table]
[table] Table 3: The correlations of HMs between surface water and sediments. [/table]
|
Dyadic parent–infant interaction patterns at 12 months: Exploring dyadic parent–infant gender compositions
This study investigates differences in dyadic mother-infant and father-infant interaction patterns at infant age 12 months, and the relation between different parent-infant gender compositions and the dyadic interaction. Data were drawn from a large-scale, population-based Norwegian community sample comprising 671 mother-infant and 337 father-infant interactions. The Early Relational Health Screen (ERHS), a screening method for observing dyadic parent-infant interactions, was used to assess the parent-infant interactions. Scores on the ERHS were employed to investigate dyadic differences in the overall interaction scores, and dyadic interaction on seven sub-dimensions between mother-infant and father-infant pairs. The relation between different parent-infant gender compositions and the dyadic interaction scores was also examined. As expected in a normative sample, most parent-infant interactions received scores in the upper rating levels. Differences between mother-infant and father-infant patterns were generally small, but mother-infant dyads tended to obtain slightly higher scores. The mother-infant dyads received higher scores on the dimensions of engagement and enjoyment, but no other significant differences between the parent-infant pairs were found for the remaining dimensions. We did not find evidence for a moderation effect of child gender. However, parent-daughter dyads received somewhat higher scores than the parent-son dyads.K E Y W O R D Sparent-infant interaction, parent-infant gender composition, child development, early childhood, parenting. By the end of the first year of life, infants' emerging capacity to initiate positive dyadic interactions with the caregiver alters the nature of the bidirectional exchanges, and the reciprocity and dyadic mutuality between the infant and the caregiver becomes more evident [bib_ref] Parent-infant synchrony and the construction of shared timing; physiological precursors, developmental outcomes,..., Feldman [/bib_ref] [bib_ref] From social expectations to social cognition in early infancy, Striano [/bib_ref]. Dyadic mutuality in the parent-infant interaction has been assumed to consist of several factors including mutual engagement, and shared pleasure [bib_ref] Parent-infant synchrony and the construction of shared timing; physiological precursors, developmental outcomes,..., Feldman [/bib_ref] [bib_ref] Shared affect and dyadic synchrony among secure and insecure parent-toddler dyads, Lindsey [/bib_ref] [bib_ref] The structure of parent-child dyadic synchrony in toddlerhood and children's communication competence..., Lindsey [/bib_ref] , mutual responsiveness, temporal coordination, contingency, and a shared focus of attention [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref] [bib_ref] Mother's early perception of her infant's difficult temperament, parenting stress and early..., Mäntymaa [/bib_ref] , turn taking [bib_ref] Parent-infant synchrony and the construction of shared timing; physiological precursors, developmental outcomes,..., Feldman [/bib_ref] , and corresponding activity level, rhythm and pace [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref]. In addition, matching of positive affective states and prosocial interactive behavior in the early parent-infant interaction might be described as interactional or dyadic synchrony consisting of smooth exchanges which involve mutual regulation, reciprocity and harmony [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref].
Although the dyadic interaction is jointly constructed, the relationship between the parent and the infant is still asymmetrical, and the main responsibility for maintaining and coordinating the dyadic interaction relies on the caregiver [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref]. It has been pointed out that the dyadic entity is more than the sum of the parent and the infant behavior [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref] ; G. A. [bib_ref] Dyadic interaction: Greater than the sum of its parts?, Moore [/bib_ref] , implying that the match between the parent and infant is of importance, and research suggest that parent-child interactions should be investigated within a dyadic framework rather than by looking at each individual's contributions [bib_ref] Mother-child and father-child mutually responsive orientation in the first 2 years and..., Kochanska [/bib_ref]. Accordingly, the present study aimed to investigate the dyadic motherinfant and father-interaction, and also the match between the parent and the infant dyad in interaction. So far, dyadic mother-child interactions have received the most attention [bib_ref] Trajectories of maternal depressive symptoms, maternal sensitivity, and children's functioning at school..., Campbell [/bib_ref] [bib_ref] Change in mother-infant interactive behavior: Relations to change in the mother, the..., Feldman [/bib_ref] [bib_ref] Infant-mother face-to-face interaction: Age and gender differences in coordination and the occurrence..., Tronick [/bib_ref] , but currently, the field of father-infant interaction has been increasingly emphasized [bib_ref] Father involvement, father-child relationship, and attachment in the early years, Cabrera [/bib_ref] [bib_ref] Do early father-infant interactions predict the onset of externalising behaviours in young..., Ramchandani [/bib_ref]. Whether mother-infant and father-infant interactions are equal or different in terms of dyadic mutuality and reciprocity has also been questioned. Findings from earlier studies suggest that fathers tend to express less positive emotions and engagement than mothers, including less responsivity to infant cues [bib_ref] The Pennsylvania infant and family development project, I: Stability and change in..., Belsky [/bib_ref] [bib_ref] Mother-and father-infant play: A developmental analysis, Power [/bib_ref] [bib_ref] Patterns of mother and father play with their 8-month-old infant-A multiple analyses..., Power [/bib_ref] , while mothers are more likely to display affection such as holding, smiling, and vocalizing [bib_ref] Mother-and father-infant interaction involving play and holding in traditional and nontraditional Swedish..., Lamb [/bib_ref]. In other reports, parents were found to be equally engaged in exploration, and communicative and relational play while interacting with their infants and toddlers [bib_ref] A comparison of fathers' and mothers' speech with their young children, Golinkoff [/bib_ref] [bib_ref] Syntactic and conversational characteristics of fathers' speech, Hummel [/bib_ref] [bib_ref] Mother-and father-infant play: A developmental analysis, Power [/bib_ref]. In addition,
## Three key findings
- The mother-infant and father-infant interactions received scores in the upper range, but mother-infant interactions tended to receive higher scores. - The mother-infant dyads obtained higher score in the ERHS dimensions engagement and enjoyment than the father-infant dyads, but there were no such differences in the remaining dimensions. - Parent-daughter interactions earned higher scores than parent-son interactions suggesting an association between different dyadic gender compositions and the parent-infant interaction.
## Statement on the relevance of this research to the field of infant and early childhood mental health
This study examines dyadic interactions as assessed by the ERHS between mothers and fathers and their infants aged 12 months. Results suggest that mother-infant dyads tended to obtain slightly higher scores compared to father-infant dyads, and that parent-daughter pairs received higher scores than parent-son pairs. Findings increase the knowledge base on gender typed differences in dyadic parent-infant interactions during infancy.
mothers were found to use more directives in interactions with daughters than sons [bib_ref] Mother-and father-infant play: A developmental analysis, Power [/bib_ref] , while fathers display rougher and tumble play with their sons than daughters. However, these studies are old and might not reflect the contemporary dyadic mother-infant and father-infant interactions, for instance, due to increased father involvement in caregiving from infancy [bib_ref] Autonomy support in toddlerhood: Similarities and contrasts between mothers and fathers, Hughes [/bib_ref]. In Scandinavian countries, for example, most mothers are employed, and dual-earner and dual-caregiver families have been facilitated by providing one full year of paid parental leave that can be shared between the parents. As part of such wage-based leave, an individualized and nontransferable paternal quota has been legislated to ensure father involvement in infants' first year of life [bib_ref] Fatherhood and social policy in Scandinavia, Haas [/bib_ref]. Reports show that extended periods of paternal leave are associated with more father involvement and care of children [bib_ref] New fathers? Residential fathers' time with children in four countries, Hook [/bib_ref] [bib_ref] Father-friendly policies and time-use data in a cross-national context: Potential and prospects..., Sullivan [/bib_ref] , and that children whose fathers are involved and engaged during infancy develop fewer behavioral problems later [bib_ref] Do early father-infant interactions predict the onset of externalising behaviours in young..., Ramchandani [/bib_ref]. Other societal changes with increasing maternal employment and a greater emphasis on fathers' involvement with their infant might influence patterns of dyadic mother-infant and father-infant interaction [bib_ref] Mother-child and father-child interactional synchrony in dyadic and triadic interactions, De Mendonça [/bib_ref] , and also in the gender-typed interactions. It has also been suggested that mothers and fathers are becoming more similar in terms of the interactional behavior they engage in with their children and in their parental roles, especially in Western countries [bib_ref] Should researchers conceptualize differently the dimensions of parenting for fathers and mothers, Fagan [/bib_ref] [bib_ref] When do fathers care? Mothers' economic contribution and fathers' involvement in child..., Raley [/bib_ref]. Thus, more knowledge is needed about dyadic mother-infant and father-infant interaction patterns in societies where both parents typically are more involved in caretaking of infants.
The present study aimed to investigate dyadic motherinfant and father-infant interaction patterns in a large low-risk community sample from Norway. Consistent with a dyadic construct, dyadic interaction in parent-infant pairs with different parent-gender compositions were also examined. The dyadic parent-infant interactions were observed in a semi-structured play situation at infant age 12 months.
## Similarities and differences between the dyadic mother-infant and father-infant interactions
Studies on dyad interaction patterns between motherinfant and father-infant pairs give mixed results, and some researchers have asked whether mothers and fathers interact in different ways with their infants [bib_ref] Fathers and mothers at play with their 2-and 3-yearolds: Contributions to language..., Tamis-Lemonda [/bib_ref]. One study showed that mothers and fathers were equally responsive in interaction with their infants, but mothers expressed more positive affect during dyadic interactions than fathers. In another report, positive and negative engagement levels within the mother-infant dyads were reported to match levels of engagement in both mothers and infants during interactions. No such associations were found for the father-infant interactions, perhaps reflecting higher mutual emotional engagement in the mother-infant dyads [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref]. Other findings have shown that both mothers and fathers display more sensitivity than intrusiveness during dyadic interactions with their infant [bib_ref] Fathers' influence on their children's cognitive and emotional development: From toddlers to..., Cabrera [/bib_ref]. Similar patterns of interactional synchrony in mother-child and father-child dyads have also been found [bib_ref] Infant affect and affect regulation during the still-face paradigm with mothers and..., Braungart-Rieker [/bib_ref] [bib_ref] Mother-child and father-child interactional synchrony in dyadic and triadic interactions, De Mendonça [/bib_ref]. Taken together, these studies suggest equal levels of sensitivity, responsiveness, and dyadic synchrony in the mother-infant and the fatherinfant dyads, but somewhat more mutual engagement and expressions of positive affect in the dyadic motherinfant than the father-infant interactions [bib_ref] Father involvement, father-child relationship, and attachment in the early years, Cabrera [/bib_ref].
In a similar vein, several studies show that mothers display higher levels of sensitivity toward infants and toddlers [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] [bib_ref] Mothers, fathers, sons and daughters: Parental sensitivity in families with two children, Hallers-Haalboom [/bib_ref] [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] , and also use more socioemotional verbal speech relative to fathers during dyadic interactions [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref]. Mothers have also been found to behave more sensitive in a dyadic play situation compared to fathers meanwhile, the infants were more cooperative in interaction with their mothers than fathers [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref] , indicating higher mutuality and easiness in the dyadic mother-infant interactions. In corroboration, one recent study showed that mothers offered more emotional support to the child in dyadic play interactions than fathers; however, no differences in the frequency of support offered by each of the parents were observed (Waldman-Levi, 2021). Similarly, fathers and mothers were assessed to be equally engaged in directing infant attention during dyadic play. Meanwhile the fathers talked less to the infant and were less accessible in the dyadic play situation compared to mothers [bib_ref] A comparison of fathers' and mothers' involvement in childcare and stimulation behaviors..., Laflamme [/bib_ref]. These results show that parent-infant dyads seem to be similarly engaged in joint play activities, but different in terms of more emotional support, verbal comments, and more sensitivity in the mother-infant interaction compared to the father-infant interaction. One suggestion is that lower levels of sensitive behavior in the dyadic father-infant interactions might be explained by less paternal experience with the dyadic play context in infant first year of life [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref]. It has also been questioned whether paternal behavior in interaction with the infant might be qualitatively different from those of mothers, and therefore should be evaluated on their own terms [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref]. Another point of view is that dissimilarities in mothers' and fathers' interactive styles in the dyadic interaction contribute differently to infant development. For instance, the more physical, arousing, and unpredictable play reported in fatherinfant interactions [bib_ref] Measurement of father-child rough-and-tumble play and its relations to child behavior, Stgeorge [/bib_ref] , is suggested to impact significantly on the child's later capacity to self-regulate [bib_ref] Measurement of father-child rough-and-tumble play and its relations to child behavior, Stgeorge [/bib_ref]. Furthermore, experiences with differential parental styles might give the infant opportunities to learn about different behaviors and in taking part in the dyadic interaction with each of the parents. These different experiences might in turn contribute importantly to child social and emotional development [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref] [bib_ref] Mothers, fathers, sons and daughters: Parental sensitivity in families with two children, Hallers-Haalboom [/bib_ref] [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref]. To sum up, there is still a lack of knowledge concerning similarities and differences between the dyadic motherinfant and father-infant interaction and more research is needed. However, in future research, the societal context of the parent-infant dyads and the parental roles should be taken into account.
## Parent-child gender compositions and dyadic interaction patterns
Some research suggests that the dyadic parent-infant interactions are influenced by the gender composition of the dyads [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref] [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. In one study mother-daughter dyads were reported to display more sensitivity, responsiveness and involvement followed by the mother-son dyads, the father-daughter dyads, and finally, the father-son dyads, who were more often assessed with poorer interactional scores compared to the other dyads [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. No association between higher levels of sensitivity in the mother-infant interactions and child gender was found in another study; however, fathers exhibited more negative intrusiveness towards sons than daughters in the dyadic interaction [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref]. According to these studies, dyadic father-son interactions may be characterized by less mutual engagement and more intrusiveness compared to other parent-infant dyads, possibly decreasing reciprocity and joint pleasure in these dyadic exchanges, and perhaps resulting in less emotional sharing and less dyadic mutuality [bib_ref] Contingency, imitation, and affect sharing: Foundations of infants' social awareness, Markova [/bib_ref]. In contrast, higher dyadic mutuality in same sex dyads has been suggested, for instance, fathers of boys are found to be more involved and positively engaged relative to fathers of girls in infancy and toddlerhood [bib_ref] Caring for infant daughters and sons in dual-earner households: Maternal reports of..., Manlove [/bib_ref] [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref]. Father-son dyads are also reported to involve more kissing and caressing than father-daughter dyads [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref] , and mother-daughter interactions are shown to be more sensitive than mother-son interactions (Schoppe- [bib_ref] Attachment and sensitivity in family context: The roles of parent and infant..., Schoppe-Sullivan [/bib_ref]. Likewise, mothers tend to use more supportive speech with their toddleraged daughters than with their sons. These results showing higher engagement and emotional involvement in mother-daughter and father-son dyads are consistent with suggestions of higher interactional synchrony in same-sex parent-infant pairs [bib_ref] Infant-mother and infant-father synchrony: The coregulation of positive arousal, Feldman [/bib_ref]. However, other studies suggest that parents display similar levels of positive engagement and emotional availability towards boys and girls during both infancy [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref] and toddlerhood [bib_ref] Emotional communication in mother-toddler relationships: Evidence for early gender differentiation, Robinson [/bib_ref]. In summary, there is some suggestion that parents display gender-differentiated behavior towards daughters and sons in dyadic interactions, but the results are inconclusive.
An alternative view is to consider gender-typed child behavior during infancy, and the implication this might have on the dyadic parent-infant interaction. The literature demonstrating gender-typed child behavior during infancy is limited. Some reports show that infant girls are more socially involved and tend to orient more to a face or a voice, and to seek more physical and relational contact with their parents than do infant boys in the dyadic interaction [bib_ref] Sex differences in neonates' cuddliness, Benenson [/bib_ref]. Infant girls have also been found to be more able to discriminate between emotional expressions [bib_ref] A meta-analytic review of sex differences in facial expression processing and their..., Mcclure [/bib_ref] , and to display stronger preferences for dolls than do boys [bib_ref] Sex differences in infants' visual interest in toys, Alexander [/bib_ref]. Similarly, girls were observed to be more responsive and involved in both mothers and fathers during dyadic interactions than boys [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. These findings suggest that girls from infancy seem to be more oriented to and actively seek emotional support from the parent during the dyadic interactions than infant boys of the same age. Accordingly, girls aged 12 months are also found to be more responsive and more positively engaged in interactions with both parents [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref] and outperform boys on social communication [bib_ref] Joint attention in term and preterm infants at 12 months corrected age:..., Olafsen [/bib_ref]. Possibly, a girl's heightened social and relational interest will impact positively on the parent's experience of reciprocity in the back -and forth interactions, that in turn results in more dyadic mutual engagement and enjoyment. It has also been suggested that minor sex differences in neurological maturity and brain organization from birth may influence the dyadic parent-infant interaction [bib_ref] Prenatal hormonal contributions to sex differences in human cognitive and personality development, Reinisch [/bib_ref]. However, since boys and girls are treated differently from birth it is difficult to disentangle the impact of biological versus social contributions (i.e., parental gender-typed expectations) to sex differences observed after birth (D. S. [bib_ref] Sex differences in normal fetuses and infants: A commentary, Moore [/bib_ref]. To sum up, the literature demonstrating infant gender-typed behavior in the dyadic parent-infant interactions in infancy is scarce and more research to clarify alternative implications of same-sex dyads or child gender-typed behavior is needed.
## The present study and study aims
The present study is conducted in a dyadic framework and focuses on the parent-infant dyad as a unit of analysis. It includes a large community-based sample comprising mothers, fathers, and their 12-month-old infants. It aims to address some gaps in the literature concerning similarities and differences between motherinfant and father-infant interactions in infancy. Moreover, the study also examines dyadic gender compositions, to provide new knowledge on a topic with limited documentation concerning the relation between different dyadic gender compositions and interaction patterns during infancy. In this study, the Early Relational Health Screen (ERHS), a tool for assessing dyadic interactions, was used to assess the overall parent-infant interaction, as well as interaction patterns on seven different dyadic ERHS-dimensions: engagement, enjoyment responsiveness, pacing, attention, imitation, and initiation. A limitation of several observational methods that do exist, is the sole focus on one person at a time within the parent-infant dyad, thus lacking information on the dyad as an entity.
The first aim of this study is to investigate differences and similarities between the mother-infant and father-infant dyads in terms of overall interaction score by using the ERHS, as well as a more fine-grained mapping of dyadic interactions of key ERHS dimensions (engagement, enjoyment responsiveness, pacing, attention, imitation and initiation). The second objective is to examine if different dyad parent-infant gender compositions (mother-daughter and mother-son, fatherdaughter and father-son) are related to the overall interaction scores.
# Method
## Design and participants
This study is part of a Norwegian community-based prospective cohort study [bib_ref] Little in Norway: A prospective longitudinal community-based cohort from pregnancy to child..., Moe [/bib_ref]. Recruitment into the study took place between September 2011 and October 2012. Midwives at nine well-baby clinics situated in different parts of the country were informed about the project and asked all pregnant women who attended their first pregnancy checkup if they wanted to be contacted by a local health nurse. These health nurses worked as research assistants, conducted informed consent, and collected all data.
Most of the pregnant women and their partners were enrolled at 16-32 weeks of gestation (range 8-32 weeks); the majority of participants were recruited before the 20th week. A total of 1041 pregnant women were enrolled in the study, but five later withdrew requesting deletion of all data. Thus, there were 1036 women, and ultimately 884 of their partners (878 fathers and six female partners), who consented to participate and contributed their data to the study. A total of 1007 women gave birth to 1017 children, 529 boys, and 480 girls (including 10 pairs of twins).
In the analyses discussed here, 671 mother-infant pairs and 337 father-infant pairs were included. Every twin coded randomly as number one in the pair took part, the other member of the pair was excluded, and a total of seven twins were included at 12 months. Information about parental socio-demographic variables, such as age, education, marital status, parity and ethnicity, was collected from each of the parents at inclusion. At infant age 12 months, the parents and their infant were invited to the well-baby clinic for data collection. The health nurse (working as a research assistant) video-recorded the semistructured parent-infant play situation. When an infant reached the age of 12 months, fathers were also invited to respond to a web-based questionnaire concerning use of the paternal leave, and how they shared this leave with the mother. This questionnaire was e-mailed to fathers after the data collection at 12 months, and fewer participants answered than who participated in other types of data collection at that point (n = 261).
As shown in , most parents were married or cohabiting, had education at the university level, and had incomes between 300.000 and 600.000 Norwegian kroner (25,000-50,000 Euro). Most of the parents had no previous children, and a high percentage was ethnic Norwegians. Most fathers reported that they had used their quota of parental leave, which is a less extensive leave than the mothers.
## Procedures
## Parent-child interaction recording procedure
The mother-infant and father-infant interactions were recorded at child age 12 months (±14 days). The dyads were video recorded in a play situation and the parents were asked to play with their child as they usually would do at home. The research assistant used a handheld camera and stayed in the room without interfering with the interaction while recording. The interaction consisted of 10 min of free play with a specified set of toys (i.e., a koosh ball, several colored blocks, a box, a spoon, and a cup). In analyzing the interactions, the first 5 min were treated as "warm up" and only the last five minutes of each video recording were used for coding. The five last minutes was coded in entirety.
Six mother-infant interactions and ten father-infant interactions were not possible to score and therefore excluded (e.g., if the child was mostly crying, the videotape was too short, the parents terminated the recording, or a sibling interrupted the interaction). The videotapes with female partners were excluded from the analysis, as not to confound being the non-childbearing partner TA B L E 1 Maternal and paternal characteristics at inclusion and paternal leave at 12 months
## The early relational health screen (erhs)
The ERHSis a short videobased coding method for assessing dyadic parent-infant 2. Observed a 1. Sometimes observed b 0. Not observed c a "Observed" (2 points): The skill is demonstrated in a positive affective state, and the skill is the predominant characteristic of the interaction. b "Sometimes Observed" (1 point): The skill is present, but it is not the predominant characteristic of the interaction. It is weakly demonstrated or present, but the interaction appears uncomfortable or anxious or the child seems wary of the parent. The overarching affect in the interaction is less than positive, or there is uncertainty about whether a skill should be scored "observed" or not. c "Not observed" (0 points): The skill is clearly absent during the interaction.
interactions at child age 6, 12, 18, and 24 months. The ERHS addresses the parent-child interaction and focuses on the dyad as a unit. The number of dyadic dimensions is increased as the child grows older and obtains more competencies. At 12 months, seven dimensions are coded, namely "engagement," "enjoyment," "pacing," "responsiveness," "attention," "initiation," and "imitation" (see for descriptions). Each dimension is scored on a 3-point scale: observed (2 points); sometimes observed (1 point); not observed (0 points), yielding a sum score from 0 to 14. The dyadic interaction must be assessed on all seven dimensions in order to obtain a sum score. The sum score has three different cut off values that may be used to give a clinical recommendation. "Pass" (sum score 11-14) indicates a "good enough" interactional quality, "suspect" (sum score 8-10) indicates some concern for the dyad, whereas "fail" (≤7) suggests significant concerns. The scoring procedure starts by noting whether the overarching affective tone of the relationship is positive or less than positive. A positive overarching affect is identified by specific behaviors observed during the interaction, for instance "smiles both on the faces of the child and the parent," "positive tone of voice," "snuggling," and "caressing." The observed behaviors suggesting absence of a positive overarching affect are for instance "wariness," "lack of parental soothing behavior," and "the child turning away from the parent." A judgement of the overarching tone as less than positive will sway all subsequent scoring (i.e., all dyadic dimensions must then be scored as either "sometimes observed" [1 point] or "not observed" [0 points]).
## 2.2.3
Assessment of inter-rater reliability A team of five trained and supervised coders scored the taped parent-infant interactions at child age 12 months. The team consisted of students and was trained part-time together as a group. After completing training, they were required to pass a test consisting of seven tapes of parentinfant interactions coded by experts, to achieve satisfactory scores before coding the data material. The scoring of the tapes was completed in approximately one year. Twenty percent of the recordings were successively scored independently by two coders to monitor inter-rater reliability. Reliability recordings were randomly drawn from the sample, with reliability status hidden from the research assistants performing the coding. To maintain a satisfactory inter-rater reliability and avoid drifting, the team attended weekly meetings with a supervisor. Tapes that were difficult to score or had discrepancies in scoring values were discussed after the coding had taken place. In addition, tapes with full consensus were included in these sessions to learn from successfully scored tapes. The coders watched the tapes twice before scoring and without consulting with each other. They had no information about the families and were oblivious to tapes that were randomly selected for reliability purposes. Inter-rater reliability estimated by weighted quadratic Kappa demonstrated that the overall coder agreement was .76. Reliability analyses for the mother-infant interaction gave Cronbach's alpha = .72 and .75 for the father-infant interaction, indicating acceptable internal consistency.
## Analyses
Background information about the participants obtained at inclusion was examined with descriptive analyses along with data concerning paternal leave at child age 12 months. Preliminary analyses included examining the mean and distribution of ERHS sum scores and cut off scores. Linear mixed-effects (lme) models with fixed effects including parental age, education and child gender were used to analyze differences in sum scores between the mother-infant and father-infant interactions. For this model, a standardized difference is also computed based on the total standard difference for the random effects. This is only included for illustration since there is no general agreement on how to define a standardized difference in models with more than one random effect. Crosstabs with marginal homogeneity tests were used to analyze differences in each of the ERHS-dimensions between the mother-infant and the father-infant pairs. Linear mixed-effects models with fixed effects including parental age, education and child gender were used to analyze differences in sum scores between the parentdaughter and the parent-son interactions, as well as differences between mother-daughter/father-daughter and mother-son/father-son interactions. The parent by child gender interaction was subsequently analyzed in a separate linear mixed-effects model.
Logistic regressions were conducted to examine differences among the parents who participated in the interactions compared to the non-participants. Covariates were parental age, education, and parity. Inter-rater reliability was estimated by weighted quadratic Kappa. The significance level was set to p < .05, and analyses were performed by using SPSS (IBM SPSS Statistics for Windows Version 26.0.) and R (The R Foundation for Statistical Computing, 2022) packages nlme for mixed-effects modeling and vcd for estimation of inter-rater reliability. , most of mother-infant and the father-infant dyads had mean sum scores in the upper end of the scale. A higher percentage of the mother-infant pairs had cut off scores designated as "pass" as well as a lower percentage of cut off scores designated as "fail" compared to the father-infant pairs. A majority of the motherinfant and the father-infant pairs obtained the highest score "observed" (2 points) in the first five ERHS dimensions (see . In the ERHS "initiation" dimension, the majority of the parental dyads were scored as "sometimes observed" (1 point).
# Results
## Erhs sum scores, cut off scores, and distribution of scores in the dimensions of the erhs
## As shown in
## Differences between mother-infant and father-infant dyads
Mixed effects models with fixed effects for adjustment of parental age, education and child gender, showed significantly lower mean scores for the father-infant interactions compared to the mother-infant interactions (see . Analysis with crosstabs showed that the mother-infant dyads earned significantly higher scores on the ERHS dimensions of "engagement" and "enjoyment" compared to the father-infant dyads.
## Parent-child gender compositions and dyadic interactions
Mixed effects models with fixed effects for adjustment for parental age, education and child gender, demonstrated that the mother-daughter pairs scored significantly higher than the mother-son pairs. The father-daughter pairs had significantly higher scores relative to the father-son pairs (see . Further analysis showed that the motherdaughter pairs scored significantly higher than the fatherdaughter pairs, and that the mother-son dyads scored significantly higher than the father-son dyads. The parent by child gender interaction was not significant.
# Discussion
The first objective of this study was to examine differences in interaction patterns between mother-infant and father-infant dyads at infant age of 12 months, as well as dyadic interactions on seven ERHS-dimensions. Second, we examined if different dyadic parent-infant gender compositions were related to the overall interaction score. We found that most of the mother-infant and fatherinfant pairs received high scores, indicating that the majority of the dyads were observed to show mutual engagement, and to have a "good enough" interactional quality. This was expected, as the participants were drawn from a community sample of low risk (e.g., most of the parents were married or cohabiting, they were highly educated, and mostly had a high income). Other studies also report an association between socioeconomic status and the quality of the dyadic parent-child interaction [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref] [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref] [bib_ref] Fathers and mothers at play with their 2-and 3-yearolds: Contributions to language..., Tamis-Lemonda [/bib_ref]. A high proportion of mother-infant (67.5 %) and fatherinfant dyads (61.4 %) obtained the highest cut-off score (pass) at 12 months, perhaps indicating a smooth dyadic flow in the observed play situations. A good dyadic quality implies a balanced two-way process with mutual initiations and reciprocal responses, and several studies have shown, that this may serve as a context for the development of later self-regulatory capacities [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref] [bib_ref] Mother-child and father-child mutually responsive orientation in the first 2 years and..., Kochanska [/bib_ref] [bib_ref] Shared affect and dyadic synchrony among secure and insecure parent-toddler dyads, Lindsey [/bib_ref] [bib_ref] Interactive skills of infants with their high-risk mothers, Savonlahti [/bib_ref].
The result showing that 9.3% of mother-infant pairs obtained cut-off scores in the "fail" area, indicate less mutuality and responsiveness in the dyadic interactions. Although not directly comparable, it is worth noting that a proportion of similar size was found in a Danish population-based study reporting "relational disturbances" in 8.5% of 18 months-old children [bib_ref] Mental health problems and psychopathology in infancy and early childhood, Skovgaard [/bib_ref]. Possibly, in a sample with higher risk than in our cohort, one would expect to find more dyads with lower scores on interactional quality. In this vein it has been shown that 48% of clinically referred toddlers had relationship disturbances [bib_ref] Infant Mental Health Assessment: The use of DC 0-3 in an outpatient..., Mothander [/bib_ref]. Furthermore, mother-infant interactions including mothers with substance abuse and mental health problems display a significantly lower dyadic mutuality compared to low-risk dyads . TA B L E 5 Differences between the mother-daughter/motherson and the father-daughter/father-son dyads This study also demonstrated that even if the differences were small, the mother-infant dyads earned significantly higher scores as compared with the father-infant ones. These findings concur with previous studies showing that mother-infant dyads tend to score better on interaction quality than do father-infant dyads [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] , and with reports of more mutual engagement in the motherchild than father-child interactions in infancy and toddlerhood [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref] [bib_ref] Mothers, fathers, sons and daughters: Parental sensitivity in families with two children, Hallers-Haalboom [/bib_ref] [bib_ref] Fathers and mothers at play with their 2-and 3-yearolds: Contributions to language..., Tamis-Lemonda [/bib_ref]. Mothers are more often involved in daily child care in the first year of life, and spend more time in close proximity, physical care and infant nursing as compared with fathers, who generally are less involved in the early phases of the child's life [bib_ref] A comparison of fathers' and mothers' involvement in childcare and stimulation behaviors..., Laflamme [/bib_ref] [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref]. Accordingly, the parent who spend more hours together with the infant will be better acquainted with infant emotional needs and signals during dyadic interactions [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref]. Although most fathers who participated in this study had used their paternal quota, they mainly had acted as the primary caregiver for a shorter period of time than the mothers, whose leave typically was more extensive. The lower scores in the present dyadic father-infant interactions might possibly reflect that these fathers and their infants had spent less time together and had less routine in playing together than the mother-infant dyads [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref]. More time spent together (and particularly more positive time) will, in keeping with the transactional framework [bib_ref] The transactional model, Sameroff [/bib_ref] , promote a sense of familiarity and ease in the dyad, since both parent and child will have a greater opportunity to learn to anticipate each other's behaviors and responses.
## Ta b l e 4 differences between mother-infant and father-infant dyads
## Erhs sum score
## Mixed effects analyses a difference b ci p
## Dyadic differences
## Mixed effect analyses a difference b ci p
Our results, hinting at somewhat higher scores in some of the dimensions reflecting mother-infant interaction as compared with father-infant interaction, suggest that mothers and fathers may interact somewhat differently with their infants, as reported elsewhere [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] [bib_ref] A comparison of fathers' and mothers' involvement in childcare and stimulation behaviors..., Laflamme [/bib_ref] [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref]. Possibly, mothers and fathers display qualitatively different behavior during the dyadic parent-infant interaction [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] , and it is further suggested that such differences in the dyadic parent-infant interactions contribute differently to infant development in the social and emotional domains [bib_ref] The effects of parental sensitivity and involvement in caregiving on mother-infant and..., Fuertes [/bib_ref] [bib_ref] Mothers, fathers, sons and daughters: Parental sensitivity in families with two children, Hallers-Haalboom [/bib_ref] [bib_ref] The impact of sociodemographic variables, social support and child sex on motherinfant..., Piccinini [/bib_ref].
The play situations observed in the present study revealed that there were significant but small differences between the mother-infant and the father-infant dyads in the ERHS dimensions of "engagement" and "enjoyment." These dimensions, tapping emotional connection and delight, as well as pleasant facial expressions and reciprocity in responding, earned higher scores in the motherinfant relative to the father-infant pairs. This is in line with other reports showing that mother-infant dyads tend to display higher levels of emotional engagement including sensitivity and affection, sharing of affects, smiling and vocalizing more, as well as using more socio-emotionally toned speech compared with father-infant dyads [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. Reciprocal emotional engagement and sharing of affect have been emphasized as salient aspects of the early dyadic parent-infant relationship and are possibly crucial for infant communication and social-emotional development [bib_ref] Parent-infant synchrony and the construction of shared timing; physiological precursors, developmental outcomes,..., Feldman [/bib_ref] [bib_ref] Dyadic synchrony: Its structure and function in children's development, Harrist [/bib_ref] [bib_ref] Shared affect and dyadic synchrony among secure and insecure parent-toddler dyads, Lindsey [/bib_ref] [bib_ref] The structure of parent-child dyadic synchrony in toddlerhood and children's communication competence..., Lindsey [/bib_ref]. It has also been suggested that shared pleasure in the dyadic interaction is important for later child healthy psychological development [bib_ref] Shared pleasure in early mother-infant interaction: predicting lower levels of emotional and..., Mäntymaa [/bib_ref] , and conversely, that lack of emotional sharing and co-regulation in the dyadic exchanges may compromise a child's social and cognitive development [bib_ref] Contingency, imitation, and affect sharing: Foundations of infants' social awareness, Markova [/bib_ref] [bib_ref] Emotions and emotional communication in infants, Tronick [/bib_ref]. It is noteworthy that no significant differences were found between the mother-infant and the father-infant pairs in the remaining dyadic ERHS dimensions, indicating equal amounts of mutual responsiveness and pacing, attending to each other, and in regard to initiatives and imitation in the joint play situation in both the mother-infant and the father-infant dyads. These findings concur with other studies showing that the interactive behaviors in mother-infant and father-infant dyads evidence more similarities than differences [bib_ref] Parental responsiveness and infant-parent attachment: A replication study with fathers and mothers, Notaro [/bib_ref] [bib_ref] Fathers and mothers at play with their 2-and 3-yearolds: Contributions to language..., Tamis-Lemonda [/bib_ref] , and with reports of similar patterns of interactional synchrony in parent-child dyads [bib_ref] Mother-child and father-child interactional synchrony in dyadic and triadic interactions, De Mendonça [/bib_ref]. In summary, the present results imply similarities between mother-infant and father-infant dyadic interactions on several dimensions. However, in line with others studies [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] , motherinfant dyads were found to display more engagement and enjoyment during interactions than the father-infant dyads. The present study, investigating interaction in different dyadic parent-infant gender compositions demonstrated that the mother-daughter and father-daughter dyads obtained higher interactional scores compared to the mother-son and father-son dyads. These results are in keeping with another report showing that both parents displayed higher levels of sensitivity in dyadic interactions with their daughters than with their sons [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] and that mother-daughter interactions includes higher levels of sensitivity (Schoppe- [bib_ref] Attachment and sensitivity in family context: The roles of parent and infant..., Schoppe-Sullivan [/bib_ref] , responsiveness, involvement and engagement [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref] , as well as more supportive speech than do mother-son interactions. However, other observations show that mother-daughter and mother-son interactions display similar levels of posi-tive engagement and emotional availability [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref].
Our results, showing differences in interaction patterns between father-daughter pairs as compared to father-son pairs, are supported elsewhere. Specifically, more negative and intrusive behaviors have been observed in father-son than in father-daughter interactions [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref] , probably decreasing opportunities for mutual enjoyment and engagement in the father-son dyadic interaction. In addition, father-son dyads are more often found to display poorer interactional quality in terms of more intrusiveness and less structuring compared to father-daughter dyads [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. Fathers' tendency to be intrusive during play is associated with a lack of attention to infant behavior, which may weaken the paternal ability to perceive signals from the infant and adjust their own behavior according to infant cues in the father-infant interaction [bib_ref] Interdependence of parenting of mothers and fathers of infants, Barnett [/bib_ref].
In contrast, other studies have shown that fathers are more positively engaged with their sons and less sensitive with their daughters during dyadic interactions, and that the level of dyadic synchrony tend to be higher in same sex dyads [bib_ref] Infant-mother and infant-father synchrony: The coregulation of positive arousal, Feldman [/bib_ref] [bib_ref] Shared affect and dyadic synchrony among secure and insecure parent-toddler dyads, Lindsey [/bib_ref] [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref] , possibly explained by a better biological match in these dyads [bib_ref] Infant-mother and infant-father synchrony: The coregulation of positive arousal, Feldman [/bib_ref]. It has also been suggested that fathers' different patterns of play with sons versus daughters might mirror their expectations of gender typed interaction patterns [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref]. Thus, fathers who believe that girls are more vulnerable may be less physically and verbally restrictive towards girls than boys in the dyadic interactions [bib_ref] Interactional synchrony and child gender differences in dyadic and triadic family interactions, De Mendonça [/bib_ref]. Viewed in tandem with the possible influence of parent gender in the parent-infant dyadic interaction, this study shows higher scores in dyadic interactions in the parentdaughter pairs than the parent-son pairs, which might also reflect gender differences between infant girls and boys. For instance, girls are usually found to be more responsive, involved and positively engaged during dyadic interactions with both parents as compared with boys [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] [bib_ref] Family concordance and gender differences in parent-child structured interaction at 12 months, Nordahl [/bib_ref]. Infant girls are also shown to orient more to a face or a voiceand to evidence a higher occurrence of relational dyadic behavior than infant boys [bib_ref] Sex differences in neonates' cuddliness, Benenson [/bib_ref] , possibly promoting higher dyadic mutuality. Similarly, another study found that infant girls may be more able to discriminate between emotional expressions [bib_ref] A meta-analytic review of sex differences in facial expression processing and their..., Mcclure [/bib_ref] , and tend to display stronger preferences to dolls than boys [bib_ref] Sex differences in infants' visual interest in toys, Alexander [/bib_ref]. In corroboration with the present results, these findings suggest that infant girls might show higher social interest that elicits more parental responsiveness and reciprocity in the parent-daughter dyadic interactions. Accordingly, some studies suggest that daughters are more encouraged to take responsibility for the creation of dyadic mutuality by their mothers, who more actively maintain relational states with their sons in the dyadic interaction [bib_ref] Gender and patterns of emotional availability in mother-toddler and father-toddler dyads, Lovas [/bib_ref] [bib_ref] Gender and emerging autonomy in development, Robinson [/bib_ref] [bib_ref] Emotional communication in mother-toddler relationships: Evidence for early gender differentiation, Robinson [/bib_ref]. It has also been pointed out that girls seemingly arrive earlier and remain longer in the developmental period that promotes face-to face orientation and dyadic social interactions than do boys. Boys seem more quickly to move on to activities that facilitate independence but provide fewer opportunities for dyadic interactions with the parents [bib_ref] Prenatal hormonal contributions to sex differences in human cognitive and personality development, Reinisch [/bib_ref]. Minor sex differences in neurological maturity and brain organization from birth may influence such bidirectional exchanges in the parent-infant dyad [bib_ref] Prenatal hormonal contributions to sex differences in human cognitive and personality development, Reinisch [/bib_ref] , but it is difficult to disentangle the biological and the social contributions to sex differences observed after birth (D. S. [bib_ref] Sex differences in normal fetuses and infants: A commentary, Moore [/bib_ref]. In accordance, the transactional model of development [bib_ref] The transactional model, Sameroff [/bib_ref] , emphasizes that gender differences in infant behavior and responsiveness are augmented through dynamic parent-infant transactions over time, resulting in different developmental, emotional trajectories for girls and boys in their first year of life. To sum up, the present results showing higher scores in the parent-daughter compared to the parent-son interactions might reflect that parents treat girls different from boys, possibly due to gender-typed expectations. Another explanation might be biological gender differences between infant girls and boys, and possibly girls more actively seek emotional support from the parent during the dyadic interactions than infant boys of the same age.
# Limitations
This study has some limitations. There is a large discrepancy between the number of mothers and fathers who participated in the present study, and this unequal participation rate may lead to bias. However, logistic regression analyses were used to investigate how non-response may be related to background characteristics. Data were from a community-based sample, and most of the dyads earned scores in the upper-level bordering on a ceiling effect, and possibly indicating a good enough dyadic parent-infant interaction in terms of dyadic mutuality. These values would probably have varied more in socially more divergent groups or in samples including clinical groups. The ERHS method used in present study has previously been little used and needs further validation. The ERHS scoring procedure, assessing the overarching affect prior to scoring the specific dimensions, might have resulted in less nuance in the scoring values. As mentioned above, a one-point score is the highest the parent-dyad could obtain on any of the dyadic dimensions provided a positive overarching affect was not observed. Consequently, possible strengths in dyads with less than positive overarching affect cannot be revealed, even if they display "pacing", "attention" or "imitation" outside the range of concern. In clinical practice, it is essential to identify both strengths and weaknesses in dyads at risk. Strengths can serve as a port of entry in a therapeutic setting and may be used to enhance other challenging aspects of the dyadic interaction.
## Conclusions, clinical implications and recommendations
Most of the parent-infant interactions received scores at the upper end of the rating scales, as one would expect in a community sample such as this. The differences between mother-infant and father-infant interaction patterns were generally small, but those involving motherinfant dyads tended to receive higher scores. The motherinfant dyads also obtained higher scores in the ERHS dimensions engagement and enjoyment, probably suggesting somewhat higher dyadic mutuality in these dimensions. The present results show somewhat different dyadic interaction patterns with infant daughters compared to infant sons. They suggest that infant girls evince somewhat differential developmental pathways in the social domain and co-create somewhat different dyadic interactions with their parents compared to boys. However, the documentation on gender typed differences in infancy are scarce, and more research is needed.
This study suggests that the ERHS may be used for the assessment of dyadic parent-infant interactions, however, there is a need for further validation and replication of results. In clinical work with infants and parents, the quality of the early parent-infant interaction is often the main target of intervention. Videotaped observations are often needed for capturing the microscopic moments of dyadic exchanges between the infant and the parent, and there is a need for screening methods in this domain.
This study further suggests that ERHS might be feasible for assessing parent-child interaction in large research samples. The substantial inter-rater reliability that was obtained supports this feasibility. To do this well, it is recommended that researchers monitor inter-rater agreement successively, and that regular meetings are held for consensus discussions and supervision of the coders. In further use of the ERHS, researchers should consider whether overarching affect should be decisive for scoring of the individual dimensions, or rather be a valuable supplement to ensure that the emotional tone of the dyad is considered. Although outside the scope of the present study, the ERHS coding system needs to be validated against other parentchild assessment tools. Also, the cut-off scores are in need of validation.
This study was restricted to dyads observed in a play situation at child age 12 months. In future studies, dyadic interactions with younger as well as older children ought to be assessed. Further, we suggest that the ERHS should also be employed in higher-risk samples and examined in studies within basic research and clinical practice contexts. In closing, this study contributes to the dearth of literature on dyadic interaction patterns in mother-infant and fatherinfant dyads in a large low-risk community sample in a contemporary framework. The present study also fills a gap in the literature regarding gender-typed differences in the dyadic parent-infant interactions in infancy.
## C o n f l i c t s o f i n t e r e s t none
## Pat i e n t c o n s e n t f o r p u b l i c at i o n
Not required.
## D i v e r s i t y s tat e m e n t
The study was conducted in Norway, and an important aim was to sample from all geographical regions in Norway to obtain a sample that reflects the population. All pregnant women who attended routine care during pregnancy at nine well-baby clinics were asked to participate, including their partners. These clinics were selected to ensure geographical diversity and representation of all health regions in Norway. There were no criteria for exclusion, but since the questionnaires were either in Norwegian or in English, this might have excluded those who spoke other languages. Most of the participants who consented to participate was ethnic Norwegians, and in future research it will be important to implement strategies to obtain a higher variability in ethnicity. Further, variability in educational level and income reflects the demographic characteristics of the study regions; however, differences between participants living in urban and rural parts of Norway ought to be investigated in future studies.
## O r c i d
Torill Sundet Siqveland https://orcid.org/0000-0001-7101-8820
## R e f e r e n c e s
[fig] n: (%) (%) (%) (%) (%) (%) [/fig]
|
Expanded CUG repeats in DMPK transcripts adopt diverse hairpin conformations without influencing the structure of the flanking sequences
[formula] U A G G C C U G G C C U A U C G G A G G C G C U U U C C C U G C U C C U G U U C G C C G U U G U U C U G U C U C G U G C C G C C G C C C U G G G C U G C A U U G G G U U G G U G G C C C A C G C C G G C C A A C U C A C C G C A G U C U G G C G C C G C C C A G G A G C C G C C C G C G C U C C C U G A A C C C U A G A A C U G U C U U C G A C U C C G G G G C C C C G U U G G A A G A C U G A G U G C C C G G G G C A C G G C A C A G A A G C C G C G C C C A C C G C C U G C C A G U U C A C A A C C G C U C C G A G C G U G G G U C U C C G C C C A G C U C C A G U C C U G U G A U C C G G G C C C G C C C C C U A G C G G C C G G G G A G G G A G G G G C C G G G U C C G C G G C C G G C G A A C G G G G C U C G A A G G G U C C U U G U A G C C G G G A A U G C U G C U G C U G C U G C U G G G G G G A U C A C A G A C C A U U U C U U U C U U U C G G C C A G G C U G A G G C C C U G A C G U G G A U G G G C A A A C U G C A G G C C U G G G A A G G C A G C A A G C C G G G C C G U C C G U G U U C C A U C C U C C A C G C A C C C C C A C C U A U C G U U G G U U C G C A A A G U G C A A A G C U U U C U U G U G C A U G A C G C C C U G C U C U G G G G A G C G U C U G G C G C G A U C U C U G C C U G C U U A C U C G G G A A A U U U G C U U U U G C C A A A C C C G C U U U U U C G G G G A U C C C G C G C C C C C C U C C U C A C U U G C G C U G C U C U C G G A G C C C C A G C C G G C U C C G C C C G C U U C G G C G G U U U G G A U A U U U A U U G A C C U C G U C C U C C G A C U C G C U G A C A G G C U A C A G G A C C C C C A A C A A C C C C A A U C C A C G U U U U G G A U G C A C U G A G A C C C C G A C A U U C C U C G G U A U U U A U U G U C U G U C C C C A C C U A G G A C C C C C A C C C C C G A C C C U C G C G A A U A A A A G G C C C U C C A U C U G C C C A A A A A A A A A A A A A A A A A A ACUG 5 G G G U C C C U A G G C C U G G C C U A U C G G A G G C G C U U U C C C U G C U C C U G U U C G C C G U U G U U C U G U C U C G U G C C G C C G C C C U G G G C U G C A U U G G G U U G G U G G C C C A C G C C G G C C A A C U C A C C G C A G U C U G G C G C C G C C C A G G A G C C G C C C G C G C U C C C U G A A C C C U A G A A C U G U C U U C G A C U C C G G G G C C C C G U U G G A A G A C U G A G U G C C C G G G G C A C G G C A C A G A A G C C G C G C C C A C C G C C U G C C A G U U C A C A A C C G C U C C G A G C G U G G G U C U C C G C C C A G C U C C A G U C C U G U G A U C C G G G C C C G C C C C C U A G C G G C C G G G G A G G G A G G G G C C G G G U C C G C G G C C G G C G A A C G G G G C U C G A A G G G U C C U U G U A G C C G G G A A U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G G G G G G A U C A C A G A C C A U U U C U U U C U U U C G G C C A G G C U G A G G C C C U G A C G U G G A U G G G C A A A C U G C A G G C C U G G G A A G G C A G C A A G C C G G G C C G U C C G U G U U C C A U C C U C C A C G C A C C C C C A C C U A U C G U U G G U U C G C A A A G U G C A A A G C U U U C U U G U G C A U G A C G C C C U G C U C U G G G G A G C G U C U G G C G C G A U C U C U G C C U G C U U A C U C G G G A A A U U U G C U U U U G C C A A A C C C G C U U U U U C G G G G A U C C C G C G C C C C C C U C C U C A C U U G C G C U G C U C U C G G A G C C C C A G C C G G C U C C G C C C G C U U C G G C G G U U U G G A U A U U U A U U G A C C U C G U C C U C C G A C U C G C U G A C A G G C U A C A G G A C C C C C A A C A A C C C C A A U C C A C G U U U U G G A U G C A C U G A G A C C C C G A C A U U C C U C G G U A U U U A U U G U C U G U C C C C A C C U A G G A C C C C C A C C C C C G A C C C U C G C G A A U A A A A G G C C C U C C A U C U G C C C A A A A A A A A A A A A A A A A A A A A1C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G C U G [/formula]
[fig] Figure S4, Figure S5: LNA-induced structure perturbation for validation of the proposed DMPKe15 RNA structure. Experiments were performed and presented as inFigure 3using a different RT primer (D2). The large variability in apparent reactivity of nucleotides 410-420 is caused by a high frequency of natural RT stops, likely due to high U-content. For this same reason, no SHAPE data was generated for this region in other experiments. Note that in this figure the numbering does not correspond to that in the other figures due to the different repeat lengths. Additional analysis and experiments using RNAse T1. A) The experiment in [/fig]
|
Species Tree Inference Methods Intended to Deal with Incomplete Lineage Sorting Are Robust to the Presence of Paralogs
Many recent phylogenetic methods have focused on accurately inferring species trees when there is gene tree discordance due to incomplete lineage sorting (ILS). For almost all of these methods, and for phylogenetic methods in general, the data for each locus are assumed to consist of orthologous, single-copy sequences. Loci that are present in more than a single copy in any of the studied genomes are excluded from the data. These steps greatly reduce the number of loci available for analysis. The question we seek to answer in this study is: what happens if one runs such species tree inference methods on data where paralogy is present, in addition to or without ILS being present? Through simulation studies and analyses of two large biological data sets, we show that running such methods on data with paralogs can still provide accurate results. We use multiple different methods, some of which are based directly on the multispecies coalescent model, and some of which have been proven to be statistically consistent under it. We also treat the paralogous loci in multiple ways: from explicitly denoting them as paralogs, to randomly selecting one copy per species. In all cases, the inferred species trees are as accurate as equivalent analyses using single-copy orthologs. Our results have significant implications for the use of ILS-aware phylogenomic analyses, demonstrating that they do not have to be restricted to single-copy loci. This will greatly increase the amount of data that can be used for phylogenetic inference.[Gene duplication and loss; incomplete lineage sorting; multispecies coalescent; orthology; paralogy.]
Species tree inference often requires us to account for the fact that the evolutionary histories of different loci can disagree with each other, as well as with the phylogeny of the species. The reasons for this incongruence include biological causes such as incomplete lineage sorting (ILS) and introgression (broadly interpreted to include all biological processes involving genetic exchange), as well as technical causes such as the misidentification of paralogs as orthologs ("hidden paralogy"; [bib_ref] Tempo, mode, the progenote, and the universal root, Doolittle [/bib_ref].
The inference of phylogenies can be carried out by concatenating all loci together or by treating each locus separately (reviewed in [bib_ref] The concatenation question, Bryant [/bib_ref]. While concatenation ignores incongruence, gene tree-based methods allow each locus to take on its own topology. Some gene tree-based methods rely on a model for how these trees evolve within the species phylogeny (in addition to probabilistic models of sequence evolution on the gene trees). The multispecies coalescent (MSC) [bib_ref] Testing the constant-rate neutral allele model with protein sequence data, Hudson [/bib_ref] [bib_ref] Gene genealogy in three related populations: consistency probability between gene and population..., Takahata [/bib_ref] [bib_ref] Bayes estimation of species divergence times and ancestral population sizes using DNA..., Rannala [/bib_ref] [bib_ref] Gene tree discordance, phylogenetic inference and the multispecies coalescent, Degnan [/bib_ref] has emerged as the most commonly employed model of such gene genealogies. Indeed, in the last two decades a wide array of methods and computer programs have been developed for species tree inference under the MSC; see [bib_ref] Coalescent methods for estimating phylogenetic trees, Liu [/bib_ref] , , and [bib_ref] Estimating phylogenetic trees from genome-scale data, Liu [/bib_ref] for recent reviews and surveys of these methods. Other gene tree-based methods are inspired by the MSC, but do not rely explicitly on this model (e.g., [bib_ref] ASTRAL: genome-scale coalescent-based species tree estimation, Mirarab [/bib_ref]. In either case, the goal is for the methods to be robust to incongruence caused by ILS.
Regardless of the method being employed, the inference of species trees usually assumes that the data consist of only orthologous sequences. Indeed, most phylogenetic methods require the identification of orthologs; see [bib_ref] New approaches for inferring phylogenies in the presence of paralogs, Smith [/bib_ref] for a review of methods that do not require orthologs. As a result of the common requirement of orthologous loci, before such inference methods are applied to a phylogenomic data set paralogs must be identified and removed from the data. One common approach for removing paralogs is to use graph-based methods to identify homologous gene families, and then to use those gene families present in exactly a single copy in each sampled genome for phylogenetic inference (e.g., [bib_ref] OrthoMCL: identification of ortholog groups for eukaryotic genomes, Li [/bib_ref]. Another approach is to use branch-cutting methods to extract orthologs from larger gene families (e.g., [bib_ref] Orthology inference in nonmodel organisms using transcriptomes and low-coverage genomes: improving accuracy..., Yang [/bib_ref]. Neither of these two approaches guarantees that the resulting data set includes only orthologous sequences [bib_ref] Orthologs, paralogs, and evolutionary genomics, Koonin [/bib_ref]. Furthermore, restricting the data to single-copy genes-which is by far the most common practice in the community-means that much of the data must be excluded from the analysis. In particular, as more species are sampled, the frequency of genes that are present in single-copy across all species will decrease.
Paralogous sequences are often modeled by a process of gene duplication and loss (GDL) [bib_ref] Genome-scale coestimation of species and gene trees, Boussau [/bib_ref]. 368 This process can also produce incongruence, as every duplication event adds a single branch not found in the species tree (losses cannot generate incongruence). Although the MSC generates a distribution of gene trees due to ILS, it is likely that GDL models induce a distribution that differs from this. An obvious way to handle data sets where ILS and GDL could have simultaneously acted on gene families is to employ models of gene evolution that go beyond the MSC in order to incorporate GDL as well. Indeed, such models are beginning to emerge [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] [bib_ref] The multilocus multispecies coalescent: a flexible new model of gene family evolution, Li [/bib_ref]. However, the more complex the models of gene family evolution, the more computationally prohibitive statistical inference under these models becomes [bib_ref] Species tree and reconciliation estimation under a duplication-loss-coalescence model, Du [/bib_ref] , rendering their applicability infeasible except for very small data sets in terms of the number of species and gene families.
Given that much progress in terms of accuracy and computational efficiency has been made on gene treebased, ILS-aware species tree inference methods, we ask in this paper the following question: are these inference methods robust to the presence of paralogs in the data? If they are, then the reach of gene tree-based inference methods is significantly extended and the exclusion of paralogous loci from phylogenomic data sets is deemed unnecessary, thus providing more signal for the inference task. To answer this question, we study the performance of five species tree inference methods, all of which use gene trees as the input data: The maximum pseudolikelihood method of [bib_ref] A maximum pseudo-likelihood approach for phylogenetic networks, Yu [/bib_ref] as implemented by the function InferNetwork_MPL in PhyloNet [bib_ref] Inferring phylogenetic networks using, Wen [/bib_ref] , ASTRAL-III [bib_ref] ASTRAL-III: polynomial time species tree reconstruction from partially resolved gene trees, Zhang [/bib_ref] , NJ st [bib_ref] Estimating species trees from unrooted gene trees, Liu [/bib_ref] , ASTRAL-Pro [bib_ref] ASTRAL-Pro: quartet-based species-tree inference despite paralogy, Zhang [/bib_ref] , and FastMulRFS (Molloy and Warnow 2020). The latter two methods were developed with paralogs in mind, and so should serve as a good baseline for comparison to the MSC-inspired methods. In particular, ASTRAL-Pro makes use of counts of quartets from speciation, but not duplication, events. Thus, there is a connection between the ASTRAL-Pro method and orthology detection.
To test these methods, we use both simulated and real data. We simulate across a wide range of GDL rates and levels of ILS, and use two genome-scale empirical data sets with thousands of loci that contain branches with very different levels of discordance. We also sample the gene family data in multiple ways, in all cases finding that the inferences made by all methods are quite accurate and are mostly identical to the accuracy of the inferences when using only single-copy orthologs. Particularly striking is the finding that these methods infer very accurate species trees when all gene tree incongruence is due to GDL, and ILS is not a factor. We find that gene tree estimation error affects the methods' performances at a similar, or even higher, level than ILS. We also find that methods designed specifically to take GDL into account, namely ASTRAL-Pro and FastMulRFS, do not generally have higher accuracy than the other methods. Overall, our results support the use of approaches that account for gene tree incongruence, regardless of its causes.
# Methods
## Species tree inference methods
For species tree inference, we use five different methods. The first three assume that the input data come from single-copy genes:
- The maximum pseudolikelihood inference function InferNetwork_MPL in PhyloNet, which implements the method of [bib_ref] A maximum pseudo-likelihood approach for phylogenetic networks, Yu [/bib_ref]. This method amounts to running MP-EST [bib_ref] A maximum pseudo-likelihood approach for estimating species trees under the coalescent model, Liu [/bib_ref] when restricted to trees with no reticulations.
- ASTRAL-III [bib_ref] ASTRAL-III: polynomial time species tree reconstruction from partially resolved gene trees, Zhang [/bib_ref] [bib_ref] Multi-allele species reconstruction using astral, Rabiee [/bib_ref] , Version 5.6.3.
- NJ st [bib_ref] Estimating species trees from unrooted gene trees, Liu [/bib_ref].
While the maximum likelihood method of [bib_ref] Maximum likelihood inference of reticulate evolutionary histories, Yu [/bib_ref] as implemented by the InferNetwork_ML function in PhyloNet [bib_ref] Inferring phylogenetic networks using, Wen [/bib_ref] is relevant here, it is much more computationally demanding than maximum pseudolikelihood, so we chose not to run it. For comparison, we also use two methods that were designed specifically with paralogs in mind:
- ASTRAL-Pro [bib_ref] ASTRAL-Pro: quartet-based species-tree inference despite paralogy, Zhang [/bib_ref].
- FastMulRFS [bib_ref] FastMulRFS: fast and accurate species tree estimation under generic gene duplication and..., Molloy [/bib_ref].
For the sake of conclusions that we draw from this study, it may be helpful to highlight the differences between these methods. InferNetwork_MPL optimizes a pseudolikelihood function that is derived based on the assumptions of the MSC. This function is very different, for example, from a likelihood function based on a model of gene duplication and loss [bib_ref] The gene evolution model and computing its associated probabilities, Arvestad [/bib_ref]. Therefore, its accuracy in inferring species trees from data with paralogs reflects directly on the performance of MSC-based methods on such data. None of the other four methods make direct use of the MSC, though ASTRAL, ASTRAL-Pro, and NJ st have all been shown to be statistically consistent under the MSC, at least when both gene lengths and the number of genes go to infinity. Their accuracy on data with paralogs therefore reflects the suitability of these methods, rather than the MSC itself, for analyzing such data. [bib_ref] Polynomial-time statistical estimation of species trees under gene duplication and loss, Legried [/bib_ref] proved that ASTRAL-ONE and ASTRAL-multi are statistically consistent under the GDL model of [bib_ref] The gene evolution model and computing its associated probabilities, Arvestad [/bib_ref] , whereas [bib_ref] Quartet-based inference methods are statistically consistent under the unified duplication-loss-coalescence model, Markin [/bib_ref] andproved that ASTRAL-ONE and ASTRAL-multi are statistically consistent under the unified GDL/ILS model (the DLCoal model) of [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref]. ASTRAL-Pro is conjectured to be statistically consistent under the DLCoal model [bib_ref] ASTRAL-Pro: quartet-based species-tree inference despite paralogy, Zhang [/bib_ref]. FastMulRFS has been proven to be statistically consistent Given a collection of trees corresponding to gene families (one tree per gene family), we generated four types of input to each of the methods:
- ONLY: The input consists of trees of only gene families that are present in exactly one copy in each of the species.
- ONLY-NoDup: The input consists of trees of ONLY gene families that have no history of gene duplication. These are canonical single-copy orthologs.
- ONE: The input consists of trees of all gene families, but where a single copy per species per gene family is selected at random and the remaining copies are removed. If a gene family has no copies at all for some species, then the resulting tree of that gene family also has no copies for that species.
- ALL: The input consists of trees of all gene families, but where all copies of a gene in a species are treated as multiple alleles from different individuals within the species. Similar to ONE, if a gene family has no copies at all for some species, then the resulting tree of that gene family also has no copies for that species.
ONLY corresponds to the practice that is followed in many phylogenomic studies, though it does not necessarily guarantee that the included genes are orthologs. Instead, "hidden paralogs" [bib_ref] Tempo, mode, the progenote, and the universal root, Doolittle [/bib_ref] or "pseudoorthologs" [bib_ref] Orthologs, paralogs, and evolutionary genomics, Koonin [/bib_ref] may occur: these are cases in which complementary losses result in single-copy paralogs present in different species. ONLY-NoDup corresponds to a scenario where researchers know which genes have a history of duplication and can exclude them from their analysis. ONE is likely to have some hidden paralogs in the input, unless GDL does not occur. By construction, ALL has all orthologs and paralogs as input, but these are effectively labeled as orthologs with multiple individuals sampled per species, since InferNetwork_MPL, ASTRAL-III, and NJ st were not originally developed with paralogs in mind.
## Simulation setup
For model species trees, we used the trees of 16 fungal species and 12 fly species reported in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref]. Note that Saccharomyces castellii has since been renamed Naumovozyma castellii (https://www.uniprot.org/taxonomy/27288), Kluyveromyces waltii has since been renamed Lachancea waltii (https://www.uniprot.org/taxonomy/1089441), and Ashbya gossypii has been renamed Eremothecium gossypii (https://www.uniprot.org/taxonomy/33169).
The 12 To generate gene trees while allowing for ILS and GDL, we used SimPhy [bib_ref] SimPhy: phylogenomic simulation of gene, locus, and species trees, Mallo [/bib_ref] with the parameters specified below (assuming all species are diploid). SimPhy uses the three-tree model developed in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] to simulate data. In this model, a locus tree is simulated within the branches of the species tree. All incongruence between the locus tree and the species tree is due to GDL. Then, a gene tree is simulated within the branches of the locus tree, where all incongruence between the locus tree and the gene tree is due to ILS. The resulting gene tree differs from the species tree due to a combination of ILS and GDL. Using the locus trees as input to an inference method amounts to using data where all incongruence is solely due to GDL (but not ILS). Setting the rates of GDL to 0 amounts to generating gene trees where all incongruence is solely due to ILS. Note that SimPhy makes two further assumptions relevant to the results presented here: first, it assumes no hemiplasy of new duplication mutations. That is, all new duplicates immediately fix before they can be lost during a polymorphic phase. [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] found that this assumption affected 5% of gene families simulated under similar conditions. Furthermore, hemiplasy results in an excess of apparent gene losses, which should not affect inferences of species trees. The second assumption is that all gene families are independent: no events duplicate or delete more than a single gene at a time. In real data, large-scale events (including whole-genome duplications) can affect many genes at a time.
For the fungal tree simulated data sets, we used five different duplication and loss rates (assuming duplication and loss rates are equal): 0 (to investigate the performance when ILS, but not GDL, acted on the gene families), 1×10 −10 , 2×10 −10 , 5×10 −10 , and 10×10 −10 per generation. We take the case where the rate is 1×10 −10 to be similar similar to the duplication rate of 7.32×10 −11 and loss rate of 8.59×10 −11 used by [bib_ref] A Bayesian approach for fast and accurate gene tree reconstruction, Rasmussen [/bib_ref] , and denote this rate as "1×". We used two effective population sizes: 10 7 and 5×10 7 , where the former was also used by [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] as the true population size. We assumed 0.9 years per generation as in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] and used 4×10 −10 as the nucleotide mutation rate per site per generation, similar to the rates of 3.3×10 −10 and 3.8×10 −10 used by [bib_ref] Coestimation of gene trees and reconciliations under a duplication-loss-coalescence model, Zhang [/bib_ref] and [bib_ref] Estimating the per-base-pair mutation rate in the yeast Saccharomyces cerevisiae, Lang [/bib_ref] , respectively.
For the fly tree simulated data sets, we used five different duplication and loss rates (assuming duplication and loss rates are equal): 0, 1×10 −10 , 2× 10 −10 , 5×10 −10 , and 10×10 −10 per generation. A GDL rate of 1.2×10 −10 was used in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] ; [bib_ref] Coestimation of gene trees and reconciliations under a duplication-loss-coalescence model, Zhang [/bib_ref] and reported by [bib_ref] Gene family evolution across 12 drosophila genomes, Hahn [/bib_ref] ; we again denote this rate as "1×". We used two effective population sizes: 10 6 and 5×10 6 , similar to the values used in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] and the estimated value of 1.15×10 6 reported in [bib_ref] Population genetics of polymorphism and divergence, Sawyer [/bib_ref] and [bib_ref] Widespread discordance of gene trees with species tree in Drosophila: evidence for..., Pollard [/bib_ref]. We assumed 10 generations per year as in [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref] and [bib_ref] Coestimation of gene trees and reconciliations under a duplication-loss-coalescence model, Zhang [/bib_ref] and used 3×10 −9 as the mutation rate per site per generation, similar to the rate of 5×10 −9 found in [bib_ref] Rates and genomic consequences of spontaneous mutational events in Drosophila melanogaster, Schrider [/bib_ref].
For each combination of GDL rate and population size, 10,000 gene families (each containing a locus tree and its corresponding gene tree) were simulated in this fashion as one data set. Ten such data sets, each with 10,000 gene families, were generated for each condition. To study the effect of using data sets of varying sizes, for each of the 10 data sets we randomly sampled 10, 50, 100, and 250 gene families from the 10,000 gene families under the ALL, ONE, ONLY, and ONLY-NoDup scenarios. In case the number of available gene families that fits ONLY or ONLY-NoDup is smaller than the desired size, that number of gene families was used (e.g., when only 6 gene family trees are available when data sets of size 10 are desired, the 6 trees are used as input).
To study the effect of GDL and ILS on species tree estimates, for each data set of trees (true gene trees or true locus trees; i.e., trees without estimation error) of a given size, we fed the data set as input to InferNetwork_MPL, ASTRAL, NJ st , ASTRAL-Pro, and FastMulRFS and computed the Robinson-Foulds distance [bib_ref] Comparison of phylogenetic trees, Robinson [/bib_ref] , normalized by the number of internal branches in the (unrooted) species tree to obtain a value between 0 and 1. This is the normalized distance between the true and inferred species trees. To study the further effect of error in the gene tree estimates on species tree estimates, we simulated the evolution of sequences of length 500 nucleotides on all gene trees under the HKY model, using Seq-gen [bib_ref] Seq-Gen: an application for the Monte Carlo simulation of DNA sequence evolution..., Rambaut [/bib_ref]. We then inferred gene trees from the simulated sequence data using IQ-TREE [bib_ref] IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies, Nguyen [/bib_ref]. Furthermore, to study the effect of error in the locus tree estimates, we treated the true locus tree as a gene tree and simulated the evolution of sequences of length 500 nucleotides on all locus trees under the HKY model, again using Seq-gen, and inferred locus trees from the simulated sequence data using IQ-TREE. It is important to note that in practice only gene trees, but not locus trees, are inferrable, as the locus tree is an artifact of the three-tree model and not a biological entity [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref]. However, conducting analysis using inferred locus trees gives a picture of the performance when all incongruence is due to GDL and gene tree error only. Finally, InferNetwork_MPL assumes that the input gene trees are rooted. In this study, we rooted the gene tree estimates by minimizing deep coalescences [bib_ref] Gene trees in species trees, Maddison [/bib_ref] [bib_ref] Species tree inference by minimizing deep coalescences, Than [/bib_ref] ; that is, we rooted each gene tree in a way that minimizes the number of extra lineages when reconciled with the true species tree.
## Biological data
For the fungal data set, we used 2932 gene trees reported in http://compbio.mit.edu/dlcoal/ and estimated with PhyML [bib_ref] A simple, fast, and accurate algorithm to estimate large phylogenies by maximum..., Guindon [/bib_ref] , where 1867 gene trees fit the ONLY setting. For the fly data set, we used 9233 gene trees from [bib_ref] Gene family evolution across 12 drosophila genomes, Hahn [/bib_ref] reconstructed using the neighbor-joining algorithm, where 6698 gene trees fit the ONLY setting. For the fly data set, we removed any gene trees containing polytomies prior to running NJ st . In neither data set did we attempt to identify single-copy orthologs. We again rooted each gene tree in the empirical data with respect to the species trees of [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] so as to minimize deep coalescences [bib_ref] Gene trees in species trees, Maddison [/bib_ref] [bib_ref] Species tree inference by minimizing deep coalescences, Than [/bib_ref] using the method of [bib_ref] Algorithms for MDC-based multilocus phylogeny inference: beyond rooted binary gene trees on..., Yu [/bib_ref] , as implemented by the function ProcessGT in PhyloNet [bib_ref] Inferring phylogenetic networks using, Wen [/bib_ref]. We estimated species trees using ASTRAL, NJ st , maximum pseudolikelihood, ASTRAL-Pro, and FastMulRFS with these gene trees as input.
# Results
## Characteristics of the simulated data
Before we describe the inference results, we discuss the characteristics of the simulated data. First, we investigated the effects of gene duplication and loss on the number of gene copies per species in each gene family. ,b and Supplementary [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] ,b available on Dryad at http://dx.doi.org/10.5061/dryad.t76hdr81d show data on the sizes (numbers of copies) of gene families in the 16-taxon and 12-taxon data sets, respectively, under the various settings of effective population sizes and duplication and loss rates.
Clearly, the higher the GDL rates, the larger the variance in size of gene families. The figure also shows
[formula] a) b) c) d) e) f) [/formula]
FIGURE 2. Characteristics of the simulated data under different settings of the duplication/loss rates and tree topologies. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate found in nature for the clade represented by each species tree topology (see Methods section). a,b) Distribution of the total number of gene copies in individual gene families in the 16-taxon and 12taxon data sets, respectively. Note that the two tree topologies also have different simulated effective population sizes in these figures (see [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] ,b available on Dryad for more conditions). c,d) Scatter plots of XL (Species tree, Locus tree), the number of extra lineages when reconciling the true locus trees with the true species tree, for the 16-taxon and 12-taxon data sets, respectively. These plots therefore represent the effects of GDL alone. e,f) Scatter plots of XL (Locus tree, Gene tree), the number of extra lineages when reconciling the true gene trees with the true locus tree, for the 16-taxon and 12-taxon data sets, respectively. These plots therefore represent the effects of ILS alone, though note that higher rates of GDL allow there to be more gene tree branches on which ILS can act. that the average size of a gene family is roughly equal to the number of species, with the largest gene families having 65 copies for the 16-taxon data sets, and 94 copies for the 12-taxon data sets (recall that these trees use different rates of GDL). We then counted the average (over the 10 data sets per setting) number of gene families for each setting that have ONLY one copy per species and the average number of gene families with no history of duplication (i.e., ONLY-NoDup). The results are shown in [fig_ref] TABLE 1: The average number of gene families that fit the ONLY/ONLY-NoDup settings out... [/fig_ref]. The table shows that as the GDL rates increase, the number of single-copy orthologs decreases. However, as predicted by theory, there appear to be very few pseudoortholog in the ONLY data set. We then set out to assess the extent of incongruence in the gene trees due to GDL and ILS. For every pair of true species tree and true locus tree, we computed the number of extra lineages [bib_ref] Gene trees in species trees, Maddison [/bib_ref] using the DeepCoalCount_tree command in PhyloNet [bib_ref] Species tree inference by minimizing deep coalescences, Than [/bib_ref] [bib_ref] Inferring phylogenetic networks using, Wen [/bib_ref] as a proxy for the amount of incongruence in the data. Here, we treated all gene copies from the same species as different individuals. Zero extra lineages mean there is no incongruence between the two trees, and the higher the value, the more incongruence there is. In particular, no incongruence means that all gene copies from the same species are monophyletic in the locus tree, and when restricted to a single arbitrary copy per species, the locus tree and species tree have identical topologies. ,d and Supplementary [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] ,d available on Dryad show data on the number of extra lineages in the simulated 16-taxon and 12-taxon data sets, respectively, under the various settings of effective population sizes and duplication and loss rates. It is important to note that all incongruence in this case is exclusively due to GDL (ILS is not a factor in the results in these two panels). The panels do not have results for the GDL rate of 0×, because in such cases there is no incongruence at all between the locus tree and the species tree, and thus there are zero extra lineages. The results show that, unsurprisingly, there is much more incongruence for the ALL scenario than the ONE scenario. For the ONLY scenario, there is very little incongruence in either data set. The incongruence in ONLY would indicate the phenomenon of hidden paralogy: single-copy genes are paralogs, so that their gene trees do not always agree with the species tree. Given the small number of hidden paralogs [fig_ref] TABLE 1: The average number of gene families that fit the ONLY/ONLY-NoDup settings out... [/fig_ref] , these results are unsurprising. The ONLY-NoDup data sets are not plotted, because the number of extra lineages in those locus trees is always zero, as expected.
We also computed the number of extra lineages when reconciling the true gene trees with the true locus trees. Here, incongruence is exclusively due to ILS (GDL is not a factor). Figure 2e,f and Supplementary Figure S1e,f available on Dryad show data on the number of extra lineages in the simulated 16-taxon and 12-taxon data sets, respectively, under the various settings of effective population sizes and duplication and loss rates. When the gene tree topology is identical to the locus tree topology, the number of extra lineages is zero, and the larger the number of extra lineages, the more ILS has an effect on the data. The figure shows that, as expected, the amount of ILS is larger for larger population sizes, and consequently there is much more ILS in the 16taxon data set than in the 12-taxon data set. One other trend to observe is that, on average, the amount of incongruence due to ILS increases with the increase in the GDL rate. This is a reflection of the fact that for higher GDL rates, the locus trees are larger (more leaves and internal branches), and this naturally results in more branches that can be affected by ILS. Finally, the amount of incongruence due to ILS is generally far lower than the amount due to GDL in the 12-taxon data set, while the levels of incongruence due to GDL and ILS are similar in the 16-taxon data set, especially when the rates of duplication and loss are high.
## Results on simulated data
We are now in position to describe the inference results. We show figures for the 16-taxon data sets in the main text, while figures for the 12-taxon data sets are all in the Supplementary Figs. S8-S11 available on Dryad. The results for the 12-taxon data sets are consistently better in terms of accuracy, so we chose to focus here on the less-optimal results.
We first ran the inference methods ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro, and FastMulRFS on the true gene trees for all four input scenarios: ALL, ONE, ONLY, and ONLY-NoDup. In this case, gene tree estimation error is not a cause of gene tree incongruence. Instead, all incongruence is due to a combination of ILS and GDL. Results on the full 16-taxon tree are shown inand Supplementaryavailable on Dryad. Note that, in all cases, using input data with GDL levels of 0 amounts to inferring a species tree from gene trees whose incongruence is solely due to ILS.
There are several observations based on these results. First, the accuracy of the inferred 16-taxon trees is much lower in general than that of the inferred 12taxon trees. In particular, for the 12-taxon data sets, the species trees are perfectly estimated in almost all casesavailable on Dryad), whereas the species tree estimation error is high, especially for the larger population sizes, for the 16-taxon data sets. As shown in and Supplementary Figure S1 . Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of both ILS and GDL (but no gene tree estimation error). Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from true gene trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance. available on Dryad, both data sets have similar gene family sizes but differ significantly in terms of the amount of ILS in the data, with the 12-taxon data sets having very little ILS. Therefore, the straightforward explanation for the observed differences species tree inference accuracy between the 16-and 12-taxon data sets is the higher level of ILS in the former. Given that the level of incongruence due to GDL is similar between the 16-taxon and 12-taxon data sets ,d and [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] ,d available on Dryad), these results point to the larger role that ILS plays in the methods' performance than GDL does.
Second, in the case of the 16-taxon data, the performance of all methods improves as the number of gene families used as input to the method increases. Note also that the largest data set used here consists of only 250 gene trees, which is much smaller than the number available in most phylogenomic data sets. While there is very little difference observed in the performance among the methods on the 16-taxon data, ASTRAL, ASTRAL-Pro, and NJ st are more similar to each other in terms of performance than either of them is to inference under maximum pseudolikelihood or FastMulRFS. This makes sense as ASTRAL, ASTRAL-Pro, and NJ st are summary methods that make inference based on statistics derived from the input gene trees, whereas maximum pseudolikelihood uses calculations based on the multispecies coalescent directly. The performance of FastMulRFS is similar to that of other methods, but its error rates remain higher than the other methods when more gene families are used. Although ASTRAL-Pro and FastMulRFS were developed with gene duplication and loss in mind, they do not appear to outperform the other summary methods.
Third, the level of ILS for a population size of 50M is higher than for a population size of 10M, and this results in lower accuracy of inferred species trees by all methods in the former caseavailable on Dryad). This behavior is expected for any method, regardless of whether GDL is acting. Notably, FastMulRFS was not developed to deal correctly with ILS and seems to have an inflated error rate with larger population sizes, but not with smaller population sizes. Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of GDL only (no ILS or gene tree estimation error). Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from true locus trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance. that ILS may be the cause of higher error rates in this method.
Lastly, we observe very little difference in the accuracy of inferred species trees across the four input scenarios: ALL, ONE, ONLY, and ONLY-NoDup. The only case in which there is a noticeable difference is in the 12taxon data sets with the duplication rate 10× that found in nature, when only ten genes are used for inference (Supplementary Figs. S8 and S9 available on Dryad). These results imply that the presence of paralogs in the data, no matter how they are treated, does not have much of an effect on the performance of the five methods, unless very few genes are used.
The results thus far raise the important question: does GDL have any effect on the performance of these five methods? To answer this question, we ran all of them on the locus trees as input to infer species trees. By the three-tree model, this amounts to feeding these methods "gene trees" whose incongruence is solely due to GDL; that is, ILS plays no role in incongruence here. It is important to point out that locus trees are mathematical constructs of the three-tree model; in practice, inferring a locus tree is not possible, unless the data has no ILS at all. We conducted this experiment to study the performance of methods when GDL, but not ILS, causes all incongruence. Results on the full 16-taxon data sets are shown inand Supplementary available on Dryad. As the results show, all methods infer the species tree perfectly accurately on almost all data sets, regardless of the parameter settings and the input scenario. In other words, when these methodssome of which have been developed based on the multispecies coalescent directly (InferNetwork_MPL), some of which were inspired by the MSC (ASTRAL, ASTRAL-Pro, and NJ st ), and one that does not deal with ILS at all (FastMulRFS)-are applied to data that have no ILS but do have paralogs in them, they have almost perfect accuracy in terms of the species tree topology they infer, under the conditions of our simulations. Combined with the results summarized inand Supplementaryavailable on Dryad, these results show, perhaps surprisingly, that methods developed to handle ILS but not GDL do much better in handling GDL than they do in handling ILS. Perhaps unsurprisingly, ASTRAL-Pro and FastMulRFS, methods designed to handle GDL, also perform well on the true . Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of ILS, GDL and gene tree estimation error. Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from estimated gene trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance. locus trees. The inflated errors seen with FastMulRFS under some settings with gene trees are absent when true locus trees are used as input, suggesting that, indeed, these errors were due to ILS. ASTRAL-Pro was designed to deal with both ILS and GDL and performs well on both true gene trees and true locus trees.
In practice, gene trees are unknown and are inferred from sequence data. Therefore, to simulate more realistic scenarios, we inferred gene trees and locus trees from simulated sequence data and fed these tree estimates as input to the five methods. In this case, gene tree estimation error is a factor in the observed incongruences. We show the extent of error in the estimated gene and locus trees for the 16-taxon data in Supplementary available on Dryad.
Gene tree estimation error is measured by the normalized RF distance between the true gene tree and the reconstructed gene tree. For the 12-taxon data set, the average gene tree estimation error ranges from 0.456 to 0.648, whereas the average locus tree estimation error is slightly lower, ranging from 0.414 to 0.627available on Dryad). For the 16taxon data set, the average gene tree estimation error ranges between 0.073 to 0.130 while the average locus tree estimation error ranges from 0.065 to 0.099. In other words, there is much less gene tree estimation error in the 16-taxon data sets than in the 12-taxon data sets. Moreover, for the 12-taxon data sets under the ALL and ONLY settings, with increased GDL rate, a decline in error was observed (the average error dropping from 0.614 to 0.477 and 0.615 to 0.489 under ALL and ONE, respectively). Such a pattern, however, was not detected for the 16-taxon data sets.
Results of species tree inference using the full 16taxon data set based on estimated gene trees are shown in and Supplementary . Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of GDL and gene tree estimation error (no ILS). Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from estimated locus trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance.
In the case of species tree inferences using data where ILS, GDL, and gene tree estimation error are involved, the error rates of all five species tree inference methods went up, as expected and available on Dryad), but only slightly. The accuracy of the species trees improves as the number of gene families increases. As discussed above, the error in gene tree estimates in the 16-taxon data sets is very low. Since gene tree estimation error in the 12-taxon data sets is much higher (because the higher substitution rates result in noisier sequence data), we observe a larger impact of this error on the performance of methods on the 12-taxon data sets [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] available on Dryad). While the methods had an almost perfect accuracy on true gene trees, species tree estimates now have as high as 50% error when 10 gene family trees are used, and close to 25% error when 250 gene family trees are used [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] available on Dryad). These results illustrate the large impact gene tree estimation error has on these methods. In the case of the 12-taxon data sets, the impact of gene tree estimation error significantly outweighs that of ILS or GDL. [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref] available on Dryad demonstrate how GDL and gene tree estimation error (but no ILS) impact species tree inference. As withand Supplementary available on Dryad, which show almost perfect performance of species tree inference from true locus trees (i.e., GDL and no ILS), we observe little reduction in performance here due to error in the estimates of gene trees. The results demonstrate that in the absence of ILS, all methods are robust to gene tree estimation error, except when the number of gene families is very small. In the case of the 12-taxon data sets, where locus tree estimation error is much higher, the five species tree inference methods also have comparable, but lower, accuracies [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] available on Dryad).
## Figure 6 and supplementary
All of these results combined point to a very small impact of GDL on the performance of the five studied species tree inference methods and given the simulation parameters used here, regardless of how the paralogs are handled. On the other hand, across all data sets it was evident that gene tree estimation error has a noticeable impact on the methods' performance, and that ILS often had a substantial impact on accuracy.
## Results on biological data
We ran all five methods used above on two empirical data sets, each consisting of thousands of gene trees. As the two data sets were the basis for the simulated data presented above, they share many of the same properties as these data.
For the 16 fungal genomes, the inferred species trees from all five methods differ from the tree shown in [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref]. ASTRAL, NJ st , ASTRAL-Pro, and FastMulRFS inferred the same topology depicted in [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref] under all three input scenarios (recall that ONLY-NoDup is not used here, since true orthologs are not known). The same phylogeny is also inferred by InferNetwork_MPL(ONE). This inferred tree is topologically different from the tree shown in [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] : in particular, the positions of Kluyveromyces waltii and Kluyveromyces lactis have been switched, as have the positions of Candida glabrata and Saccharomyces castellii [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref]. The trees inferred by InferNetwork_MPL(ALL) and InferNetwork_MPL(ONLY) differ from the reference tree of [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] in terms of the placement of Candida glabrata and Saccharomyces castellii, as shown in [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref] [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref] ,c here.
The placement of these species shown in [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref] originally comes from a concatenated analysis of 706 single-copy genes [bib_ref] Evolution of pathogenicity and sexual reproduction in eight candida genomes, Butler [/bib_ref].
For the 12 fly genomes, all three sampling schemes and all five methods inferred the exact same tree as the species tree shown in [fig_ref] FIGURE 1: The species trees reported in Rasmussen and Kellis [/fig_ref].
# Discussion
As phylogenomic data sets grow, our ability to use them within the bounds of current analysis paradigms shrinks. One of the main problems is the decreasing number of gene families that are single-copy as the number of sampled species increases. Because most current phylogenetic methods assume that only single-copy orthologs are being used, this restriction means that such methods cannot be used for data sets with even several dozen taxa without severe downsampling or other ad hoc solutions (e.g., [bib_ref] Gene content evolution in the arthropods, Thomas [/bib_ref]. Here, we set out to ask whether phylogenomic methods intended to deal with incongruence due to ILS can be applied to data containing both orthologs and paralogs, which contain incongruence due to GDL.
On simulated data sets where only ILS acted, and GDL was not a factor, all methods had the expected performance: accurate species tree estimates that improved as the number of gene trees used increases. In the case where the level of ILS was very low (the 12-taxon data), the methods had perfect performance under almost all conditions, regardless of the number of gene trees used. FastMulRFS (Molloy and Warnow 2020) sometimes had high error rates when rates of ILS were high, a result that has been found in previous studies on the accuracy of this method 378 SYSTEMATIC BIOLOGY VOL. 71 [bib_ref] ASTRAL-Pro: quartet-based species-tree inference despite paralogy, Zhang [/bib_ref]. FastMulRFS is also the only method employed here that has not been proven to be statistically consistent under the multispecies coalescent model, in which ILS is the driving forces behind incongruence.
In the cases where both ILS and GDL acted, the performance of the five methods was hardly affected by the type of data set used (ALL, ONE, ONLY, ONLY-NoDup). Within the range of simulation parameters and data sets analyzed here, our results imply that running these methods on data with paralogs will produce species tree topologies at least as accurate as those using single-copy orthologs alone. This is especially important for data sets with a large number of species or high GDL rates.
When the methods were run on the locus tree data, where ILS does not play a role and the data consist of many gene families with multiple copies, the methods produced very accurate species trees. When as few as ten gene trees were used, error rates were elevated in data sets including paralogs available on Dryad). However, with more than 10 genes, GDL alone did not appear to affect species tree inference under our simulation conditions. This further demonstrates that GDL has very little effect on the performance of these methods.
While at first it may be surprising that these methods performed very well in terms of accuracy, the majority of signal in any input gene tree reflects species relationships. Gene duplication-if random across the species tree-simply adds noise to the data, while at the same time often doubling the amount of information on the relationships among species carrying an extra gene copy. Similarly, gene loss does not positively mislead these methods, leading to accurate reconstructions of the species tree. Nevertheless, upon close inspection, some of these results are not intuitive, especially for the maximum pseudolikelihood inference. InferNetwork_MPL makes direct use of the MSC, whose assumptions are clearly violated in all data sets except when the GDL rates are set to 0, whereas all other methods are summary methods that make no direct use of the MSC. Consequently, one would have expected that InferNetwork_MPL would be very sensitive to the presence of paralogs in the data, while the others were less so. However, we largely did not observe this behavior (but see discussion of the fungal tree below). Using methods designed specifically to deal with duplication and loss (ASTRAL-Pro and FastMulRFS) also did not lead to lower error rates. In the case of ASTRAL-Pro, we find performance similar to ASTRAL, as expected given the statistical consistency of these methods, as discussed above.
In practice, gene trees are estimated from sequence data and can be erroneous. Error in the gene tree estimates, rather than ILS, could explain much of the heterogeneity observed in phylogenomic analyses, especially at deeper nodes in a species tree [bib_ref] Incomplete lineage sorting in mammalian phylogenomics, Scornavacca [/bib_ref]. We showed the gene tree estimation error can indeed impact species tree inference significantly, and that the level of its impact is similar to that of ILS, if not larger. The results from simulations including gene tree error (and from the biological data sets) should be considered the most realistic. However, as more gene trees are used, regardless of levels of ILS or GDL, species tree accuracy increased.
In analyses of two biological data sets where a species tree has been inferred using hundreds or thousands of loci, we found high accuracy of the methods using paralogs. All methods accurately inferred the published fly species tree. For the fungal species tree, no methods inferred the species tree we initially assumed to be true, which is originally based on a concatenated analysis of 706 single-copy genes [bib_ref] Evolution of pathogenicity and sexual reproduction in eight candida genomes, Butler [/bib_ref]. All methods, applied to all data sets, disagreed with this published tree with respect to the relative positions of C. glabrata and S. castellii [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref]. Interestingly, the position of S. castellii in [bib_ref] Evolution of pathogenicity and sexual reproduction in eight candida genomes, Butler [/bib_ref] was constrained prior to tree search based on several rare genomic changes; an unconstrained search produced a topology consistent with the one found here. [bib_ref] Reconstructing the backbone of the Saccharomycotina yeast phylogeny using genome-scale data, Shen [/bib_ref] , using a data set of 1233 single-copy orthologs, could not confidently determine the relationships among these species. Here, by more than doubling the number of gene trees, we find a species tree with a local posterior probability of 1.0 for the topology shown in [fig_ref] FIGURE 7: Inferred fungal species trees [/fig_ref]. Furthermore, the results of [bib_ref] Reconstructing the backbone of the Saccharomycotina yeast phylogeny using genome-scale data, Shen [/bib_ref] support the placement of K. lactis found here. The only sets of relationships that appears to differ with up-to-date fungal phylogenies are the ones inferred by InferNetwork_MPL(ALL) and InferNetwork_MPL(ONLY). This may be because InferNetwork_MPL explicitly models data according to the MSC.
As we highlighted above, we used SimPhy to generate synthetic data, and this tool makes simplifying assumptions including no hemiplasy of new duplicates and that all gene families are independent. Under the conditions of our simulations and on the two biological data sets used here, our results point to a clear message: running species tree inference methods intended to deal with ILS on gene trees with paralogs yields highly accurate results. This conclusion is powerful for at least two reasons. First, it implies that orthology assignment and paralogy removal are not necessary for running gene tree-based species tree inference; simply treating all copies as different individuals or randomly selecting a single copy would yield very accurate species tree topologies. Nevertheless, accurate orthology inference prior to species tree inference could be helpful under evolutionary scenarios not captured by our simulations. Second, in many practical cases, too few single-copy genes are available to ensure good performance of species tree inference from those data alone. In these cases, our results suggest a ready source of more phylogenetic signal. Summary methods that do not explicitly use the MSC model (i.e., ASTRAL, ASTRAL-Pro, FastMulRFS, and NJ st ) are expected to be more robust in the presence of GDL than methods that explicitly use the model-some of these methods have 379 even been found to be statistically consistent under a model of GDL and ILS, as discussed above.
While our study focused on the accuracy of the inferred species tree topology, using paralogs for inference would clearly have an impact on the estimated branch lengths of the species tree for methods designed with orthologs in mind. In particular, under the ALL setting, there could be much more incongruence due to the large number of lineages, and, consequently, methods that use incongruence (and assume all incongruence is due to ILS) to estimate branch lengths would give values that are shorter than they truly are. For this reason, branch lengths inferred by such methods should not be used. Branch lengths estimated in ASTRAL-Pro should be accurate assuming that the rooting-and-tagging algorithm used is accurate, but, to our knowledge, the accuracy of branch length estimates using this approach has not been evaluated. When users wish to estimate branch lengths using a method designed for use with paralogs, an alternative approach is needed. The results of our analyses point to the following potential approach for inferring accurate species trees (topologies and branch lengths) by utilizing as much of the phylogenomic data as possible:
1. Use all available gene trees as input, whether or not they are single-copy in all species.
2. Feed all gene trees to a gene tree-based method to obtain a species tree topology.
3. Using a smaller subset of truly single-copy genes, and fixing the species tree topology obtained from
Step (2), optimize the branch lengths of the species tree.
For Steps (1) and (2), one option is to repeat the random sampling of single copies from each species used to generate multiple "ONE" data sets. Then, these inferred species trees could be scored under some criterion that combines the MSC with a model of gene duplication/loss. This would overcome the issue of fixing a single species tree as input to Step (3), and avoids searching species tree space while evaluating a likelihood function that is very complex and computationally very demanding to compute. As an alternative to using only single-copy orthologs in
Step (3), one could also use a statistical model that combines the MSC and GDL models (e.g., [bib_ref] Unified modeling of gene duplication, loss, and coalescence using a locus tree, Rasmussen [/bib_ref]. Such methods allow for paralogy detection and orthology assignment, conditional on the fixed species tree (or species trees), by using a more detailed evolutionary model and the full signal in the sequence data. For example, the orthology assignment could be "integrated out" or sampled, depending on the desired outcomes of the analysis. Unfortunately, while full Bayesian methods exist that model GDL alone [bib_ref] Genome-scale coestimation of species and gene trees, Boussau [/bib_ref] or that model ILS alone [bib_ref] StarBEAST2 brings faster species tree inference and accurate estimates of substitution rates, Ogilvie [/bib_ref] , none that we know of can model both.
# Conclusions
In this article, we set out to study how gene treebased species tree inference would perform on data with paralogs. The motivation for exploring this question was two-fold. First, as methods for dealing with incongruence due to ILS have become commonplace, and as practitioners are almost never certain that their data contain no paralogs, it is important to understand the effect of hidden paralogy on the quality of the inference. Second, as larger phylogenomic data sets become available, insistence on single-copy genes would mean throwing away most of the data and potentially keeping a number of loci that may be inadequate for suitably complex species tree inference methods to perform well. We investigated this question through a combination of simulations and biological data analyses. Our results show that gene tree-based inference is robust to the presence of paralogs in the data, at least under the simulation conditions and on the empirical data sets we investigated.
Our results highlight the issue that gene tree-based inference could result in very accurate species trees even when ILS is not a factor or not the only factor. This finding implies that orthology detection and restricting data to single-copy genes as a requirement for employing gene tree-based inference can be mostly eliminated, thus making use of as much of the data as possible (cf. Smith and Hahn 2021b). In particular, for very large data sets (in terms of the number of species), eliminating all but single-copy genes might leave too few loci for the species tree to be inferred accurately. Our findings show that this data exclusion could be an unnecessary practice. It is important to note however, that our results do not apply to concatenated analyses, and in such cases, the presence of paralogs may indeed have a large, negative effect [bib_ref] Bayes factors unmask highly variable information content, bias, and extreme influence in..., Brown [/bib_ref]. Species tree inference from a concatenation of the sequences with gene families is challenging in the presence of paralogs for at least two reasons. First, when gene families have different numbers of copies across species, the concatenated alignment will have very large gaps. Second, correct orthology detection is still required, so that orthologous gene copies are placed in correct correspondence across the multiple genomes in the concatenated alignment. This issue is very important to examine so as to avoid aligning non-orthologous sequences in the concatenated data set.
In our simulations, we generated gene families under a neutral model and with GDL rates that were the same across all families. It is well known that the functional implications of gene duplication and the ways in which they are fixed and maintained in the genome result in much more complex scenarios than those captured in our simulations [bib_ref] Distinguishing among evolutionary models for the maintenance of gene duplicates, Hahn [/bib_ref] [bib_ref] The evolution of gene duplications: classifying and distinguishing between models, Innan [/bib_ref]. However, analyses of the two biological data sets yield results with very similar trends to those observed in our simulations. Finally, while we did not discuss or incorporate gene flow in our study, it is possible that all three processes-ILS, GDL, and gene flow-are simultaneously involved in the evolution of some clades. Studies of the robustness of gene tree-based species tree inference under some models of gene flow exist [bib_ref] Recovering the tree-like trend of evolution despite extensive lateral genetic transfer: a..., Roch [/bib_ref] [bib_ref] Identifying a species tree subject to random lateral gene transfer, Steel [/bib_ref] [bib_ref] Phylogenomic species tree estimation in the presence of incomplete lineage sorting and..., Davidson [/bib_ref] [bib_ref] Inconsistency of species tree methods under gene flow, Solís-Lemus [/bib_ref] [bib_ref] In the light of deep coalescence: revisiting trees within networks, Zhu [/bib_ref] [bib_ref] The effect of gene flow on coalescentbased species-tree inference, Long [/bib_ref] , but, to the best of our knowledge, such studies under scenarios that incorporate all the aforementioned processes do not exist yet. It is important to highlight, as well, that great strides have been made in developing methods for phylogenetic network inference in the presence of ILS, but no probabilistic methods currently incorporate gene duplication and loss (see [bib_ref] The multilocus multispecies coalescent: a flexible new model of gene family evolution, Li [/bib_ref] for a very interesting alternative approach). We believe methods along the lines described in the previous section could be promising for accurate and scalable phylogenomic inferences without sacrificing much of the data.
# Supplementary material
Data available from the Dryad Digital Repository: http://dx.doi.org/10.5061/dryad.t76hdr81d.
[fig] FIGURE 1: The species trees reported in Rasmussen and Kellis (2012), which we use as the topologies in the simulations and in the empirical data analysis. a) The species tree of 16 fungal species. b) The species tree of 12 fly species. The species tree topologies and their branch lengths in units of million years are taken from http://compbio.mit.edu/dlcoal/. under a model of either only duplication or only loss (Molloy and Warnow 2020). [/fig]
[fig] FIGURE 3: Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of both ILS and GDL (but no gene tree estimation error). Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from true gene trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance. [/fig]
[fig] FIGURE 4: Species tree estimation error for data simulated from the 16-taxon fungal tree with a population size of 5.0×10 7 and varying GDL rates; note that simulations include the effects of GDL only (no ILS or gene tree estimation error). Species tree estimation error was measured as the normalized RF distance between the true species tree and the ones inferred from true locus trees. The five inference methods used are ASTRAL, InferNetwork_MPL, NJ st , ASTRAL-Pro ("A-pro"), and FastMulRFS. The duplication/loss rates are denoted by the rate multiplier (0×, 1×, 2×, 5×, and 10×), where 1× is the rate estimated in nature for fungi. Each row corresponds to a combination of population size and GDL rates. The X-axis in each panel represents the number of gene families used and the Y-axis represents the normalized RF distance. [/fig]
[fig] FIGURE 7: Inferred fungal species trees. a) The fungal species tree inferred by InferNetwork_MPL(ALL). b) The fungal species tree inferred by InferNetwork_MPL(ONLY) c) The fungal species tree inferred by ASTRAL, NJ st ,ASTRAL-Pro, FastMulRFS, and InferNetwork_MPL(ONE). Differences between the inferred species trees and the tree inFigure 1are highlighted in red. [/fig]
[fig] FUNDING: The National Science Foundation [DBI-1355998, CCF-1514177, CCF-1800723, and DMS-1547433 to L.N.] and [DBI-1564611 and DEB-1936187 to M.W.H.]. [/fig]
[table] TABLE 1: The average number of gene families that fit the ONLY/ONLY-NoDup settings out of the 10,000 gene families. [/table]
|
Pneumocystis jirovecii diagnosed by next-generation sequencing of bronchoscopic alveolar lavage fluid: A case report and review of literature
BACKGROUNDThe advent of molecular targeted agents and immune checkpoint inhibitors has greatly improved the treatment of advanced renal cell carcinoma (RCC), thus significantly improving patient survival. The incidence of rare drug-related adverse events has gained increased attention.CASE SUMMARYWe report a patient with advanced RCC treated with multiple lines of molecular targeted agents and immune checkpoint inhibitors, who developed a pulmonary infection after treatment with everolimus in combination with lenvatinib. Determining the pathogenic organism was difficult, but it was eventually identified as Pneumocystis jirovecii by next-generation sequencing (NGS) of bronchoscopic alveolar lavage fluid (BALF) and successfully treated with trimethoprim-sulfamethoxazole.CONCLUSIONRare pulmonary infections caused by molecular targeted agents are not uncommon in clinical practice, but their diagnosis is difficult. Evaluating BALF with NGS is a good method for rapid diagnosis of such infections. Cheng QW et al. PJP diagnosed by NGS of BALF WJCC https://www.wjgnet.com 867 February 6, 2023 Volume 11 Issue 4
# Introduction
Renal cell carcinoma (RCC) is a malignant tumor originating from the renal tubular epithelium and accounts for 80% to 90% of renal malignancies . According to GLOBOCAN 2020 global cancer statistics, RCC was the 14 th most prevalent and the 15 th most deadly malignancy . Radical surgical resection is the mainstay treatment for localized renal cancer, while a combination of systemic drugs is preferred for advanced renal cancer. Sorafenib was first approved for the treatment of metastatic renal cancer in , followed by the approval of targeted drugs such as pegaptanib, sunitinib, axitinib, sorafenib, and everolimus ; while immune checkpoint inhibitors such as navulizumab, pablizumab, and ipilimumab have also been licensed for use . The combination of various molecular targeted drugs and immune checkpoint inhibitors has resulted in a rise in rare adverse effects.
Herein, we report a case of RCC that progressed after multiple lines of therapy and improved with the mammalian target of rapamycin (mTOR) inhibitor, everolimus in combination with lenvatinib; however, the patient complained of a dry cough that developed into a lung infection. Due to difficulty in identifying the infection, empirical anti-bacterial treatment was initially administered, with unsatisfactory results. Eventually, this rare case of Pneumocystis jirovecii infection was confirmed using nextgeneration sequencing (NGS) of bronchoscopic alveolar lavage fluid (BALF), and the infection improved after appropriate treatment.
## Case presentation
## Chief complaints
A 61-year-old man diagnosed with RCC for nearly 3 years presented to our department because of a dry cough with occasional expectoration on February 22, 2022.
## History of present illness
The patient was admitted to our hospital with "hematuria for 7 d" on August 5, 2019. A computed tomography (CT) scan on admission revealed right kidney and right renal pelvis occupancy, suggestive of malignancy. In addition, multiple nodules in both lungs were seen, and metastasis was suspected. A timeline of the episode of treatment is shown in [fig_ref] Figure 1: Patient treatment timeline [/fig_ref]. After consultation and discussion, the patient agreed to undergo a laparoscopic radical right nephrectomy. Postoperative pathology revealed clear cell RCC of the right kidney. Following surgery, he was treated with oral sunitinib. A follow-up CT scan after 9 mo revealed metastases in the right adrenal gland, and axitinib was given to the patient. Three months later, CT imaging found that the metastases in the right adrenal gland and both lungs were significantly larger than before. He was thus treated with nine cycles of the immune checkpoint inhibitor sintilimab. However, a subsequent CT scan showed that the right adrenal metastasis was slightly enlarged, with multiple metastases appearing in the abdominal cavity, while the size of the lung metastases was unchanged. On July 19, 2021, four cycles of sorafenib in combination with sintilimab were administered. CT scanning revealed that the lung and abdominal cavity metastases were slightly larger than before, and new subcapsular metastases were detected on the right lobe of the liver. On December 14, 2021, he was switched to everolimus 10 mg once daily in combination with lenvatinib 8 mg once daily. However, on February 22, 2022, the patient developed a dry cough with occasional and little sputum, accompanied by mild dyspnea after activity. No fever or other discomfort was observed.
## History of past illness
The patient had no past illness.
## Personal and family history
The patient had no specific personal and family history.
## Physical examination
On physical examination, the patient's basic vital signs were within normal limits, and his respiratory sounds were rough on both lungs, without dry or wet rales.
## Laboratory examinations
The initial blood investigations revealed the following: Leukocytes 5.17 × 10 9 /L (reference range: 4.0-11.0), hemoglobin 121.0 g/L, platelet count 400 × 10 9 /L, lymphocytes 15.9%, monocytes 6.8%, neutrophils 76.3%, eosinophils 0.6%, basophils 0.4%, high sensitivity C-reactive protein 37.0 mg/L (reference range: 0-5 mg/L), and procalcitonin 0.07 ng/mL (normal value < 0.052 ng/mL). Liver and kidney functions were normal, and sputum culture revealed normal oral flora.
## Imaging examinations
CT scan on February 23, 2022 [fig_ref] Figure 2: Chronological computed tomography of the chest demonstrating changes in the lungs [/fig_ref] identified multiple patches and ground glass shadows in both lungs and scattered soft tissue nodules of different sizes in both lungs, suggestive of a pulmonary infection. The lung and peritoneal metastases were slightly smaller in size. The infection was treated with cefuroxime sodium, piperacillin sulbactam, and moxifloxacin injections, but the patient's condition did not improve, and he still had a persistent dry cough. Another CT scan on the March 21, 2022 [fig_ref] Figure 2: Chronological computed tomography of the chest demonstrating changes in the lungs [/fig_ref] showed little to no resolution of the lung infection.
## Final diagnosis
A bronchoscopy lavage fluid NGS test was performed on March 25, 2022. The patient underwent bronchoscopy, and the fiberoptic bronchoscope reached the trachea and bronchi of both lungs. Irrigation with sterile saline was performed repeatedly in the left lower lobe bronchus and right lower lobe bronchus. The collected samples were then sent to the laboratory for NGS testing and analysis, which revealed Pneumocystis jirovecii infection.
## Treatment
Two cotrimoxazole tablets (sulfamethoxazole 0.4 g, trimethoprim 80 mg/tablet) were given every 6 h and 40 mg methylprednisolone succinate was given twice a day for 5 d, followed by 40 mg once daily for 5 d.
## Outcome and follow-up
Following treatment, the patient's cough disappeared. Cotrimoxazole tablets were continued for 3 wk, and the repeat CT on April 1, 2022 [fig_ref] Figure 2: Chronological computed tomography of the chest demonstrating changes in the lungs [/fig_ref] and May 10, 2022 [fig_ref] Figure 2: Chronological computed tomography of the chest demonstrating changes in the lungs [/fig_ref] showed significant improvement and resolution in both lungs. There were no treatment-related adverse effects. At present, the patient is well and has not complained of symptoms such as cough and dyspnea.
# Discussion
Kidney cancer is the third most prevalent urogenital tumor in China, accounting for 2%-3% of adult malignant tumors, and its incidence is increasing annually . Targeted therapy is the primary treatment strategy for advanced renal cancer. According to their targets, these drugs can be divided into two main categories: Vascular endothelial growth factor (VEGF)/VEGF receptor inhibitors and mTOR inhibitors. VEGF, a major factor in the angiogenesis process in RCC, is a primary target of antiangiogenic treatments . Additionally, mTOR, which is positioned downstream of phosphoinositide 3-kinase and protein kinase B and is regulated by phosphatase and tensin homolog, is heavily involved in RCC development . Inhibition of the mTOR pathway can inhibit both angiogenesis and tumor cell proliferation. Everolimus, an mTOR inhibitor, has been found to improve survival in metastatic RCC patients after TKI-targeted drug therapy failure . Everolimus is not only approved for use in advanced RCC but also for the treatment of advanced breast cancer , pancreatic neuroendocrine tumor[18], subependymal giant cell astrocytoma associated with tuberous sclerosis , and other tumors. It can also prevent immunological rejection after kidney, liver, or heart transplantation . The adverse reactions of everolimus mainly affect the digestive system, respiratory system, endocrine system, and skin mucosa , with interstitial lung disease (ILD) being the most common pulmonary associated toxicity . The incidence of noninfectious pneumonia of metastatic RCC treated with everolimus can reach up to 14%[16]. Following everolimus administration, our patient developed a dry cough, with the CT scan revealing a pulmonary infection. Repeated sputum culture and sputum smear showed no clear evidence of microbial infection. After empirical anti-bacterial treatment, the therapeutic outcome was poor. We thus suspected an uncommon bacterial or fungal infection, but identifying the pathogen was challenging. Therefore, we resorted to BALF NGS testing and revealed that Pneumocystis jirovecii was the cause of the infection. Prior to the lung infection, the patient had been treated with many lines of targeted treatments and immune checkpoint inhibitors, and his immune function was compromised. Following treatment with everolimus in combination with lenvatinib, he acquired a rare Pneumocystis jirovecii infection. Given that lenvatinib is a multi-targeted anti-angiogenic agent, its common adverse effects include hypertension, February 6, 2023
Volume 11 Issue 4 fatigue, diarrhea, palmoplantar erythroderma syndrome, proteinuria and hemorrhagic events, while everolimus is immune-suppressive, making patients susceptible to opportunistic pulmonary infections . found that the use of everolimus in patients with advanced RCC can lead to rare pathogenic infections, including Pneumocystis jirovecii infections. As a result, we speculate that the Pneumocystis jirovecii infection in this patient was associated with the use of everolimus. Pneumocystis jirovecii can accumulate on the surface of the respiratory tract of healthy humans and can proliferate when the immune system is weakened, leading to opportunistic infections [bib_ref] Treatment of Pneumocystis jirovecii pneumonia in HIV-infected patients: a review, Huang [/bib_ref]. In the 1980s, with the emergence of the human immunodeficiency virus (HIV), Pneumocystis jirovecii pneumonia (PJP) became more prevalent in HIV patients [bib_ref] Pneumocystis jirovecii Pneumonia in the Non-HIV-Infected Population, Avino [/bib_ref]. With the increased use of immunosuppressive drugs, the incidence of PJP has gradually increased in non-HIV-infected patients, especially in those with malignant tumors who are treated with immunosuppressants . The early stage of infection in non-HIV-infected patients is characterized by a repeated low-grade fever and dry cough without characteristic symptoms, and the disease can advance rapidly to fever and respiratory failure. The mortality rate of non-HIV patients with PJP is 30% to 50%, higher than that of HIV patients [bib_ref] Diagnosing Pneumocystis jirovecii pneumonia: A review of current methods and novel approaches, Bateman [/bib_ref].
The clinical features of PJP are not specific, and the diagnosis mainly depends on the detection of pathogens. Normal or reduced white blood cells are often seen while serum lactate dehydrogenase and blood fungal (1-3)-β-d-glucan are elevated, and the G test is positive; however, these markers have limited specificity and clinical translational value . X-ray and CT chest examinations lack specificity in the early stage of infection, while after disease progression, CT imaging usually shows diffuse, bilateral pulmonary "ground glass" interstitial infiltrates, which may also appear as pulmonary nodules. Nonetheless, these imaging changes are not specific, especially as they are identical to the presentation of ILD in everolimus-induced interstitial pneumonia, which can easily lead to misdiagnosis . Unlike bacterial and other fungal infections, most non-HIV individuals have a low PJP load, and standard smear microscopy has a low sensitivity; thus, early PJP diagnosis is difficult. Microscopic inspection is commonly utilized to confirm PJP, and clinical specimens typically employed include induced sputum, bronchoalveolar lavage fluid, and lung tissue biopsy, but induced sputum culture has a low positive rate, and lung tissue biopsy is traumatic and difficult to carry out clinically. The BALF technique allows targeted sampling of the lower respiratory tract with a diagnostic positivity rate of 90% to 99% compared to sputum analysis . Polymerase chain reaction (PCR) of alveolar lavage fluid samples is a reliable method for diagnosing PJP, with some studies indicating that PCR has a sensitivity of ≥ 97% and a negative predictive value of ≥ 99% [bib_ref] Pneumocystis jirovecii pneumonia PCR test on upper respiratory tract swab, Sivaraj [/bib_ref]. If the BALF PCR test is negative, PJP can be ruled out, while a positive PCR makes distinguishing between colonization and active infection difficult .
Traditional diagnostic methods, such as smear microscopy and induced sputum culture, have a low positive rate for the diagnosis of Pneumocystis jirovecii infection. Lung tissue biopsy is a traumatic examination. BALF test has a higher positive rate. As a new detection method independent of microbial culture, NGS is a second-generation gene sequencing technology with the advantages of high throughput, wide coverage and high accuracy [bib_ref] Application of Metagenomic Next-Generation Sequencing in the Diagnosis of Pulmonary Infectious Pathogens..., Chen [/bib_ref] [bib_ref] Clinical metagenomics, Chiu [/bib_ref]. NGS can directly detect the nucleic acid sequence of pathogenic bacteria in the samples and determine their type and proportion [bib_ref] Metagenomic next-generation sequencing for mixed pulmonary infection diagnosis, Wang [/bib_ref]. It can be used for the detection of not only a variety of pathogens, such as bacteria, fungi, viruses, and parasites, but also a variety of specimens, such as sputum, blood, cerebrospinal fluid, alveolar lavage fluid, and tissue. Compared with the traditional culture method, NGS has higher detection rate and higher negative predictive value. At present, NGS has been successfully used to diagnose and treat difficult and critical infectious diseases, identify unknown pathogens, monitor drug-resistant genes, carry out epidemiological follow-up investigations, etc [bib_ref] Pneumocystis Jirovecii Pneumonia Diagnosis via Metagenomic Next-Generation Sequencing, Lu [/bib_ref]. While identifying rare pathogens (e.g., Legionella pneumophila, Corynebacterium striatum, Listeria spp., Pneumocystis jirovecii, cryptococcus, Chlamydia pinioticus) is challenging clinically, NGS detection of alveolar lavage fluid can offer a rapid diagnosis and determine a correct anti-infection treatment. Notably, NGS of BALF is more sensitive and specific than the traditional approaches in diagnosing HIV-negative PJP . However, NGS still faces many problems with a widespread application. Diagnosis by NGS testing needs to be combined with host factors, and chest CT findings. NGS pathogen test can only identify the pathogen, but cannot detect antimicrobial susceptibility. In addition, the current research on NGS testing of atypical respiratory pathogens is mostly in the form of case reports and clinical studies with a small sample size. Further clinical studies with a larger sample size are required to compare its sensitivity and specificity with traditional detection methods.
Trimethoprim-sulfamethoxazole (TMP/SMZ) is the preferred drug for the treatment of PJP, and it is emphasized that early and adequate dosage yields the best outcome , and the standard course to treat PJP is 3 wk [bib_ref] High initial (1, 3) Beta-d-Glucan concentration may be a predictor of satisfactory..., Jin [/bib_ref]. The main side effects of TMP/SMZ include skin rash, drug fever, leukopenia, renal dysfunction, electrolyte disorder, and hepatotoxicity. Patients with renal insufficiency should lower their TMP dosage according to the creatinine clearance rate. For some patients with ineffective TMP/SMZ treatment or intolerable side effects, in recent years, caspofungin combined with low-dose TMP/SMZ has been used in patients with PJP infection following organ transplant. These two drugs have a synergistic effect, achieving satisfactory efficacy and a low incidence of adverse reactions . February 6, 2023
Volume 11 Issue 4
# Conclusion
With the widespread use of anti-tumor immunosuppressants, the risks of lung infection with atypical bacteria or fungi in cancer patients have increased, and existing traditional diagnostic procedures make such infections difficult to diagnose. Therefore, in the event that tumor patients experience lung infection after anti-tumor treatment and the effect of conventional diagnosis and treatment are unsatisfactory, evaluating BALF with NGS technology can be used to detect pathogens and determine the correct treatment plan for such patients. Collectively, we believe that this approach is promising in the early diagnosis of such infections and deserves more clinical attention.
[fig] Figure 1: Patient treatment timeline. [/fig]
[fig] Figure 2: Chronological computed tomography of the chest demonstrating changes in the lungs. A: Multiple patches and ground glass shadows in both lungs and scattered soft tissue nodules of different sizes in both lungs in February 2022; B: There was no significant change in pulmonary inflammation after empirical anti-infective treatment in March 2022; C: Pulmonary infection was significantly improved and resolved in both lungs in April 2022; D: Pulmonary infection almost completely disappeared in May 2022. [/fig]
|
A Parathyroid Adenoma: Benign Disease Presenting with Hyperparathyroid Crisis
Hyperparathyroid crisis is a rare manifestation of parathyroid disease. We present the case of a 53-year-old gentleman with a review of the current literature. He presented in acute renal failure with epigastric pain and vomiting. His serum-corrected calcium (CCa 2+ ) was raised at 5.2 mmol/L, in addition to a massively raised parathyroid hormone (PTH) level (3957 ng/L). Ultrasound studies of the neck revealed a 2 cm well-defined mass inferoposterior to right thyroid lobe. CT scans of the neck showed a normal mediastinum and confirmed no associated lymphadenopathy. Having undergone medical resuscitation for 9 days, a neck exploration revealed a cystic mass, which was excised. Histological investigations revealed a 9.25 g, cystic parathyroid adenoma with no features of malignancy. His PTH and CCa 2+ returned to normal postoperatively. This suspicious presentation of benign disease, including a marked elevation in PTH, highlights the challenges facing the endocrine surgeon in dealing with parathyroid disease.
# Introduction
Primary hyperparathyroidism is a rare disease, with hyperparathyroid crisis being one of its unusual manifestations. Large rises in PTH levels in benign parathyroid disease are unusual and have been associated with more sinister diseases [bib_ref] Clinical review 122: parathyroid carcinoma, Shane [/bib_ref]. We discuss a case of a patient with a benign cystic parathyroid adenoma presenting in sinister fashion with, in particular, a massively raised serum PTH level only previously seen in malignant disease [bib_ref] Parathyroid carcinoma, atypical parathyroid adenoma, or parathyromatosis, Fernandez-Ranvier [/bib_ref].
## Case report
A 53-year-old gentleman presented with a 2-week history of worsening epigastric pain, vomiting, and constipation. He reported mild confusion but no loss of consciousness. There was no history of polyuria or polydipsia. He reported longstanding gastroesophageal reflux symptoms but no other abdominal history. His medical history was negative for depression and renal calculi. There was no history of carcinoma or any radiotherapy treatment of any kind. He reported no significant family history. On examination he was pale with dry mucous membranes. His chest and abdominal examination were unremarkable. There was no evidence of any bony pain and no palpable lumps in the neck.
At admission, he was tachycardic and hypotensive with an increased respiratory rate. He was in acute renal failure, with an elevated urea (19.5 mmol/L) and creatinine (272 µmol/L). Liver function tests were normal. Laboratory tests revealed a markedly elevated serum-corrected calcium level (CCa 2+ ) of 5.20 mmol/L (normal range 2.12-2.65) and a parathyroid hormone (PTH) level of 3957 ng/L (normal range 12-75). Vitamin D levels were normal. A myeloma screen was negative. An ECG showed that he was in normal sinus rhythm and a chest radiograph was unremarkable. Given the acute presentation and massively elevated PTH levels, the major concern was a possible malignant parathyroid lesion. An ultrasound scan of his neck revealed a 2 cm welldefined oval hypoechoic mass posterior and inferior to right thyroid lobe [fig_ref] Figure 1: USS neck revealing 3 [/fig_ref]. A CT scan of the neck confirmed the 2 cm nodule as stated which abutted the right tracheal margin, with no associated lymphadenopathy and a normal mediastinum [fig_ref] Figure 2: CT neck showing large right sided neck lesion [/fig_ref].
The patient underwent initial conservative management with aggressive intravenous fluid resuscitation, vitamin D replacement, intravenous loop diuretic treatment, and intravenous bisphosphonate therapy. On day 9, biochemical markers had improved (Urea 5.3, Creatinine 179), and CCa 2+ had fallen to 2.99 mmol/L. The patient underwent a neck exploration on day 10. During the exploration, the right inferior gland was found to be large, cystic, soft, and of a brown colour. There was a no evidence of local invasion and no lymphadenopathy. As such, a diagnosis of adenoma was made. Intraoperative PTH assay is not routinely used at our unit and as such was not performed. The right inferior parathyroid gland was excised. The right superior parathyroid gland was normal, and the contralateral neck was not explored. Histological investigations revealed a 3.5 cm × 2 cm × 1.5 cm encapsulated parathyroid adenoma weighing 9.25 g, composed of chief and oxyphil cells with cystic change. The Ki-67 proliferative index was low (1-2%). There was no local, capsular, or vascular invasion and no other features suggestive of malignancy. Given the suspicious clinical picture, the report was further confirmed at another centre.
The patient recovered well postoperatively, with PTH levels falling to 45 ng/L within a day. Subsequent recovery included a period of hypocalcaemia with a raised Alkaline Phosphatase (ALP), which resolved with calcium replacement within a month.
# Discussion
Primary hyperparathyroidism is a rare disease, where the majority of patients now present asymptomatically after biochemical testing which shows a mildly raised serum CCa 2+ and marginally elevated PTH level [bib_ref] Parathyroid crisis as first manifestation of primary hyperparathyroidism, Ntaios [/bib_ref]. However, our case highlights two important aspects of parathyroid disease that the authors would like to explore further. Firstly the difficulty in diagnosis of a patient with a markedly raised PTH and secondly the difficulty faced preoperatively in the management of a patient presenting with hyperparathyroid crisis.
Preoperative and even histological differentiation between benign or malignant pathology is often difficult, and appropriate management requires judgement at the time of surgery as malignant lesions will need more radical procedures. The features suggestive of malignant lesions at the time of surgery would include adherence due to fibrosis, local infiltration, or lymph node involvement. There has been recent work into Preoperative diagnosis based on biochemical grounds with a raised CCa 2+ and PTH increasing the index of suspicion for malignant disease suggesting that serum PTH levels are mildly elevated in benign disease, but carcinoma causes levels up to 10 times that of normal [bib_ref] Clinical review 122: parathyroid carcinoma, Shane [/bib_ref]. To further stress the benefit of PTH in preoperative diagnosis, Robert et al. suggested that a PTH level <4 times the upper limit of normal excludes a malignancy [bib_ref] Primary hyperparathyroidism: can parathyroid carcinoma be anticipated on clinical and biochemical grounds?..., Robert [/bib_ref]. Other studies support the above trend-a recent case series looking at solitary adenomas causing asymptomatic disease have derived mean values of Preoperative PTH as 186 [bib_ref] Primary hyperparathyroidism: role of the preoperative oral calcium loading test in the..., Hagag [/bib_ref] and 165 ng/L [bib_ref] Correlation of intraoperative parathyroid hormone levels with parathyroid gland size, Moretz [/bib_ref] , while research into carcinoma show higher PTH levels of 714 [bib_ref] Parathyroid carcinoma, atypical parathyroid adenoma, or parathyromatosis, Fernandez-Ranvier [/bib_ref] and 1220 ng/L [bib_ref] Functional large parathyroid carcinoma extending into the superior mediastinum, Iwata [/bib_ref]. This consolidates the unusual nature of a PTH level of 3957 ng/L in an adenoma with no radiological, Intraoperative, or histological evidence of malignancy. CCa 2+ is usually no greater than 1 mmol/dl above normal ranges in benign disease compared to over 4 times the upper limit of normal in malignant lesion [bib_ref] Clinical review 122: parathyroid carcinoma, Shane [/bib_ref]. Tumour mass has also been shown to correlate with diagnosis. The average adenoma weighs 1 g, but the average values for carcinoma weight are approximately 4 g [bib_ref] Primary hyperparathyroidism: a current perspective, Delellis [/bib_ref] [bib_ref] Unusual presentation of a giant parathyroid adenoma: report of a case, Power [/bib_ref] correlating well with Robert's study (1.3 g versus 4.9 g) [bib_ref] Primary hyperparathyroidism: can parathyroid carcinoma be anticipated on clinical and biochemical grounds?..., Robert [/bib_ref]. Our case is unusual in that we report a benign adenoma with PTH, CCa 2+ , and weight that are all suggestive of parathyroid carcinoma.
There are several published cases where benign disease presents suspiciously and PTH in particular reaches levels associated with carcinoma. These include parathyroid cysts, those presenting in developing countries, patients with oxyphil adenomas, and those presenting, as demonstrated in our case, in hyperparathyroid crisis. Parathyroid cysts are rare neck lesions [bib_ref] Giant functioning parathyroid cyst presenting as a retrosternal goitre, Mckay [/bib_ref]. Case reports and series investigating these have reported a wide range of PTH levels, fluctuating from normal to 2250 ng/L [bib_ref] Hyperparathyroidism with a functioning parathyroid cyst, Ak [/bib_ref] with overlapping clinical and biochemical presentations of parathyroid cysts with carcinoma [bib_ref] Presentation of 6 cases with parathyroid cysts and discussion of the literature, Wirowski [/bib_ref] , making Preoperative differentiation difficult but important as surgical management will differ. In developing countries, benign adenomas are reported to present with a similar biochemical profile to carcinoma. Agarwal et al. reported PTH levels in benign disease similar to that in cancerous disease [bib_ref] Indian primary hyperparathyroidism patients with parathyroid carcinoma do not differ in clinicoinvestigative..., Agarwal [/bib_ref] , attributing this to the late presentation of the disease and nutritional deficiencies. Oxyphil adenomas have been suggested to cause 3% of all adenomas. Recent case reports have suggested markedly elevated PTH levels up to 1291 ng/L [bib_ref] Functioning oxyphil adenoma of parathyroid, Dewanda [/bib_ref] , although often there are no systemic features suggestive of patients in crises.
Our case of parathyroid crisis is a further example of suspiciously elevated PTH levels, suggestive of parathyroid carcinoma. A recent study published in 2008 looking at just under 300 patients undergoing procedures for hyperparathyroidism showed 2.8% present in crisis [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref]. These patients present with 3-4 times higher PTH levels than those with asymptomatic disease and over 10 times that of normal (up to 1770 ng/L), consistent with PTH levels in malignant disease. They found that 88% were caused by underlying adenoma. Bleeding into adenomas that have undergone cystic degeneration is a well-recognised cause of crisis with histopathological findings revealing a large adenoma with cystic spaces filled with haemorrhagic fluid [bib_ref] Intracystic hemorrhage in a mediastinal cystic adenoma causing parathyrotoxic crisis, Manouras [/bib_ref]. There is a suggested trend where the weight of tumour excised during a hyperparathyroid crisis is heavier than those with asymptomatic disease (means 7.5 g versus 1.6 g) [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref]. Our case is an example of this, where a patient in crisis has an adenoma weight much higher than typical benign disease.
Management of hyperparathyroid crisis has traditionally involved an emergency parathyroidectomy within 72 hours of presentation, which has a mortality of up to 14% [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref] [bib_ref] Hyperparathyroid crisis. Clinical and pathologic studies of 14 patients, Wang [/bib_ref]. However, recent evidence published by Phitayakorn and McHenry [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref] supports an early excision after a period of optimisation, rather than emergency surgery. Once volume depletion is corrected, loop diuretic therapy and intravenous bisphosphonate therapy can be initiated [bib_ref] Acute hyperparathyroidism: our experience with 36 cases, Gasparri [/bib_ref]. This relatively rapid reduction in CCa 2+ thereby acts as a "bridge to surgery," with a mean interval between presentation and operative intervention of 8 days [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref]. This is not dissimilar to our interval of 10 days, thereby allowing for appropriate operative workup and management of concurrent medical problems. Surgical management of patients presenting with hyperparathyroid crisis secondary to adenomatous disease is effective, with a reported success rate of 92% [bib_ref] Parathyroidectomy for hypercalcemic crisis: 40 years' experience and long-term outcomes, Cannon [/bib_ref]. While medical optimisation preoperatively is largely successful [bib_ref] Hyperparathyroid crisis: use of bisphosphonates as a bridge to parathyroidectomy, Phitayakorn [/bib_ref] , there are reports where medical management has failed and disease has only responded to definitive surgical management [bib_ref] Seven patients with hyperparathyroid crisis: emphasis on prompt parathyroidectomy, Iihara [/bib_ref]. Dialysis has also been used as a successful adjunct to medical therapy, in the interval between presentation and surgery [bib_ref] Hypercalcaemic crisis and acute renal failure due to primary hyperparathyroidism, Georges [/bib_ref]. With appropriate medical management prior to surgical excision, mortality rates have fallen, with rates for patients presenting in hyperparathyroid crisis reported as 2.8% [bib_ref] Acute hyperparathyroidism: our experience with 36 cases, Gasparri [/bib_ref].
In summary, our case involves a gentleman presenting in hyperparathyroid crisis with a massively raised serum PTH level that the authors believe to be one of the highest reported in benign disease published to date. It represents the presentation of benign cystic adenoma mimicking malignant disease. It also reinforces the importance of prompt initial medical management, Preoperative diagnostic and localising studies, and sound operative judgement, highlighting the difficulties facing the endocrine surgeon when dealing with lesions of the parathyroid gland.
[fig] Figure 1: USS neck revealing 3.69 cm × 2.11 cm lesion in right side of neck. [/fig]
[fig] Figure 2: CT neck showing large right sided neck lesion. [/fig]
|
Microcurrent Stimulation Triggers MAPK Signaling and TGF-β1 Release in Fibroblast and Osteoblast-Like Cell Lines
Wound healing constitutes an essential process for all organisms and involves a sequence of three phases. The disruption or elongation of any of these phases can lead to a chronic or non-healing wound. Electrical stimulation accelerates wound healing by mimicking the current that is generated in the skin after any injury. Here, we sought to identify the molecular mechanisms involved in the healing process following in vitro microcurrent stimulation-a type of electrotherapy. Our results concluded that microcurrents promote cell proliferation and migration in an ERK 1/2or p38-dependent way. Furthermore, microcurrents induce the secretion of transforming growth factor-beta-1 (TGF-β1) in fibroblasts and osteoblast-like cells. Interestingly, transcriptomic analysis uncovered that microcurrents enhance the transcriptional activation of genes implicated in Hedgehog, TGF-β1 and MAPK signaling pathways. Overall, our results demonstrate that microcurrents may enhance wound closure through a combination of signal transductions, via MAPK's phosphorylation, and the transcriptional activation of specific genes involved in the healing process. These mechanisms should be further examined in vivo, in order to verify the beneficial effects of microcurrents in wound or fracture healing.
# Introduction
The ability of an organism to respond to injury, in order to achieve tissue repair and to maintain homeostasis, is of high significance. Acute wounds normally heal in a steady and efficient way, consisting of three distinct but overlapping phases: inflammation, tissue formation and remodeling. In each phase, different cell types, such as macrophages, neutrophils, keratinocytes, fibroblasts, endothelial cells etc., are recruited and interact with each other to restore the damaged tissue area. Impairment in one or more of these phases leads to delayed wound healing or to a chronic, non-healing wound. In chronic wounds, parts of the injured areas are found in different stages of healing or are "trapped" in a certain phase for a prolonged period. However, the molecular mechanisms that lead to ineffective wound healing are not clearly elucidated.
When a skin injury occurs, a flow of current, immediately, generates an electric field through the wound, known as "the current of injury", which activates and recruits the cell types involved in the healing process, to the site of the wound. These electric fields (EFs) stimulate division, proliferation and the directional migration of cells, in order to initiate the wound healing process. Electrical stimulation (ES) has been introduced as a treatment for chronic wounds, as it mimics the "current of injury" and has beneficial effects in every phase of wound healing. ES has been shown to promote the migration of keratinocytes and macrophages, to stimulate fibroblasts
# Materials and methods
## Cell culture
NIH3T3 and MG-63 cells (ATCC, Rockville, MD, USA) were cultured in Dulbecco's modified Eagle's medium (DMEM, Biosera, Nuaillé, France), supplemented with 10% fetal bovine serum (FBS, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) and 1% penicillin/streptomycin (Biosera). Cells were cultured without any stimulatory supplements or additional vitamins and were incubated at 37 - C in a 5% CO 2 humidified atmosphere. For passaging, the cells were detached with trypsin/ethylenediaminetetraacetic acid (EDTA) (Biosera) and subsequently re-plated. The cells were routinely subjected to mycoplasma testing and found to be negative.
## Generation of microcurrents
For the generation of microcurrents, the commercially available WMCS W200 device (Wetlinghealth, Fredensborg, Denmark) was used. To avoid the loss of microcurrents, the stimulation of cells was performed in glass cell culture dishes. Regarding the setting up of the experiment, the return wire of the device was placed under the glass cell culture dish and the treatment head of the device was positioned in the appropriate distance from the dish (approximately 10-12 cm from the bottom of the plate), until the continuous beeping sound stopped emitting (Supplementary Materials,. This warning sound occurs when the distance is too great and the transferred microcurrent does not reach the target. The thickness of the culture medium was 3 mm and the electric conductivity of DMEM was determined to be~1.6 S/m. The required microcurrent amount and duration of the treatment were set in the control panel. For our experiments, the amplitude of the applied currents was constantly at 2 µA. During the exposure, the pH and the temperature of the culture medium were monitored by using a pH-meter and an infrared thermometer. Total RNA was measured in NanoDrop (ND1000 Spectrophotometer, PEQLAB, Erlangen, Germany). The samples were diluted accordingly to a mean concentration of approximately 100-150 ng/µL and their quality assessed in a Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) using the Agilent RNA 6000 Nano Kit reagents and protocol (Agilent Technologies, Santa Clara, CA, USA). For the library preparation, the 3 mRNA-Seq Library Prep Kit Protocol for Ion Torrent (QuantSeq-LEXOGEN Vienna, Austria) was used according to the manufacturer's instructions. Up to 500 ng of RNA was used for the first and second strand synthesis, followed by 13 cycles of amplification. Library quality and quantity were assessed in Bioanalyzer using the DNA High Sensitivity Kit reagents and protocol (Agilent Technologies, Santa Clara, CA, USA). The quantified, barcoded libraries were pooled together at a final concentration of 7pM. The pools were templated and enriched on an Ion Proton One Touch system. Templating was performed using the Ion PI Hi-Q OT2 200 Kit (Thermo Fischer Scientific), followed by sequencing with the Ion PI Hi-Q Sequencing 200 Kit on Ion Proton PI V2 chips (Thermo Fischer Scientific) according to the commercially available protocols, on an Ion Proton System, according to the manufacturer's instructions.
## Short read mapping
The obtained FASTQ files were mapped on the UCSC hg19 reference genome using a two-phase mapping procedure. Firstly, the short reads were mapped using tophat2, with default settings and using additional transcript annotation data for the hg19 genome from Illumina iGenomes. Then, the reads which remained unmapped were submitted to a second round of mapping using Bowtie2 with the -local and -very-sensitive local switches turned on.
# Differential expression analysis
Differential expression analysis was performed using the Bioconductor package metaseqR. The BAM files obtained after short read mapping, one for each RNA-Seq sample, were summarized to a 3 UTR read counts table, using the Bioconductor package GenomicRanges. In the final read counts table, each row represented each column one RNA-Seq sample and each cell, the corresponding read counts associated with each row and column. The gene counts table was normalized for inherent systematic or experimental biases (e.g., sequencing depth, gene length, GC content bias etc.) using the Bioconductor package DESeq after removing the genes that had zero counts over all the RNA-Seq samples. Prior to the statistical testing procedure, the gene read counts were filtered for possible artifacts that could affect the subsequent statistical testing procedures. Genes presenting any of the following were excluded from further analysis: (i) genes with a total length less than 500, (ii) genes whose average reads per 100 bp was less than the 25th quantile of the total normalized distribution of average reads per 100 bp, (iii) genes with read counts below the median read counts of the total normalized count distribution, (iv) genes whose Ensembl biotype matched the following: rRNA, TR_V_pseudogene, TR_J_pseudogene, IG_C_pseudogene, IG_J_pseudogene, IG_V_pseudogene, and (v) the genes where 50% of the samples did not present more than five normalized counts across all samples. The resulting gene counts table was subjected to differential expression analysis for the contrasts WM (cells stimulated with wireless microcurrents) vs. the Ctrl. (control cells), using a combination of the Bioconductor packages DESeq, edgeR, limma, NBPSeq and NOISeq. In order to combine the statistical significance from the multiple algorithms, the PANDORA weighted p-value across the results method was calculated and applied. .
## Real-time pcr
# Statistical analysis
Data are expressed as the means ± standard error of the mean (SEM) of at least three independent experiments. The statistical analysis was performed using one-way ANOVA. A value of p < 0.05 was considered statistically significant.
# Results
## Stimulation with microcurrents activates erk 1/2 and p38 map kinases
To identify whether the microcurrents activate specific signaling pathways in mammalian cells, we examined the phosphorylation of ERK 1/2 and p38 kinases in two different cell lines: NIH3T3 and MG-63. NIH3T3 cells are mouse embryonic fibroblasts, which participate in all three phases of wound healing by mediating several important activities for wound closure. Osteoblasts are involved in fracture healing. Therefore, MG-63 were chosen as osteosarcoma cells sharing certain osteoblastic features. NIH3T3 and MG-63 cell cultures were serum starved and subsequently exposed to microcurrentsuntil different charges of ionized O 2 of −414, −916, −1672 and −3100 µC were transferred. Treatment with microcurrents had no cytotoxic effect and did not induce changes in the temperature and pH of the culture medium, as shown in-D. Protein extracts were collected and analyzed using specific antibodies for the phosphorylated forms of ERK 1/2 and p38. As shown in, the maximum phosphorylation of ERK 1/2 and p38 in NIH3T3 cells was evident when −916 µC O 2 − were transferred. Regarding the MG-63 cells, higher levels of ERK 1/2 and p38 phosphorylation were detected following the transfer of −414 µC O 2 − and started to decline afterwards. Taken together, these data suggest that the microcurrent stimulation activates MAPKs ERK 1/2 and p38, via phosphorylation, in osteoblasts and fibroblasts, following the transfer of −414 µC and −916 µC of O 2 − , respectively.
Cells 2020, x FOR PEER REVIEW 6 of 15 cells was evident when −916 μC O2 − were transferred. Regarding the MG-63 cells, higher levels of ERK 1/2 and p38 phosphorylation were detected following the transfer of −414 μC O2 − and started to decline afterwards. Taken together, these data suggest that the microcurrent stimulation activates MAPKs ERK 1/2 and p38, via phosphorylation, in osteoblasts and fibroblasts, following the transfer of −414 μC and −916 μC of O2 − , respectively. were separated by SDS-PAGE and immunoblotted to detect the phosphorylation levels of ERK 1/2 and p38. Graphs depict the phosphorylation levels of ERK 1/2 and p38 normalized to total-ERK 1/2 and total p38, respectively. Actin was used as the loading control. (* p < 0.05, ** p < 0.01, *** p < 0.005, treated vs. control, n = 3).
## Microcurrents induce wound closure in an erk 1/2-or p38-dependent manner in vitro
To directly examine the effects of microcurrent stimulation on the healing process, wound closure was monitored in monolayer cultures. For this purpose, the scratch wound assays were performed in NIH3T3 and MG-63 cells and the rate of gap closure was determined upon stimulation with microcurrents. The percentage of wound closure was measured daily until the surface of the wound had been fully "healed". When the microcurrents were applied and the optimal number of electric charges was transferred (−916 μC O2 − for NIH3T3 and −414 μC O2 − for MG-63), both NIH3T3and MG-63 cellsshowed increased migration and proliferation rates compared to the untreated cells (control). As a result, the stimulation with microcurrents enhances the wound closure in NIH3T3 and MG-63 cells. In order to investigate whether microcurrentdependent wound closure requires MAPKs ERK 1/2 or p38 activation, we repeated the experiments, in the presence of inhibitors, U0126 for ERK 1/2 or SB203580 for p38. Treatment with ERK 1/2 or p38 inhibitor in stimulated NIH3T3 and MG-63 cells caused reduced wound closure rate.
These results indicate the significance of ERK 1/2 or p38 MAPKs activation during wound closure induced by microcurrents. To validate the specificity of the inhibitors, U0126 and SB203580 regarding were separated by SDS-PAGE and immunoblotted to detect the phosphorylation levels of ERK 1/2 and p38. Graphs depict the phosphorylation levels of ERK 1/2 and p38 normalized to total-ERK 1/2 and total p38, respectively. Actin was used as the loading control. (* p < 0.05, ** p < 0.01, *** p < 0.005, treated vs. control, N = 3).
## Microcurrents induce wound closure in an erk 1/2-or p38-dependent manner in vitro
To directly examine the effects of microcurrent stimulation on the healing process, wound closure was monitored in monolayer cultures. For this purpose, the scratch wound assays were performed in NIH3T3 and MG-63 cells and the rate of gap closure was determined upon stimulation with microcurrents. The percentage of wound closure was measured daily until the surface of the wound had been fully "healed". When the microcurrents were applied and the optimal number of electric charges was transferred (−916 µC O 2 − for NIH3T3 and −414 µC O 2 − for MG-63), both NIH3T3and MG-63 cellsshowed increased migration and proliferation rates compared to the untreated cells (control). As a result, the stimulation with microcurrents enhances the wound closure in NIH3T3 and MG-63 cells. In order to investigate whether microcurrent-dependent wound closure requires MAPKs ERK 1/2 or p38 activation, we repeated the experiments, in the presence of inhibitors, U0126 for ERK 1/2 or SB203580 for p38. Treatment with ERK 1/2 or p38 inhibitor in stimulated NIH3T3 and MG-63 cells caused reduced wound closure rate.
These results indicate the significance of ERK 1/2 or p38 MAPKs activation during wound closure Cells 2020, induced by microcurrents. To validate the specificity of the inhibitors, U0126 and SB203580 regarding the blockage of ERK 1/2 or p38 activation, we analyzed the protein extracts from NIH3T3 and MG-63 cells, treated with U0126 or SB203580 and stimulated with microcurrents. The analysis revealed that the inhibitors U0126 and SB203580 blocked MAPKs' phosphorylation, both in the untreated and in the microcurrent-treated cells. In general, our results demonstrate that stimulation with microcurrents induces cellular migration and/or proliferation through the activation of ERK 1/2 or p38 MAPKs, leading to enhanced wound closure.
Cells 2020, x FOR PEER REVIEW 7 of 15 the blockage of ERK 1/2 or p38 activation, we analyzed the protein extracts from NIH3T3 and MG-63 cells, treated with U0126 or SB203580 and stimulated with microcurrents. The analysis revealed that the inhibitors U0126 and SB203580 blocked MAPKs' phosphorylation, both in the untreated and in the microcurrent-treated cells. In general, our results demonstrate that stimulation with microcurrents induces cellular migration and/or proliferation through the activation of ERK 1/2 or p38 MAPKs, leading to enhanced wound closure.
## Microcurrents enhance cell proliferation through erk 1/2 and p38 activation
Cellular migration and proliferation are crucial events of the wound healing process. To test whether stimulation with microcurrents could enhance cell proliferation, we determined the proliferation rate of NIH3T3 and MG-63 cells, in response to treatment with microcurrents. NIH3T3 and MG-63 cells were serum starved and then stimulated with microcurrents. The number of cells was determined 24 and 48 h following the stimulation with microcurrents. The analysis revealed that microcurrents significantly increased the proliferation rate of both NIH3T3 and MG-63 cells. Moreover, in the presence of an ERK 1/2 or p38 inhibitor (U0126 and SB203580, respectively), or both the inhibitors, the microcurrent-induced proliferation was abolished. Overall, these data suggest that the stimulation with microcurrents enhances cellular proliferation in an ERK 1/2-and p38-dependent way, as examined in the cell types participating in wound healing. images recorded under identical conditions with a 100× magnification are shown. (* p < 0.05, *** p < 0.005, n = 3, n = 3, microcurrent stimulation vs. control).
## Microcurrents enhance cell proliferation through erk 1/2 and p38 activation
Cellular migration and proliferation are crucial events of the wound healing process. To test whether stimulation with microcurrents could enhance cell proliferation, we determined the proliferation rate of NIH3T3 and MG-63 cells, in response to treatment with microcurrents. NIH3T3 and MG-63 cells were serum starved and then stimulated with microcurrents. The number of cells was determined 24 and 48 h following the stimulation with microcurrents. The analysis revealed that microcurrents significantly increased the proliferation rate of both NIH3T3 and MG-63 cells. Moreover, in the presence of an ERK 1/2 or p38 inhibitor (U0126 and SB203580, respectively), or both the inhibitors, the microcurrent-induced proliferation was abolished. Overall, these data suggest that the stimulation with microcurrents enhances cellular proliferation in an ERK 1/2and p38-dependent way, as examined in the cell types participating in wound healing.
## Treatment with microcurrents increases tgf-β1 secretion
During the healing process, cells interact with various ECM components, such as collagens, fibronectin, proteoglycans, etc. This process is mediated by cytokines and growth factors. TGF-β1 is an essential component for matrix formation, as it stimulates the synthesis of matrix proteins, such as collagen I, and the receptors associated with these proteins. Our goal was to investigate
## Treatment with microcurrents increases tgf-β1 secretion
During the healing process, cells interact with various ECM components, such as collagens, fibronectin, proteoglycans, etc. This process is mediated by cytokines and growth factors. TGF-β1 is an essential component for matrix formation, as it stimulates the synthesis of matrix proteins, such as collagen I, and the receptors associated with these proteins. Our goal was to investigate. As depicted in, TGF-β1 levels were elevated in the supernatants from stimulated NIH3T3 cells compared to the control. Furthermore, TGF-β1 levels were increased in the supernatants collected from MG-63 cells, at 6, 8, 24 and 48 h post-treatment with microcurrents. To unravel the possible implication of ERK 1/2 and p38 kinases in TGF-β1 release upon stimulation with microcurrents, we repeated the experiments using the U0126, SB203580 or the combination of both inhibitors. Treatment with the ERK 1/2 and p38 inhibitors blocked the TGF-β1 release in the microcurrent-stimulated NIH3T3and MG-63 cells. In summary, these data suggest that microcurrents enhance the secretion of TGF-β1 in an ERK 1/2and p38-dependent manner, which may contribute to the healing process through the synthesis of ECM components.. As depicted in, TGF-β1 levels were elevated in the supernatants from stimulated NIH3T3 cells compared to the control. Furthermore, TGF-β1 levels were increased in the supernatants collected from MG-63 cells, at 6, 8, 24 and 48 h post-treatment with microcurrents. To unravel the possible implication of ERK 1/2 and p38 kinases in TGF-β1 release upon stimulation with microcurrents, we repeated the experiments using the U0126, SB203580 or the combination of both inhibitors. Treatment with the ERK 1/2 and p38 inhibitors blocked the TGF-β1 release in the microcurrent-stimulated NIH3T3and MG-63 cells. In summary, these data suggest that microcurrents enhance the secretion of TGF-β1 in an ERK 1/2-and p38-dependent manner, which may contribute to the healing process through the synthesis of ECM components.
## Upregulation of genes participating in tgf-β, mapk and hedgehog signaling pathways upon stimulation with microcurrents
To decipher the total transcriptomic profile of MG-63 cells, following the treatment with −414 μC O2 − , RNA seq was applied to the total RNA extracted from the stimulated and control cells at 8 h post stimulation. The RNA seq data analysis revealed that 202 genes were downregulated, 121 were upregulated, while 816 genes did not show significant differences in the expression levels between
## Upregulation of genes participating in tgf-β, mapk and hedgehog signaling pathways upon stimulation with microcurrents
To decipher the total transcriptomic profile of MG-63 cells, following the treatment with −414 µC O 2 − , RNA seq was applied to the total RNA extracted from the stimulated and control cells at 8 h post stimulation. The RNA seq data analysis revealed that 202 genes were downregulated, 121 were upregulated, while 816 genes did not show significant differences in the expression levels between the stimulated and control cellsand,B, Tables S1 and S3). Gene ontology analysis revealed that several upregulated genes participate in cellular processes, mediated by TGF-β, MAPK and Hedgehog signaling pathways, as well as in cell cycle progressionand . On the other hand, 25 of the 202 downregulated genes were found to participate in metabolic pathways, with five of them, e.g., ENTPD6: ectonucleoside triphosphate diphosphohydrolase 6 and PDE5A: phosphodiesterase 5A, implicated in the metabolism of purineand . Real-time PCR was performed to verify the differences in the gene expression of selected target genes identified by RNA seq analysis. As shown in, NR1D1, KIF13b and ST6GALNAC2 were highly expressed in the stimulated compared to the control MG-63 cells. Furthermore, we examined the expression levels of Tgf-β1, Col1A1 and Mmp19 in NIH3T3 cells, treated with −916 µC O 2 − . As shown in, Col1A1 expression was significantly higher at 1.5 h post stimulation compared to the control, while expression levels of Tgf-β1 were significantly increased at 2 h after stimulation. Mmp19 expression was significantly elevated both at 1.5 and 2 h following stimulation. The expression levels of these genes started to decline in the samples collected 3 h post-stimulation. Hedgehog (Hh) signaling pathway is implicated in wound healing through the activation of cell proliferation and angiogenesis. To investigate the potential activation of the Hh signaling pathway upon microcurrent stimulation in NIH3T3 cells, the expression levels of Smo, Ptch1, Gli3 were examined by real-time PCR. The analysis revealed that Smo, Ptch1, Gli3 genes were upregulated in stimulated compared to control cells. To sum up, these data suggest that stimulation with microcurrents leads to the transcriptional activation of multiple genes participating in MAPK-, Hedgehog-or TGF-β1-signaling pathways.
Cells 2020, 9, x FOR PEER REVIEW 10 of 15 the stimulated and control cells,B, Tables S1 and S3). Gene ontology analysis revealed that several upregulated genes participate in cellular processes, mediated by TGF-β, MAPK and Hedgehog signaling pathways, as well as in cell cycle progressionand . On the other hand, 25 of the 202 downregulated genes were found to participate in metabolic pathways, with five of them, e.g., ENTPD6: ectonucleoside triphosphate diphosphohydrolase 6 and PDE5A: phosphodiesterase 5A, implicated in the metabolism of purineand . Real-time PCR was performed to verify the differences in the gene expression of selected target genes identified by RNA seq analysis. As shown in, NR1D1, KIF13b and ST6GALNAC2 were highly expressed in the stimulated compared to the control MG-63 cells. Furthermore, we examined the expression levels of Tgf-β1, Col1A1 and Mmp19 in NIH3T3 cells, treated with −916 μC O2 − . As shown in, Col1A1 expression was significantly higher at 1.5 h post stimulation compared to the control, while expression levels of Tgf-β1 were significantly increased at 2 h after stimulation. Mmp19 expression was significantly elevated both at 1.5 and 2 h following stimulation. The expression levels of these genes started to decline in the samples collected 3 h post-stimulation. Hedgehog (Hh) signaling pathway is implicated in wound healing through the activation of cell proliferation and angiogenesis. To investigate the potential activation of the Hh signaling pathway upon microcurrent stimulation in NIH3T3 cells, the expression levels of Smo, Ptch1, Gli3 were examined by real-time PCR. The analysis revealed that Smo, Ptch1, Gli3 genes were upregulated in stimulated compared to control cells. To sum up, these data suggest that stimulation with microcurrents leads to the transcriptional activation of multiple genes participating in MAPK-, Hedgehog-or TGF-β1-signaling pathways.
# Discussion
Wound healing is a very efficient process, which starts immediately after the tissue has been injured. Chronic ulcers-occurring when the healing process is disturbed-constitute a substantial socioeconomic burden, as they may lead to severe morbidity (e.g., amputation) and mortality. In recent decades, electrotherapy was employed either with the use of contact electrodes or with the exposure to current flow, and has been applied for the ailment of chronic ulcers, as an alternative to biochemical compounds. ES is a safe, cost-effective and painless wound healing method that mimics the natural "current of injury". According to various studies, ES stimulates many cell types involved in wound or fracture healing, promoting their migration, proliferation and DNA synthesis, as well as the upregulation of TGF-β transcription, which mediates the synthesis of ECM proteins. It has been shown that in response to exogenous EFs, keratinocytes, macrophages, epidermal and epithelial cells migrate to contribute to the wound closure. Moreover, the ES of low amplitude induces angiogenesis both directly, through the activation of endothelial cells and indirectly, by stimulating the production of VEGF from various cell types. The present study outlines the effects of microcurrents, which is another type of electrotherapy, in the cell types participating in the healing process.
Data from clinical cases suffering from chronic ulcers showed that treatment with microcurrents resulted in wound closure at 90-95% of the ulcer size. The treatment duration and the number of sessions required to achieve wound healing was found to be proportional to the severity of the injury. Based on these data, we aimed to identify the molecular mechanisms underlying the microcurrent-stimulated healing process by using two different cell lines participating in wound and fracture healing: fibroblasts and osteoblast-like cells. The induction of cellular proliferation and migration via the activation of MAPKs ERK 1/2 and p38 is suggested as a possible mechanism by which microcurrents exert their action. These MAPKs are activated in response to a variety of extracellular stimuli, and regulate crucial cellular processes such as cell proliferation, differentiation and migration. Moreover, it has been reported that the application of EFs triggered the phosphorylation of MAPKs ERK 1/2 and p38, in vitro. In our study, in order to generate microcurrents, we used a commercially available WMCS device. NIH3T3 and MG-63 cell lines were stimulated with five different amounts of charges of −414, −916, −1672 and −3100 µC O 2 − , which did not induce cytotoxic effects. Our results verified that MAPKs ERK 1/2 and p38 are activated upon microcurrent stimulation, while maximal ERK 1/2 and p38 phosphorylation was observed when −916 µC O 2 − and −414 µC O 2 − were transferred to NIH3T3 and MG-63 cells, respectively. In accordance to previous studies, we proved that ERK 1/2 and p38 phosphorylation upon stimulation with microcurrents is specific, as indicated with the use of specific inhibitors for these kinases. As aforementioned, it is well established that the activation of MAPKs ERK 1/2 and p38, through phosphorylation, induces cell migration and proliferation, whilst both of these cellular responses are involved in the wound healing process. To address whether microcurrent stimulation enhances wound closure through cellular migration and/or proliferation, we applied a scratch-wound and cell proliferation assay on NIH3T3 and MG-63 cell lines. A highly significant increase in both the wound closure and proliferation rate was observed in the microcurrent-stimulated cells. Furthermore, these effects were dependent on the activation of MAPKs ERK1/2 or p38 in both cell lines, suggesting that microcurrent-induced cellular proliferation and migration is mediated by MAPKs signaling.
It is worth mentioning that during the exposure to microcurrents, the temperature of the culture medium was monitored and was found to be constantly stable. Thus, we excluded the possibility that the WMCS device generates heat, which could delay the wound healing process by reducing the re-epithelialized area. As a result, the effect of microcurrents is not affected by the heat.
TGF-β plays a major role in the healing process by stimulating the proliferation of fibroblasts and induces the synthesis of proteins which constitute the ECM. Our results indicate that treatment with microcurrents promotes the secretion of TGF-β1 in NIH3T3 and MG-63 cells, which is consistent with results from other studies where ES-induced TGF-β1 secretion in human fibroblasts. Furthermore, according to our results, microcurrents induce TGF-β1 release in an ERK 1/2-and p38-dependent way. Except from affecting the activation of MAPKs signaling pathways, we examined whether treatment with microcurrents could, also, alter the transcriptome profile of osteoblasts. RNA seq revealed that from the 1139 genes analyzed; 121 were upregulated, 202 were downregulated and 816 were not differentially expressed in the MG-63 cells treated with −414 µC O 2 − compared to the controls. Gene annotation ontology analysis uncovered that several upregulated genes participate in the TGF-β, MAPK or Hh signaling pathways in the MG-63 cells. On the other hand, downregulated genes participate mostly in metabolic pathways, such as the metabolism of purines and pyrimidines, which are involved in many biochemical processes, including DNA and RNA synthesis. Similarly, we detected the upregulation of genes which are involved in TGF-β and Hh signaling cascades in NIH3T3 cells. The Hh signaling pathway was identified as a novel target, which is activated by stimulation with microcurrents in both cell lines and thus, may enhance wound healing. Indeed, research data support that the Hh pathway is involved in the healing process by inducing cell proliferation, migration and angiogenesis.
To summarize, our data show that ES, in the form of microcurrents, promotes the activation of ERK 1/2 and p38 in the cell types implicated in wound or fracture healing, increases the expression levels of Col1A1 and Mmp19, which are critical molecules for ulcers' healing and upregulates TGF-β1, MAPKs and Hedgehog signaling pathways. Therefore, we confirmed that stimulation with microcurrents accelerates the healing process in vitro by triggering the phosphorylation of MAPKs ERK 1/2 and p38 and alters the transcriptome profile of the cells. Herein, we describe a simple experimental setup for the study of microcurrent stimulation effects in different cell lines and the identification of signaling pathways involved in the wound-healing process in vitro. Our study provides valuable insights in the mechanistic function of microcurrents. However, further investigation using different cell types, such as macrophages or keratinocytes, is needed. Furthermore, in vivo studies or experiments on human samples or organoids are required to verify the beneficial effects of microcurrent stimulation. Moreover, it is of high significance that future studies will eliminate the possibility of the persistent activation of MAPKs, Hh and TGF-β signaling pathways, or the downregulation of genes that are implicated in the metabolism during the treatment with microcurrents, as these pathways may be associated with tumorigenic phenotypes and severe adverse effects. Indeed, the p38 signaling pathway has been shown to enhance prolonged inflammation in wounds.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4409/9/9/1924/s1,: Experimental setup, temperature and pH values, MTT assay,: Specificity of the U0126 and SB203580 inhibitors,: Heatmap of the differentially expressed genes between the stimulated with microcurrents and the control cells,: Gene ontology analysis of the downregulated genes,: Transcriptional activation of the genes following the stimulation with microcurrents, : RNA seq results, : Gene annotation ontology analysis, : Upregulated and downregulated genes, : the sequences of primers used in this study. |
Quantifying Positional Isomers (QPI) by Top-Down Mass Spectrometry
CorrespondenceGraphical Abstract 5 17 26 70 −2 0 2 0 1 C C AC AC AC Modifications Sequence position p-val: 0.97 (n.s.) Ratio (log2) 0 2 4 6 Expected ratio Detected ratio (log2) 1 : 1 4 : 1 1 6 : 1 6 4 : 1 [email protected] BriefHere we determine the quantities of the same protein modified differentially by PTMs, creating so-called positional isomers (i.e., double phosphorylation, one isomer with positions Y64 and T98, and one isomer with positions Y64 and Y120).Knowledge of the relative abundances of these separate positional isomers is highly relevant, as researchers uncovered many thousands of phosphorylation sites many of which have no clear biological function. Uncovering their relative abundance will assist in determining the relative importance of each modification site.Highlights- ETD fragmentation reveals the presence of positional isomers.- For proteins up to 40 kDa these positional isomers can accurately be quantified.- For in-vitro phosphorylated BoraNT a wide array of positional isomers is revealed.- Use of Fragment ion FDR levels improve the quality of extracted stoichiometries.
## In brief
Here we determine the quantities of the same protein modified differentially by PTMs, creating so-called positional isomers (i.e., double phosphorylation, one isomer with positions Y64 and T98, and one isomer with positions Y64 and Y120).
Knowledge of the relative abundances of these separate positional isomers is highly relevant, as researchers uncovered many thousands of phosphorylation sites many of which have no clear biological function. Uncovering their relative abundance will assist in determining the relative importance of each modification site.
## Highlights
- ETD fragmentation reveals the presence of positional isomers.
- For proteins up to 40 kDa these positional isomers can accurately be quantified.
- For in-vitro phosphorylated BoraNT a wide array of positional isomers is revealed.
- Use of Fragment ion FDR levels improve the quality of extracted stoichiometries.
## Quantifying positional isomers (qpi) by top-down mass spectrometry
Andrea M. Brunner 1,2 , Philip Lössl 1,2 , Paul P. Geurink 3 , Huib Ovaa 3, † , P. Albanese 1,2 , A. F. Maarten Altelaar 1,2 , Albert J. R. Heck 1,2 , and Richard A. Scheltema 1,2,* Proteomics has exposed a plethora of posttranslational modifications, but demonstrating functional relevance requires new approaches. Top-down proteomics of intact proteins has the potential to fully characterize protein modifications in terms of amount, site(s), and the order in which they are deposited on the protein; information that so far has been elusive to extract by shotgun proteomics. Data acquisition and analysis of intact multimodified proteins have however been a major challenge, in particular for positional isomers that carry the same number of modifications at different sites. Solutions were previously proposed to extract this information from fragmentation spectra, but these have so far mainly been limited to peptides and have entailed a large degree of manual interpretation. Here, we apply high-resolution Orbitrap fusion top-down analyses in combination with bioinformatics approaches to attempt to characterize multiple modified proteins and quantify positional isomers. Automated covalent fragment ion type definition, detection of mass precision and accuracy, and extensive use of replicate spectra increase sequence coverage and drive down false fragment assignments from 10% to 1.5%. Such improved performance in fragment assignment is key to localize and quantify modifications from fragment spectra. The method is tested by investigating positional isomers of Ubiquitin mixed in known concentrations, which results in quantification of high ratios at very low standard errors of the mean (<5%), as well as with synthetic phosphorylated peptides. Application to multiphosphorylated Bora provides an estimation of the so far unknown stoichiometry of the known set of phosphosites and uncovers new sites from hyperphosphorylated Bora.
Posttranslational modifications (PTMs) are critical in regulating protein stability, localization, activity, and interactions with other biomolecules. In the cellular environment, the number of PTMs and associated sites are often enzymatically modulated to fine-tune these activities. This creates distinct PTM configurations termed proteoforms [bib_ref] How many human proteoforms are there, Aebersold [/bib_ref] , potentially carrying the same number of modifications at different sites resulting in the same mass termed positional isomers [bib_ref] Deciphering the interplay among multisite phosphorylation, interaction dynamics, and conformational transitions in..., Lössl [/bib_ref]. Dysregulation can however occur, which has been linked to inter alia neurodegenerative diseases and tumorigenesis [bib_ref] post-translational modification: Deregulated in tumorigenesis, Dai [/bib_ref] , creating a pressing need for methods to decipher and quantify the occurring combinations of PTMs. Bottom-up proteomics has uncovered several thousands of PTM sites, but quantification of proteoforms and positional isomers is much more difficult to achieve [bib_ref] Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and Ser/Thr-based..., Sharma [/bib_ref]. Detection occurs at the peptide level giving a single value per modification site, which does not provide the number and co-occurrence of PTMs located on a given proteoform [fig_ref] FIG. 1: Schematic illustration of the limitations of bottom-up MS and the benefits of... [/fig_ref]. Top-down proteomics focuses on intact proteins, ensuring that all proteoforms are measured in parallel and can be further investigated independently [bib_ref] Progress in top-down proteomics and the analysis of proteoforms, Toby [/bib_ref] [bib_ref] Top-down proteomics: Ready for prime time?, Chen [/bib_ref]. The intact protein masses can be extracted from the data, from which the type and number of modifications can be inferred by the mass shifts they induce [bib_ref] Dissecting ribosomal particles throughout the kingdoms of life using advanced hybrid mass..., Van De Waterbeemd [/bib_ref]. In some cases, when using chromatographic separation, different peptide isomers have been demonstrated to separate in elution time and can be inspected individually to uncover the ratios between the different isomers [bib_ref] A mixed integer linear optimization framework for the identification and quantification of..., Dimaggio [/bib_ref]. However, in cases where this is not achieved or chromatography is not employed, pinpointing the position of the modifications for a proteoform can be achieved by mass selection and fragmentation of the isoform mixture. The resulting sequence-specific fragment ions in combination with the modification-induced mass shifts indicate the PTM locations [fig_ref] FIG. 1: Schematic illustration of the limitations of bottom-up MS and the benefits of... [/fig_ref]. While characterization of known proteoforms has been described [bib_ref] Quantitative analysis of modified proteins and their positional isomers by tandem mass..., Pesavento [/bib_ref] [bib_ref] Unabridged analysis of human histone H3 by differential top-down mass spectrometry reveals..., Zheng [/bib_ref] [bib_ref] Analyses of histone proteoforms using front-end electron transfer dissociationenabled Orbitrap instruments, Anderson [/bib_ref] [bib_ref] Top-down and middle-down protein analysis reveals that intact and clipped human histones..., Tvardovskiy [/bib_ref] [bib_ref] Protein species-specific characterization of conformational change induced by multisite phosphorylation, Pan [/bib_ref] , applications to discovery type experiments aimed at positional isomer characterization have been limited due to the challenges involved in separating different isomers and interpreting the acquired data [bib_ref] Comprehensive characterization of AMP-activated protein kinase catalytic domain by top-down mass spectrometry, Yu [/bib_ref] [bib_ref] Benchmarking multiple fragmentation methods on an Orbitrap fusion for top-down phospho-proteoform characterization, Brunner [/bib_ref]. Excellent work has already been done on developing algorithms capable of extracting the sequence identity of proteins from fragmentation spectra from full LC-MS runs, with well-described approaches to control the false discovery rates (FDRs) at the spectrum-, protein-, isoform-, and proteoform level, and additional algorithms are proposed to facilitate the identification of many different PTMs on highly modified proteins [bib_ref] Accurate estimation of contextdependent false discovery rates in top-down proteomics, Leduc [/bib_ref] [bib_ref] Informed-proteomics: Open-source software package for top-down proteomics, Park [/bib_ref]. As these algorithms offer no FDR control at the fragment annotation level [bib_ref] ProSight PTM 2.0: Improved protein identification and characterization for top down mass..., Zamdborg [/bib_ref] [bib_ref] Identification of ultramodified proteins using top-down tandem mass spectra, Liu [/bib_ref] [bib_ref] MASH suite pro: A comprehensive software tool for top-down proteomics, Cai [/bib_ref] , identification of positional isomers to date requires mostly manual interpretation. This has led researchers to approach the problem from an expert user-directed angle, where the presence of specific satellite ions is used to uniquely assign the different PTMs on peptides and small proteins [bib_ref] Data processing algorithms for analysis of high resolution MSMS spectra of peptides..., Guan [/bib_ref].
The fragmentation scans of positional isomers (i.e., proteins with the same set of modifications distributed at different locations) are highly multiplexed with specific fragments simultaneously present with and without the modification. This situation prevents automatic software to identify modification sites occluded by the modification site producing the largest number of modification carrying fragments. Even though the multiplexed spectra are hard to interpret, they also provide an opportunity to quantify the different positional isomers provided the set of positional isomers can be mass selected without interferences from other proteoforms, and the used fragmentation method does not induce structural biases in the fragment ion intensity depending on the position of the modification. This requires an ideal fragmentation technique that additionally does not incur modification losses that potentially lead to unspecific effects in the detected intensity levels of the fragment peaks, lowering the precision of the calculations or potentially even obfuscating the presence of the positional isomers. In addition, the specific modification can potentially induce unpredicted or even unpredictable effects, and modifications other than phosphorylation as investigated here should be verified in terms of accuracy. Techniques such as collision induced dissociation and higherenergy collisional dissociation readily produce such losses, and in previous studies these were also observed for ultraviolate photodissociation [bib_ref] Implementation of ultraviolet photodissociation on a Benchtop Q exactive mass spectrometer and..., Fort [/bib_ref]. Apart from this, the ideal fragmentation technique also leads to excellent protein sequence coverage for proteins of any size, as missing parts in the sequence coverage prevent detection of positional isomers completely. Currently, electron capture dissociation (ECD) / electron transfer dissociation (ETD) represents the best option for this set of requirements. Indeed, it has previously been reported for ECD spectra that fragment intensities correspond to their respective intact ion intensities, allowing determination of the stoichiometry of distinct histone proteoforms with known acetylation modifications [bib_ref] Quantitative analysis of modified proteins and their positional isomers by tandem mass..., Pesavento [/bib_ref]. Additionally, it was demonstrated that ECD predominantly retains the expected intensity levels for modifications such as phosphorylation, except close to the modified site [bib_ref] The impact of phosphorylation on electron capture dissociation of proteins: A top-down..., Chen [/bib_ref]. In those cases where at least three amino acids are present between the positional isomers, provided those further away from the site have the correct ratio, this suggests that the ratio expressing the stoichiometry of each positional isomer can be interpreted for many positional isomers in completely unknown systems. Of note, this is only true if the indicated preconditions for ETD apply for the modification in question. To perform this task we developed a novel workflow termed "Quantifying Positional Isomers" (QPI), which utilizes a unique combination of ETD, previously reported to have minimal loss of PTMs when fragmenting peptides or proteins, and extensive bioinformatics approaches. Additionally, the technique works best with highly charged ions and full proteins provide many opportunities to carry charges. The data analysis component is combined in the application QPI, which assigns fragments and localizes modifications and uses that information to automatically calculate the stoichiometry of the positional isomers. Using Ataxin as a model system, we show that ETD fragmentation is sufficiently "soft" to leave the phosphorylation modifications in place during fragmentation, producing no losses and opening the possibility to quantify positional isomers with this fragmentation technique. With three Ubiquitin molecules differentially modified in distinct positions and mixed in well-defined ratios, we show that stoichiometries can be obtained from fragmentation spectra for nonlabile modifications representing the minimal case where the method should work. Of note, modifications carrying charges or of large size can have influence on the ratios; e.g., it was demonstrated that the presence of a phosphorylation site can have a suppressing effect on the fragmentation intensity of the fragment containing this site [bib_ref] The impact of phosphorylation on electron capture dissociation of proteins: A top-down..., Chen [/bib_ref]. We go on to show that ETD on phosphorylated peptides produces the expected ratio from the fragmentation spectra when mixed in predefined ratios for five amino acids (or fragment intensities) between the positional isomers; we anticipate distortions when this number drops below 5, and it becomes impossible to calculate a correct ratio when the number drops below 3. Finally, we apply the approach to uncover the positional isomers of in vitro phosphorylated Bora (a mitotic regulator) and detect their relative abundances, for which we go up to three phosphorylation sites. Incredibly, we can detect for doubly phosphorylated Bora six distinct positional isomers and provide estimations for the quantities for each. For triply phosphorylated Bora the situation is even more complex, as we find 13 distinct positional isomers. As it currently stands, the technique still has limitations with the complexity introduced with such numbers of isomers, and we are not (yet) able to extract estimations of all individual abundances. It is evident that the current generation of deisotoping procedures introduce additional variation at the level of incorrectly assigned monoisotopic masses, splitting peaks, and other problems leading to incorrect intensity values. We deal with the aforementioned problems by introducing grouping of different positional isomers to a single ratio. Even though from this the exact contributions of each positional isomer cannot be extracted, we can narrow the highest abundant positional isomer as one with two positions with high confidence combined with one of four positions. Such knowledge can help to direct and specifically limit mutation studies to relatively few positions to help determine the biological relevant combination of sites.
## Experimental procedures
## Chemicals and reagents
Formic acid (FA) was purchased from Merck; acetonitrile (ACN) and methanol (MeOH) were purchased from Biosolve. The catalytic subunit of cAMP-dependent Protein Kinase (PKA) used to in vitro phosphorylate proteins was purchased from NEB (Ipswich). We utilize three different, combinatorially expressed proteins that were previously characterized in detail. Ataxin 562-815 was provided by Prof. Luc Brunsveld (Technische Universiteit Eindhoven) and was prepared as previously described [bib_ref] Expression and purification of ataxin-1 protein, Husain-Ponnampalam [/bib_ref]. Heavy labeled ubiquitin proteoforms were synthesized by solid-phase peptide synthesis according to the reported procedure [bib_ref] Chemical synthesis of ubiquitin, ubiquitin-based probes, and diubiquitin, El Oualid [/bib_ref]. Fmoc-labeled heavy valine ( 13 C 5 , 15 N) (Cortecnet) was introduced at the appropriate positions. And finally, for Bora we utilized the recorded data as presented in Lössl et al.
## Kinase reactions
Ataxin 562-815 was phosphorylated through incubation with PKAc in 50 mM Tris-HCl, 10 mM MgCl2, 1 mM ATP pH7 for 1 h at 30 - C. Reactions were quenched by addition of 20 mM EDTA pH 8.0. Prior to top-down MS analysis, phosphorylated proteins were denatured by buffer exchange to 0.1% FA using 5 kDa molecular weight cutoff centrifugal filter units with polyethersulfone membranes (Sartorius).
## Mass spectrometry
All data was acquired on a Thermo Scientific Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific) in direct infusion experiments. Ataxin 562-815 was diluted to a final concentration of 5 μM in 50% ACN, 1% FA, and ubiquitin to 2.5 μM in 50% MeOH, 1% FA. Protein samples were sprayed at a flow rate of 1 μl/min. Data were acquired in the Orbitrap mass analyzer at a resolution of 120,000 for Ataxin 562-815 and 240,000 for ubiquitin (full width at half-maximum, FWHM) in intact protein mode (2 mTorr ion-routing multipole (IRM) pressure). The presence and number of modification(s) were determined based on the monoisotopic mass of the isotopes detected in the MS1 spectra. Of each sample and each setting change, a set of triplicate measurements was recorded; each replicate was recorded as follows. Specific proteoforms were isolated with the mass selecting quadrupole and an isolation width of 1 Th for Ataxin 562-815 and 2 Th for ubiquitin. We selected the more narrow isolation window 1 Th for Ataxin to prevent any coisolation of peaks in close proximity due to the elevated charge states, which represent a different phosphorylation state. The 37+ charge state of intact, nonphosphorylated Ataxin 562-815 at 748Th and the 38+ charge state of doubly phosphorylated Ataxin 562-815 at 733 Th were subjected to ETD fragmentation with 2, 4, and 6 ms ion/ion reaction time. The 11+ charge state of ubiquitin at 778 Th was subjected to ETD fragmentation with 6 ms reaction time. MS2 spectra were acquired using a mass range of 150 to 2000 m/z. A total of 500 microscans were summed for Ataxin 562-815 and 100 for ubiquitin spectra.
# Data analysis
Deconvolution of intact protein and top-down spectra was performed in Protein Deconvolution 4.0 (Thermo Scientific) using XTRACT with a signal-to-noise ratio (S/N) threshold of 1.1, a fit factor of 40%, and a remainder threshold of 15%. Deconvoluted spectra were analyzed with QPI, a software we developed in-house for proteoform characterization written in C#. It takes as input (i) the protein sequence and (ii) the deconvoluted top-down spectra. Through the described procedures the fragment ions are assigned and identified at FDR rates <1%. Assignments are organized in PTM ladders, which are constructed by progressively adding the delta mass of the desired modification to the fragment masses. Visualization in a heatmap format allows the operator to inspect and manually pinpoint the modification sites. After calculating the ratios between n and n + 1 modifications in the heatmap, a simple t-test can be applied to determine whether a positional isomer exists between two possible locales for the PTM under investigation.
# Results and discussion
## Improving the confidence of fragment assignment in top-down mass spectra
Confident PTM localization is ideally based on protein sequencing data with no false fragment assignments and high sequence coverage. This requirement is needed as, in extreme cases, incorrectly assigned fragments potentially obfuscate the correct position leading to an incorrect assignment. Furthermore, missing fragments can make it impossible to pinpoint the site when multiple acceptor residues are closely spaced in the sequence. False positives however cannot be readily excluded and, to our knowledge, FDR calculation at the fragment ion level was not addressed in standard top-down MS software. To benchmark the FDR of fragment ion assignments in top-down fragmentation spectra we used the well-characterized AXH domain of human Ataxin-1 (Ataxin 562-815 ) and searched its fragment spectra against 200 different scrambled sequences to calculate the percentage of the sequence that can randomly be explained (defined as the median sequence coverage of the 200 searches). This can readily be automated with standard string randomization approaches accessible in any programming language. By matching the theoretical fragments from these randomized sequences to the deconvoluted spectra, the sequence coverage can be calculated. The fragment level FDR is then defined as the percentage of the full sequence length that can on average be covered by 200 randomized sequences. Based on this FDR metric, we find that using standard parameters (i.e., setting mass tolerance to 10 ppm, combining multiple spectra, and allowing fragment ions a, b, c, x, y, and z with neutral losses water, ammonia, and hydrogen activated) results in an unacceptably high FDR of 9% at 70% sequence coverage for unmodified Ataxin 562-815 , partially resolvable by using a more stringent mass cutoff for fragment matching . For optimization of the FDR, we initially sought to use a fragment ion intensity cutoff; however, searches against scrambled databases revealed no substantial correlation between false-positive assignments and intensity .
Exclusively using the fragment ion types actually present in the fragmentation spectrum is a more effective measure. To determine the relevant ion types, we utilized a "frequent flyer" algorithm, which determines those ion types actually encountered in each individual spectrum [bib_ref] Ab initio prediction of metabolic networks using Fourier transform mass spectrometry data, Breitling [/bib_ref]. This strategy works well for top-down fragmentation spectra as each spectrum contains thousands of annotatable fragment ions. For instance, ETD fragmentation of Ataxin 562-815 , which at 27 kDa is a mid-sized protein in top-down analyses, gener-ated~2000 fragment ions. The ion types c, c − 1, y, z, and z + 1 were found to be significantly present based on a significance A test (31) as the fragment ion types of interest . This is in line with previous reports of peptide ETD MS that showed that c, y, and z ions were predominantly generated by ETD fragmentation [bib_ref] Statistical analysis of peptide electron transfer dissociation fragmentation mass spectrometry, Chalkley [/bib_ref]. Re-searching the Ataxin 562-815 data allowing only for the defined ion types c, c − 1, y, z, and z + 1 reduces the FDR to 4% for the same sequence coverage. A further reduction of the search space, and thus the FDR, is achieved by higher stringency on the mass error tolerance used for fragment ion matching. This however requires mass calibration of the recorded mass spectra to remove systematic errors introduced during data acquisition. Analogous to previously reported post data acquisition approach for mass calibration [bib_ref] Software lock mass by twodimensional minimization of peptide mass errors, Cox [/bib_ref] [bib_ref] Increasing the mass accuracy of high-resolution LC-MS data using background ions -a..., Scheltema [/bib_ref] , QPI calculates the mass deviations from the entirety of annotatable fragment ions. This results in a normally distributed population demonstrating no m/z range dependency, and any systematic mass deviation within the population can be removed by correcting the median value by this shift (supplemental [fig_ref] FIG. 1: Schematic illustration of the limitations of bottom-up MS and the benefits of... [/fig_ref]. Additionally, the normally distributed full population of mass deviations allows for the calculation of the allowed mass tolerance, which we define as 3x standard deviation . Generally this results in an acceptable mass precision of <3 ppm, providing a very stringent filter for fragment ion assignment, which further reduces the FDR to~2% for a decreased sequence coverage of 62%. Finally, we hypothesized that false-positive assignments of fragment ions can be reduced by considering technical replicate scans. Accordingly, a true-positive fragment should be found in more replicate scans than false-positive random matches, which can be exposed by a "majority voting" algorithm filtering the annotations of all technical replicates based on a minimum occurrence rate. We find that combining replicate ETD scans, i.e., subsequent scans from a single run, increases protein sequence coverage in a similar manner as reported previously for combining scans of multiple different fragmentation methods [bib_ref] Benchmarking multiple fragmentation methods on an Orbitrap fusion for top-down phospho-proteoform characterization, Brunner [/bib_ref]. However, we find here that this practice also increases the FDR, potentially impeding unambiguous modification site localization. Indeed, starting with ã 2% FDR and 65% sequence coverage for single ETD fragmentation scans, the FDR initially almost doubled to~4% when fragment ions of three replicate scans were combined. Only considering fragment ions that occur in at least two out three technical replicates decreased the FDR rapidly to~1% and still improved the sequence coverage as compared with single scans from 65% to 70%. Filtering for fragment ions that are present in all replicates further reduced the FDR to <1% but also caused a substantial drop in sequence coverage to 55% . Therefore, we chose to accept fragment ions present in at least two out of three replicates, as this represents the best compromise between boosting sequence coverage and reducing FDR.
The improved sequence coverage at low false-positive rates of this compromise improves the likelihood of correctly localizing PTMs as misassignments are very rare and holes in the sequence coverage will not complicate localization. As such, automated approaches are expected to generate more reliable results.
## Using ptm ladders for site localization
In contrast to other approaches implementing automatic modification site assignment [bib_ref] ProSight PTM 2.0: Improved protein identification and characterization for top down mass..., Zamdborg [/bib_ref] [bib_ref] MASH suite pro: A comprehensive software tool for top-down proteomics, Cai [/bib_ref] [bib_ref] ProSight lite: Graphical software to analyze top-down mass spectrometry data, Fellers [/bib_ref] , in QPI, an increasing number of modifications are added to both the N and C termini of the input protein sequence. This leads to PTM ladders for the N and C termini individually, defined as sequence stretches covered by fragment ions with no modification, followed by stretches covered by ions with one modification, with two modifications, and so forth up to the maximum number of possible modifications determined from the precursor mass. The resulting modification ladder shows at which amino acid residue(s) a new modification occurs (see supplemental . Demonstrating this procedure for doubly phosphorylated Ataxin 562-815 , where the modification ladders show no signs of modification loss during ETD fragmentation, both phosphorylation sites can be unambiguously assigned to Ser214 and Ser236 [fig_ref] FIG. 3: left panel [/fig_ref].
Automatic site localization however requires a likelihood score to separate the correct site from all possibilities, which can result into many thousands of combinations. To illustrate, for our simpler system of doubly phosphorylated Ataxin 562-815 (considering that S, T, and Y can be phosphorylated) a total of 703 unique combinations of phosphorylatable amino acids can be made. To implement such a score we utilize a probability calculation established for peptide fragmentation spectra [bib_ref] Improved peptide identification in proteomics by two consecutive stages of mass spectrometric..., Olsen [/bib_ref] and modify it for the PTM ladders generated by top-down fragmentation spectra. The adaptation consists of application of the score to each ladder-or stretch between modifiable residues, resulting in stretches approximately the length of the peptides the score originally was developed forand finally combination of all the individually calculated scores. The basis for this calculation is a probability for a theoretical fragment to match by chance, which we approximate by calculating the average number of peaks (X) per 100 Da in the spectrum, as opposed to using the number from Data curation and fragment assignment in QPI results in low false-positive assignments and high protein sequence coverage. A, effect of mass tolerance as a parameter on the FDR rate. B, intensity (in-)dependence of forward and scrambled searches. C, frequent flyer fragment-type definition automatically extracts which fragment ions to consider in fragment assignments. D, the raw mass distribution of a data set generated by ETD top-down MS of intact Ataxin 562-815 is shown in blue, entirely falsified spectra created by adding and subtracting, respectively, 5 Da to/from every monoisotopic peak in every real spectrum in green and red (left panel). Data-driven mass calibration corrects the median value by the systematic mass deviation (mid panel) and reduces the accepted mass tolerance to <3 ppm (right panel). E, extensive data curation and majority voting for combination of multiple spectra result in FDRs <1% and the replicates provide the benefit of higher sequence coverage.
the TopX filtering. The theoretical chance then equates to "X/ 100" as previously described. To determine the probability of a particular combination of modification sites, we calculate the match probability for each stretch at the number of modifications based on the number of matched and theoretical peaks for that stretch. For example, in the case of the combination of 214 and 236, we calculate the probabilities: for 0 modifications, residues 0 to 214; for one modifications, residues 215 to 236; and for two modifications, residues 237 to 255. All these probabilities are multiplied to get to the final result. This however ignores small stretches "in-between" a main stretch. For example, instead of residue 236 the residue 231 can also be modified. As the difference is relatively small and potentially underrepresented with matches it can happen that this skews the end score to favor the shorter stretch. To resolve this, we correct each individual p-value by dividing it by the match probability p-value of the previous stretch prior to multiplication. Application to Ataxin 562-815 indeed finds the best p-value for the combination serine 214 and serine 236. This combination scores 1.8e-142 and is followed by 214 + 231 at 1.0e-141, clearly separating the two, and thereby the correct position pair is selected [fig_ref] FIG. 3: left panel [/fig_ref].
## Quantifying positional isomers
The ultimate goal is the quantification of positional isomers. They are visible in the PTM ladders as sequence positions carrying different numbers of modifications and their relative ratios potentially reveal the stoichiometry of each of the positional isomers. To demonstrate this, we used synthetic ubiquitin with a heavy isotope-labeled Valine positioned at residue 17, 26, or 70 (A, B, and C respectively) as a model system. Measuring these positional isomers independently results in the anticipated sequence ladders for positional isomers A and C individually and localizes the modifications at the second modifiable position for protein A (Val17) and the fourth modifiable position for protein C (Val70) (supplemental [fig_ref] FIG. 3: left panel [/fig_ref]. To verify that the position of the heavy labeled Valine does not result in differences in the intensity levels at their location point in the sequence, we investigated the fragmentation spectra. From the mirror plot, it is evident that the overall fragmentation behavior is highly similar for both positions even though these are two separate fragmentation scans (supplemental [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. After matching the peaks for the same sequence position, we observe an excellent correlation coefficient of 0.97 between the detected fragment intensities (supplemental [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. To verify whether this holds for the combination of ETD and phosphorylation, we tested the quantitation on synthesized phosphopeptides. From the mixing ratio and the detected ratio from the fragment spectra, we conclude that the fragment ion intensities are not significantly affected by either position or phosphorylated residue (supplemental .
Mixing positional isomer A and C in a 1:1 ratio in our Ubiquitin model system results in modified and unmodified fragments in the sequence stretches 17 to 26 and 26 to 70, with the N-terminal modified fragments belonging to isomer A and the unmodified to isomer C [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. Their respective fragment intensity values are indicative of their amount and were used to calculate a ratio expressing the stoichiometry of the positional isomers [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref] result in a value of 1 (standard error of the mean (SEM) < 0.05%), reflecting the expected 1:1 mixing ratio. Extending the mixing ratios of the isomers up to 64:1 results in an experimental calibration curve matching the theoretically expected quantitative results, exhibiting stable SEM values (~2%) up to a ratio of 32:1 [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. For the 64:1 mix, ratio compression starts to occur, a phenomenon previously described for MS-based relative quantitation of peptides [bib_ref] Evaluation of the variation in sample preparation for comparative proteomics using stable..., Zhang [/bib_ref] [bib_ref] Comparing SILAC-and stable isotope dimethyl-labeling approaches for quantitative proteomics, Lau [/bib_ref] , and the detected fragment abundances of the low concentration positional isomer become less reliable.
In many cases, positional isomers will involve other residues than the first and last modified residue of the modification ladder, e.g., a heavy labeled Val26 in addition to the modified Val17 and Val70 in case of ubiquitin. The presence of additional isomers cannot be determined based on the maximal number of assigned fragments. However, as their presence will affect the ratio of modified and unmodified fragments in the individual stretches between modified sites, this ratio can be used to identify whether positional isomers are present. To achieve this, a t-test compares the fragment ion ratios of all potentially modified sites. A statistically significant difference (p-value < 0.05) between the fragment ion ratios of individual sequence stretches is indicative for an additional modification site, hence the presence of additional positional isomers. In the case of the 1:1 mixture of positional isomers A and C, we find a p-value of 0.97 and conclude that Val26 is not modified, and no further isomers are present [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. To illustrate the reverse, we added positional isomer B (modified at Val26) to the ubiquitin protein mixture. The resulting situation is visualized for N-terminal fragments, showing that the fragment ion complexity has substantially increased [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. Modified fragments in the sequence stretch 17 to 25 belong to isomer A and unmodified fragments to isomers B and C, whereas modified fragments in the sequence stretch 26 to 69 belong to isomers A and B and unmodified fragments to isomer C. The contribution of A, B, and C can thus again be extracted from the fragment ion intensities. When considering a 1:1:1 mixture the detected ratios display, as expected, different behavior over the stretches 17 to 26 and 26 to 70. Comparing both stretches using the t-test results in a p-value of 4.5e-11, showing that their ratios are significantly different and therefore Val26 is modified [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref] ; bottom-left panel). In our data with triple mixing ratios (1:1:1, 1:4:8, 4:1:8, and 8:1:4), this approach retrieves all the expected ratios [fig_ref] FIG. 4: Quantification of positional isomers in the Ubiquitin model system [/fig_ref]. It should be noted that this procedure is not limited to two or three positional isomers. Rather, every additional positional isomer gives rise to a new unknown variable, for which a new stretch is introduced. The ability to solve this mathematical problem thus mainly depends on a sufficiently differing ratio and on the number of provided input parameters and the quality of the experimental data (see supplemental .
## Application to phosphorylated bora
Next, we reinvestigated data from the 17.5 kDa N-terminal domain of the mitotic regulator Bora (BoraNT), which we reported before [bib_ref] Deciphering the interplay among multisite phosphorylation, interaction dynamics, and conformational transitions in..., Lössl [/bib_ref]. As shown previously, BoraNT can be phosphorylated in vitro by Aurora kinase A, resulting in a variety of positional isomers starting with its doubly phosphorylated proteoform [bib_ref] Benchmarking multiple fragmentation methods on an Orbitrap fusion for top-down phospho-proteoform characterization, Brunner [/bib_ref]. With 30 phosphorylatable residues in the 156 amino acid sequence, some in close proximity, proteoform assignment is a difficult task, which we attempt to resolve with our approach. The site of the singly phosphorylated proteoform was correctly localized to Ser64 (see and supplemental . As the PTM ladder shows a single step, without exhibiting overlap of non-modified and singly modified fragments, we conclude 100% occupancy for this phosphorylation site. The doubly phosphorylated proteoform revealed 100% occupancy of Ser64, but the second phosphate moiety is distributed over five additional sites (Ser48, Ser91, Thr120, Ser128, and Thr149). QPI resolved and quantified these five distinct positional isomers with relative abundances ranging from 15% to 29% (see and supplemental . From the overlapping no, singly, and doubly phosphorylation carrying fragments, we conclude the indicated six modified sites (based on the procedure described above). The combination of these results in 15 theoretically possible proteoforms representing 15 unknowns. The seven data points achieved experimentally do not allow proteoform quantification; however, from the singly phosphorylated Bora it is evident that S64 has 100% occupancy. Applying this constraint reduces the number of unknowns to five potential proteoforms, which can be quantified with the seven experimentally determined data points. Tackling the next level of complexity, 3x phosphorylated BoraNT, an additional site was observed at Ser12, resulting into 13 distinct positional isomers for the triply phosphorylated Bor-aNT (see . The combination of these results in 35 theoretically possible proteoforms, i.e., 35 unknowns. The seven data points achieved experimentally do not allow proteoform quantification. As before, the data show that S64 is always phosphorylated and phosphorylation at S12 and S48 and S128 and T149 are mutually exclusive. The exclusivity between S12 and S48 arises from the requirement that S64 is always present and only fragments with three modifications after S48 are detected in the set of Cterminal fragments. Similarly, exclusivity between S149 and S128 arises from the requirement that S64 is always present and only fragments with one modification up to S120 are detected (i.e., either S149 or S128 is present). Applying these constraints leads to 13 potential proteoforms that cannot be quantified with the seven experimentally determined data points, although they could be quantified if all 14 theoretically possible data points had been achieved. However, grouping positional isomers that cannot be disentangled indicates the maximum relative abundance of the specific proteoforms . The SEM of 10% indicates that for this system, less accurate estimations can be made for this sample, which also exceeds our ability to quantify all positional isomers. The combination of PTMs and their cross talk are critical in cellular processes. To fully live up to its potential, proteomics needs to decipher which combinations of modifications occur on a protein in a defined condition. This would allow researchers to elucidate how physiological and pathological states impact the dynamic regulation of PTMs and the cross talk among PTM sites. QPI attempts to address precisely this need with its ability to identify proteoforms, including positional isomers, and-within the theoretical limitations-quantify them. The proof-of-concept applications to isotopically labeled ubiquitin and (hyper)phosphorylated BoraNT demonstrate that the QPI approach can provide estimations of the quantities of positional isomers, but also reveal the intrinsic complexity due to the exploding numbers of co-occurring isoforms when the number of attached PTMs increases. In addition, it is worth noting that some PTMs such as phosphorylation can distort the fragment intensities close to the modification site. Even though our approach is expected to be robust in those cases where at least five amino acids provide fragment intensities next to the modification site due to the use of the median, the ratios with SEMs exceeding 10% (determined from the high noise, triply phosphorylated Bora experiment presented here) should be treated with caution. This also means that the method is not generally applicable, e.g., in cases where the number of amino acids or fragment intensities becomes too low to calculate a reliable ratio. In cases where the ratio cannot be calculated or where the SEM is excessively large, orthogonal methods are required to calculate the correct ratios. Further investigations into advancing the investigation into positional isomers are therefore required as, as far as we are aware, no methods exist to date.
The theoretical limits of the approach apply to how much data can be extracted from the spectrum and how many "unknowns" are present as positional isomers. When the number of unknowns exceeds the number of data points that can be extracted from the spectrum, the calculation cannot be resolved and additional data must be collected. We envision that this is possible with higher-resolution separation methods able to disentangle positional isomers and/or with MS instrument improvements such as on-the-fly targeted MS3 sequencing steps of internal fragments, which together may solve the quantification of proteoforms that exceed the current decipherable number of PTM sites and combinations. To partly deal with the current level of data quality, we introduced grouped quantitation where for a group of positional isomers the stoichiometry can be extracted. How each individual positional isomer behaves will need to be disentangled by other methods, but the required work is severely limited to a small set of positional isomers. Another limitation of the approach is that only a single modification can be taken into account. As for multiple modifications, the theoretical limits are even more readily achieved. There are software solutions available that can handle multiple modifications, but restrict there on the polypeptide sequence length. As QPI is designed to deal with full proteins, this currently is not feasible.
## Data availability
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http:// S91/T120/S128/T149 S91/T120/S128/T149 S91 S12 S48 S12 S48 S64 S12 S48 S64 S12 S48 S128 T149 T120 S64 S64 T120/S128/T149 S128/T149 S91 T120 3x phosphorylated Bora proteoforms 67-76%
## 26-33%
FIG. 5. Quantification of positional phospho-isomers of Bora NT . A, in the singly modified protein, one proteoform is present with a phosphorylation site at S64. B, in the doubly modified protein, five additional sites are phosphorylated in combination with S64. All positional isomers are present in roughly similar relative abundances. C, in the triply phosphorylated protein, the exact relative abundance of the individual positional isomers cannot be resolved; however, grouped quantification allows extraction of a maximum relative abundance of each proteoform. Here 67 to 75% are proteoforms comprising pS64 and either pS12 or pS48 and a C-terminal modification (bold gray lines); and 26 to 33% are proteoforms comprising pS64 and either pS91 or pT120 and a C-terminal modification.
proteomecentral.proteomexchange.org) via the PRIDE partner repository with the data set identifier PXD012084. Furthermore, the software is available in the supplementary material as zip file containing the executables, a manual, and an example data set.
Supplemental data -This article contains supplemental data [bib_ref] High-Throughput assessment of kinome-wide activation states, Schmidlin [/bib_ref] Acknowledgments -The Netherlands Organization for Scientific Research (NWO) supported this research through funding of the large-scale proteomics facility Proteins@Work Conflict of interest -The authors declare no competing interests.
Abbreviations -The abbreviations used are: ECD, electron capture dissociation; ETD, electron transfer dissociation; FDR, false discovery rate; SEM, standard error of the mean.
[fig] FIG. 1: Schematic illustration of the limitations of bottom-up MS and the benefits of top-down MS for proteoform characterization. The defined site(s), type(s), and number of modification(s) on a protein result in a specific proteoform. Proteoforms with the same number of a certain modification type but different site(s) represent positional isomers. A, proteoforms cannot be characterized by traditional bottom-up proteomics, as their enzymatic digestion into peptides makes it impossible to relate peptide fragment information back to their specific proteoforms. B, by topdown MS of intact proteoforms, the type and number of modification(s) can be extracted from the MS1 precursor spectrum, the modification site(s), and abundance of positional isomers from the MS2 fragment spectrum. [/fig]
[fig] FIG. 3: left panel). Considering both N-and C-terminal fragment ions separately adds a level of confidence to the ratio determination. The median ratios of the N-terminal fragment ion pairs (eight out of nine for stretch 17-25 and 32 out of 44 for stretch 26-69) and the C-terminal fragment ion pairs (seven out of nine for stretch 18-26 and 35 out of 44 for stretch 27-70PTM localization with sequence ladders. A, PTM ladders localize modification sites from the fragmentation spectra. B, PTM localization probabilities for doubly phosphorylate Ataxin. [/fig]
[fig] FIG. 4: Quantification of positional isomers in the Ubiquitin model system. A, for positional isomers modified and unmodified fragments are observed in the PTM ladders for individual sequence positions (top-left), the ratio of which is reflective of the mixing ratio (in this case 1-to-1; bottom-left). Performance on a range dilution of mixing ratios shows correct ratios up to extreme levels (right). B, introduction of an extra positional isomer causes higher complexity (top-left) and shifts in the ratios that can be detected by Student's t-test (bottom-left). Performance of various mixing ratios shows correct values also for more complex positional isomer arrangements (right; white line is the expected value). [/fig]
|
Effects of short-term fructooligosaccharide intake on equol production in Japanese postmenopausal women consuming soy isoflavone supplements: a pilot study
Background: Recent studies suggest that some of the clinical effectiveness of soy or daidzein, which is a type of isoflavone, may be attributed to a person's ability to produce equol from daidzein. Equol, which is a metabolite of one of the major soybean isoflavones called daidzein, is produced in the gastrointestinal tract by certain intestinal microbiota where present. Habitual dietary patterns may alter the intestinal bacterial profile, and influence the metabolism of isoflavones and the production of equol. Fructooligosaccharides (FOS) have a prebiotic activity as well as being a dietary fibre. The purpose of the present study was to determine whether FOS supplementation increases equol production in equol producers and stimulates equol production in equol non-producers in Japanese postmenopausal women. Methods: A soy challenge was used to assess equol-producer status prior to the start of the study in healthy postmenopausal Japanese women. The study involved 4 separate groups in randomised crossover design. First, subjects were classified as equol producers (n = 25) or non-producers (n = 18), and then they were randomly assigned to the FOS or control group. All subjects received a daily dose of 37 mg isoflavone conjugates in the capsule (21 mg aglycone form) and either FOS (5g/day) or sucrose as control, in a randomised crossover study design. Equol -production was assessed by testing the serum and urine before and after the 2-week supplementation period.Results: The analyses were conducted on 34 subjects completed the study, 21 (61.8%) were classified as equol producers, and 13 (38.2%) as non-producers. Significant differences were observed in the interaction effect of time × equol state after 1 week of intervention (p = 0.006). However there were no effects after 2 weeks of intervention (p = 0.516). Finally, in both equol producers and non-producers, FOS supplementation did not affect the serum equol concentration or the urinary equol to daidzein concentration ratios.Conclusions:We have reported that FOS intervention (5 g/day for 2 weeks) does not significantly modulate the capacity of intestinal microbiota to produce equol in postmenopausal Japanese women, in either equol producers or non-producers in this pilot study. Further larger investigations that explore the roles of specific intestinal microbiota in equol production will enable the establishment of dietary conditions that are required to enhance equol production.
# Background
Soybean isoflavones are structurally similar to oestrogen; they exhibit a weak affinity for oestrogen receptors, ERα and ERβ, eliciting mild oestrogenic-like activity in various tissues [bib_ref] Assaying the estrogenicity of phytoestrogens in cells of different estrogen sensitive tissues, Schmitt [/bib_ref]. On the other hand, the isoflavones genistein and daidzein have been shown previously to possess anti-oestrogenic activity in human breast cancer cells in vitro [bib_ref] Comparative study of oestrogenic properties of eight phytoestrogens in MCF7 human breast..., Matsumura [/bib_ref]. Compared with ERα, ERβ exhibits a greater binding affinity for isoflavone, whereas oestrogen binds to ER with equal affinity [bib_ref] Differential response of estrogen receptor alpha and estrogen receptor beta to partial..., Barkhem [/bib_ref]. The relatively selective binding of isoflavones to ERβ indicates that isoflavones may confer distinct clinical effects compared with oestrogen [bib_ref] Isoflavones made simple -genistein's agonist activity for the beta-type estrogen receptor mediates..., Mccarty [/bib_ref]. Moreover Hwang et al. reported that isoflavones may exert their effects as oestrogen antagonists in a high oestrogen environment, or they may act as oestrogen agonists in a low oestrogen environment [bib_ref] Isoflavone metabolites and their in vitro dual functions: they can act as..., Hwang [/bib_ref].
Isoflavones have received considerable attention because of their potential to prevent postmenopausal conditions such as cardiovascular disease [bib_ref] The differential effect of the phytoestrogen genistein on cardiovascular risk factors in..., Villa [/bib_ref] , osteoporosis [bib_ref] Effect of soy isoflavone extract supplements on bone mineral density in menopausal..., Taku [/bib_ref] , and hormone-dependent cancers [bib_ref] Soy isoflavone supplementation for breast cancer risk reduction: a randomized phase II..., Khan [/bib_ref]. Daidzein, a major soybean isoflavone, is metabolized to equol in the gastrointestinal tract by intestinal microbiota. Recent studies suggest that some of the clinical effectiveness of soy or daidzein, which is a type of isoflavone, can be attributed to a person's ability to produce equol from daidzein via their intestinal bacteria, because its biological activities differ from those of its precursor [bib_ref] The clinical importance of the metabolite equol-a clue to the effectiveness of..., Setchell [/bib_ref]. However, not all healthy humans produce equol. The ability to produce equol depends on the presence of certain intestinal microbiota. Many studies have provided that only 25-30% of the adult population of Western countries produce equol when fed soy foods containing isoflavones [bib_ref] Urinary equol excretion with a soy challenge: influence of habitual diet, Lampe [/bib_ref] [bib_ref] Gut bacterial metabolism of the soy isoflavone daidzein: exploring the relevance to..., Atkinson [/bib_ref] [bib_ref] Interindividual variation in metabolism of soy isoflavones and lignans: influence of habitual..., Rowland [/bib_ref]. This is significantly lower than the reported 50-60% of equol producers in adults from Asian countries [bib_ref] Age-stratified serum levels of isoflavones and proportion of equol producers in Japanese..., Fujimoto [/bib_ref] [bib_ref] Comparisons of percent equol producers between prostate cancer patients and controls: casecontrolled..., Akaza [/bib_ref] [bib_ref] Comparison of isoflavones among dietary intake, plasma concentration and urinary excretion for..., Arai [/bib_ref]. The reasons for these differences are unclear, but dietary changes can alter the microbiotal profile of the intestine. Therefore, habitual dietary patterns may influence the metabolism of isoflavone and the production of equol [bib_ref] Dietary and lifestyle correlates of urinary excretion status of equol in Japanese..., Nagata [/bib_ref]. Several studies comparing the habitual diets of equol producers and non-producers who consume Western diets have reported that equol producers tend to have a higher intake of carbohydrate and dietary fibre, a higher percentage of energy as carbohydrate and lower percentage of energy as fat, and intake of soy, plant protein [bib_ref] Urinary equol excretion with a soy challenge: influence of habitual diet, Lampe [/bib_ref] [bib_ref] Interindividual variation in metabolism of soy isoflavones and lignans: influence of habitual..., Rowland [/bib_ref] [bib_ref] Long-term dietary habits affect soy isoflavone metabolism and accumulation in prostatic fluid..., Hedlund [/bib_ref] ; however, this has not been a consistent observation.
On the other hand, soy products are traditionally used in many Asian countries. The intake of traditional soybased foods is high in Japan, and the mean total intake of isoflavones is estimated between 19.4 and 33.6 mg/d according to the National Nutritional Survey in Japan [bib_ref] National Nutrition Survey in Japan-its methodological transition and current findings, Katanoda [/bib_ref]. In Western populations, the consumption of isoflavones from traditional soy foods is substantially lower (between 0.5 and 3 mg) than that of Japan [bib_ref] Intake of dietary phytoestrogens is low in postmenopausal women in the United..., De Kleijn [/bib_ref] [bib_ref] Intakes and sources of soya foods and isoflavones in a UK population..., Mulligan [/bib_ref]. In Western countries, a portion of total isoflavone consumption is derived from soy protein and soy flour added to a variety of foods.
While it is not clearly understood how bacterial species produce equol in the intestine, certain Lactobacillus and Bifidobacteria species have been suggested to play a role in the metabolism of daidzein to equol [bib_ref] Lactobacillus rhamnosus JCM 2771: impact on metabolism of isoflavonoids in the fecal..., Tamura [/bib_ref] [bib_ref] Bioconversion of soy isoflavones daidzin and daidzein by Bifidobacterium strains, Raimondi [/bib_ref]. Fructooligosaccharides (FOS) have prebiotic activity in addition to being a form of dietary fibre [bib_ref] Relationship of prebiotics and food to intestinal microflora, Blaut [/bib_ref] [bib_ref] The capacity of short-chain fructo-oligosaccharides to stimulate faecal bifidobacteria: a dose-response relationship..., Bouhnik [/bib_ref]. FOS are poorly digested in the human small intestine and are fermented in the colon by Lactobacillus and Bifidobacteria species; moreover, FOS stimulates their growth.
To date, a few studies have reported the effects of dietary fibre and/or prebiotics on isoflavone bioavailability in Western population [bib_ref] Wheat bran and soy protein feeding do not alter urinary excretion of..., Lampe [/bib_ref] [bib_ref] Dietary combination of soy with a probiotic or prebiotic food significantly reduces..., Larkin [/bib_ref] [bib_ref] Increased probiotic yogurt or resistant starch intake does not affect isoflavone bioavailability..., Larkin [/bib_ref] , but no intervention studies have examined the effect of FOS on equol production in Japanese. The purpose of the present study was to determine whether FOS supplementation increases equol production in equol producers and stimulates equol production in equol non-producers in Japanese postmenopausal women.
# Methods
# Ethics statement
The protocol was approved by the Institutional Review Board of the National Institute of Health and Nutrition of Japan, and the study was performed in accordance with the guidelines of the Declaration of Helsinki. All women provided their written informed consent to participate in the study.
## Subjects and design
In this study, 43 healthy postmenopausal women who fulfilled the required criteria were recruited and screened. The exclusion criteria for this study were history of chronic renal or hepatic disease, history of hormone replacement therapy or breast cancer, serum oestrogen hormone (E2) concentration > 50 pg/mL, soy allergies, and any other medication known to affect bone. A soy challenge was used to assess equol-producer status prior to the start of this study. The day before sample collection, subjects ingested soy food containing 22 mg soy isoflavones. On sample collection day, morning urine samples were collected from the subjects, and they were stratified into equol producers or nonproducers based on their urinary equol to daidzein concentration ratio [Equol producer: Log (urinary equol/ daidzein) > −1.70].
The study comprises 4 separate groups in a randomized crossover design . Subjects were identified by a single randomisation number using a computergenerated random permutation procedure in SPSS software version 11.0J. The study design involved two 2-week dietary periods separated by a 2-week washout. During each 2-week intervention period, fermented foods and additional soy isoflavone supplements that could potentially affect intestinal bacteria were excluded from the diet. Subjects were instructed to otherwise maintain their normal diets for the duration of the study. The washout period required the same dietary exclusions as the intervention periods. Studies have shown that antibiotic-associated changes in faecal flora are normally reversed in 10-14 days after the stopping of treatment [bib_ref] Ecology of antibiotic resistance determinants, Levy [/bib_ref] [bib_ref] The survival of R plasmids in the absence of antibiotic selective pressure, Richmond [/bib_ref]. Therefore, a 2-week washout period was set in our study, although antibiotics were not used by any of the participants during the study. The subject randomisation codes were allocated sequentially in the order in which the subjects were enrolled. After completion of all the analyses, the randomisation code was disclosed to the investigators and subjects.
All subjects received a daily dose of 37 mg isoflavone conjugates in the capsule (Fujiflavone P40; Fujicco Co., Ltd., Kobe, Japan; 21 mg aglycone form) and 5 g of either FOS (Meioligo; Meiji Seika Kaisha Ltd., Tokyo, Japan) or sucrose (white sugar; Mitsui Sugar Co., Ltd., Tokyo, Japan) as control. Each subject consumed the isoflavone capsule and FOS or sucrose at breakfast. Digestive products may stimulate the growth of certain microbiota, and FOS may selectively stimulate the growth of certain microbiota. As sucrose is digested readily, we used that as the control in this study. Dietary assessments of dietary soy isoflavones and nutrient intake based on 3-day dietary records were obtained at baseline. Fasting blood samples were collected at baseline and after 2 weeks of intervention. Spot urinary samples were collected at baseline and after 1 and 2 weeks of intervention.
## Isoflavones supplement
The crude isoflavone conjugate (40% isoflavones) in the capsule were daidzin (20.4%), malonyldaidzin (0.1%), acetyldaidzin (1.1%), daidzein (0.3%), genistin (4.6%), acetylgenistin (0.3%), genistein (0.1%), and glycitin plus glycitein (13%). As aglycones, 33 mg daidzein, 8.5 mg genistein, and 15 mg glycitein were included in the 100 mg of conjugates. Thus, 37 mg of conjugated isoflavones equivalent to 21.1 mg aglycone, which consisted of 12.3 mg daidzein, 3.2 mg genistein, and 5.6 mg glycitein, existed in each capsule.
## Measurement of serum and urinary daidzein and equol concentrations
The concentrations of daidzein and equol in blood and urine were analysed by the time-resolved fluoroimmunoassay method; the detection limits of daidzein and equol in these assays were 2.0 nM and 3.3 nM, respectively [bib_ref] Time-resolved fluoroimmunoassay of plasma daidzein and genistein, Wang [/bib_ref] [bib_ref] Time-resolved fluoroimmunoassay for equol in plasma and urine, Brouwers [/bib_ref].
# Statistical analysis
All values are expressed as mean ± SD, and P < 0.05 was considered significant. The differences in isoflavone and nutritional intake between equol producers and nonproducers were examined by using unpaired Student's ttest. The differences in the serum equol concentrations at baseline and after 2 weeks of intervention were examined by using paired Student's t-test. Three-way ANOVA (factors of FOS intervention, time, and equol status) was performed to determine the effect of dietary intervention on the urinary equol to daidzein concentration ratios. The changes in the urinary equol to daidzein concentration ratios in each group were evaluated by repeatedmeasures ANOVA and the Tukey post hoc test. All analyses were performed by using SPSS (SPSS 11.0; SPSS Inc., Chicago, IL, USA).
# Results and discussion
At the beginning of the study, 43 subjects (25 equol producers, 18 non-producers) were recruited. Nine subjects Study design. Subjects were classified as equol producers or non-producers and were further assigned to 2 groups as follows: ISO + control, intervention with soy isoflavone supplement and sucrose; ISO + FOS, intervention with soy isoflavone supplement and fructooligosaccharides. ISO, isoflavones; FOS, fructooligosaccharides.
withdrew because of personal reasons (5 subjects) and because they had taken oral antibiotics (4 subjects). The remaining 34 subjects completed the study, 21 (61.8%) were classified as equol producers, and 13 (38.2%) as non-producers. Characteristics of the subjects and their daily intake of isoflavones and nutrients at baseline are shown in [fig_ref] Table 1: Equol producers and non-producers [/fig_ref]. There were no major differences in age, height, weight, and body mass index between the equol producers and non-producers. The average daily intake of isoflavones from soy foods (other than isoflavone supplements) in each group ranged from 41.6-54.8 mg. Except for the fat intake, there were no significant differences between the daily intake of isoflavones and nutrients in the equol producers or non-producers. The average daily fat intake in equol producers (56.3 ± 13.7 g) was significantly higher compared to non-producers (49.1 ± 13.6 g). Epidemiological studies have been performed investigating the effects of habitual diet on equol production. Aldercreutz et al. found that the intake of total fat and meat and the dietary ratio of fat to fibre correlated with the urinary excretion of equol in a Japanese population [bib_ref] Urinary excretion of lignans and isoflavonoid phytoestrogens in Japanese men and women..., Adlercreutz [/bib_ref] , our result is consistent with this previous study. Whereas in a Western population, Rowland et al. reported that equol producers consumed significantly less energy as fat and significantly more energy as carbohydrate than equol non-producers [bib_ref] Interindividual variation in metabolism of soy isoflavones and lignans: influence of habitual..., Rowland [/bib_ref]. The effects of habitual diet on equol production are controversial.
The serum equol concentrations in the equol producers and non-producers were 38.0 ± 24.8 nmol/L and 23.2 ± 19.0 nmol/L, respectively, and that of the equol producers was significantly higher than that of the nonproducers (p = 0.008). Among the equol producers, serum equol concentrations in both the FOS and control groups significantly increased after 2-week of intervention . However, there were no significant differences in serum equol levels between the FOS and control groups in the equol producers. The similarly increased serum equol concentrations in the FOS and control groups among equol producers may have been stimulated by the isoflavone supplement. Among the equol non-producers, there were no significant differences in serum equol concentrations between the FOS and control groups at baseline and after 2 weeks of intervention. Thus, FOS intervention did not affect serum equol production as compared to sucrose intervention as control in the equol producers or non-producers who were treated with isoflavones in this study.
Setchell et al. have reported that expressing the productprecursor relationship as the urinary equol to daidzein concentrations ratio is a more reliable indicator of the conversion of daidzein to equol [bib_ref] Method of defining equol-producer status and its frequency among vegetarians, Setchell [/bib_ref]. Among the equol producers and non-producers, 1-or 2-week FOS intervention had no effect on the capacity of colonic microbiota to produce equol [fig_ref] Table 2: Changes in urinary equol to daidzein concentration ratios in postmenopausal women 1... [/fig_ref]. Among the equol producers, urinary equol to daidzein concentration ratios in both the FOS and All values are mean ± SD. [bib_ref] Comparative study of oestrogenic properties of eight phytoestrogens in MCF7 human breast..., Matsumura [/bib_ref] Differences in age, height, weight, body mass index, and daily isoflavone and nutrient intake between equol producers and non-producers were examined by using unpaired t-tests.
## Figure 2
Serum equol concentrations of different study groups at baseline and after 2 weeks of intervention. EQ (+), equol producers; EQ (−), equol non-producers. Data is expressed as mean ± SD. Data was analyzed using a paired t-test. * Statistically significant (P < 0.05). The differences in the serum equol concentrations at baseline and after 2 weeks of intervention were examined using a paired Student's t-test. Three-way ANOVA (factors of FOS intervention, time, and equol status) was performed to determine the effect of dietary intervention on the urinary equol to daidzein concentration ratios. [bib_ref] Isoflavones made simple -genistein's agonist activity for the beta-type estrogen receptor mediates..., Mccarty [/bib_ref] The changes in the urinary equol to daidzein concentration ratios in each group were evaluated by repeated-measures ANOVA and the Tukey post hoc test.
control groups was not significantly different after 1 week (p = 0.343 and p = 0.120, respectively) or 2 weeks (p = 0.578 and p = 0.432, respectively) compared with baseline. Similarly, in equol non-producers, FOS intervention did not affect the urinary equol to daidzein concentration ratios after 1 week (p = 0.794) or 2 weeks (p = 0.564) compared with baseline. In equol non-producers, no significant differences in urinary equol to daidzein concentration ratios in control group between baseline and after 1 week (p = 0.890) or 2 weeks (P = 0.383) were observed. Moreover, among the equol producers, urinary equol to daidzein concentration ratios in both the FOS and control groups was not significantly different between 1 and 2 weeks (p = 0.690 and p = 0.921, respectively). Similarly, among equol non-producers, urinary equol to daidzein concentration ratios in both the FOS and control groups was not significantly different between 1 and 2 weeks (p = 0.166 and p = 0.091, respectively). However, significant differences were observed in the interaction effect of time × equol state after 1 week of intervention (p = 0.006); however, there were no effects after 2 weeks of intervention (p = 0.516). One important reason for this observation could be the intervention time. Lampe et al. reported that isoflavone excretion did not differ according to the duration of soy intervention, whether 4 days or 1 month [bib_ref] Wheat bran and soy protein feeding do not alter urinary excretion of..., Lampe [/bib_ref]. On the other hand, other studies have suggested that long-term exposure to isoflavone may change the usual plasma concentrations and urinary excretion of these compounds as result of altered metabolism [bib_ref] Underlying biochemical and pharmacological issues, Barnes [/bib_ref] [bib_ref] Altered kinetics and extent of urinary daidzein and genistein excretion in women..., Lu [/bib_ref]. After an initial increase, plasma concentrations of daidzein and genistein decreased when individuals consumed soy daily over a 2-week period [bib_ref] Underlying biochemical and pharmacological issues, Barnes [/bib_ref]. Lu et al. reported that urinary recovery of genistein and daidzein decreased progressively over 4 weeks of daily soy ingestion but increased for equol [bib_ref] Altered kinetics and extent of urinary daidzein and genistein excretion in women..., Lu [/bib_ref]. Thus, such findings have not always been consistent among studies, and the intervention period that would elicit an isoflavone response is controversial.
In present study, there were no effects of 2-week FOS intervention on the capacity of colonic microbiota to produce equol. Supplementation with FOS and isoflavones for more than 2 weeks may be required in order to increase the capacity of colonic microbiota to produce equol. Another reason for this observation could be the diet. It is possible that a dietary intake of isoflavone may affect equol production [bib_ref] Dietary and lifestyle correlates of urinary excretion status of equol in Japanese..., Nagata [/bib_ref] [bib_ref] Prevalence of daidzein-metabolizing phenotypes differs between Caucasian and Korean American women and..., Song [/bib_ref]. Japanese diet typically contains higher amounts of soy products than the western diets, but the consumption of soy foods during the study was not controlled. The average daily total isoflavones (aglycone and conjugates) in this study ranged from 41.6 -54.8 mg and isoflavone supplement contains 21 mg aglycone form. If dietary isoflavones may be sufficient isoflavone to saturate equol producer capacity, the supplement cannot elicit a response. Hence, the increased equol production in the equol producers after 1 week of intervention may correlate with isoflavones intake during this period. If a diet had low isoflavone content in our study, FOS supplementation may elicit a response by increasing the capacity of colonic microbiota to produce equol. FOS intervention and time and the interaction effects of FOS × time, FOS × equol state, and FOS × time ×equol state did not influence the urinary equol to daidzein concentration ratios significantly. Hence, our results suggest that a 2-week intervention with an FOS dosage of 5 g/day does not significantly modulate equol production both in equol producers and non-producers among postmenopausal Japanese women. Although we did not assess the gastrointestinal microbial activity in the present study, a daily FOS intake of 2.5-10 g for 2-4 weeks has been reported to have prebiotic effects [bib_ref] The capacity of short-chain fructo-oligosaccharides to stimulate faecal bifidobacteria: a dose-response relationship..., Bouhnik [/bib_ref] [bib_ref] Four-week short chain fructo-oligosaccharides ingestion leads to increasing fecal bifidobacteria and cholesterol..., Bouhnik [/bib_ref]. Additionally, FOS have been reported to increase the number of faecal bifidobacteria [bib_ref] The capacity of short-chain fructo-oligosaccharides to stimulate faecal bifidobacteria: a dose-response relationship..., Bouhnik [/bib_ref] [bib_ref] Four-week short chain fructo-oligosaccharides ingestion leads to increasing fecal bifidobacteria and cholesterol..., Bouhnik [/bib_ref] , affecting the intestinal microbiotal balance [bib_ref] Relationship of prebiotics and food to intestinal microflora, Blaut [/bib_ref]. Ohta et al. reported that FOS increased equol production from daidzein in OVX rats [bib_ref] A combination of dietary fructooligosaccharides and isoflavone conjugates increases femoral bone mineral..., Ohta [/bib_ref]. Thus, it is likely that FOS intake produced similar effects in the present study, although the effects on equol production were not significant. One possible reason for this observation could be that FOS intake of 5 g/ day may not be sufficient to change and increase the number of intestinal microbiota required for a significant increase in equol production in humans. It was reported that the gastrointestinal tolerances of healthy male and female volunteers to FOS were 0.30 g/kg and 0.40 g/kg, respectively, and a FOS daily intake of up to 30 g was safe [bib_ref] The relation of fructo-oligosaccharide intake and gastrointestinal symptom -The observations of no..., Hata [/bib_ref]. In this study, 5 g FOS was ingested per day. Therefore, if a much higher level of FOS were used, there may be effects on the intestinal microbiota of equol production. On the other hand, previous studies have reported that probiotics or prebiotic supplementation with soy does not affect equol production in American and Australian populations [bib_ref] Wheat bran and soy protein feeding do not alter urinary excretion of..., Lampe [/bib_ref] [bib_ref] Dietary combination of soy with a probiotic or prebiotic food significantly reduces..., Larkin [/bib_ref] [bib_ref] Plasma phytoestrogens are not altered by probiotic consumption in postmenopausal women with..., Nettleton [/bib_ref]. Although certain Lactobacillus and Bifidobacteria species have been suggested to play a role in the metabolism of daidzein to equol [bib_ref] Bioconversion of soy isoflavones daidzin and daidzein by Bifidobacterium strains, Raimondi [/bib_ref] [bib_ref] Metabolism of dietary soy isoflavones to equol by human intestinal microflora-implications for..., Yuan [/bib_ref] , the relevance of dietary-induced changes in gastrointestinal microbiota activity to isoflavone bioavailability is poorly understood. This suggests that further research addressing the roles of intestinal microbiota is required to determine the effects of FOS on equol production. This is a pilot study, so there are several limitations in our study. Firstly, with regard to our Japanese subjects, dietary isoflavones during the study was not controlled. The Japanese diet contains higher amounts of soy products compared to Western diets. It is possible that a high dietary intake of isoflavone may have an effect on the capacity of equol production. Secondly, this study was carried out with small populations in each group. Sample size calculation, performed before the start of the study, showed that with 24 subjects in each group, a difference of 30 nmol/L between the means of urinary equol concentration could be shown with a power of 0.80 and two-side type 1 error of 0.05. Thus, it was seemed that 24 subjects in each group were enough for the study. However, this study included only 21 equol producers and 13 equol non-producers. One possible explanation for the non-significant effects of FOS treatment was the insufficient power as a result of the small sample size. Large scale studies with a similar design on statistically significant populations are essential to confirm our findings. Thirdly, it is a short-term analysis, and the subjects under review were adults whose gastrointestinal microbiotal composition is relatively stable. Some studies have suggested that long-term exposure to isoflavones may change the usual plasma concentrations and urinary excretion of these compounds as result of altered metabolism [bib_ref] Altered kinetics and extent of urinary daidzein and genistein excretion in women..., Lu [/bib_ref]. Hence, it is possible that the 2week duration of this study was insufficient to alter the intestinal microbiota for which a longer trial period may be required. Additionally, the washout period was only 2 weeks, thus carryover effect might exist in our study. Based on the results derived, we intend to carry out a large-scale study in which the treatment period and FOS dose are modified, and dietary isoflavones are controlled.
# Conclusion
In conclusion, we have shown that a 2-week intervention with a daily dose of 5 g FOS does not significantly modulate the capacity of intestinal microbiota to produce equol in postmenopausal Japanese women, in either the equol producers or the non-producers treated with isoflavones in this pilot study. However, it may be possible to stimulate equol production with dietary conditions, especially probiotics such as Lactobacillus and Bifidobacteria species are added to a diet may be useful, on long-term intervention. Further large investigations that explore the roles of specific intestinal microbiota in equol production will enable the establishment of dietary conditions that are required to enhance equol production.
Abbreviations FOS: Fructooligosaccharides; ANOVA: Analysis of variance.
[table] Table 1: Equol producers and non-producers: subject characteristics and daily intake of isoflavones and nutrients 1 [/table]
[table] Table 2: Changes in urinary equol to daidzein concentration ratios in postmenopausal women 1 Urinary equol to daidzein concentration ratios; (equol/daidzein) log. All values are expressed as mean ± SD. [/table]
|
Dynamical time-reversal symmetry breaking and photo-induced chiral spin liquids in frustrated Mott insulators
The search for quantum spin liquids in frustrated quantum magnets recently has enjoyed a surge of interest, with various candidate materials under intense scrutiny. However, an experimental confirmation of a gapped topological spin liquid remains an open question. Here, we show that circularly polarized light can provide a knob to drive frustrated Mott insulators into a chiral spin liquid, realizing an elusive quantum spin liquid with topological order. We find that the dynamics of a driven Kagome Mott insulator is well-captured by an effective Floquet spin model, with heating strongly suppressed, inducing a scalar spin chirality S i · (S j × S k ) term which dynamically breaks time-reversal while preserving SU(2) spin symmetry. We fingerprint the transient phase diagram and find a stable photo-induced chiral spin liquid near the equilibrium state. The results presented suggest employing dynamical symmetry breaking to engineer quantum spin liquids and access elusive phase transitions that are not readily accessible in equilibrium.
C ontrol of quantum materials out of equilibrium represents one of the grand challenges of modern condensed matter physics. While an understanding of general nonequilibrium settings beyond heating and thermalization is still in its infancy, a loophole concerns considering instead the transient quantum states of quasi-periodic perturbations such as wideenvelope laser pulses. Here, much of the intuition and language of equilibrium survives in a distinctly non-equilibrium setting within the framework of Floquet theory. While recently enjoying much attention and experimental success in the manipulation of singleparticle spectraand band topology or short-range entangled topological states, a natural extension regards pumping of strongly correlated systems. Here, the essence of Floquet physics lies not merely in imbuing one-1 and two-particle 2 responses with the pump frequency as an additional energy scale, but in reshaping the underlying Hamiltonian to stabilize phases of matter that might be inaccessible in equilibrium.
Indeed, initial investigations suggest that the notion of effective low-energy physics persists in certain high-frequency regimes of time-periodic perturbations, leading for instance to enhancement of correlated hopping, strong-field sign reversal of nearestneighbor Heisenberg exchange in a 1D magnet, or enhancement of Cooper-pair formation. Similar ideas are being pursued in the field of ultracold atoms to simulate artificial gauge fields, to dynamically realize topological band structuresor even propose fractional quantum Hall effects and spin liquids in optical lattices. At the same time, recent advances in Floquet thermodynamics indicate that, while driven nonintegrable closed systems are in principle expected to heat up to infinite temperature, heating can be exponentially slow on pre-thermalized time scalesor altogether avoided via manybody localizationor dissipation. An ideal condensedmatter realization hence entails a charge gap to limit absorption, as well as a delicate balance of competing phases, such that timedependent perturbations and dynamical symmetry breaking can be expected to have an outsized effect and phase boundaries can be reached on pre-thermalized time scales with moderate effort.
Frustrated quantum magnetsare prime candidates for such ideas. Strong local Coulomb repulsion between electrons freezes out the charge degrees of freedom, whereas the spin degrees of freedom are geometrically obstructed from ordering, hosting a delicate competition of conventionally ordered phases as well as quantum spin liquids (QSLs) with long-range entanglement and exotic excitations. The chiral spin liquid (CSL) constitutes one of the earliest proposals of a topologically ordered QSL; it breaks time-reversal symmetry (TRS) and parity, while preserving SU(2) spin symmetry, and can be regarded as a bosonic ν = 1/2 fractional quantum Hall state of spins with zero net magnetization and gapped semion excitations. While an unlikely ground state in unperturbed microscopic models, recently the Photo-induced Kagome chiral antiferromagnet. a Starting from a Kagome Mott antiferromagnet, pumping with circularly polarized light dynamically breaks time-reversal and parity while preserving SU(2) spin symmetry, photo-inducing scalar spin chirality S i · (S j × S k ) contributions on elementary equilateral (χ) and isosceles (χ′) triangles. Pump strength and frequency provide knobs to tune χ,χ′ as well as Heisenberg exchange J, J′, J 3 , as described in the main text. Examples of b nearest-neighbor and c three-site Floquet virtual hopping processes including absorption of photons, in the Mott-insulating regime. Boxes graphically depict the example initial, virtual intermediate and final states for second-and fourth-order virtual hopping processes, in terms of the product space of electronic degrees of freedom and the Floquet index. TRS is broken in c, inducing scalar spin chiralities χ on triangles of the lattice, whereas b solely induces nearest-neighbor Heisenberg exchange CSL was found to be a competing state, in particular after explicit breaking of TRS and parity. However, TRS breaking in experiment is realized canonically via external magnetic fields, necessarily entailing a Zeeman shift as the dominant contribution, which breaks SU(2) symmetry and disfavors CSLs.
[formula] m =2 m =1 m =0 m =-1 m =-2 m =2 m =1 m =0 m =-1 m =-2 m =2 m =1 m =0 m =-1 m =-2 m=2 m=1 m=0 m=-1 m=-2 m=2 m=1 m=0 m=-1 m=-2 m =2 m =1 m =0 m =-1 m =-2 m =2 m =1 m =0 m =-1 m =-2 m=2 m=1 m=0 m=-1 m=-2 b a c [/formula]
Here, we show that pumping a Mott insulator with circularly polarized light below the Mott gap can dynamically break TRS without breaking of SUor translation symmetry, providing a knob to drive a frustrated quantum magnet into a CSL. Starting from a prototypical Hubbard model, the key questions posed by this work are three-fold: First, how does optically induced TRS breaking manifest itself in a Mott insulator; second, can the ensuing effective Floquet spin model support a transient CSL and what are its signatures; and finally, does such an effective Floquet steady-state description capture the many-body time evolution of an optically driven Hubbard model? In the following, we answer all three questions affirmatively.
# Results
Floquet-Hubbard model. Our focus lies on Kagome antiferromagnets, which have recently garnered much attention due to candidate materials herbertsmithite, kapellasite, and others 34 with putative spin-liquid behavior at low temperatures. Experiments 54 and first-principles calculationsindicate that the ground state and low-energy excitation spectra of these materials are well-captured by antiferromagnetic Heisenberg exchange between d 9 spins localized on Cu 34 . However, as photons couple to charge, a microscopic modeling of the light-matter interaction in principle must account for the multi-orbital structure at higher energies 57 , above the~2 eV charge gap. Here, we take a phenomenological approach, and, as an effective starting point that captures the essential physics but without pretense of a direct materials connection, start from a driven single-orbital Hubbard model at half fillinĝ
[formula] HðtÞ ¼ Àt h X ij h iσ e i e h r ij ÁAðtÞĉ † iσĉ jσ þ U X in i"ni#ð1Þ [/formula]
Here, t h , U, e denote nearest-neighbor hopping, Coulomb interaction, and electron charge, r ij denotes vectors between sites i,j, and A(t) = A(t)[cos(Ωt), sin(Ωt)] T models a circularly polarized pump beam with wide-pulse envelope A(t), coupling to electrons via Peierls substitution. Comparison of nearest-neighbor exchange J % 4t 2 h =U with first-principles predictions for herbertsmithitesuggests U/t h of up to 40 due to the exceedingly narrow width of Cu d-orbital derived bands.
If A(t) varies slowly with respect to the pump period, then the Hamiltonian becomes approximately periodic under a translation H t þ 2π=Ω ð Þ¼ĤðtÞ. Floquet theory then dictates that the behavior near the pump plateau is completely determined via many-body eigenstates of the form Ψ n ðtÞ j i¼ e Àiϵnt P m e imΩt Φ m j i with ϵ n the Floquet quasi-energy, where the Φ m j i conveniently follow as eigenstates of the static Floquet-Hubbard Hamiltonian
[formula] H ¼ Àt h P ij h iσ mm′ J mÀm′ A ð Þe i mÀm′ ð Þ argr ijĉ † iσĉ jσ m j i m′ h j þU P in i"ni# 1 À P m mΩ 1 m j i m h jð2Þ [/formula]
where A denotes the dimensionless field strength at the pump plateau, such that A(t) ≈ Aħ/(ea 0 ) with a 0 the nearest-neighbor distance, m 2 Z is the Floquet index, and J m ðÁÞ denotes the Bessel function of the first kind (Methods section). Note that the apparent Hilbert space expansion is merely a gauge redundancy of Floquet theory, as eigenstates with energy ϵ n + mΩ identify with the same physical state ∀m.
Floquet Chiral spin model. Physically, Eq. (2) describes photonassisted hopping in the presence of interactions, where electrons can enlist m photons to hop at a reduced energy cost U − mΩ of doubly occupying a site. Deep in the Mott phase the formation of local moments persists out of equilibrium as long as the pump remains off resonance and red-detuned from the charge gap. However, photon-assisted hopping reduces the energy cost of virtual exchange, pushing the system closer to the Mott transition and enlarging the range of virtual hopping paths that provide non-negligible contributions to longer-ranged exchange or multispin processes. Second, electrons acquire gauge-invariant phases when hopping around loops on the lattice for circular polarization. Crucially, and in contrast to an external magnetic field, an optical pump precludes a Zeeman shift, retaining the SU(2) spin rotation symmetry that is shared by CSL ground states. Symmetry considerations dictate that a manifestation of TRS breaking must to lowest-order necessarily involve a photo-induced scalar spin chirality χ ijk term, with:
[formula] H spin ¼ X ij J ij S i Á S j þ X ijk χ ijk S i Á S j S k À Áð3Þ [/formula]
This Floquet Chiral Spin Hamiltonian is the central focus of the paper; to derive it microscopically from the driven Kagome-Hubbard model (1), it is instructive to first consider the highfrequency limit Ω ) U; t h . Here, circularly polarized pumping induces complex nearest-neighbor
[formula] hoppingst ¼ t h 1 À A 2 =4 ð Þþ i ffiffi ffi 3 p =4 À Á t 2 h A 2 =Ω as well as purely-complex next-nearest-neighbor hoppingst 0 ¼ Ài ffiffi ffi 3 p =4 À Á t 2 h A 2 =Ω, [/formula]
analogous to a staggered magnetic flux pattern in the unit cell (Supplementary Note 1). To third order int;t 0 , a spin description then includes scalar spin chirality contributions, with χ ¼ 9 ffiffi ffi 3 p t 4 h A 2 =2U 2 Ω of equal handedness for both equilateral triangles per unit cell, as depicted in , and six isosceles triangles of opposite handedness with χ′ = χ/3, such that the total chiral couplings in the unit cell sum to zero. Now consider sub-gap pumping Ω < U. Starting from Eq. (2), a microscopic derivation of the Floquet spin Hamiltonian proceeds via quasi-degenerate perturbation theory, where care must be taken to simultaneously integrate out m ≠ 0 Floquet states and many-body states with doubly occupied sites (Supplementary Note 2).. While every process contributes to Heisenberg exchange, a scalar spin chirality contribution appears for multi-hop processes that enclose an area. Naïvely, to third order, an electron could simply circumnavigate the elementary triangles of the Kagome lattice; however, these processes interfere destructively and cancel exactly to all orders in A even though TRS is broken, and in contrast to an external magnetic field (Supplementary Note 2). This is consistent with results on the resonant A 2g Raman response of Mott insulators, that connect to the A → 0, m = 1 limit. Instead, TRS breaking first manifests itself to fourth order in virtual hopping. Here, processes can either encompass an elementary triangle, or virtually move an electron back and forth two legs of a hexagon, inducing scalar spin chirality contributions as shown in , with
[formula] J ¼ 4 P m J m A ð Þ j j 2 t 2 h = U À mΩ ð Þ 8,χ ¼ 3 X m sin 2π m 1 Àm 2 þm 3 ð Þ 3 h i Λ ð1Þ m À sin 2πm2 3 Â Ã Λ ð2Þ m h i ð4Þ χ 0 ¼ X m sin π m1Àm2þm3 ð Þ 3 h i Λ ð1Þ m À sin πm 2 3 Â Ã Λ ð2Þ m h ið5Þ [/formula]
where m = {m 1 , m 2 , m 3 } are Floquet indices, and
[formula] Λ ð1Þ m ¼ 8t 4 h J m 1 A ð ÞJ m 2 Àm 1 A ð ÞJ m 2 Àm 3 A ð ÞJ m 3 A ð Þ Q 3 i¼1 U À m i Ω ð Þ ð6Þ Λ ð2Þ m ¼ 2 1 À δ m2 ð Þ UÀm 2 Ω m 2 Ω À1 ð Þ m1Àm3 cos 2 m 2 π 2 Â Ã Λ ð1Þ mð7Þ [/formula]
Here, Λ ð1Þ m and Λ ð2Þ m parameterize fourth-order virtual hopping processes for which the second intermediate virtual state retains a single double-occupied site or returns to local half filling (albeit with non-zero Floquet index), respectively. Furthermore, nextnearest-neighbor Heisenberg exchange
[formula] J 0 ¼ X m Γ m À cos π m1Àm2þm3 ð Þ 3 h i Λ ð1Þ m 2 þ cos πm 2 3 Â Ã Λ ð2Þ m 2 ! J 3 ¼ X m Γ m À Λ ð1Þ m 2 þ Λ ð2Þ m 2 !ð8Þ [/formula]
and corrections to nearest-neighbor Heisenberg exchange
[formula] J ¼J þ P m 2 þ cos π m 1 Àm 3 ð Þ 3 h i þ cos 2π m 1 Àm 3 ð Þ 3 h i h n þcos π m1Àm2þm3 ð Þ 3 h i þ 1 2 cos 2π m1Àm2þm3 ð Þ 3 h i i Λ ð1Þ m À 1 þ 2 cos πm 2 ½ þcos πm2 3 Â Ã þ 1 2 cos 2πm2 3 Â Ã Â Ã Λ ð2Þ m À 7Γ m Éð9Þ [/formula]
appear at the same order, with
[formula] Γ m ¼ 4t 4 h δ m3;0 U À m 1 Ω þ δ m3;0 U À m 2 Ω ! Y 2 i¼1 J mi A ð Þ ½ 2 U À m i Ω ð Þð10Þ [/formula]
parameterizing a twofold virtual nearest-neighbor exchange process.
Steady-state phase diagram. Having established the effective steady-state physics for the duration of the pump pulse, the next question concerns whether the photo-induced Floquet spin model (Eq. (3)) can indeed stabilize a CSL. Consider its parameter space as a function of A, Ω, depicted infor U = 20t h . Adiabatic ramping up of the circularly polarized pump then corresponds to horizontal trajectories with fixed Ω., b show that TRSbreaking scalar spin chiralities develop with increasing field strength, whereas the effect on longer-ranged Heisenberg exchange ((c) and (d)) is comparatively weak. This immediately suggests that circularly polarized pumping grants a handle to , whereas χ′ changes sign when Ω approaches a two-photon resonance (Ω = U/2). We analyze the steady-state phase diagram using exact diagonalization of the Floquet spin model (Eq. (3)), parameterized by pump strength and frequency. In equilibrium (A = 0), the ground state of Eq. (3) is gapped and TRS invariant. Absence of conventional spin order is evidenced by a rapid decay of spin-spin and chiral-chiral correlation functions on a 36-site cluster, consistent with density-matrix renormalization group simulations that find a gapped Z 2 QSL 60-63 . We adopt this view for the thermodynamic limit, but note that the ground state degeneracy of a Z 2 QSL remains inaccessible in exact diagonalization of finite-size clusters (Methods section). Upon pumping (A ≠ 0), the spin correlator displays no propensity for ordering; however, chiral correlations develop smoothly. Importantly, a twofold ground state quasi-degeneracy develops continuously, with a gap to many-body excitations, indicative of a CSL. To track the phase boundary as a function of A, Ω, we determine the parameter space region within which the ground state degeneracy as well as the gap Δ CSL to the many-body excitation manifold above the CSL survives insertion of a flux quantum through the torus (Methods section). As shown in, a robust photo-induced CSL develops already for weak A, with excited states well-separated in energy. Finally, a proper verification of the CSL necessitates characterizing its ground state topological order. We therefore fingerprint the photo-induced phase by determining a basis of minimally entangled states from combinations of the two degenerate ground states ψ 1;2 , minimizing the Rényi entropy for their reduced density matrices in two distinct bipartitions(1)).
To this end, we consider a circularly polarized optical pump pulse with a slow sinusoidal ramp-up and a wide pump plateau, and simulate the exact many-body dynamics of driven 12-site U = 30 Kagome Hubbard clusters for long times t 1000 t À1 h . Conceptually, the transient state can then be thought of as dynamically following the instantaneous Floquet eigenstate Ψ τ; T slow À Á % e Àiϵ T slow ð Þτ P m e imΩτ Φ T slow À Á , with the time variable t "separating" into fast (τ) and slow T slow À Á moving components. Reaching the pump plateau, the timeevolved state will nevertheless retain a finite quasi-energy spread ΨðtÞ j i¼ P α ρ α e Àiϵαt P m e imΩt Φ α ðtÞ j i (with α indexing the Floquet eigenstates). While dephasing of these constituent Floquet eigenstates should ultimately thermalize the system to infinite temperature, the system nevertheless matches the effective
[formula] S i Á S j S k À Á S i ¼ĉ † iσ s ! σσ 0ĉ iσ 0 À Á [/formula]
on elementary triangles of the Kagome cluster. Vanishing in equilibrium due to TRS, the pump-period average of χ ijk (t) should saturate to its Floquet expectation value at the pump plateau.compares χ ijk (t), time-averaged over the pump plateau, to corresponding static χ ijk expectation values of the Floquet spin model (3) ground state. The latter follows from choosing χ, χ′, J, J′, J 3 via Eqs. (4)-(10), with A, Ω the pump parameters of the Hubbard time evolution. Intriguingly, the electronic time evolution is in excellent qualitative agreement with predictions for the Floquet spin model, even when driven close to the Mott transition.
Quantitative discrepancies predominantly originate from deviations of the local moment S z ð Þ 2 <1=4; additionally, the 12-site cluster with periodic boundary conditions permits weak ring exchange contributions from loops of virtual hopping around the cluster. Importantly, the transient increase in double occupancies is not an indication of heating-instead, this follows from a reduction of the effective U of the transient Floquet-Hubbard Hamiltonian, a consequence of the photo-assisted hopping processes depicted in , c. Quantitative differences between spin and fermionic observables are therefore analogous to differences between canonical spin and fermionic descriptions of equilibrium quantum magnets for a finite Hubbard-U.
To analyze this in detail, we focus on pumping the system across the charge resonance with the upper Hubbard band, where the photo-induced scalar spin chirality contribution is expected to be largest.-c show the period-averaged double occupancyn "n# as a function of pump strength and detuning from the charge resonance ≈U − 5.5t h . Upon resonant charge excitation, the system heats up rapidly and the double occupancy approaches its infinite-temperature limitn "n# ! 1=4. Importantly, this entails that thermalization at long times is independent of the pump strength A.
Conversely, in the off-resonant regime, one observes a pumpstrength dependent saturation of the double occupancy. Here, proper heating is strongly suppressed and the system instead realizes the effective Floquet chiral quantum magnet with a transient reduction of the Hubbard interaction U. To verify that the driven steady state indeed follows the ground state of the effective Floquet Hamiltonian adiabatically, consider a period-shifted Floquet "fidelity measure" F ¼ Ψ t þ T ð ÞΨðtÞ j h i j j , where T = 2π/Ω is the pump period. At the pump plateau with discrete time-translation symmetry, F is time-independent and quantifies the Floquet quasi-energy spread of the transient steady state. For a pure Floquet state, 1 À F ! 0, suggesting that the driven state below resonance adiabatically follows a Floquet eigenstate, whereas adiabaticity is lost when crossing the absorption edge.
To distinguish residual heating on these pre-thermalized time scales of interest 25, 26 from a transient increase in energy in the chiral quantum magnet due to modulation of the Hamiltonian, consider the period-averaged stroboscopic energy operator h b Ei (Methods section). On the pump plateau, both the double occupancy and h b
Ei saturate to their pre-thermalized steady-state expectation values; however, a minuscule residual gradient over thousands of pump cycles remains. To good approximation for the time scales considered here, we can linearize the energy on the pump plateau h b
EiðtÞ % E 0 þ tΔE, and extract the heating rate ΔE from simulations.depicts the absorbed energy per pump cycle on the pump plateau, as a function of pump strength and detuning from the absorption edge. Remarkably, residual heating is largely suppressed close to resonance, with an absorbed energy on the order of 10 −6 t h per pump cycle. Naïvely, this extraordinary meta-stability suggests that it could take on the order of tens of thousands of pump cycles for heating to dominate the dynamics, absorbing a total energy~J the exchange coupling.
A more microscopic analysis of photo-excitation for realistic materials will likely lead to a less optimistic upper bound on the time scales of interest. First, a materials-specific modeling of electron-photon coupling and multi-band effects will modify the effective photo-induced spin Hamiltonian, albeit necessarily retaining the salient symmetry properties and scalar spin chirality contributions that stabilize the CSL. Second, an intriguing followup question regards the role of coupling to-and heating of-the lattice. While magneto-elastic coupling to phonons is weak in most materials and the optical frequencies under consideration are far from resonance with infrared-active phonon modes, electron phonon coupling will nevertheless indirectly heat the lattice due to Raman-assisted hopping processes. Conversely, the separation of time scales for electrons and phonons suggests that the phonon bath could similarly play the role of a dissipative channel, effectively "cooling" the electronic system. While initial investigations have already studied the case of free or weaklyinteracting electrons 30-32 , a proper understanding of the confluence of strong interactions, external drive and dissipation remains an interesting topic for future study.
Conclusions. In summary, we have shown that pumping a frustrated Mott insulator with circularly-polarized light can dynamically break TRS while preserving SU(2) symmetry of the underlying spin system, by augmenting its effective dynamics with a transient scalar spin chirality term. Remarkably, on the Kagome lattice this effective Floquet spin model was found to stabilize a transient CSL in a broad parameter regime. Our results suggest that wide-pulse optical perturbations can provide an intriguing knob to tune the low-energy physics of frustrated quantum magnets, shedding light on regions of their phase diagram hitherto unexplored.
# Methods
Floquet theory. Consider a generic time-dependent many-body Hamiltonian with discrete time-translation invarianceĤðtÞ ¼Ĥ t þ 2π=Ω ð Þ . Instead of solving the many-body time evolution, one can reexpress the time-dependent Schrödinger equation in a Floquet eigenbasis Φ α ðtÞ j i¼e Àiϵα t P m e imΩt u α;m , where α indexes the basis wave function and ϵ α is its respective Floquet quasi-energy. Then, determination of the time-dependent eigenstates of the driven system reduces to finding the time-independent eigenstates of the Floquet Hamiltonian Lines are guides to the eye. On resonance (gray region), the system heats rapidly, withn "n# thermalizating independent of the pump strength A and approaching its infinite-temperature expectation value 1/4. Below resonance (white region), the system transiently realizes the TRS-breaking chiral quantum magnet, with a tunable Hubbard interaction U (as well as correspondingly tunable magnetic J, χ, seeas a function of pump strength. d Floquet fidelity FðTÞ ¼ Ψ t þ T ð ÞΨðtÞ j h i j j on the pump plateau-below charge resonance, (1 − F) → 0, indicating the controlled preparation of a Floquet eigenstate. e Extracted stroboscopic heating rates per pump cycle and below resonance. Remarkably, heating is strongly suppressed close to the charge resonance, with the driven system requiring many thousands of pump cycles to finally absorb energy on the order of the Heisenberg exchange J NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00876-y ARTICLE conditions along one direction and tracking the many-body spectral flow as a function of twist angle, as depicted in the inset of. A fine sampling of flux insertion, as depicted in, was performed for 30-site clusters, spanned by vectors R 1 = 2a 1 + a 2 , R 2 = −2a 1 + 4a 2 , and checked against the 36-site cluster. The winding of the quasi-degenerate ground states upon flux insertion is a signature of CSLs, with the two quasi-degenerate ground states exchanging once under flux insertion, or remaining separated, depending on whether they lie in different (30-site cluster) or the same (36-site cluster) momentum sectors.
[formula] H floquet ¼ X mm 0Ĥ mÀm 0 À mΩδ m;m 0 Â Ã m j i m 0 h jð12Þ [/formula]
We furthermore consider two bipartitions A, B of the 36-site cluster with periodic boundary conditions, as depicted in the inset in, and calculate the Rényi entropies S α ðθ; ϕÞ ¼ Àlog tr ρ 2 α ðθ; ϕÞ
[formula] È É ; α ¼ A; Bð13Þ [/formula]
where ρ α ðθ; ϕÞ ¼ tr α Ψðθ; ϕÞ j iΨðθ; ϕÞ h j f g is the reduced density matrix on bipartition α = A, B for superpositions Ψðθ; ϕÞ j i¼ cosðθÞ ψ 1 j iþ sinðθÞe iϕ ψ 2 j i of the quasi-degenerate ground states ψ 1 j i; ψ 2 j i, as a function of θ, ϕ. Figure 3c depicts S A and S B calculated from twofold quasi-degenerate ground states of the Floquet chiral spin model. For a CSL, S α (θ, ϕ) is expected to display two entanglement minima; the two corresponding minimally entangled states Ψðθ; ϕÞ j i permit extraction of the modular matrices 64 , which match expectations for a Kalmeyer-Laughlin CSL and are quoted in the main text.
Time evolution. The electronic many-body time evolution was simulated for a 12site Kagome-Hubbard cluster (4 unit cells) with periodic boundary conditions, spanned by vectors R 1 = 2a 1 , R 2 = 2a 2 , and with the time propagation employing adaptive step size control. We note this is the minimum cluster size to faithfully host all permutations of virtual hopping processes that give rise to the effective Floquet chiral spin Hamiltonian discussed in the main text (Eq. (3)).
To model broad circularly polarized pump pulses, we consider a pulsed field AðtÞ ¼ AðtÞ cosðΩtÞ; sinðΩtÞ
[formula] ½ Tð14Þ [/formula]
with a smooth sinusoidal pump envelope
[formula] AðtÞ ¼ 0; t 0 A sin 2 π 2 t tplateau ; 0< t< t plateau A; t ! t plateau 8 > < > :ð15Þ [/formula]
where t plateau ¼ 700t À1 h for the results of the main text. Details on pump envelope dependence can be found in Supplementary Note 3.
Finally, to quantify energy absorption in the driven system, we compute the period-averaged energy operator
[formula] E ¼ 1 T R tþT t dt 0Ĥ t 0 ð Þ ¼ Un "n# À J 0 ðAÞt h P ij h iσĉ † iσĉjσð16Þ [/formula]
where J 0 ðAÞ denotes the zeroth Bessel function of the first kind. Note that h b Ei is time-independent for a pure Floquet state, in theory. Instead, the finiteness of the pump envelope entails a residual quasi-energy spread, with the resulting dephasing of the driven state leading to residual heating on the pump plateau. While the driven state is ultimately expected to thermalize to an infinite-temperature state at infinite times, the results of the main text demonstrate a long-lived and remarkably stable pre-thermalized regime with negligible absorption.
Data availability. The data that support the results presented in this study are available from the corresponding authors on request.
Received: 27 January 2017 Accepted: 1 August 2017 |
Environmental regulation of carbon isotope composition and crassulacean acid metabolism in three plant communities along a water availability gradient
Expression of crassulacean acid metabolism (CAM) is characterized by extreme variability within and between taxa and its sensitivity to environmental variation. In this study, we determined seasonal fluctuations in CAM photosynthesis with measurements of nocturnal tissue acidification and carbon isotopic composition (d 13 C) of bulk tissue and extracted sugars in three plant communities along a precipitation gradient (500, 700, and 1,000 mm year -1 ) on the Yucatan Peninsula. We also related the degree of CAM to light habitat and relative abundance of species in the three sites. For all species, the greatest tissue acid accumulation occurred during the rainy season. In the 500 mm site, tissue acidification was greater for the species growing at 30% of daily total photon flux density (PFD) than species growing at 80% PFD. Whereas in the two wetter sites, the species growing at 80% total PFD had greater tissue acidification. All species had values of bulk tissue d 13 C less negative than -20%, indicating strong CAM activity. The bulk tissue d 13 C values in plants from the 500 mm site were 2% less negative than in plants from the wetter sites, and the only species growing in the three communities, Acanthocereus tetragonus (Cactaceae), showed a significant negative relationship between both bulk tissue and sugar d 13 C values and annual rainfall, consistent with greater CO 2 assimilation through the CAM pathway with decreasing water availability. Overall, variation in the use of CAM photosynthesis was related to water and light availability and CAM appeared to be more ecologically important in the tropical dry forests than in the coastal dune.
# Introduction
Crassulacean acid metabolism (CAM) is characterized by a nocturnal CO 2 uptake by the enzyme phosphoenolpyruvate carboxylase (PEPc), which represents phase I within the four phases of CAM as defined by [bib_ref] Crassulacean acid metabolism: a curiosity in context, Osmond [/bib_ref]. This nocturnal uptake of CO 2 via PEPc forms C 4 acids, which are decarboxylated during the daytime and generate an elevated intercellular CO 2 concentration when stomata are closed (phase III). Between these two phases, there are transitional periods of net CO 2 uptake at the beginning (phase II) and at end (phase IV) of the day, when both PEPc and RUBISCO can contribute to CO 2 assimilation [bib_ref] Carbon dioxide and water demand: crassulacean acid metabolism (CAM), a versatile ecological..., Lüttge [/bib_ref] [bib_ref] Crassulacean acid metabolism. A plastic photosynthetic adaptation to arid environments, Cushman [/bib_ref] [bib_ref] Crassulacean acid metabolism: plastic, fantastic, Dodd [/bib_ref]. Environmental factors such as light intensity, relative humidity, nocturnal temperature, and water availability affect the proportion of CO 2 uptake at night via PEPc or directly during the day via RUBISCO [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref] [bib_ref] Induction of Crassulacean acid metabolism by water limitation, Cushman [/bib_ref] [bib_ref] Environmental hormonal and circadian regulation of crassulacean acid metabolism expression, Taybi [/bib_ref] [bib_ref] Correlation between foliar stable carbon isotope composition and environmental factors in desert..., Jian-Ying [/bib_ref].
The nocturnal CO 2 uptake mechanism found among CAM species generally improves water-use efficiency because of reduced transpiration rate during the night . One hallmark of CAM plants is the remarkable flexibility of the basic metabolic framework described above, which varies in terms of the proportion of CO 2 assimilated through the CAM and C 3 pathways, among taxa, and in response to environmental conditions [bib_ref] Degrees of crassulacean acid metabolism in tropical epiphytic and lithophytic ferns, Holtum [/bib_ref] [bib_ref] Crassulacean acid metabolism: plasticity in expression, complexity of control, Holtum [/bib_ref] [bib_ref] Environmental hormonal and circadian regulation of crassulacean acid metabolism expression, Taybi [/bib_ref] [bib_ref] Induction of Crassulacean acid metabolism by water limitation, Cushman [/bib_ref] [bib_ref] Crassulacean acid metabolism: plastic, fantastic, Dodd [/bib_ref] [bib_ref] Environment or development? Lifetime net CO 2 exchange and control of the..., Winter [/bib_ref] [bib_ref] On the nature of facultative and constitutive CAM: environmental and developmental control..., Winter [/bib_ref] , and may confer an ecological advantage for surviving in habitats with seasonal variation in resource availability [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref] [bib_ref] Induction of Crassulacean acid metabolism by water limitation, Cushman [/bib_ref] [bib_ref] Ecophysiology of crassulacean acid metabolism (CAM), Lüttge [/bib_ref]. Therefore, CAM plants are likely to vary in the balance of CO 2 assimilation by RUBISCO versus PEPc along environmental gradients, and such variation may provide some advantage over other species within the community.
Investigations of carbon isotopic composition (d 13 C) can yield information regarding long-term and daily changes in the proportion of CO 2 fixed during the night or day [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref]. The 13 C/ 12 C ratio is an indicator of these changes in the assimilation of CO 2 because the enzyme responsible for net CO 2 uptake in the dark, PEPc, discriminates less against 13 C than RUBISCO, the enzyme responsible for most net CO 2 uptake during the day. The changes in d 13 C in CAM plants thus depend on: (1) the contribution of CO 2 fixed directly by RUBISCO in the light (Phase II and IV); (2) the proportion of respiratory CO 2 fixed during phase I in the dark by PEPc; and (3) the extent of leakage during the decarboxylation (phase III), allowing RUBISCO discrimination to be expressed [bib_ref] Short-term changes in carbon-isotope discrimination identify transitions between C3 and C4 carboxylation..., Griffiths [/bib_ref] [bib_ref] Discrimination in dark. Resolving the interplay between metabolic and physical constraints to..., Griffiths [/bib_ref] [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref]. CAM plants with CO 2 uptake almost exclusively at night would, thus, be expected to show d 13 C values around -11%, whereas if all the CO 2 is fixed directly by RUBISCO the values of d 13 C would be approximately -27% (O'Leary 1988). A shortcoming of using whole-tissue carbon isotope surveys to obtain an integrated value of the contribution of dark and light CO 2 fixation to the total carbon gain is that, in the absence of measurements of acidity or CO 2 exchange, it is unclear where C 3 ends and CAM begins. Species with values of d 13 C characteristic of C 3 plants have been reported to obtain up to one-third of their carbon through CAM activity [bib_ref] How closely do the d 13 C values of crassulacean acid metabolism..., Winter [/bib_ref] [bib_ref] Distribution of crassulacean acid metabolism in orchids of Panama: evidence of selection..., Silvera [/bib_ref] [bib_ref] Discrimination in dark. Resolving the interplay between metabolic and physical constraints to..., Griffiths [/bib_ref].
Many studies document a wide range in the magnitude of CAM expression that can be induced by water limitation, high temperatures, and high light availability in different species [bib_ref] Implications of genotypic diversity and phenotypic plasticity in the ecophysiological success of..., Kluge [/bib_ref] [bib_ref] Carbon isotope ratio and extent of daily CAM use by Bromeliaceae, Pierce [/bib_ref] [bib_ref] ) d 13 C values and crassulacean acid metabolism in Clusia species..., Holtum [/bib_ref] [bib_ref] Discrimination in dark. Resolving the interplay between metabolic and physical constraints to..., Griffiths [/bib_ref] [bib_ref] Crassulacean acid metabolism and fitness under water deficit stress: if not for..., Herrera [/bib_ref] [bib_ref] Crassulacean acid metabolism and epiphytism linked to adaptive radiations in the Orchidaceae, Silvera [/bib_ref] [bib_ref] Carbon isotope composition and mode of photosynthesis in Clusia species from Mexico, Vargas-Soto [/bib_ref]. Moreover, ecophysiological surveys are a powerful tool to obtain insights leading to a better understanding of the adaptive significance of CAM [bib_ref] Ecophysiological studies on orchids of Madagascar: incidence and plasticity of crassulacean acid..., Kluge [/bib_ref] [bib_ref] Implications of genotypic diversity and phenotypic plasticity in the ecophysiological success of..., Kluge [/bib_ref]. In this paper, we present a comparative survey of CAM activity in three plant communities of the Yucatan Peninsula along a gradient of water availability. This is to our knowledge one of the few comparative ecophysiological investigations of CAM activity of different species in different ecosystems. The aims of the present study were: (1) to determine the relative ecological importance of CAM species as a function of precipitation;
(2) to determine seasonal fluctuations in CAM photosynthetic activity through nocturnal acid accumulation in plants growing on a precipitation gradient and in different light microenvironments; (3) to evaluate the extension and degree of CAM activity through isotopic variation in d 13 C of bulk photosynthetic tissue and sugars along a precipitation gradient.
# Materials and methods
## Study site and species
The Yucatan Peninsula experiences a seasonally-dry tropical climate and contains a north-south precipitation gradient of 500-2,000 mm. A marked dry season from March to May, when most of the trees are leafless, is separated from the rainy season between June and October, and an early dry season from November to February, which is often characterized by up to 3-day events of strong winds ([80 km h -1 ), little rainfall (20-60 mm) and low temperatures (\20°C;.
This study was conducted in three sites: (1) the coastal dune scrubland of San Benito (21°19 0 10 00 N, 89°30 0 40 00 W), which receives approximately 500 mm of annual rainfall, and has a mean annual temperature of 26°C;
(2) Dzibilchaltún, National Park (21°05 0 N, 89°35 0 W), a tropical dry deciduous forest with a mean annual rainfall of 700 mm and temperature of 25.8°C [bib_ref] Phenology of five tree species of a tropical dry forest in Yucatan,..., Thien [/bib_ref] ; and (3) Cuxtal Ecological Reserve (20°47 0 N, 89°49 0 W), a tropical dry deciduous forest, with a mean annual rainfall of approximately 1,000 mm, and a mean annual temperature of 25.5°C [bib_ref] Carbon-isotope ratios and photosynthetic pathways in the Rapataceae, Chnaid [/bib_ref]. The sites were selected because of high CAM species diversity and relatively little disturbance compared to ecosystems of the northern Yucatan [bib_ref] Composition and structure of a tropical dry forest in Yucatan, Mexico, Rico-Gray [/bib_ref] [bib_ref] Vegetation and soil properties in two tropical dry forest of differing regeneration..., Ceccon [/bib_ref] [bib_ref] Vegetation pattern and environmental gradients in tropical dry forest of northern Yucatan..., White [/bib_ref].
At each site, species in families with previous reports of CAM [bib_ref] An introduction to crassulacean acid metabolism: biochemical principles and biological diversity, Winter [/bib_ref] , such as Agavaceae, Bromeliaceae, Cactaceae and Orchidaceae, were selected for investigation. Only one epiphyte species was selected, Myrmecophila christinae, and all other species were terrestrial [fig_ref] Table 1: Seasonal tissue acidity [/fig_ref]. Community structure and relative importance value of CAM species
The point-centered quarter method was used to determine community structure and captures the most abundant species in a rapid and reliable way (Mueller-Dombois and Ellenberg 1974). We used two 100-m-long transects and selected 20 sampling points at random in each transect. At each sampling point, four quarters were established through a cross formed by two lines. The distance to the midpoint of the nearest individual from the sampling point was measured in each quarter. For each sampling point, the species identity and the cover or dominance of the individual were recorded. The individuals considered were those that had a minimum height of 10 cm; Poaceae and annual species were not considered. The relative importance value (RIV) was used to estimate the relative importance of plant species in the three communities defined as: RIV = Relative density ? Relative coverage ? Relative frequency [bib_ref] An extension of the concept of importance value in analyzing forest communities, Skeen [/bib_ref] [bib_ref] New York Nobel PS (1985a) PAR, water, and temperature limitations on the..., Mueller-Dombois [/bib_ref].
## Nocturnal tissue acidification
At each study site, ten individuals of each species were randomly selected, five growing at ca. 80% of total daily photon flux density (PFD) and five growing at ca. 30% of total daily PFD, which was measured with gallium arsenide phosphide photodiodes (Hamamatsu, Bridgewater, NJ, USA), previously calibrated against a quantum sensor (LI190S; LI-COR, Lincoln, NE, USA) connected to a datalogger (CR21X; Campbell Scientific, Logan, UT, USA) and placed 5 m above the ground in a non-shaded location. The minimal distance among individuals was 5 m. Two samples of one leaf or stem were taken with a cork borer (1.5 cm diameter) at dusk, and before dawn the following day during the rainy (15-18 October 2005), early dry (7-10 February 2006) and dry seasons. Environmental conditions at the time of sampling were: during the rainy season, a daily average precipitation of 27.4 mm in the three sites (15 October 2005), minimum temperature 20°C (San Benito), maximum temperature 34°C (Dzibilchaltún and Cuxtal); during the early dry season no rainfall during the sampling period and previous week, minimum temperature 10°C (San Benito), maximum temperature 32°C (Cuxtal); during the dry season little rainfall at Cuxtal (3.5 mm) during the sampling period; minimum temperature 22°C (San Benito); maximum temperature 35°C (Cuxtal). Samples were frozen until transported to the laboratory, then boiled in 10 ml of distilled water for 10 min, and, after adding 50 ml of distilled water, ground with a mortar and pestle, and titrated with 0.01 N NaOH to pH 7 [bib_ref] Crassulacean acid metabolism, Osmond [/bib_ref] measured with an electronic pH meter (Model 744; Metrohm, Herisau, Switzerland). Nocturnal tissue acidification (DH ? ) was estimated from the hydrogen ion concentration (H ? ) at dawn minus H ? at dusk.
## Carbon isotopic composition
During the rainy season (18 October 2005), five samples of leaf or stem were taken from the same individuals at the time when the predawn samples for tissue acidity were taken. The samples were dried in an oven at 80°C for 48 h and homogenized to a fine powder using a mortar and pestle. Samples for carbon isotopic composition of bulk tissue were prepared as 3-mg samples rolled in tin capsules [bib_ref] Carbon isotope ratio and extent of daily CAM use by Bromeliaceae, Pierce [/bib_ref] [bib_ref] How closely do the d 13 C values of crassulacean acid metabolism..., Winter [/bib_ref].
Whereas carbon isotopic composition (d 13 C) of bulk tissue represents the time-integrated proportion of CO 2 assimilated through the CAM versus C 3 pathway in CAM plants, we also measured the extractable sugar fraction, which more closely reflects the photosynthetic discrimination against 13 CO 2 over the previous 1-2 days allowing an indication of how plants are responding to environmental conditions over a shorter period [bib_ref] University of Utah, Stable isotope ecology course, Salt Lake City Espejel I..., Ehleringer [/bib_ref] [bib_ref] Weather and climate controls over the seasonal carbon isotope dynamics of sugars..., Hu [/bib_ref]. The sugar extraction and purification was performed according to [bib_ref] Correlation between the carbon isotope discrimination in leaf starch and sugars of..., Brugnoli [/bib_ref] and modified as follows: approximately 50 mg of oven-dried leaf material was mixed with 5 cm 3 of distilled de-ionized water (vortex for few seconds), and centrifuged for 20 min at 12,000g. The supernatant was then decanted through ion-exchange resins (Dowex 50WX8-100, 1X2, Cl --form; Sigma-Aldrich, St. Louis, USA) to isolate the neutral fraction (mostly sugars). After freeze-drying the samples, 0.8-2.0 mg of dried sugars were prepared in sample tins for isotopic analysis. d 13 C was determined on an elemental analyzer (ANCA-SL; PDZ Europa, Crewe, UK) interfaced with a continuous flow isotope ratio mass spectrometer (model 20/20; PDZ Europa) at the University of California, Berkeley, Center for Stable Isotope Biogeochemistry (CSIB). Long-term external precision for these analyses are ±0.22% for d 13 C. The abundance of 13 C in each sample was calculated relative to the abundance of 13 C in internal standards that had been calibrated against Pee Dee belemnite. Relative abundance was determined using the relationship [bib_ref] University of Utah, Stable isotope ecology course, Salt Lake City Espejel I..., Ehleringer [/bib_ref] : d 13 C (%) = [( 13 C/ 12 C of sample)/( 13 C/ 12 C of standard)-1] 9 1,000.
# Statistical analysis
A three-way analysis of variance (ANOVA) with 'repeated measures' (ANOVAR; [bib_ref] The statistical analysis of ecophysiological response curves obtained from experiments involving repeated..., Potvin [/bib_ref] was used to compare the tissue acidity among light levels, seasons and species within each site. Differences in carbon isotopic composition among sites and between light levels were tested using a nested ANOVA, where the different light levels were nested within each site, and using mean species values for bulk tissue and sugar d 13 C. Means were compared with a Tukey test and all analyses were performed using Statistica v. 7.0 (StatSoft, Tulsa, OK, USA). Linear regression was utilized to describe the relationship of both bulk tissue and extracted sugar d 13 C values for Acanthocereus tetragonus as a function of annual rainfall.
# Results
## Relative importance of cam species
In the 500 mm site (coastal dune scrubland, San Benito), 10% of species had CAM, which represented 19% of the relative importance value (RIV) in the plant community. In this site, one CAM species (Agave angustifolia) was 1 of the 15 species with the highest RIV (16% RIV; , while Acanthocereus tetragonus and Opuntia dillenii contributed only 3% of the RIV in the community. In both tropical dry deciduous forest communities, 15% of species were CAM and represented 35 and 20% of the RIV for the 700 mm site (Dzibilchaltún) and the 1,000 mm site (Cuxtal), respectively. Four CAM species were among the 16 species with the highest importance value in these two communities: in the 700 mm site, these were Agave angustifolia, Nopalea inaperta, Bromelia karatas and Stenocereus eichlamii; whereas in the 1,000 mm site, these were S. eichlamii, N. inaperta, B. karatas, and Pereskiopsis scandens. Acanthocereus tetragonus was not within the 15 species with the highest RIV, but it was always among the 24 species of most importance in both forest communities .
## Nocturnal tissue acidification
Tissue acidification was significantly different among seasons in the three sites, with the greatest values during the rainy season (P \ 0.0001 for the three sites; [fig_ref] Table 1: Seasonal tissue acidity [/fig_ref]. In the 500 mm site, the lowest tissue acidification values were during the dry season, and in both wetter sites, the lowest values were during the early dry season. During the rainy season, tissue acidification was significantly different for plants growing in the two light microclimates in the three communities (P \ 0.0001 for the 500 mm site, and P \ 0.05 for the 700 mm and the 1,000 mm sites, respectively; [fig_ref] Table 1: Seasonal tissue acidity [/fig_ref]. At the 500 mm site, plants growing at 30% of total daily ambient PFD showed higher acidity values than those growing at 80% of total PFD (P \ 0.0001; [fig_ref] Table 1: Seasonal tissue acidity [/fig_ref]. In both forest communities, tissue acidification during the rainy season was greater for plants growing at 80% PFD than plants growing at 30% PFD (P = 0.020 and P = 0.047, for the 700 and 1,000 mm sites, respectively).
## Carbon isotopic composition of bulk tissue fractions
All plants in all sites showed bulk tissue leaf d 13 C values typical of strong CAM, less negative than -20% , with the least negative bulk tissue d 13 C values (-11 to -15%) in the 500 mm site and the most negative values in the 1,000 mm site . Plants growing in the 500 mm site were more enriched in 13 C (P \ 0.05), with mean d 13 C of bulk tissue approximately 2% less negative than plants growing in the forest communities. There was a significant negative relationship between annual rainfall and leaf d 13 C values for Acanthocereus tetragonus, the only species found in all three sites . No differences in bulk tissue d 13 C values were found between plants growing at 30 and 80% total daily ambient PFD in any of the three sites (P [ 0.05).
## Carbon isotopic composition of the sugar fraction
Similar to bulk tissue d 13 C values, in the 500 mm site, plants had sugar d 13 C values 2% less negative than plants in the other two sites (P \ 0.05). The standard deviation from the mean of the d 13 C values of the sugar fraction in each species showed that Acanthocereus tetragonus had greater d 13 C variability than the rest of the species . Similar to bulk tissue, d 13 C values of the sugar fraction for Acanthocereus tetragonus showed decreasing enrichment of 13 C of sugar with increasing precipitation . Moreover, A. tetragonus was the only species that showed significant differences in the d 13 C values of sugar between the plants growing at 80 and 30% of total daily PFD in the coastal dune scrubland; plants that had the least negative values of d 13 C were those growing at 80% PFD in comparison with plants growing at 30% PFD (P \ 0.0001; . In the other two communities, this species did not show differences in d 13 C of sugar between light microenvironments (P [ 0.05). Tissue acidification was significantly different between individuals of A. tetragonus growing at 80 and 30% of PFD in the coastal dune scrubland ; individuals with the lowest acidity values were those growing at 80% PFD (P \ 0.0001). No differences in tissue acidity were found between the plants of this species growing in both light conditions in the other two communities (P [ 0.05; .
# Discussion
Our data demonstrate that carbon gain by CAM plants along this precipitation gradient was greatest in the wet and Cuxtal (c, f). Each bar represents a 2% range of d 13 C season, indicating that seasonal water limitation has the potential to reduce carbon gain by approximately 75% at the relatively dry coastal dune site and by 50% in the wetter forest sites, when comparing tissue acidification in the dry season versus wet season. In addition, at the driest site, plants growing at lower light had greater rates of carbon gain, whereas in the two wetter sites, plants growing in the higher light microhabitat had greater rates of carbon gain. These results highlight a shift in the interaction between light and water availability along this gradient, in which we observed that, when water availability increases, plants were able to increase their tissue acidification at higher PFD. In addition, our d 13 C data showed greater proportional use of the nocturnal CO 2 uptake of the CAM species at the driest site . Moreover, species importance values and the proportion of CAM species in the communities increased from the coastal dune to the tropical dry forests, suggesting a more favorable balance of light and water availability for CAM performance in the wetter, forested communities. Tissue acidity is an indirect way to measure nocturnal CO 2 uptake, including direct CO 2 uptake and CO 2 produced during respiration and re-assimilation. Therefore, an increase in tissue acidity would indicate an increase in the CO 2 uptake [bib_ref] Achievable productivities of certain CAM plants: basis for high values compared with..., Nobel [/bib_ref] [bib_ref] Crassulacean acid metabolism, Osmond [/bib_ref]. Other CAM species also show an enhancement in their net CO 2 uptake and increase tissue acidity when well watered and a reduction of these two processes when droughted . The reduction of tissue acidity during the dry season can be caused by higher nocturnal temperatures and greater water deficits together with higher PFD levels than during the rainy season [bib_ref] Microenvironments, water relations, and productivity of CAM plants, Andrade [/bib_ref]. To avoid excess transpiration during the dry season, CAM species tend to show maximal stomatal opening later in the night reducing the amount of nocturnal increase of tissue acidity [bib_ref] Microhabitat and diel tissue acidity changes for two sympatric cactus species differing..., Nobel [/bib_ref] [bib_ref] Microhabitats and water relations of epiphytic cacti and ferns in a lowland..., Andrade [/bib_ref] [bib_ref] PSII photochemistry, thermal energy dissipation, and the xanthophyll cycle in Kalanchoe daigremontiana..., Lu [/bib_ref] [bib_ref] Light microhabitats, growth and photosynthesis of an epiphytic bromeliad in a tropical..., Cervantes [/bib_ref] [bib_ref] Photosynthesis and optimal light microhabitats for rare cactus, Mammillaria gaumeri, in two..., Cervera [/bib_ref]. Conversely, the reduction of tissue acidity during the early dry season could be caused by the reduction in the PFD (30 mol m -2 day -1 ) in comparison with that of the rainy and dry seasons (35 and 40 mol m -2 day -1 , respectively). Although it has been reported that, for several desert CAM species, nocturnal acid accumulation increases up to a total daily PFD of about 30 mol m -2 day -1 and any reduction in PFD reduces net CO 2 uptake, lower water vapor pressure deficits in our study sites than in the desert could increase that PFD limit, allowing the plants to use more light without losing more water.
In the 500 mm site, plants growing at 30% of the total daily ambient PFD had greater tissue acidification than plants growing at 80% ambient PFD [fig_ref] Table 1: Seasonal tissue acidity [/fig_ref]. Nevertheless, in both forest communities, the greatest increase in tissue acidity was for plants growing at 80% PFD. These differences among plants of the three sites are likely due to differences in soil water potentials and water vapor pressure deficits. For example, after 30 days of drought during the dry season, soil water potential 15 cm below the surface was -22.4 ± 2.3 MPa for a coastal dune scrubland and -19.8 ± 1.6 MPa for a tropical dry deciduous forest ). In addition, in some CAM species, under light-saturating conditions, an increase in the photosynthetic capacity as a response to growth in high light without a proportional increase in the malate supply would predispose some plants to photoinhibition, causing drought stress to increase the negative effects of the high radiation [bib_ref] High photosynthetic capacity in a shade-tolerant crassulacean acid metabolism plant, Skillman [/bib_ref]. In our study, during the dry season, most species of the coastal dune site showed the lowest tissue acidity values and less negative d 13 C values than all other species for the two communities, indicating stomatal closure during phases II and IV of the CAM cycle and no contribution of RUBISCO to CO 2 uptake.
CAM species that showed significant differences in the tissue acidity growing in the two light microenvironments also had the highest relative importance value in the three communities. One exception was Acanthocereus tetragonus , which was the only species in the three communities with strong CAM photosynthesis and a high photosynthetic plasticity in the coastal dune. It has been proposed that, given the range in the magnitude of CAM expression that can be induced by water limitation in different species, the ecological significance of CAM induction would depend on its relative importance as a survival mechanism, rather than the direct contribution to growth and productivity of the species [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref] [bib_ref] Induction of Crassulacean acid metabolism by water limitation, Cushman [/bib_ref] [bib_ref] Ecophysiology of crassulacean acid metabolism (CAM), Lüttge [/bib_ref]. [bib_ref] Implications of genotypic diversity and phenotypic plasticity in the ecophysiological success of..., Kluge [/bib_ref] found that CAM species of Madagascar with the highest photosynthetic plasticity were the species most widely distributed across the island, and thus occupied a greater variety of environments, compared to CAM species with low photosynthetic plasticity, which occupied a more restricted set of environments.
CAM plants in the three communities showed bulk tissue d 13 C values typical of strong CAM (d 13 C values less negative than -20%; [bib_ref] ) d 13 C values and crassulacean acid metabolism in Clusia species..., Holtum [/bib_ref] [bib_ref] Distribution of crassulacean acid metabolism in orchids of Panama: evidence of selection..., Silvera [/bib_ref] , but the d 13 C values in plants from the coastal dune scrubland were 2% less negative than those from plants of the forest communities . The bulk tissue and sugar d 13 C values for Acanthocereus tetragonus, which occurred in the three sites, showed a significant negative relationship with rainfall . These data suggest that plants at drier sites decreased transpiration and stomatal conductance relative to photosynthesis, which would lead to the measured increase in d 13 C values. Moreover, the duration of the phase II of CAM decreases under water stress [bib_ref] Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial..., Griffiths [/bib_ref] , which would also decrease the isotope discrimination, leading to less negative d 13 C values. Similar results have been found in terrestrial plants from desert and tropical dry forest, and for epiphytes in a tropical cloud forest and three different secondary tropical dry forests where the carbon isotope ratios also increased with decreasing water availability [bib_ref] Correlations between carbon isotope ratio and microhabitat in desert plants, Ehleringer [/bib_ref] [bib_ref] Carbon isotope ratios of a tropical dry forest in Mexico, Mooney [/bib_ref] [bib_ref] Stable isotopic composition of carbon and nitrogen and nitrogen content in vascular..., Hietz [/bib_ref] [bib_ref] ) d 13 C soluble sugars in Tillandsia epiphytes vary in response..., Goode [/bib_ref].
It has been proposed that the flexibility of CAM provides an advantage for acquisition of ecological niches [bib_ref] Ecophysiology of crassulacean acid metabolism (CAM), Lüttge [/bib_ref] [bib_ref] Crassulacean acid metabolism and fitness under water deficit stress: if not for..., Herrera [/bib_ref]. In the present study, although all the species investigated in the three communities showed d 13 C values typical of strong CAM, one species, A. tetragonus, which inhabited all three communities, showed the highest variability in d 13 C values . Members of the family Cactaceae have been considered as constitutive or obligate CAM species even at their early stages (Hernández-González and Briones-Villarreal 2007); although some studies suggest that the contribution of CAM to total carbon gain is small during the first developmental stages [bib_ref] On the nature of facultative and constitutive CAM: environmental and developmental control..., Winter [/bib_ref]. The high variability of the strong CAM species, A. tetragonus, suggests that some cacti are able to respond to changes in the environment and inhabit more environments than other cacti. More detailed field and laboratory studies on different life stages of these strong CAM plants would be required to fully understand their ecological role in these tropical plant communities.
[fig] Figure 3, Figure 4, Figure 5: Relationship between bulk tissue (a) and sugar (b) d 13 C values for Acanthocereus tetragonus and mean annual rainfall in San Benito (500 mm rainfall), Dzibilchaltún (700 mm) and Cuxtal (1,000 mm). Each point represents a sample for an individual during the rainy season (n = 9) Mean d 13 C values (closed symbols) of sugar ± standard deviation and range (open symbols) of 10 CAM species. Key for the species code: Acanthocereus tetragonus (At), Agave angustifolia (Aa), Bromelia karatas (Bk), Nopalea inaperta (Ni), Stenocereus eichlamii (Se), Pilosocereus gaumeri (Pg), Pereskiopsis scandens (Ps), Opuntia dillenii (Od), Selenicereus donkelaarii (Sd), and Tillandsia dasyliriifolia (Td) Values of d 13 C of sugars (a) and tissue acidity (b) in three populations of Acanthocereus tetragonus in San Benito (circles), Dzibilchaltún (triangles) and Cuxtal (squares) growing at 80% of total daily photon flux density (PFD) (open symbols) or at 30% total PFD (closed symbols) [/fig]
[table] Table 1: Seasonal tissue acidity (DH ? ; mmol H ? m -2 ) of CAM species growing at 80 and 30% of total daily photon flux density (PFD) in a coastal dune and in two tropical dry deciduous forests [/table]
|
Reaching the Challenging Diagnosis of Complicated Liver Hydatid Disease: A Single Institution’s Experience from an Endemic Area
Citation: Christodoulidis, G.; Samara, A.A.; Diamantis, A.; Floros, T.; Sgantzou, I.-K.; Karakantas, K.-S.; Zotos, P.-A.; Koutras, A.; Janho, M.B.; Tepetes, K. Reaching the Challenging Diagnosis of Complicated Liver Hydatid Disease: A Single Institution's Experience from an Endemic Area. Medicina 2021, 57, 1210. https://doi.Abstract:Background and Objectives: Hydatid disease (HD) remains a significant public health issue causing morbidity and mortality in many Mediterranean countries. Material and Methods: The present cohort study included 50 consecutive patients with liver hydatid disease who underwent surgery in a tertiary University Hospital. A total of 18 patients (36%) had a case of complicated HD, including simple communication of the cyst with the biliary tree (6 cases), rupture of the cyst into the biliary tree (6 cases), presence of a bronco-biliary fistula (2 cases), rupture of the cyst in the peritoneal cavity (2 cases), and rupture of the cyst and formation of a hepatic abscess (2 cases). Endoscopic retrograde cholangiopancreatography (ERCP) was pre-operatively performed on six patients. Results: The main clinical symptom presented was right upper quadrant pain in 16 patients (88%), which was associated with high fever (>39 - C) in 14 patients (78%). C-reactive protein (CRP) was the primary indicator of a complicated HD (p = 0.003); however, it was only elevated in 67% of cases. CRP was a more sensitive indicator of a rupture in the biliary tree cyst (p = 0.02). Computer tomography (CT) detected more cases (44%) of a complicated HD than ultrasonography (US) (25%); however, the difference was not statistically significant. Conclusions: For prevention and control of HD, a high suspicion of the disease leading to early referral to specialized centers, mainly in endemic areas, is required. Prior to surgical or percutaneous intervention, a combination of imaging and laboratory findings are essential in diagnosing a complicated case and avoiding unnecessary interventions.
# Introduction
Hydatid disease (HD) or echinococcosis is a chronic parasitic disease caused by the larval stage of the tapeworm Echinococcus granulosus. To date, it remains a severe public health issue with a considerable economic impact. This disease contributes to morbidity and mortality in many parts of the world, primarily Mediterranean countries, the Middle East, New Zealand, Australia, India, and South America, due to the close association between sheep, dogs, and humans [bib_ref] Unusual imaging characteristics of complicated hydatid disease, Turgut [/bib_ref] [bib_ref] Hydatid Disease of the Liver: A Continuing Surgical Problem, Safioleas [/bib_ref] [bib_ref] World review of laparoscopic treatment of liver cystic echinococcosis-914 patients, Tuxun [/bib_ref] [bib_ref] Diagnosis and surgical treatment of complicated liver echinococcosis, Agayev [/bib_ref] [bib_ref] Laparoscopic management of hydatid cyst of the liver, Yagmur [/bib_ref]. Humans are considered intermediate hosts that become infected by Echinococcus inadvertently. Intermediate hosts are infected by ingesting eggs within the faeces of the definite hosts. The eggs liberate their larvae into the duodenum, crossing the intestinal wall and, via the portal system, reach the liver, where they develop into cysts [bib_ref] Treatment options for hepatic cystic echinococcosis, Smego [/bib_ref] [bib_ref] Hydatid disease of the liver, Shaw [/bib_ref] [bib_ref] Human Echinococcosis: A Neglected Disease, Silva [/bib_ref] [bib_ref] Management of hepatic hydatidosis by open versus laparoscopic surgery, Darbari [/bib_ref].
HD can involve almost every part of the body; however, the liver is the most commonly affected organ (75%), followed by the lungs (15%) and other locations, such as the brain (2%) or the spine (1%) [bib_ref] Unusual imaging characteristics of complicated hydatid disease, Turgut [/bib_ref] [bib_ref] World review of laparoscopic treatment of liver cystic echinococcosis-914 patients, Tuxun [/bib_ref] [bib_ref] Surgical management of complicated hydatid cysts of the liver, Malik [/bib_ref]. Without treatment, hydatic cysts gradually grow and can develop fistulas with adjacent organs or the biliary tree. The hydatic cysts may also rupture into the peritoneal cavity causing an anaphylactic reaction and the seeding of secondary cysts, develop daughter cysts within, or die [bib_ref] Surgical management of complicated hydatid cysts of the liver, Malik [/bib_ref]. The rupture of a hydatid cyst can occur in 50-90% of cases, leading to an allergic response due to the presence of antigenic cyst content in the surrounding structures. These clinical syndromes may lead to a patient's death, but treatment does not ensure improvement of the patient's quality of life [bib_ref] Controversies in the laparoscopic treatment of hepatic hydatid disease, Acarli [/bib_ref] [bib_ref] Management of liver hydatid cysts-current perspectives, Anand [/bib_ref] [bib_ref] Surgical Treatment of Liver Echinococcosis-Open or Laparoscopic Surgery?, Busić [/bib_ref].
The aim of the present study is to evaluate the diagnostic importance of several biochemical and radiological values in the diagnosis of complicated liver hydatid disease.
# Materials and methods
## Study design-settings
All consecutive patients who were admitted to a tertiary University Hospital diagnosed with hydatid liver disease during a 7-year period (between 2009 and 2016) were included in the present study. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist was used [bib_ref] The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines..., Von Elm [/bib_ref]. Ethical review and approval were waived for this study, due to the retrospective study design based primarily on medical records and laboratory results, as well as the anonymity of patients in the statistical analysis.
## Variables of the participants
In this retrospective analysis, we reviewed the medical records of 50 patients that underwent surgery due to echinococcal liver disease. A total of 18 (36%) patients were presented as a complicated case, with the final diagnosis set intra-operatively.
Parameters considered for the evaluation of these patients included the following: age, gender, co-morbidities, clinical presentation, physical examination, laboratory findings, imaging modalities (such as ultrasonography (US), computer tomography (CT), magnetic resonance (MRI), and other interventions, including Endoscopic Retrograde Cholangiopancreatography (ECRP) or Magnetic Resonance Cholangiopancreatography (MRCP)), duration of hospitalization, complications, mortality, type of surgery performed, and follow-up. Laboratory findings included serum leucocyte and eosinophile count, C-reactive protein (CRP), aspartate and alanine transaminases (AST and ALT, respectively), γ-glutamic transaminase (γGT), alkaline phosphatase (ALP), total and direct bilirubin (TB, DB respectively), and anti-echinococcal antibodies. Data regarding the number, size (maximum diameter), and location of the cyst were also evaluated.
Communication or rupture of the cyst within the biliary tree was suspected preoperatively due to clinical presentation (i.e., obstructive jaundice or acute cholangitis), laboratory findings (i.e., increased levels of bilirubin or cholestatic enzymes), and/or positive imaging findings (i.e., dilatation of common biliary duct with or without the presence of material within it). A broncho-biliary fistula was suspected due to the presence of yellowish sputum and dyspnea on an otherwise fit and well patient. However, in all 18 cases, the final diagnosis was set intra-operatively and was confirmed by post-operative histological examination.
# Statistical analysis
Data were processed using SPSS V26 (IBM, Chicago, IL, USA). Qualitative variables are presented as absolute (N) and relative (%) frequencies, and quantitative variables are presented as means with standard deviation (SD). Statistical analysis for categorical variables was performed using χ 2 and Fischer exacts test, and student's t-test for descriptive variables. A p-value of <0.05 was considered statistically significant.
# Results
A total of 18 of the 50 (36%) patients that underwent surgery for a hydatic cyst located in the liver were diagnosed with a complicated hydatic cyst and were included in the present study. The patients' demographic and basic characteristics are displayed in [fig_ref] Table 1: Basic characteristics of patients regarding co-morbidities, clinical presentation and cyst characteristics [/fig_ref]. The mean patient age was 56.3 years (range 25-82 years); 14 (78%) patients were male and 4 (22%) were female. In 12 cases (67%), there was a communication of the cyst with the biliary tree, either as simple a communication of the cyst with the biliary tree (CBT, 6 cases) or as a rupture of the cyst into the biliary tree (RBT, 6 cases). There were 2 cases of bronco-biliary fistula (BBF, 11%), 2 cases of a cyst rupture in the peritoneal cavity (RPC, 11%), and 2 cases of cyst rupture and formation of a hepatic abscess (11%). Furthermore, the mean number of cysts was 1.3 (range 1-4) with a mean maximum diameter of 9.6 cm (range 4.5-20 cm). In one case, the cyst was enormous and occupied the whole right lobe and the left medial part of the liver [fig_ref] Figure 1: An enormous hydatid cyst occupying the whole right lobe and the left... [/fig_ref] , while in another case a large right liver lobe cyst compressed the hilum resulting in left liver lobe biliary dilatation [fig_ref] Figure 2: A large hydatid cyst of the right lobe of liver compressing the... [/fig_ref].
It is noteworthy that in 9 cases (50%), the diagnosis of a complicated liver hydatid cyst was unclear at the time of referral. One case of broncho-biliary communication was initially regarded as an exacerbation of Chronic Obstructive Pulmonary Disease (COPD), while two further cases were considered as an exacerbation of tuberculosis. A total of 4 patients (22%) had previously received medical treatment for liver hydatid disease, but had missed their follow-up. Moreover, two cases were misdiagnosed as Klatskin or adrenal tumor.
The main clinical symptom presented was right upper quadrant pain in 16 patients (88%), which was associated with high fever (>39 - C) in 14 patients (78%). A total of 2 patients (16%) with involvement of the biliary tree presented with the typical signs of acute cholangitis, whereas both patients with BBF presented with yellowish sputum along with upper right quadrant pain and fever. One patient with RPC presented with acute abdomen, without specific signs of hydatid disease. All patients had non-typical symptomatology [fig_ref] Table 1: Basic characteristics of patients regarding co-morbidities, clinical presentation and cyst characteristics [/fig_ref]. [fig_ref] Table 2: Cases of elevated biochemical values classified by definitive intra-operative diagnosis [/fig_ref] presents the number of cases with increased biochemical values, classified by the type of complicated hydatic disease. Anti-echinococcal antibodies were measured in only 5 patients (28%) and were positive in 4 of them (80% of measured cases). When comparing the main biochemistry examination, CRP was the main indicator of a complicated HD (p = 0.003) as it was elevated in 12 cases (67%), followed by ALP/γGT that was elevated in 11 patients (61%), eosinophils count that increased in 10 cases (56%), and total bilirubin levels that were elevated in only 2 cases (11%) of rupture in the biliary tree. Furthermore, CRP was a more sensitive indicator of a rupture in the biliary tree cyst (increased in 5 cases, 83%) when compared to the increase in cases of simple communication with the biliary tree (increased in only 1 case, 17%) (p = 0.02).
## Total (18)
Bilirubin 2/6 (33%) 0/6 (0%) 0/2 (0%) 0/2 (0%) 0/2 (0%) 2/18 (11%) ALP/γGT 4/6 (66%) 4/6 (66%) 2/2 (100%) 1/2 (50%) 0/2 (0%) 11/18 (61%) CRP 5/6 (83%) 1/6 (17%) 2/2 (100%) 2/2 (100%) 2/2 (100%) 12/18 (67%) Eosinophiles 2/6 (33%) 2/6 (33%) 2/2 (100%) 2/2 (100%) 2/2 (100%) 10/18 (56%) [fig_ref] Table 3: Diagnostic value of preoperative image modalities classified by definitive intra-operative diagnosis [/fig_ref] summarizes the diagnostic value of preoperative image modalities classified by definitive intra-operative diagnosis. More specifically, an ultrasound scan was performed in 16 patients (89%) (excluding patients with a broncho-biliary fistula) and was diagnostic of intrabiliary rupture in 2 cases, which was 16% of total cases that involved the biliary tree and 33% of all cases with a cyst ruptured into the biliary tree. Ultrasound scanning was also diagnostic in 2 CBT cases (33%). Overall, the ultrasound scan was diagnostic in 33% of cases with a cyst involved in the biliary tree (4/12 patients). An ultrasound scan was also performed in RPC patients, with no diagnostic value; therefore, an abdominal CT scan was performed to establish the diagnosis. A CT scan was performed in all (18) complicated cases, of which 8 cases (44%) were diagnosed with complicated HD. More precisely, the CT scan was diagnostic in 2 RBT cases (33%). It should be noted that, in one case, the hydatid mass was reported as an adrenal mass. CT scanning also failed to identify the simple communication of the cyst with the biliary tree, as it was non-diagnostic in 5 out of 6 cases (83%). For BBF and RPC cases, a CT scan was diagnostic in all cases, which, apart from the hepatic lesion, showed the diaphragmatic erosion, the infiltration of the lung and the bronchi, and the presence of pericystic abdominal fluid. In cases of hepatic abscess, a CT scan was diagnostic in one case only, presenting air within the cyst. Dilatation of the bile ducts were observed in 2 cases (11%) and paracystic fluid collection was present in all cases of complicated HD and hydrothorax, in both cases with a broncho-biliary fistula.
Additional imaging modalities included MRI/MRCP. These modalities were performed as a means of further investigation when US and CT could not lead to a definitive diagnosis in two cases of RBT. In both cases (100%), a definitive diagnosis of a common bile duct cyst was set with MRI/MRCP. When comparing the diagnostic accuracy of CT and US, CT detected a greater number of complicated HD cases (8/18, 44%) than US (4/16, 25%). However, the difference was not statistically significant (p = 0.2). A statistically significant difference (p = 0.045) was found in cases in which the cyst was ruptured in the peritoneal (PRC), with CT being diagnostic in all cases (2/2) and US diagnostic in none (0/2). Moreover, CT detected a case of AF (50%), while US could detect none (0/2), but the difference was not statistically significant (p = 0.08).
Pre-operative ERCP was performed in a total of 6 cases (33%). This procedure was conducted in both cases of BBF and in the four cases in which a rupture was highly suspected by imaging findings.
# Discussion
Hydatid disease remains a serious public health challenge contributing to morbidity and mortality worldwide, particularly in Mediterranean countries where the disease is endemic. In these regions, the annual incidence of HD ranges from less than 1 to 200 per 100,000 inhabitants [bib_ref] Human Echinococcosis: A Neglected Disease, Silva [/bib_ref]. Greece is a highly endemic area for echinococcosis, with an annual incidence of 0.3 per 100,000 inhabitants [bib_ref] Cystic echinococcosis in Greece. Past and present, Sotiraki [/bib_ref]. Echinococcosis is not only a severe disease that may result in death (mortality rate 1-2%), but also a disease that can lead to severe economic losses due to treatment costs, lost wages, and losses in the livestock industry. It has been estimated that disability-adjusted life year (DALY) losses are approximately US $193,529,740 globally [bib_ref] Global Socioeconomic Impact of Cystic Echinococcosis, Budke [/bib_ref] [bib_ref] Economic Effects of Echinococcosis, Torgerson [/bib_ref].
Unexpectedly, over a 7-year period, only 50 cases of hydatid disease were referred to our University hospital, despite the hospital serving a largely rural population of approximately 700,000. Eighteen of these patients presented to our unit as complicated cases of hydatid disease. Their delayed presentation at the University hospital can be attributed to either an underestimation by primary healthcare physicians of initial presenting symptoms, or failure on the part of the patients to follow medical advice received. This highlights both the low level of suspicion and the population's lack of knowledge related to the disease, despite the region of Thessaly being a leading Mediterranean endemic area. Public education/awareness campaigns should be developed to inform and educate populations living in endemic areas for the early recognition of HD symptoms and to seek timely medical advice. Furthermore, for prevention and control of the disease the World Health Organization (WHO) suggests periodic dog deworming, improving hygienic conditions in slaughterhouses, and the vaccination of livestock with an E. granulosus recombinant antigen (EG95).
Liver hydatid disease is frequently silent, with approximately 75% of patients experiencing an asymptomatic abdominal mass incidentally diagnosed during abdominal investigation for another pathology. In symptomatic cases, the most common symptom displayed is epigastric pain or pain located in the right upper abdominal quadrant, and typical findings include an enlarged liver and a palpable mass. In many cases, patients seek medical advice at a late stage when complications are already established. These complications include rupture in the biliary tree, infection, rupture into viscera, intraperitoneal or intrathoracic rupture, and external compression leading to portal hypertension [bib_ref] Hydatid disease of the liver, Shaw [/bib_ref] [bib_ref] Surgical management of complicated hydatid cysts of the liver, Malik [/bib_ref] [bib_ref] The role of omentoplasty in the surgical management of remnant cavity in..., Muftuoglu [/bib_ref]. Rupture can occur because of trauma or relentless expansion of the cyst resulting rupture into the peritoneal cavity, the pleural cavity or the bile duct. When located in the superior and posterior liver segments, HD can grow upwards towards the chest, eroding through the diaphragm and resulting in fever, cough, and bile-stained sputum. Rupture into the peritoneal cavity is accompanied by shock and signs of diffuse peritonitis [bib_ref] Surgical management of complicated hydatid cysts of the liver, Malik [/bib_ref]. Laboratory findings can be useful in the diagnosis of liver hydatid disease, although they are not always conclusive as indicators for biliary tree involvement. Differential diagnosis of simple communication from rupture into the biliary tree cannot rely on bilirubin levels or other liver tests [bib_ref] Diagnosis and surgical treatment of complicated liver echinococcosis, Agayev [/bib_ref]. Moreover, the use of the anti-echinococcal antibodies in diagnosing the disease remains unclear [bib_ref] Antibody responses of patients with cystic hydatid disease to recombinant myophilin of..., Martin [/bib_ref]. In our series, anti-echinococcal antibodies were tested in only five patients with a complicated disease. The eosinophile count was slightly elevated in most complicated cases, but failed to differentiate rupture from simple communication. CRP provided a diagnostic indicator in RBT cases. This may be due to the fact that the rupture of the cyst is a prerequisite for the contents to be released into the biliary tree, and the rupture results in an inflammatory reaction. Furthermore, Gram negative bacteria may infect the ruptured cyst causing an inflammatory reaction, which is unlikely to happen when simple communication occurs. In our study, CRP was the main indicator of a complicated HD (p = 0.003); however, it was elevated in only 67% of cases. Therefore, without a specific blood test for the diagnosis of liver HD, a wide range of biochemical values should be investigated to support the diagnosis [bib_ref] Postoperative changes in serum cytokines profile and nitric oxide levels in patients..., Refik [/bib_ref] [bib_ref] Advances in Liver Echinococcosis: Diagnosis and Treatment, Filippou [/bib_ref].
Radiology imaging is the main diagnostic tool for the diagnosis of HD. This tool can detect the disease even in asymptomatic cases, or cases with mild and non-specific symptoms, which may not be easily evaluated via clinical examination. Ultrasonography and a CT scan are typically used in diagnosis of liver cystic lesions. In our study, CT detected more cases (44%) of a complicated HD than US (25%); however, the difference was not statistically significant. In cases of broncho-biliary fistula and free intraperitoneal rupture, CT was proven to be adequate for diagnosis; however, it was inconclusive in terms of biliary tree involvement. Moreover, CT offers a better view of the hydatid cysts regarding location and relationship with surrounding structures, while US is a non-invasive, costeffective method that can be used as an initial diagnostic tool for cystic lesions of the liver. HD cysts appear as lobulated structures containing daughter vesicles or membranes and septa, while the wall may appear partially or heavily calcified [bib_ref] Hydatid Disease of the Liver: A Continuing Surgical Problem, Safioleas [/bib_ref] [bib_ref] Hydatid disease of the liver, Shaw [/bib_ref] [bib_ref] The role of omentoplasty in the surgical management of remnant cavity in..., Muftuoglu [/bib_ref]. Furthermore, our experience indicates that a preoperative MRI/MRCP should not be performed as a routine, but in cases with an obvious communication with the biliary tree. Prior to a surgical or percutaneous intervention, a combination of imaging and laboratory findings are essential in diagnosing a complicated case and avoiding unnecessary interventions.
Endoscopic retrograde cholangiopancreatography (ERCP) remains an important diagnostic tool in cases of major biliary communication, while simultaneously allowing the clearance of the common bile duct and reducing the risk of a post-operative fistula [bib_ref] Hydatid disease of the liver, Shaw [/bib_ref] [bib_ref] Endoscopic retrograde cholangiography for intrabiliary rupture of hydatid cyst, Galati [/bib_ref]. Many authors suggest that ERCP should be performed in every case with clinical or biochemical suspicion of biliary involvement, before or after any surgical procedure [bib_ref] Endoscopic retrograde cholangiography for intrabiliary rupture of hydatid cyst, Galati [/bib_ref]. ERCP is a safe and effective procedure in means of avoiding strictures, removal of endocystic material, closure of fistulas, and reduction of the possibility of another operation [bib_ref] Endoscopic stenting for selected cases of biliary fistula after hepatic hydatid surgery, Akcakaya [/bib_ref] [bib_ref] Effectiveness of Endoscopic Treatment Modalities in Complicated Hepatic Hydatid Disease after Surgical..., Saritas [/bib_ref]. The majority of our cases included involvement of the biliary tree, which was indicated by both biochemical and imaging findings. In our study, pre-operative ERCP was performed in 6 cases (33%) of biliary communication, with favorable results for the patients.
The most common complication in the management of hydatid disease is cyst rupture into the biliary tree or the thoracic cavity. Biliary communications can occur with a frequency ranging from 3.5% to 19% [bib_ref] Hydatid Disease of the Liver: A Continuing Surgical Problem, Safioleas [/bib_ref] [bib_ref] Controversies in the laparoscopic treatment of hepatic hydatid disease, Acarli [/bib_ref]. Less common complications involve bronchobiliary fistulas and a rupture of the cyst in the peritoneum. Broncho-biliary fistula may present with a simple cough, bile-stained sputum, or even hydatidemesia [bib_ref] Treatment of Hydatid Cyst of the Liver: Where Is the Evidence?, Dziri [/bib_ref] [bib_ref] Open drainage versus overlapping method in the treatment of hepatic hydatid cyst..., Yol [/bib_ref]. A rupture of the cyst in the abdomen, which result from trauma or high intra-abdominal pressure, is a severe situation and should always be treated vigorously with emergency laparotomy [bib_ref] Successful treatmne tof intractable bronchobiliary fistula using long-term biliary stenting, Katsinelos [/bib_ref] [bib_ref] Treatment of postoperative bronchobiliary fistula by nasobiliary drainage, Patrinou [/bib_ref].
To the best of our knowledge, the present study identified the diagnostic value of several biochemical markers and evaluated radiological modalities to diagnose complicated liver hydatid disease for the first time in the literature. However, prior to the appraisal of these results, several limitations should be considered. Data were collected retrospectively depending on the availability and accuracy of the data records, and information bias may have occurred. Due to the limited number of patients, the results of the statistical analysis should be interpreted cautiously.
# Conclusions
In conclusion, echinococcosis remains a serious public health issue leading to impaired quality of life for affected people and even death. Health providers should be highly suspicious of liver hydatid disease in endemic areas, and should refer any suspected case to a specialized center for definitive diagnosis and treatment. The diagnosis of a complicated liver HD can be established through clinical, biochemical (mainly with CRP), and imaging findings (US and CT).
## Institutional review board statement:
The manuscript meets the all the criteria of the Declarationof Helsinki. Ethical review and approval were waived for this study, due to the retrospective study design based primarily on medical records and laboratory results, as well as the anonymity of patients in the statistical analysis.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
[fig] Figure 1: An enormous hydatid cyst occupying the whole right lobe and the left medial part of the liver. [/fig]
[fig] Figure 2: A large hydatid cyst of the right lobe of liver compressing the hilum, resulting in left liver lobe biliary dilatation. [/fig]
[fig] Author: Contributions: Conceptualization, K.T.; formal analysis, A.A.S.; investigation, G.C., A.A.S., A.D., T.F., I.-K.S., K.-S.K., P.-A.Z., A.K. and M.B.J.; supervision, K.T.; writing-original draft, G.C., A.A.S., A.D. and T.F.; writing-review & editing, I.-K.S., K.-S.K., P.-A.Z., A.K., M.B.J. and K.T. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research received no external funding. [/fig]
[table] Table 1: Basic characteristics of patients regarding co-morbidities, clinical presentation and cyst characteristics. [/table]
[table] Table 2: Cases of elevated biochemical values classified by definitive intra-operative diagnosis. [/table]
[table] Table 3: Diagnostic value of preoperative image modalities classified by definitive intra-operative diagnosis. [/table]
|
Analyzing the Effect of Time in Migration Measurement Using Georeferenced Digital Trace Data
Georeferenced digital trace data offer unprecedented flexibility in migration estimation. Because of their high temporal granularity, many migration estimates can be generated from the same data set by changing the definition parameters. Yet despite the growing application of digital trace data to migration research, strategies for taking advantage of their temporal granularity remain largely underdeveloped. In this paper, we provide a general framework for converting digital trace data into estimates of migration transitions and for systematically analyzing their variation along a quasi-continuous time scale, analogous to a survival function. From migration theory, we develop two simple hypotheses regarding how we expect our estimated migration transition functions to behave. We then test our hypotheses on simulated data and empirical data from three platforms in two internal migration contexts: geotagged Tweets and Gowalla check-ins in the United States, and cell-phone call detail records in Senegal. Our results demonstrate the need for evaluating the This is an open access article distributed under the terms of a Creative Commons license (CC BY-NC-ND 4.0).
# Introduction
Issues of data availability and comparability have long hampered quantitative migration research. Missing or incomplete data is a common problem, and inconsistencies in how migration is defined by institutions can make migration estimates from traditional sources such as population censuses, administrative records, or survey data difficult to synthesize across context [bib_ref] Overcoming the problems of inconsistent international migration data: A new method applied..., De Beer [/bib_ref] [bib_ref] Reconciling and translating migration data collected over time intervals of differing widths, Rogers [/bib_ref]. Faced with a persistent lack of comprehensive migration data, demographers and other social scientists have begun to draw on new sources of digital data for the study of migration. Among the promising new sources are digital trace data [bib_ref] Dynamic population mapping using mobile phone data, Deville [/bib_ref] [bib_ref] Characterizing urban landscapes using geolocated tweets, Frias-Martinez [/bib_ref] [bib_ref] Geo-located twitter as proxy for global mobility patterns, Hawelka [/bib_ref]. Generated as a byproduct of everyday information technology use, digital trace data consist of individual-level records of digital behavior, which may include information on a person's physical location [bib_ref] Collecting vertical trace data: Big possibilities and big challenges for multi-method research, Menchen-Trevino [/bib_ref]. With the global proliferation of digital technology, digital trace data are increasingly common and are available in a wide range of forms that are potentially useful to migration scholars, such as metadata associated with cellular calls and texts, GPS information captured passively by smartphone applications, and geotags posted to social media or other location-based social networks (LBSNs) [bib_ref] Digital footprinting: Uncovering tourists with user-generated content, Girardin [/bib_ref].
However, in their raw, unprocessed state, locational digital trace data do not correspond to any meaningful measure of migration. Instead, they represent millions upon millions of high-resolution locational traces, each of which is a record of a unique individual at a particular place at a particular time. This fundamental characteristic of digital trace data both poses challenges and offers new opportunities for research. There are no standards or best practices for how to convert highly granular locational trace information into estimates of migration transitions or events. Not all moves are migrations. Thus, how researchers choose to operationalize the largely ambiguous distinction between migration and other kinds of movement (e.g., long-distance commuting, tourism, seasonal travel) will greatly affect the consistency of migration estimates generated from digital trace data. Although it is wellestablished that these new forms of data come with bias, linking digital trace estimates of migration with traditional survey or administrative estimates is not necessarily the only validation strategy. As we demonstrate in this paper, much can be learned about the bias and coverage of these data by systematically assessing the consistency of migration estimates with respect to definition parameters.
Although analyzing digital trace data poses methodological challenges, the unique granularity of these data also provides researchers with a novel opportunity to address substantive questions about the spatial and temporal dimensions of migration phenomena. Much of this work is well underway. For example, research on internal migration has historically been limited by inconsistent and arbitrary subnational administrative geographies, such as state or province borders that bisect metropolitan regions or important rural-urban gradients contained within a single administrative region. By mining for patterns in individual mobility traces or activity spaces, researchers have developed a number of approaches that exploit the high spatial resolution of digital trace data to reveal more meaningful spatial scales that define different kinds of movement [bib_ref] Redefining neighborhoods using common destinations: Social characteristics of activity spaces and home..., Jones [/bib_ref] [bib_ref] New approaches to human mobility: Using mobile phones for demographic research, Palmer [/bib_ref]. But as we argue in this paper, the granularity of digital trace data is not limited to spatial granularity. The temporal definitions used in migration research can also be inconsistent or arbitrary. Thus, much can be learned about the temporal scales that define certain kinds of movement by leveraging the temporal granularity offered by these new kinds of data.
In this paper, we offer a general framework for converting digital trace data into migration estimates. Conceptually coherent and simple to implement, the goal of the framework is to produce migration estimates that follow the logic and structure of migration transition data -that is, to estimate the share of individuals in a population who undergo a transition from one place of residence to another over a given time interval.
When implemented systematically, our framework has two applications. First, it can be used to evaluate the consistency of migration estimates derived from a given source of digital trace data. Second, provided that the digital trace migration estimates are deemed to be of sufficient quality, this framework can be used to investigate substantive issues related to population movement and time. The strategy in both applications is to generate many different estimates of migration from the same data source and then to assess how the estimates vary with respect to a quasi-continuous time parameter.
This kind of analysis should look familiar to demographers because it resembles a survival function-that is, the proportion of a population who become migrants (or remain nonmigrants) with increasing time-and it can be used to test simple hypotheses. Migration researchers have long theorized that migration rates should go up with increased exposure to the risk of moving [bib_ref] The measurement of migration, from census data and other sources, Rees [/bib_ref]. A five-year rate is almost always higher than a one-year rate in surveys that collect migration data using both intervals. However, by systematically applying our framework to digital trace data, we can produce many intermediate estimates and empirically investigate the cumulative effect of exposure to the risk of migration for different populations and geographies. This kind of approach is novel in migration research. Moreover, if we find that our migration estimates are inconsistent with respect to time (e.g., if a migration flow between two regions defined by a span of 12 months differs considerably from the flow defined by the same 12-month span plus two weeks), we can consider this evidence that our data and our definition of migration are producing problematic estimates and that we need to do more to identify and parse out short-term moves (e.g., travel) in the estimation procedure. We show how this can be done by adjusting the temporal place of residency criterion.
In the sections that follow, we begin by outlining the conceptual difficulties involved in defining migration with respect to time as well as the related challenges that can arise when studying migration with digital trace data. We introduce our framework for converting digital trace data into migration transition data and discuss our hypotheses. To provide a heuristic against which to compare our empirical results, we propose a simple stochastic model and evaluate its properties with micro-simulation. We then apply our method to three unique data sets in two internal migration contexts: call detail record (CDR) data in Senegal, and Twitter and Gowalla data in the United States. Finally, we briefly demonstrate one way that our method could be used to compare and contrast the geographic patterns of short-term mobility and long-term migration flows.
Because digital trace research is novel and data coverage remains an issue, we do not attempt to draw definite substantive conclusions from our findings. Instead, by applying our method to three data sets of varying quality from different contexts, we focus on the kinds of insights that this framework can reveal. Nevertheless, our findings should be of interest to researchers involved in developing standards for inferring migration from digital trace data and in deepening our understanding of the spatial and temporal dimensions of migration phenomena and migration measurement. In this era of continuously changing and increasingly heterogeneous spatial-temporal patterns of population movement, it is unlikely that the data problems migration researchers face will be resolved easily. However, when analyzed using the framework introduced in this paper, digital trace data can provide timely insights at a level of detail and with a degree of flexibility that can greatly improve efforts to infer migration patterns and advance our understanding of complex migration phenomena.
# Background
Studying migration entails measuring the movement of people in space and time. However, distinguishing migration from other kinds of mobility can be difficult and ultimately depends on the purpose of the measurement. In terms of space, defining migration is complicated by the varying social, political, and economic meanings of the different geographic units between which people move. International migration is important for political reasons and is simple to conceptualize as a relocation across a national border. However, internal migration is more common globally and can have causes and effects similar to those of international migration [bib_ref] Mind the gap! Integrating approaches to internal and international migration, King [/bib_ref]. The task of identifying meaningful geographies of migration at subnational scales is not straightforward, and comparisons of internal migration patterns across contexts are often hampered by differing standards [bib_ref] Migration distances: An international comparison, Long [/bib_ref].
In terms of time, defining migration is complicated by the varying frequencies with which people move. People may engage in return or onward migration or move around for short periods as evidenced by the growing phenomena of circulation and short-term or temporary mobility. There is no theoretically grounded definition of permanence [bib_ref] Tourism, migration, circulation and mobility: The contingencies of time and place, Williams [/bib_ref]. Even individuals who have lived away from their country of birth for many years might someday return [bib_ref] Theorising return migration: The conceptual approach to return migrants revisited, Cassarino [/bib_ref] , and increasing numbers of people split their time between multiple locations. To determine when a person becomes a migrant, rather than temporary visitor, governments often rely on a length-of-stay criterion, such as 12 months. However, such criteria are arbitrary and differ from context to context. Scholars of migration have long been aware of the complexities of measuring migration, but their ability to systematically investigate this issue has been constrained by a lack of highquality longitudinal data. Because most survey data are cross-sectional, these data can generally be used to estimate migration at no more than one or two intervals (e.g., the place of residence at the time of the survey is compared with the reported place of residence one year ago or five years ago). This limitation has the effect of masking the degree and the character of short-term mobility and repeat migration behavior.
In some cases, researchers have used panel survey data to study the characteristics of repeat migrants. For example, using data from the Panel Study on Income Dynamics, [bib_ref] Repeat migration in the United States: Who moves back and who moves..., Davanzo [/bib_ref] found that migrants with higher levels of education are less likely to return and more likely to move onward to a third location. Although panel survey data are valuable for understanding cohort and life course migration dynamics, they typically cannot be used to provide estimates of population-level migration trends because of their small sample sizes. Administrative data such as address registries or tax records can be used to investigate the spatial and temporal dynamics of migration measurement. For example, [bib_ref] The extent of repeated migration: An analysis based on the Danish population..., Goldstein [/bib_ref] used Danish registry data to demonstrate that repeat migration behavior is common among young adults. More recently, Weber and Saarela (2019) used linked registry data from Finland and Sweden to show that many moves among young adults are temporary and short term. In general, however, access to population registry data is difficult to obtain, and not all countries keep accurate administrative records.
## Challenges of measuring migration with digital trace data
Given the limitations of traditional migration data, georeferenced digital trace data are a boon to migration scholars because of their size, simple structure, and high spatial and temporal resolution. These data come from many sources and have become increasingly abundant with the growing adoption of telecommunication technologies across the globe. Digital trace data can be collected actively when people post their location using LBSNs, such as Yelp, Foursquare, or Instagram; or passively when people use telecommunication technology to make calls, send text messages, or use smartphones and web applications. The structure of these data-that is, tuples consisting of individual ID, timestamp, and locationis the same regardless of the platform or service from which they originate, and the levels of spatial and temporal detail that they provide afford researchers a high degree of flexibility in deciding how to measure migration. However, using digital trace data in migration research poses conceptual challenges that reflect both the longstanding definitional issues discussed earlier and the indirect manner in which the data are collected.
## Defining place of residence
An obvious problem that can arise when estimating migration using georeferenced digital trace data is that these data contain no direct information on an individual's place of residence. Unlike a survey (which might ask respondents about their residential history) or an administrative records system (which requires individuals to register each change of address), digital trace data simply log the location of an individual at a particular moment in time. Without further context-specific information, it is difficult to determine whether the position of a given individual corresponds to place of residence. Thus, to determine whether an individual is a migrant, it is first necessary to make some inferences about where the individual typically spends time.
In recent years, published studies have provided techniques for inferring residency from georeferenced digital trace data. [bib_ref] Understanding individual human mobility patterns, Gonzalez [/bib_ref] used a sample of 100,000 mobilephone users to demonstrate the regularity with which most individuals spend time at home and work. These regular patterns made it possible to assign individuals to well-defined areas. The authors offered one method for doing so: namely, summarizing the location of each individual during a specified period and calculating radius of gyration and center of mass defined by their movement trajectory. Others have taken this idea further, attempting to use digital trace data to generate separate inferences regarding an individual's home, work, and other ancillary locations. For example, to estimate the home location of individuals from a sample of 3 million mobile-phone users in Singapore, [bib_ref] Activity-based human mobility patterns inferred from mobile phone data: A case study..., Jiang [/bib_ref] restricted each individual's set of positions to those occurring at night (from 7 p.m. to 7 a.m.). Then, having linked each mobile-phone user to a home location, the authors demonstrated how to identify different kinds of daily mobility trajectories, ranging from staying home to moving between several places throughout the city.
The decision about how locational information should be summarized depends on the quality of the data and the research objectives. Although researchers would ideally have access to information about the daily activity spaces of individuals in a given sample, this may not be feasible if individuals are irregularly observed or if the geographic information is coarse, as is the case with many social media-generated geotags. Moreover, for the purposes of migration research, daily activity information is often unnecessary and can make the desired migration behavior more difficult to parse. Not all administrative units are arbitrary, and there are ways to make use of the spatial and temporal granularity that these new data provide without using them to estimate activity spaces. Nevertheless, because digital trace data contain no direct information on an individual's place of residence, researchers will always need to group multiple observations to infer the place of residence regardless of whether the strategy is to map activity spaces or simply to assign each individual to their most frequented administrative unit.
## Issues with coverage bias
The more commonly discussed challenge that can arise when using digital trace data is coverage bias. The penetration of various digital technologies and platforms is uneven. In many cases, digital trace data are not accompanied by additional demographic information. Moreover, it can be difficult to assess the one-to-one relationship between a user ID number and an individual. Cell-phone sharing is common [bib_ref] Who's calling? Demographics of mobile phone use in Rwanda, Blumenstock [/bib_ref] , and social media data can contain bots and business accounts. Complicating matters further, digital trace data can also produce biased estimates of migration if the platform or service through which they are generated is associated with certain kinds of mobility behavior. For example, [bib_ref] Choosing the right home location definition method for the given dataset, Bojic [/bib_ref] showed that because geotags from Flickr (a photo-sharing social media platform) tend to capture travel and vacation activity, it is difficult to use these data to accurately identify the user's home location.
Separating issues of bias with respect to digital trace data from the conceptual ambiguities that complicate migration measurement, however, is important regardless of the data source. Even if it were possible to obtain digital trace data containing the accurate location of every individual at every minute of the day, a person's migration behavior would not be selfevident. In this scenario of total surveillance, there would still be a need to apply methods, rooted in migration theory, to determine which kinds of mobility behavior in the data meet which definitions of migration.
A potential upside of digital trace data is that they can still be useful in migration research even if they are of lower quality. For many applications, the individual-level accuracy of digital trace data is arguably less important than the population-level migration signal. The field of demography has a long history of working with biased or incomplete data, and as we argue in the following section, the bias of a given digital trace data set can be evaluated by assessing the extent to which it produces migration estimates that are consistent.
## Conceptual framework and method
In this section, we introduce our framework for converting digital trace data into estimates of migration. The overarching goal of the framework is to isolate three interrelated but distinct temporal dimensions that define a migration transition estimate: the start or reference point, the temporal buffer or residency window, and the interval or exposure period. By systematically altering each dimension, researchers can assess the consistency or sensitivity of the migration estimates generated from a given digital trace data set. Based on migration theory, we develop two simple hypotheses for how we expect migration estimates to behave. Testing these hypotheses has applications for both data quality assurance and substantive analysis. Before we introduce the framework, we explain our decision to estimate migration transitions.
## Event data and transition data
In the migration literature, researchers make a distinction between event data and transition data. Event data consist of information pertaining to relocations that have occurred over a given period. Transition data consist of information on the population that has relocated over a given period. As we have discussed, unprocessed digital trace data meet neither of these definitions. Each observation is simply a record of an individual in a specific location at a specific time. Because transition data are more commonly estimated in survey data and used by demographers to study population change [bib_ref] Concept, measurement, and data in migration analysis, Haenszel [/bib_ref] , we propose a framework for estimating transitions. Going forward, unless otherwise specified, a migration refers to a transition, and a migrant refers to a person who has transitioned. A migration rate refers to the proportion of people who have migrated.
We estimate transitions rather than events to simplify the conceptual scope of the problem. Estimating a transition involves determining whether an individual's place of residence at time t is different from place of residence at some interval u in the future. By making this our goal, we avoid trying to determine how many migrations have occurred within a digital trace data set over a specified period. If an individual moves more than once over an interval, a given estimate of migration will capture one move at most. If this individual is a return migrant, having moved away and come back over the interval, then we would count that person (incorrectly) as a nonmigrant. Our argument is that by varying the specificationproducing many migration estimates with a range of intervals-we can assess the effect of return migration on the population-level migration signal.
## The start, buffer, and interval approach
The logic of our framework is simple. First, we specify a reference date or start. Then, for all individuals in the data, we infer the person's residency for some specified window or temporal buffer around the start. Next, we select as a second reference date some specified period or interval in the future and infer the person's residency around that date using a temporal buffer of the same size. Finally, with estimates of each individual's place of residence at two distinct points in time to compare-one at the start and the other at the end of the interval-we determine whether the individual is a migrant or nonmigrant. [fig_ref] Figure 1: Fig. 1. [/fig_ref] illustrates the implementation of our framework to a hypothetical timeline corresponding to an individual who travels back and forth between two U.S. states, New York (NY) and Florida (FL). In the top row of each panel, the specification of the start, the buffer and the interval are the same. In the bottom row, we show how one dimension can be changed while the other two are held fixed.
The strength of this framework is its flexibility. As long as digital trace data exist for a population and period of interest, many different estimates of migration can be calculated by systematically changing the start-buffer-interval specification. The only rule that must be followed when using this approach is that the interval must be of greater length than the buffer size. If, for example, we want to estimate the number of transitions over a threemonth interval, we cannot use six months of data to estimate the place of residence at the beginning and the end of the interval. To do so would result in a double counting of observations and a conflation of the effects of interval and the buffer size. Nevertheless, by isolating three distinct temporal dimensions of migration measurement, our framework makes it possible to assess the consistency of estimates generated from a given set of digital trace data in conceptually coherent way: (1) the start measures the effects of seasonal or period trends; (2) the buffer measures the sensitivity to temporal residency criteria; and (3) the interval measures the effects of exposure to the risk of migrating.
## Evaluating the consistency of digital trace migration estimates
Now that we have established how our framework can be used to generate many different estimates of migration from the same set of digital trace data, the question becomes how to evaluate the output. Our strategy is to analyze migration estimates in a manner equivalent to that of a survival function. This strategy involves assessing how the number of estimated migrants changes as the interval changes. In other words, we identify the specific set of individuals observed living in place i at reference point t and then evaluate the proportion of individuals in this population who have left place i as the interval, u, increases. This simple
## Author manuscript
Author Manuscript
## Author manuscript
Author Manuscript analytic strategy leads us to propose two interrelated hypotheses about regularities in migration measurement.
## Consistency and interval
First, we expect to find that migration estimates increase as the interval increases. The logic behind this expectation is straightforward. As the population that resides in a particular place is exposed to the risk of migrating, the number of people who migrate away should also increase. Although this hypothesis might seem so obvious as to be of little value, testing it on a given set of digital trace data is useful for assessing data quality. If the data coverage is poor or if the underlying behaviors that generate the data are biased toward other kinds of mobility (such as travel), then the migration estimates will likely be irregular with respect to interval. Thus, instead of observing a slow increase in the rate of migration, we might see sharp spikes or a multimodal trend line as the migration signal is obscured by periodic shortterm mobility and returns.
Moreover, if the digital trace data are deemed to be of sufficient quality, then analyzing the relationship between the migration estimates and the interval will provide useful and novel information for characterizing migration dynamics. Although migration estimates have long been theorized to increase with increased exposure to the risk of moving, empirical data on the precise functional relationship between migration estimates and intervals are scarce. Based on data from surveys that estimated migration using both a one-year and a five-year interval, migration scholars have observed that the relationship is nonlinear: that is, five-year estimates tend to be higher than one-year estimates but are not five times as high [bib_ref] The one year/five year migration problem, Kitsul [/bib_ref] [bib_ref] The measurement of migration, from census data and other sources, Rees [/bib_ref]. Moreover, considerable variation in the relationship between one-and five-year estimates has been observed across contexts [bib_ref] Reconciling and translating migration data collected over time intervals of differing widths, Rogers [/bib_ref] , and evidence of inconsistencies between the spatial structure of migration measured with oneyear and five-year intervals suggests that there are different patterns of return and onward migration [bib_ref] Migration analysis using data with time intervals of differing widths, Rogerson [/bib_ref]. By leveraging the temporal granularity of digital trace data, our framework can provide insight into this so-called one-year/five-year problem in migration estimation.
## Consistency and buffer
Second, we expect to find that migration estimates are higher and more inconsistent when they are produced with smaller temporal buffers. The logic behind this expectation follows that of the first. Measuring migration transitions using digital trace data requires us to infer each individual's place of residence at two points in time and then to determine how many individuals have relocated over the interval. If we use a very small window of time on either side of the interval to infer each individual's place of residence, we would expect to capture both short-term moves (i.e., tourism; long-distance commuting; or travel for work, education, or family) and long-term moves (i.e., migration) in our estimate. This would make the estimate higher than if we use a larger buffer size to screen out short-term moves. Moreover, because short-term moves are characterized by return behavior-it is only a short-term emigration if the person comes back-we expect to find that small buffer estimates are multimodal with respect to the interval. For example, the number of people observed at a location other than their place of residence might spike during a holiday period and then decline as most of these people return to their place of residence when the holiday is over.
This hypothesis might seem self-evident, but it also has valuable applications for assessing data quality. As researchers have begun exploring the use of digital trace data in migration research, a common validation goal has been to compare digital trace estimates with traditional survey or administrative estimates of migration. However, the best way to produce comparable estimates has yet to be established. For example, if a survey conducted on March 1, 2015, asked respondents where they currently reside and where they resided one year prior, producing a similar digital trace estimate for validation would entail inferring the place of residence on March 1, 2014, and March 1, 2015, for each individual in the data set. Although this may seem straightforward, it is unclear how much data on either of the intervals is needed to sufficiently screen out short-term moves occurring around those two dates. Using only the locational information from one day at either end of the interval-March 1, 2014, and March 1, 2015-would likely be insufficient. Would it be better to use a week? Two weeks? A whole month? The answer to this line of inquiry will depend on the quality of a given data set, and we argue that investigating the relationship between temporal buffer size and the consistency of migration estimates will help make these kinds of determinations.
Studying the effect of the buffer size also provides a useful framework for evaluating different residency criteria and residency inference methods. As we stated previously, because of the fundamentally atomistic nature of digital trace data, we can make inferences about each individual's place of residence only by grouping some of her observations. This is true regardless of whether the geography of residence is predefined (e.g., national borders) or mined from the data (such as an activity space, established by individual commuting trajectories). The simplicity of the buffer concept means that any number of functions can be used to infer residency within a buffer, or to compare buffers to assess whether a migration has occurred. For example, following [bib_ref] Migration as a spatial and temporal process, Roseman [/bib_ref] , we could extract the spatial polygon defining each user's activity space within a buffer and then identify migrants as those whose activity spaces at either end of the interval fail to overlap. How migration estimates vary with respect to buffer size can provide information on how different residency inference methods perform. It is possible, for example, that with only one day's worth of information on either end of an interval, most methods for inferring residency will perform the same. However, as the buffer size is increased and more data is incorporated in the inference procedure, the differences between methods should become more pronounced. Because the quality of digital trace data can vary considerably, this kind of analysis would be useful for justifying a particular analytical approach.
## Research questions
Our two research questions extend from our discussion of interval and buffer size. These research questions, which are simple and easy to evaluate, represent the primary application of our framework. We expect to find that the number of people who migrate from their place of residence will increase as the interval increases because of their added exposure to the risk of migration. Although the strength of this relationship should diminish at long intervals because of return migration, we expect to observe a largely positive relationship between the interval and the migration estimate.
- Research Question 2: Do migration estimates decrease and become more consistent as the buffer size increases? We expect to find that the number of people who have migrated from their place of residence will decrease as buffer size increases. With larger buffer sizes comes a larger amount of data that can be used to infer location at either end of the interval, increasing our ability to accurately estimate long-term relocations by screening out short-term moves.
How a particular digital trace data set performs with respect to these research questions will provide useful information on the suitability of the data for migration research. Moreover, once it has been established that the data are of sufficient quality, studying the specific relationship between migration estimates and interval or buffer size will deepen our understanding of the temporal complexities of migration phenomena and their measurement. Empirical data on how migration estimates change with respect to a quasi-continuous time interval could be used to address the one-year/five-year problem by allowing researchers to chart the specific functional relationship between population-level migration behavior and exposure to the risk of migrating. At the same time, empirical data on how migration estimates change with respect to buffer size could be used to evaluate different techniques for inferring residency and provide a basis for further analysis of the relationship between patterns of short-term mobility and patterns of long-term migration.
Answering these research questions does not, however, entail validating digital trace migration estimates using traditional estimates or assessing bias using some external source of more trusted data. Although this kind of validation can be seen as an essential component of any substantive digital trace migration research, we argue that it unnecessary for the purposes of demonstrating the utility of our framework. This argument rests on two points. First, given the lack of standards governing how highly granular digital trace data should be converted into migration estimates, we suggest that evaluating the consistency of such estimates should be considered a preliminary step to validating with traditional survey or administrative estimates. Only after it has been established that a given data set produces consistent estimates should an attempt be made to link these data to any other kind of data. Second, we expect that most digital trace data biases would not hamper our ability to validate the internal consistency of the migration estimates they produce. For example, even if the users on a particular social media platform are disproportionately young, we would still hypothesize that estimates of their migration activity will increase as the interval increases. It is for this latter reason that we apply our method to simulated data and empirical data of varying quality and from different contexts. Although we expect that both of our hypotheses will be confirmed using these different kinds of data, precisely how well they perform is a key insight that will be provided by the application of our framework.
## Author manuscript
Author Manuscript
## Author manuscript
Author Manuscript
## The simulation model
Having outlined our framework and our research questions, we present a simple simulation model to produce data that will meet our stated expectations. The goal of the model is not to simulate a specific context or replicate precise patterns in our empirical data. Instead, the purpose of the model is to explore, using simple behavioral assumptions, how short-term mobility and migration might be manifested within individual-level time-and-place data. The simulated data will provide a point of reference against which we can evaluate the patterns observed in the empirical data introduced in the next section.
## Strategy
We simulate data that take the form of tuples (individual ID, timestamp, and location). The structure of each tuple is simple and mimics the format of the locational digital trace data discussed in this paper. A single tuple does not provide much information about where a given individual resides. However, a series of these tuples over time for the same individual offers insights into patterns of residency and into patterns of mobility and migration between different places. The underlying assumption of our model is that each individual has a latent characteristic: namely, a home location (such as a U.S. state) that conditions the individual's mobility behavior. Individuals will be observed most often in their home locations; however, individuals who are observed away from their homes for a sufficiently long period may be assumed to have changed their home locations.
In our approach, we simulate timelines for a population of m individuals. Each individual has known location l at each unit of time 1, 2, … , t such that individual i can be represented by a vector:
[formula] l i, 1 , l i, 2 , … , l i, t , [/formula]
where l i,t is the location of individual i at time t.
We then build a model in which units of time are equivalent to one week (i.e., an individual is observed only once per week), with only two possible locations, 1 or 0. The probability that individual i is observed at either 1 or 0 at time t is represented by a simple Bernoulli random variable conditional on the individual's "home" attribute, which can also only be 1 or 0. This gives us two conditions: P l i, t | ℎome = 1 = p, for l i, t = 1 1 − p, for l i, t = 0 and P l i, t | ℎome = 0 = p, for l i, t = 0 1 − p, for l i, t = 1 . Although the decision to use independent Bernoulli random trials to model short-term mobility rests on strong assumptions, we chose this method for its simplicity. Empirical evidence has demonstrated that the duration of temporary moves skews heavily toward shorter lengths of time, and that the probability of observing consecutive Bernoulli values reasonably replicates the smaller likelihood of taking extended trips (e.g., three months) relative to taking shorter trips (e.g., one week). In future research, more realistic distributions of short-term mobility could be inferred directly from the empirical data. But given that here we are using our model only as a heuristic for evaluating our empirical data, we argue that relying on a Bernoulli distribution will suffice for now.
To model long-term relocation, we add an additional feature. If an individual is observed "away" from "home" for k consecutive weeks, then the probabilities associated with being observed in the location designated as "away" become those previously associated with being in the location once designated as "home":
if l i, t + 1 = … = l i, t + k = 0 | ℎome = 1, then 0 ℎome if l i, t + 1 = … = l i, t + k = 1 | ℎome = 0, then 1 ℎome .
For example, take a scenario in which we observe a set of individuals for whom the probability of being home in a given week, p, is equal to .7, and the threshold of relocation, k, is equal to 4. If we observe these individuals for 100 weeks, then the rate of transition should be approximated by the probability of a streak of four or more consecutive weeks away occurring in 100 Bernoulli trials. This value can be obtained using recursion with the following formula:
[formula] S(N, K) = 1 − p K + ∑ j = 1 K 1 − p j − 1 p S N − j, K , [/formula]
where (1 − p) is the probability of being observed away from home on a given week, S(N,K) is the probability of being observed K or more consecutive weeks away from home out of N weeks, and j is the position of the first week an individual is observed at home [bib_ref] The first occurrence of n successes in N trials, Greenberg [/bib_ref]. Either we observe an individual away from home K consecutive weeks in the first K weeks (which has the probability (1 − p) K ), or we observe the individual at home at least once in the first K weeks (at position j). In the latter case, the probability of going away for K or more weeks is equal to the probability of doing so following the jth week. Using the values from our example, this formula returns a value of .433.
## A simulated outcome
Continuing with the preceding example, we simulate 1,000 individual timelines with the probability of being home on a given week, p, equal to .7; and the long-term move threshold, k, equal to 4. Each individual is observed for 100 weeks. We then derive many different migration rates from the simulated data by systematically changing the start-buffer-interval specification. As we would do when using empirical data, we estimate that a simulated individual is a migrant if place of residence at the start of the interval is not the same as place of residence at the end of the interval. In this case, we infer the place of residence by calculating the modal location-either at home or away from home-during the buffer. If there is a tie, we take the first location to hit the maximum.
In each of the three panels in [fig_ref] Figure 2: Fig. 2. [/fig_ref] , we track how the migration rate changes as the interval increases, while holding the buffer fixed at one of three different values: 1, 4, or 12 "weeks." The y-axis is the proportion of movers-the migration rate-and the x-axis is time. A line represents a set of rates derived using a common start, which, when followed left to right, tracks the proportion of migrants as the interval grows. (For a schematic, refer to the right-hand panel of [fig_ref] Figure 1: Fig. 1. [/fig_ref] The lines are plotted such that their position over the xaxis corresponds to the date associated with the end of the interval, and shading indicates start date, with later starts being darker. The start value is also plotted at the base of each line. [fig_ref] Figure 2: Fig. 2. [/fig_ref] illustrates how we expect migration estimates to vary as we systematically change the buffer and interval size. When the buffer is small, at either 1 (left panel) or 4 (center panel), the observed rates of migration are high and multimodal. We set the conditions of our model (p = .7) such that individuals exhibit a high degree of short-term mobility.
Although the overall rates appear to increase slightly as the interval increases, the rate of long-term relocation is somewhat masked by the short-term noise. When the buffer is increased to 12 (right panel), the signal associated with short-term mobility is mostly removed. Very few individuals are observed away from home more than 6 times in 12 tries unless they have relocated; thus, less short-term return migration is observed. In this plot, the trend lines start lower but rise consistently as the interval increases. Taken together, the three panels of [fig_ref] Figure 2: Fig. 2. [/fig_ref] illustrate what we expect to find in our empirical data. As the buffer size increases, the high levels of migration and mulitmodality resulting from short-term mobility are reduced. Thus, we see a consistent, positive relationship with respect to interval.
## Data
The empirical portion of the analysis is conducted on three data sets: call detail records (CDR) in Senegal from the telecommunication company Orange-Sonatel; and two sets of social media data in the United States, one from Twitter and other from Gowalla. In this section, we describe the three data sets and how the differences between them could affect our estimation. Because a major motivation for developing our framework is to take advantage of the temporal granularity of digital trace data, we downplay the potential uses of the spatial granularity of these data in our subsequent demonstration and analysis. However, as we noted previously, any number of sophisticated spatial techniques could be used within a buffer to infer location. For all three data sets, we infer the place of residence by simply assigning individuals to the administrative unit in which they most frequently observedtheir modal administrative unit. In the case of a tie, we assign the individual to the first administrative unit to achieve the maximum level observed.
## Orange-sonatel
Our analysis makes use of anonymized CDRs obtained via the Orange-Sonatel 2014 Data for Development (D4D) Challenge. These data consist of phone calls and SMS exchanges between more than nine million Orange-Sonatel customers in Senegal between January 1, 2013, and December 31, 2013. The specific data set used in this paper is a subset of the CDRs corresponding to a random sample of 150,000 subscribers. Only the 146,352 users meeting the following criteria were included in the sample:
## 1.
Users having interactions on more than 75% of days in the given period.
## 2.
Users having had an average of fewer than 1,000 interactions per week (given that users with more than 1,000 interactions per week were presumed to be machines or shared phones).
The data consist of roughly 561 million CDRs representing all the interactions (voice call or SMS) placed or received by these users in the 2013 calendar year . Each record provides a numerical pseudonym representing the user, the timestamp of the call, and the arrondissement (third administrative level) in which the user was located when the interaction took place. The D4D data set divides Senegal's territory into 123 arrondissements, which can be grouped into 45 départements or 14 régions. The latter, the 14 régions of Senegal, are the geographic units used in this paper. We define migrant as an individual who changes régions over a given interval. Because these administrative units are so small, especially relative to the U.S. administrative units used in the Twitter and Gowalla analysis, we may expect much higher estimates of migration and mobility in Senegal. On the other hand, the consistency with which the cell-phone users are observed might make the estimates derived from this data set more stable.
## Twitter
We also analyze a large set of Twitter data extracted from a long-term archive of the 1% Twitter stream sample. Only Twitter accounts with at least one tweet geotagged within the United States between January 2011 and December 2014 are included. Because of top-down changes Twitter made in early 2015 to the kinds of locational information contained within tweets, the data collected after 2014 are not directly comparable and are therefore excluded from the analysis [bib_ref] State of the geotags: Motivations and recent changes, Tasse [/bib_ref]. We use the latitude and the longitude associated with each tweet to place it in one of nine U.S. Census divisions, 1 and we define migration as a change in residency with respect to these divisions.
The Twitter data consist of roughly 447 million geolocated tweets from 1.9 million Twitter users spanning 2011-2014 . The mean number of geotagged tweets per user is 267, but the users are not necessarily observed every week. On average, a user's tweets appear in 24 weeks, spread over a range of about one year. Sporadic tweeting means that under certain migration specifications, some user timelines will exhibit either left-or rightcensoring. In cases in which data are missing at either or both ends of an interval given a particular buffer, the user is excluded from both the numerator and the denominator for that particular migration estimate. Unlike for the Orange-Sonatel CDR data, we make no effort to limit our sample to those individuals who were consistently observed over the period. Instead, we take a maximalist approach to see how well our hypotheses hold up when our framework is applied to a very imperfect data set. Thus, we are interested in determining the extent to which we can use increased buffer size to screen out short-term mobility.
## Gowalla
Finally, we analyze geolocated data captured through the check-in feature of the mobile geosocial network Gowalla. Similar to the more familiar check-in app Foursquare, Gowalla was a short-lived platform on which users shared their locations. For each post, Gowalla stored user ID information, a timestamp, and latitude/longitude coordinate data-sufficient information for estimating how many users changed location over time. These lesser-known Gowalla data have also been used by other researchers, including [bib_ref] Friendship and mobility: User movement in location-based social networks, Cho [/bib_ref] , who used the check-ins to conduct an analysis of short-term mobility and social networks. This data set includes 6,442,890 check-ins generated from 107,092 unique users between November 2010 and October 2011. The mean number of check-ins per user is 60. On average, a user's check-ins appear in nine weeks, spread across three months. As with the U.S. Twitter data, we use U.S. Census divisions as the geographic unit of analysis, and we estimate internal migration that occurred between divisions. Like for the Twitter data, we make no exclusions based on how consistently the users are observed in the data set. Instead, we are interested in determining whether our hypotheses still hold when applied to a small, noisy data set.
# Empirical results
In the following section, we first compare broad patterns from our empirical results with the patterns from our simulation model. We then describe the differences between the empirical results. Finally, we briefly discuss how short-term mobility estimates might be used to indirectly infer long-term migration.
## Comparison with simulated results
We perform the same start-buffer-interval analysis on our empirical data that we previously performed on our simulated data. Results are presented in [fig_ref] Figure 4: Fig. 4. [/fig_ref]. The rows correspond to the three data sets: CDRs from Orange-Sonatel in Senegal; geotagged tweets from Twitter in the United States; and Gowalla check-ins, also from the United States. Each column shows a set of migration estimates produced while holding the buffer fixed at 1 week, 4 weeks, or 12 weeks, respectively. In each panel, the y-axis is migration rate, and the x-axis is the time in weeks. Followed left to right, each line tracks the proportion of movers as the interval grows from a common start. Although there are 9 U.S. Census divisions and 14 régions in Senegal, we calculate a total migration rate as the proportion of all individuals observed at a given start who changed location (division or région) over a given interval.
We are encouraged to report that the patterns that emerge from all three data sets are broadly similar and that our two expectations are generally confirmed. First, the level of migration observed increases as the interval increases. Although the trend is less obvious with a smaller buffer, it is still apparent. Second, increasing the buffer size both lowers the rates and reduces their multimodality with respect to time.
There is one obvious difference between these findings and our simulated results. In our model, the levels of migration estimated with a small buffer are consistently elevated, and the effect of increasing the buffer size is dramatic. Although we find some evidence of lower rates due to increases in the buffer size in the empirical data, smaller buffer estimates are not as consistently high. This discrepancy is partially due to the unrealistic assumption in the simulation of a uniform likelihood of short-term mobility. There are clear seasonal patterns, particularly in the Twitter estimates, that show when temporary mobility is more likely. If further simulation models are to be developed, more sophisticated strategies for modeling short-term mobility will be needed.
## Comparisons across data sets
Given the differences in the contexts, the platforms, and the quality of the data in the three data sets, some interesting comparisons between them can be made. The smoothness of the estimates from Orange-Sonatel is surprising given that the administrative geography of the Senegalese régions is more fine-grained than that of the U.S. Census divisions. If short-term mobility is conditioned by distance, then we would expect to observe higher and more multimodal patterns of migration when the total distance an individual needs to cross a boundary is smaller. This is clearly not the case when we look at the results from Orange-Sonatel. Unlike the noisier estimates derived from Twitter and Gowalla, the Orange-Sonatel estimates lose their multimodal pattern as the buffer increases from one week to four weeks, likely because the regularity with which Orange-Sonatel users are observed makes it easier to remove the short-term mobility signal from the migration estimates.
The nearly four-year span of the Twitter data demonstrates that it is possible to analyze longer-term patterns and seasonal regularities. Across all three panels of Twitter estimates, it is apparent that there are seasonal bumps in short-term mobility. The consistency of these patterns is especially notable considering that the total number of Twitter users generating the data grew steadily over span of the data. A visual inspection of the broader trend in the Twitter data indicates that the relationship between migration level and the length of the interval is positive and almost linear, with the longest observed migration rates increasing as the interval grows, regardless of the choice of buffer size. This near-linear trend suggests little long-term return migration among our sample of Twitter users. As some individuals return to the place where they were first observed, we would expect the rate of increase with respect to interval to diminish. After all, survey and administrative data suggest that fiveyear migration estimates are almost always less than five times the corresponding one-year estimates [bib_ref] The measurement of migration, from census data and other sources, Rees [/bib_ref]. In comparison, the smooth hump in the Orange-Sonatel estimates [fig_ref] Figure 4: Fig. 4. [/fig_ref] , top center and top right) resembles the pattern we would expect to observe as return migration diminishes the rate of increase, ultimately lowering the rate of migration observed with increased interval. Because the Orange-Sonatel data span just one calendar year, we can draw no conclusions about longer-term trends in the Senegalese context. However, results from this analysis show how short-term seasonal migration behavior in Senegal impacts the relationship between the migration estimate and the interval.
Meanwhile, although the Gowalla estimates do not prove any additional insight per se, they do suggest that even sparse, low-quality digital trace data can contain an internally consistent migration signal. [fig_ref] Figure 5: Fig. 5. [/fig_ref] examines the geography of short-term and long-term mobility across the three data sets. Bilateral emigration rates are estimated for each possible pair of administrative units in the U.S. and Senegal contexts, respectively, using two temporal specifications: one monthover-month rate (e.g., place of residence in January vs. place of residence in February) and one six-month-over-six-month rate (e.g., place of residence from January to June vs. place of residence from July to December). This gives a short-term and a long-term estimate for 72 bilateral emigration flows in the United States (nine divisions × eight divisions) and for 182 bilateral emigration flows in Senegal (14 régions × 13 régions). When we plot the six-month rates against the one-month rates, we see a positive linear trend across all three data sets. If relatively few people in region i moved to region j over a one-month span, it is likely that relatively few people would have moved from i to j over a six-month span. Although this finding is preliminary, it suggests the novel possibility of modeling long-term rates between regions, which are harder to observe, using short-term rates. Our framework provides the conceptual flexibility necessary for investigating the relationship between short-term and long-term patterns of migration.
## Discussion and next steps
Migration data are expensive to collect, and different institutions use different temporal definitions of migration to meet their various needs. The harmonization of migration values estimated using different temporal definitions is difficult, which hampers the study of migration phenomena. These harmonization challenges arise in large part from complications due to return migration (i.e., the one-year/five-year problem) or temporary mobility (i.e., difficulties reconciling migration data defined by a different lengths of stay). Although our understanding of the temporal complexities of migration phenomena and measurement has historically been limited by scarce data, the availability of novel and increasingly prevalent forms of digital data creates opportunities for new research. In this paper, we have provided a series of methodological and theoretical advances that will make it easier to use digital trace data in migration research while leveraging their temporal granularity.
We developed a flexible method for converting georeferenced digital trace data, such as LBSN data or CDR data, into migration transition data. To address the growing number of ad hoc approaches in the literature, our method contributes a much-needed general terminology-start, buffer, interval-for this kind of inference procedure. We argue that these concepts refer to three distinct temporal dimensions of migration measurement: start measures seasonal-or period-specific effects; buffer measures residency criteria effects; and interval measures cumulative temporal effects. Because of the atomistic and indirect quality of digital trace data, all three concepts must be addressed to produce migration transition estimates, regardless of the quality or the geographic coverage of the data.
We then proposed an approach to assessing the quality and the characteristics of the migration signal present in a data source by systematically changing the temporal dimensions of the migration definition. In essence, this approach can be used to produce sets of estimates that track levels of migration along a quasi-continuous time scale analogous to a survival function. We applied this approach to simulated data as well as three empirical data sets from three platforms in two internal migration contexts. Our analysis addressed two research questions, premised on two hypotheses related to migration measurement.
First, we asked how measures of migration vary with respect to the time interval. Based on the literature on the one-year/five-year problem in migration measurement, we expected to find that migration estimates with larger intervals would be higher than the estimates with smaller intervals. Our simulation model illustrated this intuition, and our empirical analysis of three data sets confirmed it. At first blush, this finding may seem self-evident. However, because estimating migration along a pseudo-continuous time scale has never been done before, this result has important implications for further research. Now that we have demonstrated that migration rates increase as the interval increases, we can begin asking questions about the specific functional form of this relationship and about the extent to which it varies by geography or population characteristics. This approach also provides a means for assessing the temporal stability of migration estimates drawn from digital trace data. Technology platforms and their user bases are in constant flux. Any unexpected patterns with respect to the interval size should be treated as a red flag when they are found in specific data.
Second, we asked how measures of migration vary with respect to temporal residence criteria, which we called the buffer. Any research that uses digital trace data to measure migration will have to deal with questions about how to account for travel or temporary mobility, which may complicate the validity of migration estimates. We expected that migration data estimated with larger buffers would be less multimodal (i.e., noisy) than data estimated with smaller intervals. As in the case of our first research question, our simulation model illustrated this idea, and our empirical analysis confirmed it. Short-term travel is more common than long-term travel. Thus, as we increased the size of the buffer used to infer residency at either end of the interval, our migration estimates became smoother and more stable. Again, this finding may seem self-evident at first, but it demonstrates that by systematically varying the buffer, migration researchers can evaluate where the cut-off between travel and long-term relocation lies-a cutoff that will vary based on context and the purpose of the measurement. In our empirical analysis, we showed that the high degree of accuracy and the volume of CDR data from Orange-Sonatel makes it relatively easy to isolate long-term mobility from travel. As we increased the buffer from one week to four, the short-term spikes in migration disappeared from our plot. For the CDR data, we recommend using a buffer of at least four weeks to estimate migration. By contrast, the irregularity of the Twitter and Gowalla data made it more difficult to distinguish the migration signal from travel-related noise. Even when a buffer of 12 weeks was used, the plots from Twitter and Gowalla still exhibited some instability. For unprocessed social media data, we recommend using a buffer greater than 12 weeks to estimate migration. More research is needed to determine what the acceptable level of travel-related noise should be, but our method clearly provides a conceptual and methodological framework for systematically addressing this issue. Our findings point to an emergent research agenda using digital trace data to map the relationship between different kinds of migration estimates. Yet further work must be done to validate mobility and migration estimates by linking them to traditional data. Because our conceptual focus was on identifying internal consistencies within a series of migration estimates generated from the same data, we did not attempt to assess the external validity of these estimates. However, such an analysis is clearly needed. In the U.S. context, interdivisional migration rates from Twitter or Gowalla data could be linked to Internal Revenue Service or American Community Survey estimates. Overall, we are encouraged by other studies that have demonstrated the validity of migration estimates from digital trace data by linking them to traditional sources (e.g., [bib_ref] Dynamic population mapping using mobile phone data, Deville [/bib_ref].
Additionally, more research is needed to confirm the fundamental results reported here. Although the simplicity of our framework and the intuitiveness of our findings suggest strong consistencies in how measurements of migration vary with the temporal definition, we tested for these consistencies using just three sets of data. Because the analysis presented here was limited to internal migration data, collecting and analyzing relevant data from international contexts in particular is essential. Moreover, research should be conducted to assess the robustness of the approach by leveraging migration corridors with well-known features. For example, flows between the United Kingdom and Spain certainly contain a very high proportion of travelers. Examining the sensitivity of international migration into and out of the United Kingdom estimated from digital trace data would both validate our method and characterize U.K. migration in-and out-flows in new and interesting ways. Three interlocking but distinct dimensions of migration measurement: start, buffer, and interval. By changing one while holding the other two fixed, we can assess how migration estimates are affected by seasonality (start), residency criteria (buffer), and cumulative exposure to migration risk (interval). Migration rates derived from simulated data. Each line tracks the rate of individuals classified as migrants as the interval increases from a specific reference point, while the buffer is held fixed. Estimates from the three empirical data sets of change in the rate of migration with increases in the interval from a specific start, while the buffer is held fixed at 1 week, 4 weeks, and 12 weeks Each dot represents a bilateral flow or corridor, such as New England to the Mountain West (in the U.S. context). The migration rate corresponding to each bilateral flow is estimated at two different time scales: one-month-to-one-month and six-month-to-six-months. The plot shows the correlation of these two temporal specifications for each bilateral flow/corridor. The findings presented here provide fertile ground for future research and suggest that (easier to measure) short-term flows may be useful in modeling (harder to measure) longterm flows.
[fig] Figure 1: Fig. 1. [/fig]
[fig] Figure 2: Fig. 2. [/fig]
[fig] Figure 4: Fig. 4. [/fig]
[fig] Figure 5: Fig. 5. [/fig]
|
The Study of Health-Related Fitness Normative Scores for Nepalese Older Adults
Physical fitness tests are important to maintain and promote the health status of people. The purpose of this study was to develop health-related fitness evaluation norms according to the age and gender of Nepalese older adults. One thousand nine subjects (449 males, 560 females) above 60 years, residing in 19 wards (rural and urban) of Dhangadhi Sub-Metropolitan City participated in this study. The test included the PAR-Q (Physical Activity Readiness Questionnaire), social aspects questionnaire, blood pressure test, height, weight, BMI (body mass index), percent body fat, and four physical fitness components (grip strength, 1-minute sit-to-stand, sit and reach, and 2-minute step tests). Mean, SD, and fitness evaluation norms for each component were obtained after the main test and statistical analyses. This study showed higher BMI and percent body fat in female age groups than in male age groups. Grip strength, relative grip strength, sit-to-stand, and 2-minute steps scores were better in male age groups than in female age groups, but in contrast, flexibility was better in female age groups. This study may help the related sectors to assess physical fitness, identify fitness levels, and develop appropriate physical activities or exercise programs for older adults based on age.Physical fitness is the ability of body systems that function efficiently to perform activities of daily living, respond in emergency situations, and enjoy sports and leisure activities[6]. It is one of the important indices of health that helps people to carry out daily activities safely and independently without undue fatigue[7]. A new fitness concept, health-related fitness, has been introduced on the basis of relationships between physical activities, fitness, and health[8]. Health-related fitness refers to the components of fitness (cardiorespiratory endurance, muscular strength, muscular endurance, and flexibility) [7] that are affected by habitual physical activity and are related to various health outcomes. Lower levels of fitness are associated with all-cause mortality, a high risk of cardiovascular disease, cancer, functional disability, and arterial stiffness [9-16]. Individuals having an appropriate level of health-related fitness may work efficiently, reduce the risk of disease and injury, look their physical best, and participate and enjoy physical activity[17]. Fitness, in various forms, increases performance and improves safety [18] and leads to greater reserve capacity to resist physical stress[19]. Improvements in physical fitness through appropriate physical activities in an older adult should enable them to maintain and promote health and well-being[20,21]into later years of life.Physical fitness testing is a highly visible and important part of physical fitness programs. It is important to monitor fitness to evaluate risk factors for health hazards in older adults and to ensure the safety of starting new activities. In Nepal, information on the health condition of older adults is still insufficient and no study on health-related fitness tests of older adults was found in the literature. The purpose of this study was to develop health-related fitness norms according to age and gender of older adults, with the expectation that our research might provide useful information for targeting effective physical activity interventions towards people at risk for declining health and to improve the health status of the general population.
# Introduction
Aging is a natural phenomenon and an inevitable process. Population ageing has been considered as one of the important demographic phenomena that impact directly or indirectly on health-care expenditure and the kinds of formal and informal care services. Developing nations like Nepal have not yet succeeded in managing appropriate health-care systems and establishing physical activity to ensure health and a high quality of life for older adults.
Nepalese society is in a phase of modernization that is changing the age structure of the country with a shift towards older ages through declining fertility, relatively controlled adult mortality, and improved health interventions. According to the national census, there were 1.5 million people in Nepal over 60 years of age in 2001 and 2.1 million in 2011, which constitutes 6.5% and 8.1% of the total population in the country, respectively. During the years 1991-2001, the annual elderly population growth rate was 3.39%, higher than the annual population growth rate of 2.3%, which indicates the start of aging dynamics in Nepal. Because of this, declining health is an issue. It is imperative to examine quantifiable outcomes of habitual physical activity if the contribution of physical activity to the health of the aging population is to be better understood.
# Materials and methods
Prior to the study commencing, approval for the research with an ethical code 2076/077 was obtained by the Dhangadhi Sub-Metropolitan Urban Health Centre and Dhangadhi Sub-Metropolitan City, Nepal. Thirty volunteers were recruited in the combined efforts of the Sub-Metropolitan City and Sudur Paschimanchal Campus. This study was conducted from the last week of July to the third week of August in 2019. Four days of training were provided to the volunteers. A permanent health assistant for the study period was provided by the Urban Health Centre, Dhangadhi Sub-Metropolitan City. A local health assistant and volunteers were provided by the health post of each ward of the Sub-Metropolitan City. Advertising was run through local FM radio and TV spots in the representative sectors sharing information about the test, dates, places, and times starting a week before the tests commenced. On test day, after fully briefing participants about the nature and purpose of the study, a completed Physical Activity Readiness Questionnaire and written informed consent were received from each participant. Completion of the social aspects questionnaires and overseeing of the tests were performed by trained volunteers. Before each test, a trained volunteer explained and demonstrated the task and made sure that the task could be completed without any physical risk to the participant. Volunteers helped the participants in doing a warm-up consisting of head turns and half circles, single arm crossovers and chest stretches, and calf and hamstring stretches after the body composition tests and before the physical fitness tests. The test unit was always prepared for immediate first aid for injuries and sudden illness during the tests.
## Characteristics of the fitness test
Physical factors necessary to carry out the activities of daily living were included. The test included mainly health-related fitness factors (body composition, muscle strength and endurance, flexibility and cardiorespiratory endurance). It was intended to be field-based and low cost. Fitness test components were selected for having a low risk of injury during testing.
## Participants
A field-based cross-sectional study including 1009 subjects, (449 males, 560 females) above 60 years, residing in 19 wards (rural and urban) of Dhangadhi Sub-Metropolitan City participated in the health-related physical fitness test. Participants completed the blood pressure test after resting 5-10 min. Participants with any problems identified in the Physical Activity Readiness Questionnaire (PAR-Q); having systolic pressure greater than 160 mmHg and diastolic pressure greater than 100 mmHg; or having an implantable electrical device such as a pacemaker, defibrillator, or nerve stimulator were excluded from the other tests.
After completing the blood pressure test, the five remaining tests were completed in the following order to minimize fatigue: body composition analyses, 1-min sit-to-stand, grip strength, and flexibility.
## Body composition tests
The Body Composition Analyzer (Inbody 470, Korea) was used to assess weight, body mass index (BMI), and percent body fat. BMI is body weight in kilograms (kg) divided by height in meters squared (m 2 ). Percent body fat is the total mass of fat divided by total body mass and multiplied by 100, which includes essential body fat and storage body fat. A stadiometer was used to measure height.
## Hand grip strength test
A digital grip strength dynamometer (TKK 5401; Takei, Tokyo, Japan) was used to assess upper body strength. Participants were asked to stand and hold the dynamometer by their testing hand, which was held slightly away from the other parts of the body, and with the grip meter indicator facing outward. The measurement was recorded after squeezing the handle of the dynamometer as hard as they could without swinging the dynamometer.
## Sit and reach test
A trunk flexibility box was used to assess body flexibility. Participants were asked to sit with bare feet, legs extended, toes pointing up, and feet approximately hip-wide, and with the soles of the feet against the base of the measuring device. Then, they were asked them to push the slide slowly forward, as far as they could, by placing one hand on top of the other and without lifting their knee from the ground. The measurement was recorded while in a static position for a couple of seconds at the point of greatest reach.
## Sit-to-stand test
A standard chair without arm supports was used to assess lower body strength and muscle endurance. Participants were asked to stand up from and sit down on the chair with their feet resting on the floor and their hands crossed at the wrists and held stationary on their shoulders; they were asked to repeat this procedure as many times as possible within one minute at a participant-defined pace. The participant was informed when 30 and 15 s were left in the test. The number of completed repetitions was counted considering standing as being in full extension (knee and hip extension) and sitting as being in a position with their knees at 90 - flexion.
## Two-minute step test
A two-minute step test was preformed to assess cardiorespiratory endurance. Participants were asked to stand up straight behind poles with rubber bands. A rubber band was adjusted at the level corresponding to midway between the patella and iliac crest. The subject was then asked to march in place for two minutes, lifting their knees to the height of the rubber band. Resting or sitting on a chair placed behind the participant was allowed if needed. The participant was notified when 1 min was completed and when 30 seconds were remaining, and participants were given verbal motivation.
The test was stopped after two minutes of stepping, and the total number of repetitions of the right knee reaching the rubber band level in two minutes was recorded.
## Statistical procedure
Descriptive statistical analysis was used to analyze average and standard deviation by age group and gender of older adults. The Shapiro-Wilk test was completed for the normality test. Cajori's 5-grade evaluation normswere used to develop health-related physical fitness norms for the tested components on the basis of age group and gender. Percentile was used to develop fitness awards and certificates of recognition.
# Results
The general characteristics of the participants are presented in. Mean and standard deviations for anthropometric measurements height, weight, body mass index (BMI), and percent body fat are presented in Tables 2-6 and the health-related physical fitness components grip strength, relative grip strength (grip strength calculation considering weight), 1-min sit-to-stand, sit and reach, and 2-minute steps are presented in Tables 7-12.shows 55.50% of the study population were female and 45.29% were janjati. Nearly 73% of the participants were illiterate and only 0.69% were graduates.also shows that 23.29% of the participants were widowed and 85.13% were living in joint families. Nearly 95% of study group were non-smokers and nearly 86% were non-drinkers. Nearly 70% of the participants were from rural areas of Dhangadhi Sub-Metropolitan City.also shows that nearly 78.79% of the study population never had a health check-up and only 8.62% had a health check-up within the last six months.
As indicated in, gradual declines in scores were observed over the 5-year age spans for both males and females in all the tests except height and weight in both males and females and percent body fat in males. Height was similar in both male and female age groups. Weight was similar in both males and females from the 60-64 and 70-74 year age groups but females' average weight was higher in the 75-79 and 80+ age groups than the that of the same male age groups. BMI and percent body fat were higher in the female age groups than those in the male age groups and showed gradual declines in the female age groups as age increased. The percent body fat remained almost constant for all age groups of males. Grip strength and relative grip strength scores showed a gradual decline in the 5-year age spans and were higher in all male age groups than those of the female age groups. The result showed a gradual decline in sit-to-stand scores in both male and female age groups as age increased. Male showed better sit-to-stand scores in all age groups than those of female age groups. We also saw a gradual decline in 2-minute step scores in both male and female age groups as age increased. Male scores were better in all age groups than those of female age groups. Flexibility also declined gradually in all age groups as the age increased. In contrast with other test results, flexibility was better in all female age groups than that of male age groups. Reported diagnoses were mainly high blood pressure, diabetes, and arthritis, but some participants had heart disease.
## Fitness evaluation norms
As mentioned above, Cajori's 5-grade evaluation norms were used to develop older adults' health-related fitness norms for the tested components on the basis of age group and gender. For the anthropometric measurements, such as BMI and percent body fat, the 5-grade relative evaluation was obesity (7%) if there was mean + 1.5σ or above and over weight (24%) if there was mean + (0.5σ~1.5σ). It was normal (38%) if there was mean ± 0.5σ. It was evaluated as thin (24%) when included in the range of mean + (−0.5σ to −1.5σ) and very thin (7%) when included below the range of mean −1.5σ. In contrast, for health-related fitness components such as grip strength, sit and reach, sit-to-stand, and 2-minute steps, the 5-grade relative evaluation was excellent (7%) if there was mean + 1.5σ or above and very good (24%) if there was mean + (0.5σ~1.5σ). It was normal (38%) if there was mean ± 0.5σ. It was poor (24%) when included in the range of mean + (−0.5σ to −1.5σ) and very poor (7%) when included below the range of mean −1.5σ. The evaluation norms calculated by applying Cajori's 5-grade evaluation norms are presented in Tables 13-19.
## Stages of fitness awards or certificates for older adult
Recognition for achievement is considered as a motivation for those who want improvement in physical fitness. One of the main reasons for the development of the fitness awards was to provide motivation for those using the test. It is a means of self-assessment that may assure a greater degree of personal safety and the proper determination of exercise intensity. It can motivate individuals to take the test, use the test for the development of a personal strategy to improve physical fitness, and commit themselves to continue the physical activities to maintain their well-being.
Health-related physical fitness awards for Nepalese older adults was developed on the basis of the percentiles (70th, 50th, and 30th) of each test item by gender and age as shown in Table 20. Participants will get a medal award or certificate of recognition only after meeting the fitness award criteria. Body composition scores (BMI and percent body fat) were excluded in the gold award. For the gold award, all the health-related fitness test components should be at least in the 70th percentile. Except body composition, all health-related fitness test components must be equal or above the 50th percentile and below the 70th percentile for the silver award. The bronze medal is certified if the body composition is in the recommended range. BMI must be 18 kg/m 2 to 25 kg/m 2 in both males and females, and percent body fat should be 7%-35% for males and 16%-32% for females. All four health-related fitness tests must be above the 30th percentile and below the 50th percentile.
# Discussion
Maintaining independence in older adults becomes important due to the challenges of the demographic shift. A fitness test is a combination of a sequence of exercises that assesses overall health and physical status of an individual. It is known to be the initial phase in the design of a suitable exercise program based on a baseline score for general health and fitness purposes.
In this study, we determined norms of health-related physical fitness components based on gender and age groups for older adults of Nepal. Selected test components are easy and safe to assess for inactive and very active older adults. These test components are the best to assess health-related physical fitness in a country like Nepal because they are inexpensive, space efficient, and transportable as they do not require advanced devices or separate room. This is the first study in the physical fitness field in Nepal. It is centered only in the single Sub-Metropolitan City of the far western development region of Nepal. The analysis shows an interesting factor that BMI data in specific age groups (60-64, 65-69, 70-74), and sit and reach data in all age groups were not in normal distribution in both genders. Further studies in every age group, children, adolescents, adults, and even in older adults, are needed. In addition, studies covering all altitudes and areas of Nepal should be completed in the future.
The basic body functions, such as strength, endurance, flexibility, and balance on extremities, are all important to sustain physical independence at older ages. A reduction in lean body mass and an increase in fat mass may increase mortality risk in older adults. Thus, maintaining body a mass index (BMI) at a healthy level with more muscular mass than fat mass is essential.
Generally, fat and lean soft tissues change gradually with age and are believed to be associated primarily with imbalances between energy intake and expenditure due to an increasingly sedentary lifestyle. Our study shows relative stability of percent body fat in older adult males but a decrease with age in female older adults, similar to Baumgartner et al.. The World Health Organization (WHO) had recommended body mass index (BMI) greater than 23 kg/m 2 and less than 28 kg/m 2 as normal BMI for older adults; which is higher than most BMI values found in this study. Overweight is not associated with an increased risk of mortality. In contrast, risk of mortality increases in older adults with BMI values less than 23 kg/m 2, which shows a poor health status in the older adults of Nepal.
In this study, analysis of grip strength, sit-to-stand (STS), and 2-minute steps by gender shows higher grip strength in males at all ages, and analysis by age group shows a gradual decline in grip strength in both genders, similar to other studies. Our results show lower grip strength in Nepalese older adults than in older adults in other countriesand show higher relative grip strength in older females in Nepal than in Korean female older adults. Relative grip strength is calculated considering weight, that is the reason for a higher relative grip strength in Nepalese females. The sit-to-stand test was done for 1 min similar to. Nepal's life expectancy is 70.2 years (68.8 years and 71.6 years for males and female, respectively)and it is hard to find healthy 80+ older adults. The 1-minute STS test is better to determine lower body muscular strength and endurance in an older adult below the age of 80 years. In addition, STS performance was significantly associated with standing, good balance, coordination, and aerobic capacity that cover several components of physical fitness. This study does not asses the balance test separately. Thus, the 1-minute sit-to-stand test was selected for lower body strength and endurance, balance, and coordination. Lower body muscular strength and endurance were also found poorer than in other countries.
Aerobic exercise capacity is also one of the fundamental components of health-related physical fitness tests. Among many modes of a cardiorespiratory endurance test, the 2-minute step test introduced by Rikli and Jones as part of the Senior Fitness Test in 1999 is easy to conduct in almost any setting. In comparison to other studies, cardiorespiratory endurance was also found to be poor in Nepalese older adults. Sit and reach, one of the popular field-test components of Health Related Physical Fitness Test (HRPFT) to measure flexibility in the lower back and hamstring muscles, was assessed in this study. Female flexibility was similar to that of Korean female older adults and poorer than that of male older adults.
Some important findings for the participants from Nepal are as follows: a) very poor BMI, even lower than the baseline of the average recommended by the WHO for older adults; b) with increasing age, both genders showed a declining score in muscular strength and endurance, flexibility, and cardiorespiratory endurance with poor results in comparison to other countries; c) in every test except flexibility, males had higher scores with better results in comparison to women from the same age group; d) 72.45% of people were illiterate; e) 78.79% of people did not get regular health check-ups and more of the older adult population lived in rural areas than urban areas. Poor fitness scores compared to other countries shows a lower level of physical activity and exercises in older adults of Nepal. A higher percentage of illiteracy and lack of regular health check-ups suggests a poor awareness status about health and physical activities. Poor BMI values show a higher risk of mortality in older adults of Nepal. Considering the finding of this study: a) Administrators, physical educators, instructors, and health instructors should encourage older adults to do physical activities and exercise as recommended by the WHO; b) Regular awareness programs about physical activities, good health, and quality of life should be providedl c) A wide range of activities should be identified and arranged according to older adults' interests and needs; and d) Fitness should be tested regularly as per norms.
Cross-sectional design, covering only health-related fitness components, and testing only in the Terai region of Nepal are the limitations of this study. Despite these limitations, this study included both urban and rural areas of Sub-Metropolitan City and all the major ethnic groups of Nepal. A body composition analyzer (Inbody, Korea) was use to assess BMI and percent body fat.
The results of this study could be used as approximate indicative values for comparisons among the same health-related fitness scores of older adults from other developing Asian countries similar to Nepal. In addition, the scores can be used as benchmark values for health-related fitness evaluations of older adults of Nepal.
# Conclusions
This study may help the health and physical education sector to develop appropriate physical activities or exercise programs for older adults. It may help to identify the fitness level of an individual and help fitness instructors to develop suitable physical activity interventions for older adults based on their age. In addition, an awards system may encourage individuals to participate in physical activities or exercise continuously, which helps lead to a better and more independent life in upcoming years. Longitudinal studies along with functional fitness studies on older adults and adult fitness studies are necessary in the future. |
Minimally invasive devices for treating lower urinary tract symptoms in benign prostate hyperplasia: technology update
Benign prostatic hyperplasia (BPH) represents a spectrum of related lower urinary tract symptoms (LUTS). The cost of currently recommended medications and the discontinuation rate due to side effects are significant drawbacks limiting their long-term use in clinical practice. Interventional procedures, considered as the definitive treatment for BPH, carry a significant risk of treatment-related complications in frail patients. These issues have contributed to the emergence of new approaches as alternative options to standard therapies. This paper reviews the recent literature regarding the experimental treatments under investigation and presents the currently available experimental devices and techniques used under local anesthesia for the treatment of LUTS/BPH in the vast majority of cases. Devices for delivery of thermal treatment (microwaves, radiofrequency, high-intensity focused ultrasound, and the Rezum system), mechanical devices (prostatic stent and urethral lift), fractionation of prostatic tissue (histotripsy and aquablation), prostate artery embolization, and intraprostatic drugs are discussed. Evidence for the safety, tolerability, and efficacy of these "minimally invasive procedures" is analyzed.
# Introduction
Benign prostatic hyperplasia (BPH) represents a spectrum of disease, with some patients having more bothersome lower urinary tract symptoms (LUTS) than others. Age, sexual activity, fertility, medical comorbidities, past medical and surgical history, prostate volume, urodynamic parameters, and ultrasound findings are other clinical variables that influence the therapeutic decision-making. Given patients expectations and their willingness to toleratevarious treatment risks and benefits, this range of benign disease necessitates a range of treatment options. Current treatments for BPH include watchful waiting, lifestyle modifications, phytotherapy, medical therapy, and surgical intervention. Watchful waiting is recommended for mild symptomatic and uncomplicated BPH but carries the risk of progression for a substantial number of patients. [bib_ref] Benign prostatic hyperplasia: when to 'watch and wait,' when and how to..., Levy [/bib_ref] A longitudinal study found that over a 4-year period, 87% of men who were mildly symptomatic experienced worsening of their symptoms. [bib_ref] Longitudinal study of men with mild symptoms of bladder outlet obstruction treated..., Djavan [/bib_ref] Although well tolerated with mild and infrequent adverse events, there is at present no convincing evidence supporting the use of phytotherapy for BPH. [bib_ref] Phytotherapy of benign prostatic hyperplasia. A mini review, Pagano [/bib_ref] The currently recommended medications remain first-line therapy for bothersome LUTS. [bib_ref] EAU guidelines on the treatment and follow-up of non-neurogenic male lower urinary..., Oelke [/bib_ref] However, cost and early discontinuation due to side effects are significant drawbacks limiting their longterm use in clinical practice. Conventional interventions considered as the definitive treatment for BPH carry a significant risk of treatment-related complications and even submit your manuscript | www.dovepress.com
## Dovepress
## Dovepress
## 126
Aoun et al death in high-risk patients. [bib_ref] Complications of transurethral resection of the prostate (TURP) -incidence, management, and prevention, Rassweiler [/bib_ref] These issues have contributed to the development of "minimally invasive procedures" as an alternative to standard therapies. The concept of a minimally invasive approach to BPH dates back to the 19th century, but early experience with these procedures was disappointing and their use steadily faded over time. However, recent technical advances combined with translational lines of research allowed refinements of pre-existing procedures and modifications of surgical protocols to improve the potential therapeutic value of these new technologies.
This review provides an update on currently available devices and techniques used under local anesthesia for the treatment of BPH/LUTS. The evidence for their safety, tolerability, and efficacy in clinical practice is also analyzed.
## Devices for delivery of thermal treatment
Heat applied to the prostate gland causes tissue necrosis and sloughing, thereby relieving obstruction by decreasing the bulk of the prostate. Several techniques are commonly available, and use different heat sources to induce thermoablation (microwaves, radiofrequency, ultrasound, and water-induced thermotherapy), different route of administration (intraurethral, interstitial, transrectal, and recently transperineal), and different levels of energy (low versus high). Further, different mechanisms of action may be involved (hyperthermia versus thermotherapy). The evolution of devices for delivery of thermal treatment as well as the concept of urethral cooling to avoid harm to non-target tissues had been marked over the past 15 years.
## Microwave thermotherapy
Microwave radiation emitted via an antenna results in marked generation of heat sufficient in magnitude and duration to cause tissue coagulation and necrosis. In 1985, Yerushalmi et al were the first to use microwaves in the treatment of BPH in high-risk surgical candidates. [bib_ref] Localized deep microwave hyperthermia in the treatment of poor operative risk patients..., Yerushalmi [/bib_ref] Early experiences indicated inadequate tissue ablation due to an inappropriate thermal dose, defined as the temperature achieved inside the prostate multiplied by duration of exposure to that temperature. [bib_ref] Transrectal microwave hyperthermia for benign prostatic hyperplasia: long-term clinical, pathological and ultrastructural..., Montorsi [/bib_ref] Since then, advances in microwave treatment systems with respect to antenna design, route of administration, heating patterns, and treatment protocol have allowed delivery of temperatures higher than 45°C. A higher thermal dose is associated with larger intraprostatic zones of necrosis, a higher peak urinary flow rate (Qmax), and a trend toward improvement of symptoms. [bib_ref] Detailed interstitial temperature mapping during treatment with a novel transurethral microwave thermoablation..., Larson [/bib_ref] Theses results were confirmed in a prospective randomized clinical study, [bib_ref] Traitement de l'hypertrophie bénigne de prostate par Prostatron ® : étude des..., Netto [/bib_ref] and were balanced by the risk of complications and tolerability due to heating of non-target tissue. [bib_ref] Contrasting heating patterns and efficiency of the Prostatron and Targis microwave antennae..., Larson [/bib_ref] The transurethral approach facilitates targeting the prostatic transition zone. [bib_ref] Contrasting heating patterns and efficiency of the Prostatron and Targis microwave antennae..., Larson [/bib_ref] A high thermal dose and the transurethral approach led to development of cooling systems to protect the urethra. [bib_ref] Transurethral microwave thermotherapy for benign prostate hyperplasia: separating truth from marketing hype, Walmsley [/bib_ref] Following the same concept, manufacturers have continued to develop higher energy systems, more complex and efficient cooling devices, and more accurate monitoring and targeted devices, leading to third-generation systems. A Cochrane review published in 2012 demonstrated that transurethral microwave thermotherapy provides lasting symptomatic relief (for up to at least 36 months, even in patients with acute urinary retention), increased Qmax, and enhanced quality of life for patients with LUTS/BPH. [bib_ref] A high-efficiency microwave thermoablation system for the treatment of benign prostatic hyperplasia:..., Larson [/bib_ref] [bib_ref] Microwave thermotherapy for benign prostatic hyperplasia with the Dornier Urowave: results of..., Roehrborn [/bib_ref] [bib_ref] A novel transurethral microwave thermoablation system to treat benign prostatic hyperplasia: results..., Ramsey [/bib_ref] [bib_ref] Microwave thermotherapy for benign prostatic hyperplasia, Hoffman [/bib_ref] A recent meta-analysis highlights the heterogeneity of the data reported in the literature. [bib_ref] High-energy transurethral thermotherapy with Core Therm approaches transurethral prostate resection in outcome..., Kaye [/bib_ref] For example, improvement in symptom scores ranges from 0% to 60% at 3 months, with an average of 25%. However, the maximal effects of microwave therapy are observed within 3-6 months. [bib_ref] Microwave thermotherapy for benign prostatic hyperplasia with the Dornier Urowave: results of..., Roehrborn [/bib_ref] [bib_ref] A novel transurethral microwave thermoablation system to treat benign prostatic hyperplasia: results..., Ramsey [/bib_ref] Meanwhile, both temporary placement of an intraurethral prostatic bridge catheter and neoadjuvant or adjuvant alpha-blocker treatment were shown to be effective in alleviating symptoms and improving quality of life. [bib_ref] A novel intraurethral prostatic bridge catheter for prevention of temporary prostatic obstruction..., Djavan [/bib_ref] [bib_ref] Neoadjuvant and adjuvant alpha-blockade improves early results of high-energy transurethral microwave thermotherapy..., Djavan [/bib_ref] The improved energy targeting systems are better tolerated under topical local anesthesia and are associated with fewer sexual complications, such as erectile dysfunction and/or ejaculation disorders. [bib_ref] Tolerability of high energy transurethral microwave thermotherapy with topical urethral anesthesia: results..., Djavan [/bib_ref] Potent systemic sedoanalgesic medications were necessary when using the oldest systems. [bib_ref] Pain sensation in transurethral microwave thermotherapy for benign prostatic hyperplasia: the rationale..., Cormio [/bib_ref] The relatively limited acceptance of this technology in day-to-day surgical practice is due to its high cost, steep learning curve, and effectiveness and efficacy over time, and durability. Another limitation stems from the limited repeatability of this modality. The uncontrolled back heating effects associated with conventional microwave applicators may result in uncertain and not fully repeatable ablative lesion size and shape. Recently, Bartoletti et al reported promising results for a Phase I clinical study of a new microwave antenna for transperineal thermoablation. This new system AMICAprobe (Hospital Services SpA, Aprilia, Italy), comprises an integrated hydraulic circuit for cooling of the applicator and a miniaturized device (mini-choke) for entrapment of reflected waves. It allows maximum control over radial and longitudinal ablative lesion size and overcomes back heating effects, while keeping the applicator size at a minimum. [bib_ref] Microwave-induced thermoablation with Amica-probe is a safe and reproducible method to treat..., Bartoletti [/bib_ref]
## Radiofrequency thermotherapy
Transurethral needle ablation (TUNA) uses radiofrequency energy to cause heating and thereby ablation of prostate tissue. Two antennae placed within the prostate deliver the
## 127
Minimally invasive devices for treating LUTs in BPH radio waves. Low level radiofrequency generates energy up to 110°C, leading to coagulation, necrosis, and cell death only in a localized area, because the signal is transmitted into tissue by direct contact. Preservation of the urethra minimizes the short-term complications. The technology for performing TUNA has been available since the early 1990s. The first studies in the USA began in 1994, and the US Food and Drug Administration approved TUNA for the treatment of BPH in 1996. [bib_ref] Transurethral needle ablation Prostiva for treating symptomatic benign prostatic hyperplasia: a review, Xylinas [/bib_ref] A meta-analysis published in 2006 assessed the effectiveness of TUNA. [bib_ref] Systematic review and meta-analysis of transurethral needle ablation in symptomatic benign prostatic..., Bouza [/bib_ref] The technique provides lasting symptomatic relief despite high retreatment rates (7.4 times higher than transurethral resection of the prostate [TURP]). [bib_ref] Transurethral ethanol ablation of the prostate for symptomatic benign prostatic hyperplasia: long-term..., El-Husseiny [/bib_ref] However, symptom and quality of life scores were all higher with TURP, while the complication rates were lower in the TUNA group, with fewer reports of retrograde ejaculation, erectile dysfunction, and strictures. [bib_ref] Transurethral needle ablation versus transurethral resection of the prostate for the treatment..., Hill [/bib_ref] The technique is well tolerated under local anesthesia. [bib_ref] Anesthesia-free procedures for benign prostate obstruction: worth it?, Ebbing [/bib_ref] Another advantage is the recent improvements in electronic technology incorporating thermocouples in the needle to accurately monitor the temperature delivered to the prostate. In addition, in the last decade, the effects of electromagnetic fields on living organisms have been extensively studied. It was demonstrated recently in a canine model that electromagnetic fields can have anti-inflammatory effects and alter the cell proliferation rates and cell membrane permeability. [bib_ref] Effect of pulsed electromagnetic field therapy on prostate volume and vascularity in..., Leoci [/bib_ref] The first study that investigated the effects of pulsed electromagnetic fields for the treatment of BPH was published in 2011. [bib_ref] Effects of pulsed electromagnetic fields on benign prostate hyperplasia, Giannakopoulos [/bib_ref] Ten patients with BPH were exposed to a pulsed electromagnetic field. The authors reported a statistically significant improvement of symptoms and Qmax as well as a significant reduction in prostate volume and post-void residual compared with ten patients treated with alpha-blockers. The improvement was durable at 1 year of follow-up. The authors concluded that pulsed, short wave duration electromagnetic field at low intensity radiofrequencies seems to exert beneficial nonthermal effects on BPH.
## High intensity focused ultrasound
This technique is based on generation of a focused ultrasound field, which is generated by a piezoelectric transducer coupled into the body and aimed at a target region. The focused waves are absorbed by tissues throughout their propagation and their acoustic energy is converted to heat with a high energy density inside the focal zone. The result is a rapid temperature rise in the target zone to a level exceeding the threshold level of protein denaturation (approxiamtely 60°C), leading to coagulative necrosis. In contrast, there is a low acoustic energy density in the surrounding structures, keeping them spared. This creates elliptical small volume lesions of 50-300 mm [bib_ref] Phytotherapy of benign prostatic hyperplasia. A mini review, Pagano [/bib_ref]. A larger volume can be ablated without gaps by combining single lesions. High intensity focused ultrasound (HIFU) could there be utilized as a minimally invasive debulking method to treat BPH. The parameters for treating prostate were defined in 1992, and two devices were developed, ie, the Ablatherm ® (EDAP TMS, Vaulx-en Velin, France) and the Sonoblate ® (Sonacare Medical, LLC, Charlotte, NC, USA); and in 1994, Madersbacher et al published the first clinical study of HIFU for BPH in humans. [bib_ref] Tissue ablation in benign prostatic hyperplasia with high intensity focused ultrasound, Madersbacher [/bib_ref] Early studies established the safety of this technique and improved our understanding of the biological interaction between focused ultrasound and prostatic tissue. However, the treatment was not effective in terms of Qmax and symptoms. To improve the efficiency, a larger volume of prostate and the bladder neck were included. Also, more power was needed to disintegrate the urethra. The presence of a silicone urinary catheter increased the amount of energy delivered at the urethra. The advent of thermocouples helped to establish the temperature profile throughout the duration of the procedure, and development of a longer focal length probe allowed treatment of a larger prostatic volume, including the anterior fibromuscular stroma. [bib_ref] Noninvasive surgery of prostate tissue by high intensity focused ultrasound: an updated..., Sanghvi [/bib_ref] These technical modifications in association with updated software lead to substantial improvement in symptom scores, Qmax, complication rates, and quality of life. [bib_ref] Clinical outcome of high intensity focused ultrasound for treating benign prostatic hyperplasia:..., Uchida [/bib_ref] The introduction of circulating cool water kept the rectal temperature lower than body temperature and shortened the treatment time. New developments in monitoring systems allowed more accurate targeting. Magnetic resonance imaging (MRI) and ultrasound elastic fusion allowed accurate definition of the area to be treated. High resolution ultrasound permits live follow-up of HIFU shots and treatment planning that is adjustable in real time. A novel device known as Focal One ® (EDAP TMS) was recently commercialized. It combines the latest imaging and treatment technologies of electronic displacement of the focal point associated with contrast-enhanced ultrasound, and is a significant advance in the treatment of BPH and localized prostatic carcinoma. Despite these improvements, the cost of the machine and the procedure as well as the presence of a learning curve has limited its widespread use. [fig_ref] Table 1: Characteristics of HiFU devices in use for the treatment of benign prostatic... [/fig_ref] summarizes the technical characteristics of the currently available systems. In parallel, a new concept based on focal hyperthermia is emerging. [bib_ref] Catheter-based ultrasound applicators for selective thermal ablation: progress towards MRI-guided applications in..., Diederich [/bib_ref] In contrast with HIFU, which implies high power and high temperatures for a short duration, focal hyperthermia relies on feedback control to maintain a much lower temperature increase (39°C-45°C) in larger contiguous regions over time frames ranging from 15 to 60 minutes. Catheter-based ultrasound applicators have been developed for delivery of hyperthermia or high-temperature thermal ablation of cancer and benign disease of the prostate. [bib_ref] Effect of applicator diameter on lesion size from high temperature interstitial ultrasound..., Tyréus [/bib_ref] These devices allow for control of heating during delivery of therapy. Using a canine model, it was demonstrated that catheter-based ultrasound devices combined with magnetic resonance thermal monitoring can achieve relatively rapid and accurate thermal ablation of the prostate. [bib_ref] Catheter-based ultrasound devices and MR thermal monitoring for conformal prostate thermal therapy, Diederich [/bib_ref] Transurethral water vapor therapy: the Rezum system
The Rezum system (NxThera, Maple Grove, MN, USA) is a novel, minimally invasive ablative transurethral therapy. It delivers targeted and controlled doses of stored thermal energy directly to the transition zone of the prostate gland to treat BPH by using sterile water vapor. A narrow sheath, similar in size and shape to a cystoscope, is inserted via the urethra and positioned between the bladder neck and the seminal colliculus within the prostatic urethra. A thin needle is deployed through the urethra into the hyperplastic transition zone. Water vapor is delivered rapidly (in 8-10 seconds) and directly into the hyperplastic zone, and is immediately dispersed through the tissue interstices. When the water vapor comes into contact with the prostatic tissue, it condenses into its liquid state and releases its stored thermal energy. This thermal energy denatures the cell membranes immediately, causing instantaneous cell death. The first-inman trial was a dose-ranging study [bib_ref] Transurethral water vapor therapy for BPH; initial clinical results of the first..., Dixon [/bib_ref] in which the energy delivered ranged from 190 to 289 calories per injection.
In the Rezum I pilot study, all patients received the same dose of 208 calories per injection. For 30-80 mL prostates, a total of 2-6 administrations were delivered, with a total treatment time of ,90 seconds. Gadolinium-enhanced MRI T2 imaging, performed in all subjects (n=30) at 1 week, 1 month, and 3 months following the procedure showed that the lesions, transition zone, and overall prostate size decreased by a mean of 89%, 32%, and 26%, respectively. The authors reported a significant improvement in International Prostate Symptom Score (IPSS), quality of life, Qmax, and post-void residual at 3 months, which remained stable at follow-up 1 year later. They concluded that this outpatient or office-based procedure is rapid, safe, and effective. Controlled ablation of prostatic tissue was also demonstrated. Another single-center study with a limited follow-up period (3 months) confirmed these promising data. [bib_ref] What's truly minimally invasive in benign prostatic hyperplasia surgery?, Chung [/bib_ref] At present, this system is considered to be an investigational device, and there is one prospective, randomized, controlled, single-blind clinical trial underway in the USA (ClinicalTrials.gov identifier NCT01912339) evaluating the efficacy and safety of the Rezum system and assessing its effect on urinary symptoms secondary to BPH. The estimated study completion date is in June 2019.
## Mechanical devices
Another attractive mechanism for treating BPH/LUTS is to maintain urethral patency using specific designed mechanical devices. The first mechanical device to relieve urethral obstruction was used in 1980.It was made of a metal alloy and initially called a "partial catheter". Subsequently, manufacturers developed various types of catheters, described as prostatic stents, by modifying the design, the compounds used, the biocompatibility, the rigidity, and ease of insertion and removal. Recently, a novel device was invented to mechanically "hold open" the prostatic urethra, ie, the prostatic urethral lift. [bib_ref] Safety and feasibility of the prostatic urethral lift: a novel, minimally invasive..., Woo [/bib_ref] [fig_ref] Table 2: Characteristics of commonly used prostatic stents [/fig_ref] summarizes the most commonly used prostatic stents.
## Prostatic stents
The use of metallic stents to maintain luminal patency is a well established concept in surgical practice. Its first reported use in urology dates back to the 19th century. Various metallic stents now exist and can be classified into two groups, ie, permanent epithelializing stents and temporary nonepithelializing stents. The latter have the advantage of being able to be inserted in a compressed state, minimizing the risk of urethral injury and the associated pain. The radial force exerted by these stents should theoretically prevent stent migration, and the wider lumen should provide a better urinary flow or even further endourethral maneuver. Their ease of removal makes these stents the first choice for diagnostic purposes and for the treatment of elderly frail patients, as well as use as an adjunct to other minimally invasive treatments for BPH such as thermotherapy, to allow voiding while procedure-related edema resolves. On the other hand, epithelialization of the spiral coil stent had the theoretical advantage of reducing the rate of migration, infection, and encrustation.The same results were reported with the Prostakath ® (Engineers and Doctors, Copenhagen, Denmark). [bib_ref] Intraprostatic metal stents (Prostakath and Urospiral): more than 6 years' clinical experience..., Braf [/bib_ref] A second-generation of metal stents was developed to prevent encrustation and migration and to allow a much longer indwelling time. [bib_ref] Comparison between first-generation (fixed caliber) and second-generation (self-expanding, large caliber) temporary prostatic..., Yachia [/bib_ref] Refinements to these stents resulted in the development of thermoexpandable stents, such as the Memokath ® stent (Doctors and Engineers, Kvistjaard, Denmark), which is a nickel-titanium alloy spiral stent with the main advantage of ease of removal based on its physical properties at different temperatures. [bib_ref] The Memokath™ stent, Staios [/bib_ref] The ability to correctly insert this stent under fluoroscopy and transrectal ultrasound without the need for cystoscopy was also demonstrated.Another self-expandable metallic device, the Z-stent ® (Wilson-Cook Medical, Winston-Salem, NC, USA) was developed to be inserted under fluoroscopic guidance only. [bib_ref] Self-expandable metal stents in high risk patients with benign prostatic hyperplasia: long-term..., Song [/bib_ref] Clinical studies had demonstrated safety, tolerability, and short-term efficacy for the Memokath stent. [bib_ref] The thermo-expandable metal stent for managing benign prostatic hyperplasia: a systematic review, Armitage [/bib_ref] However, migration was still a major limitation to widespread acceptance of these stents. Manufacturers tried to change the shape of these cylindrical stents. However, the hourglassshaped and bell-shaped Memokath stents also had the same rate of migration. [bib_ref] Hourglass-shaped nitinol prostatic stent in treatment of patients with lower urinary tract..., Van Dijk [/bib_ref] [bib_ref] The bell-shaped nitinol prostatic stent in the treatment of lower urinary tract..., Van Dijk [/bib_ref] To avoid this, Markovic et al designed the triangular prostatic stent, which conforms well to the prostatic urethral lumen shape, and reported no migration at 9 months of follow-up. A plastic coating was also added to prevent encrustation. [bib_ref] A new generation of urethral stents: allium in the therapy of symptomatic..., Marković [/bib_ref] Refinements in temporary prostatic stents resulted in the development of biodegradable stents and polyurethane intraurethral catheters. The polyurethane Spanner ® prostatic stent (Abbey Moor Medical, Parkers Prairie, MN, USA) was initially developed for temporary use only. [bib_ref] Prostatic stents for the treatment of benign prostatic hyperplasia, Vanderbrink [/bib_ref] The early experimental and clinical test outcomes provided reasonable assurance that the Spanner stent is safe and effective when used as an adjunct to minimally invasive therapies. [bib_ref] Experience with the spanner prostatic stent in patients unfit for surgery: an..., Grimsley [/bib_ref] Corica et al easily inserted and removed the stent under topical anesthesia in 30 patients. They reported a significant improvement in IPSS, post-void residual, and Qmax, with no changes in voiding and continence. The authors confirmed the stability, patency, and lack of migration of the device on imaging during up to 12 weeks of use. However, its primary use in patients unfit for surgery was unsatisfactory, with only 37% of the patients continuing to have a stent in situ after a mean of five changes (every 3 months) or being stent-free after a successful voiding trial. [bib_ref] A novel temporary prostatic stent for the relief of prostatic urethral obstruction, Corica [/bib_ref] More recently, the properties of stents had been manipulated by using biodegradable materials such as polylactic acid, polyglycolic acid, and copolymers of lactide and glycolide to allow a longer degradation time. The main advantage of biodegradable stents is their spontaneous degradation and absorption without the need of removal. The most clinical utility of these stents is seen when they are used in combination with 5-alpha reductase inhibitors. [bib_ref] New developments in the use of prostatic stents, Papatsoris [/bib_ref] The spiral configuration of the first biodegradable stents was modified later on by researchers for a tubular mesh stent to decrease the risk of displacement and collapse as well as obstruction associated with breakage into large particles. Kotsar et al successfully treated ten patients with acute urinary retention by inserting a biodegradable braided urethral stent combined with dutasteride. All men were able to void after insertion of the stent and at 3 months 50% of patients were voiding with no problems. 54
## Prostatic urethral lift
The US Food and Drug Administration recently approved the marketing of a new device (UroLift ® , Neo Tract Inc., Pleasanton, CA, USA) offering an attractive novel mechanism to treat the enlarged prostate. Nonabsorbable monofilament sutures and a nitinol capsular tab are positioned from the urethra to the outer fibrous capsule in order to pull the lateral lobes of the prostate toward the capsule through permanent tensioning. A recent comprehensive review concluded that the UroLift system is superior to medical therapy for the improvement of LUTS but inferior to surgery. [bib_ref] Prostatic urethral lift improves urinary symptoms and flow while preserving sexual function..., Perera [/bib_ref] Apart from being a short, well tolerated procedure performed under local anesthesia and sedation in an outpatient setting, the major advantage of this system is the absence of erectile or ejaculatory dysfunction after treatment. [bib_ref] Prostatic urethral lift improves urinary symptoms and flow while preserving sexual function..., Perera [/bib_ref] In contrast, sexual function improved significantly at 1 year. [bib_ref] Preservation of sexual function with the prostatic urethral lift: a novel treatment..., Woo [/bib_ref] This could be understandably attributed to the absence of neurovascular bundle thermal injury and preservation of the bladder neck. Another attribute of this technology is the significant rapid and sustained relief from obstruction without the need for postoperative catheterization and low complication rates, as demonstrated by two multicenter, prospective, randomized blind clinical trials. [bib_ref] Treatment of LUTS secondary to BPH while preserving sexual function: randomized controlled..., Mcvary [/bib_ref]
## 131
Minimally invasive devices for treating LUTs in BPH first study, ie, the Luminal Improvement Following Prostatic Tissue Approximation trial, patient treated with the UroLift experienced a AUASI (American Urological Association Symptom Index) reduction from 22.1 at baseline to 18.0, 11.0, and 11.1 at 2 weeks, 3 months, and 12 months, respectively (P,0.001). Peak urinary flow rate increased to 4.4 mL per second at 3 months and was sustained at 4.0 mL per second at 12 months (P,0.001). Adverse events were typically mild and transient, with no occurrence of de novo ejaculatory or erectile dysfunction. [bib_ref] Treatment of LUTS secondary to BPH while preserving sexual function: randomized controlled..., Mcvary [/bib_ref] The second study had a different design, being a crossover study. It confirmed the outcomes of the first study, and showed that an UroLift procedure has no impact on subsequent use of TURP or laser therapy for BPH. [bib_ref] Multicentre prospective crossover study of the 'prostatic urethral lift' for the treatment..., Cantwell [/bib_ref] Of note, patients included in the two clinical trials had a prostatic volume ,80 mL, and long-term data beyond 3 years are lacking.
## Fractionation of prostatic tissue
This novel concept is based on the development of new technologies able to deliver high intensity energy enough to cause tissue emulsification without thermal effects. Two cutting edge technologies are under development, and are considered as experimental treatments under investigation, ie, histotripsy and aquablation.
## Emulsification of prostate tissue by histotripsy
The word histotripsy is derived from the Greek, meaning breakdown of soft tissue. This technique was initially conceived and developed at the University of Michigan. It uses acoustic energy in the form of short (,50 microseconds) very high intensity pulses that induce controlled cavitation clouds to mechanically homogenize the targeted tissue. The physical mechanism is similar to shockwave lithotripsy; each histotripsy pulse creates a localized highly dynamic cluster of microbubbles, the energetic interaction of which with tissue produces a fraction of the required tissue fragmentation or emulsification with cellular and tissue disruption, and subsequent pulses for bubble regeneration. In an in vivo rabbit model, by increasing the numbers of histotripsy pulses, small foci of homogenized tissue enlarged and merged with adjacent foci to finally produce a liquefied core of cytoplasm and fractionated debris with no discernible cellular structure surrounded by sharp smooth boundaries only a few millimeters wide. [bib_ref] Pulsed cavitational ultrasound: a noninvasive technology for controlled tissue ablation (histotripsy) in..., Roberts [/bib_ref] This process is self-limited after 1,000 pulses. [bib_ref] Pulsed cavitational ultrasound: a noninvasive technology for controlled tissue ablation (histotripsy) in..., Roberts [/bib_ref] However, its efficiency could be increased further by using techniques to remove the bubble nuclei that persist between pulses, as recently demonstrated by Wang et al. [bib_ref] An efficient treatment strategy for histotripsy by removing cavitation memory, Wang [/bib_ref] Unlike ther-mal ablative techniques, this cavitational mechanical effect is independent of temperature, as evidenced by an ex vivo porcine kidney model. [bib_ref] Refining histotripsy: defining the parameter space for the creation of non-thermal lesions..., Kieran [/bib_ref] The lack of heating as a mechanism for tissue destruction is the major distinction between histotripsy and HIFU. This principle has been confirmed with another recently proposed method of tissue emulsification by shock wave heating, ie, boiling histotripsy, which has the advantage to be less stochastic in its occurrence than cavitation, which makes the method utilizing boiling more predictable and repeatable. [bib_ref] Histological and biochemical analysis of mechanical and thermal bioeffects in boiling histotripsy..., Wang [/bib_ref] It had also been demonstrated in a canine model that the liquid consistency of the tissue after prostate histotripsy facilitates drainage via the urethra and results in minimal damage to surrounding tissues and blood vessels. [bib_ref] Histological and biochemical analysis of mechanical and thermal bioeffects in boiling histotripsy..., Wang [/bib_ref] This high precision was demonstrated in a phantom prostate composed of red blood cells and agar. By applying preconditioning pulses to delete bubble nuclei at the periphery of the targeted volume, Wang et al were able to further increase their accuracy on the target. [bib_ref] Active focal zone sharpening for highprecision treatment using histotripsy, Wang [/bib_ref] Of note, these microbubble clouds are excellent sound reflectors and allow transrectal monitoring of treatment progression by ultrasound in real time. Lake et al investigated the potential of histotripsy as a minimally invasive therapy for BPH, and reported the initial results for ablation of prostatic tissue using histotripsy in an in vivo canine model. [bib_ref] Histotripsy: minimally invasive technology for prostatic tissue ablation in an in vivo..., Lake [/bib_ref] Histotripsy was capable of precise prostatic tissue destruction and produced prostatic urethral damage, thereby facilitating drainage of fractionated material per the urethra and producing immediate debulking, as demonstrated by retrograde urethrography and histology. However, the periurethral prostatic tissues were more resilient to emulsification than their glandular counterparts and required higher doses. This lead Schade et al to develop urethral sparing histotripsy. [bib_ref] Endoscopic assessment and prediction of prostate urethral disintegration after histotripsy treatment in..., Schade [/bib_ref] They demonstrated effective homogenization and a 12% decrease in prostatic volume. In another study of nine anticoagulated canine subjects, Wheat et al suggested that histotripsy is safe and may induce hemostatic effects. [bib_ref] Prostate histotripsy in an anticoagulated model, Wheat [/bib_ref] Another study investigated the local and systemic effects of histotripsy in 18 canine subjects, [bib_ref] Histotripsy fractionation of prostate tissue: local effects and systemic response in a..., Hempel [/bib_ref] and reported the treatment to be safe and well tolerated, except in one case where a rectourethral fistula developed, with the rectum being inadvertently treated on several occasions. The relative susceptibility of the rectum when compared with the sphincter, neurovascular bundles, the bladder neck and the ureteral orifices was demonstrated in two other studies aiming to determine the thresholds of injury for surrounding tissues. [bib_ref] Histotripsy homogenization of the prostate: thresholds for cavitation damage of periprostatic structures, Styn [/bib_ref] [bib_ref] Histotripsy effects on the bladder trigone: functional and histologic consequences in the..., Allam [/bib_ref] An important consideration in translating histotripsy for the treatment of BPH in humans is the availability of an acoustic window to access and monitor the prostate. For most canines, the prostate is located cranial to the pubis, which makes the submit your manuscript | www.dovepress.com
## Dovepress
## Dovepress
## 132
Aoun et al transabdominal approach relatively straightforward. The human prostate is located deeper within the pelvis and closer to the perineum, which is why a transperineal approach is superior, as demonstrated by a recent study. [bib_ref] Acoustic access to the prostate for extracorporeal ultrasound ablation, Hall [/bib_ref] In general, the transrectal approach is preferred by the urologist. However, the transducer surface area needed to initiate cavitation clouds is larger than the size of the rectum. Future evolution of piezoelectric technology should result in miniaturization of the probe, allowing the transrectal approach. The concept of this new technology is appealing, and in order to move beyond the proof of concept, a human prototype device has been developed (Vortx Rx™, HistoSonics Inc., Ann Arbor, MI, USA). This device was approved by the US Food and Drug Administration for a human pilot study (ClinicalTrials. gov identifier NCT01775488) designed to assess and monitor the safety and efficacy of the Vortx Rx for the treatment of BPH. However, this study has been suspended because of poor recruitment.
## Aquablation using water jet hydrodissection
The Procept aquablation system (Procept Biorobotics Corp, Redwood Shores, CA, USA) is a newly developed cuttingedge technology. It is an integrated system using a highvelocity saline stream delivered under precise motion control to selectively ablate prostatic glandular tissue while sparing the blood vessels and capsule. Once the ablation is complete, a laser built into the water jet (Aquabeam™) can be used to produce surface hemostasis with a very low laser power (2-4 W). The extent, depth of ablation, and real time monitoring is predetermined by endoscopic and transrectal ultrasound guidance. The hypothetical advantage is a rapid fractionation of tissue without thermal injury and with minimal bleeding. [bib_ref] Image-guided robot-assisted prostate ablation using water jet-hydrodissection: initial study of a novel..., Faber [/bib_ref] This new technology has been established as safe in dogs, and there is one ongoing pilot study on humans (Australian New Zealand Clinical Trials Registry identifier ACTRN12613000167763).
## Percutaneous transluminal prostate artery embolization
Prostate artery embolization (PAE) involves the introduction of microparticles (polyvinyl alcohol, trisacryl gelatin microspheres, or other synthetic biocompatible materials) into the prostatic arteries via a percutaneous transfemoral approach, thereby blocking blood flow to the prostate and inducing partial necrosis and shrinkage. [bib_ref] Does polyvinyl alcohol particle size change the outcome of prostatic arterial embolization..., Bilhim [/bib_ref] [bib_ref] The histology of prostate tissue following prostatic artery embolization for the treatment..., Camara-Lopes [/bib_ref] In 1980, Darewicz et al demonstrated ischemic changes occurring in the prostate gland following internal iliac artery embolization in the dog. [bib_ref] The effect of internal iliac artery embolization of the prostatic gland in..., Darewicz [/bib_ref] They first defined the indication for internal iliac artery embolization to be used in clinical practice as a method for treating hemorrhage from the prostatic gland. DeMeritt et al described a patient with BPH who had severe gross hematuria and underwent prostatic arterial embolization. [bib_ref] Relief of benign prostatic hyperplasia-related bladder outlet obstruction after transarterial polyvinyl alcohol..., Demeritt [/bib_ref] The patient stopped bleeding and the size of the prostate decreased to 52% and 62% of its initial size on follow-up 5 and 12 months later. In 2008, an experimental study of PAE in pigs showed a significant reduction in prostate volume without compromising sexual function in the animals. [bib_ref] Benign prostatic hyperplasia: transcatheter arterial embolization as potential treatment -preliminary study in..., Sun [/bib_ref] In 2010, Carnevale et al were the first to describe primary treatment of BPH with PAE. [bib_ref] Prostatic artery embolization as a primary treatment for benign prostatic hyperplasia: preliminary..., Carnevale [/bib_ref] They successfully treated two patients with acute urinary retention, and reported a 47.8% and 27.8% reduction in prostate size for bilateral PAE and unilateral PAE, respectively, at the 6-month follow-up. Subsequently, one prospective randomized controlled trial and several uncontrolled pilot studies with limited numbers of patients had been reported, with promising results. [bib_ref] Society of Interventional Radiology position statement: prostate artery embolization for treatment of..., Mcwilliams [/bib_ref] [bib_ref] Prostatic artery embolization to treat lower urinary tract symptoms related to benign..., Golzarian [/bib_ref] [bib_ref] Early findings from a prospective US trial: prostatic artery embolization (PAE) in..., Bagla [/bib_ref] [bib_ref] Early results from a United States trial of prostatic artery embolization in..., Bagla [/bib_ref] The largest prospective non-randomized case series was reported for 255 patients in Portugal. [bib_ref] Embolisation of prostatic arteries as treatment of moderate to severe lower urinary..., Pisco [/bib_ref] Technical success was achieved in nearly all these patients, with a clinical success rate of 82% and 72% after the 1st month and 3 years, respectively. Both clinical and urodynamic parameters improved with PAE. The authors described the technique well, ie, that it involves assessment of the prostatic and pelvic arterial anatomy by performing computed tomographic angiography before the intervention. [bib_ref] How to perform prostatic arterial embolization, Pisco [/bib_ref] This allows identification of anatomical variants, avoidance of collaterals to adjacent organs, planning for targeted superselective catheterization of small prostatic arteries, and avoidance of unsuitable cases with small size vessel, excessive kinking, and/or atherosclerosis. Development of small microcatheters (2.4 French or even smaller) and catheters with a pear-shaped tip allowed experienced operators to overcome most of these anatomical difficulties. Non-target embolization had been demonstrated to cause transient ischemic proctitis and bladder wall necrosis. [bib_ref] Transient ischemic rectitis as a potential complication after prostatic artery embolization: case..., Moreira [/bib_ref] The authors suggested performing a proximal embolization first and then a second step of distal embolization to complete occlusion and stasis of blood flow to the prostatic tissue. [bib_ref] PErFecTED technique": proximal embolization first, then embolize distal for benign prostatic hyperplasia, Carnevale [/bib_ref] Clinical improvement after PAE is not fully predictable, and up to 15% of patients may have a poor outcome. [bib_ref] Prostatic arterial embolization: post-procedural follow-up, Fernandes [/bib_ref] Recently, de Assis et al reported the results of their prospective, single-center, single-arm study conducted in 35 patients. [bib_ref] Prostatic artery embolization for treatment of benign prostatic hyperplasia in patients with..., De Assis [/bib_ref] The development of infarcts and volume reduction following embolization is reported to be higher than 70% and is well assessed by MRI. [bib_ref] MRI findings after prostatic artery embolization for treatment of benign hyperplasia, Frenk [/bib_ref] The poor correlation between prostatic volume and clinical improvement suggests potential contribution of a mechanism other than reduction of prostatic volume.
## 133
Minimally invasive devices for treating LUTs in BPH In addition, the prostate volume may not be related to the severity of obstruction. This explains why unilateral embolization in some studies may achieve significant symptomatic improvement even with lower volume reduction. [bib_ref] PErFecTED technique": proximal embolization first, then embolize distal for benign prostatic hyperplasia, Carnevale [/bib_ref] Recently, the results of a prospective randomized trial comparing the efficiency and safety of PAE with that of TURP were published. [bib_ref] Benign prostatic hyperplasia: prostatic arterial embolization versus transurethral resection of the prostatea..., Gao [/bib_ref] The authors found no significant differences in functional urinary outcomes (P,0.001). IPSS, quality of life, Qmax, and post-void residual volume showed improvement in both groups. However, the improvement occurred significantly earlier in the TURP group than in the PAE group. The decreases in prostate-specific antigen (PSA) and prostate volume were significantly greater in the TURP group (P,0.05), and the PAE group showed more overall adverse events and complications (P=0.029), ie, acute urinary retention, post-embolization syndrome, and technical failures.
At present, PAE is still considered to be an experimental treatment modality for BPH and patients should be encouraged to participate in clinical trials. To our knowledge, there are currently three ongoing prospective randomized trials
recruiting. An open-label interventional trial recruiting since June 2014 in the USA (ClinicalTrials.gov identifier NCT02167919) is aiming to evaluate the efficacy of PAE in decreasing the volume of the prostate (.80 mL) in patients with refractory LUTS/BPH (IPSS .18). The UK-ROPE (Registry of Prostate Embolization) and a Swiss trial (ClinicalTrials.gov identifier NCT02054013) are designed as prospective, randomized, non-inferiority trials comparing the effects of treatment and adverse events for PAE and TURP.
## Intraprostatic drugs
The concept of intraprostatic injection dates back to 1832. In 1910, intraprostatic injection was used for the first time as a minimally invasive alternative to treat BPH.In 1966, Talwar and Pande obtained favorable results using a solution of glycerin, glacial acetic acid, and carbolic acid in 188 consecutive patients with acute urinary retention. [bib_ref] Injection treatment of enlarged prostate, Talwar [/bib_ref] Subsequently, these researchers used several chemical compounds in both animal and clinical studies with encouraging outcomes. The mechanism of action of intraprostatic injection was also partially described. It was demonstrated that these injections decrease the prostate volume by inducing apoptosis and tissue necrosis and act on afferent nerves to alleviate obstruction and improve LUTS. [bib_ref] Injectables in the prostate, Saemi [/bib_ref] The technique lost popularity until the last decade. This is partially explained by a shift of focus in the clinical understanding and management of LUTS from the prostate to the bladder.
Growing experience first developed interest in intraprostatic ethanol injections, which is to date the most widely investigated intraprostatic therapy. A recent review showed significant improvement in symptom scores, Qmax, post-void residual, and quality of life. [bib_ref] Injectables in the prostate, Saemi [/bib_ref] The post-injection perineal pain associated with extraprostatic extravasations substantially diminished after the advent of transrectal ultrasound and understanding of the anatomical landmarks. [bib_ref] Injectables in the prostate, Saemi [/bib_ref] Changing the route of administration to transrectal and inclusion of periprostatic nerve block makes the procedure more tolerable under local anesthesia. Use of anhydrous ethanol instead of liquid ethanol made control of distribution more accurate, thereby decreasing extravasations and complication rates. [bib_ref] Intraprostatic injection of alcohol gel for the treatment of benign prostatic hyperplasia:..., Larson [/bib_ref] The first investigational new drug study assessed the safety and efficacy of a novel system using a curved cystoscopic needle delivery device ( Prostaject, American Medical Systems) to administer anhydrous ethanol transurethrally. [bib_ref] Phase I/II examination of transurethral ethanol ablation of the prostate for the..., Plante [/bib_ref] The safety, tolerability, and initial efficacy were demonstrated. However, rare serious events and complications and considerable retreatment rates (41% after 3 years) preclude its current use in clinical practice. [bib_ref] Transurethral ethanol injection therapy for prostatic hyperplasia: 3-year results, Goya [/bib_ref] In the last decade, there has been growing interest in intraprostatic injection of botulinum toxin for the treatment of LUTS/BPH. Although the mechanism of action was attractive with promising results reported initially, the largest, placebocontrolled, double-blind trial on the efficacy of botulinum toxin in men with LUTS/BPH showed no significant difference between botulinum toxin and placebo in terms of IPSS (41% versus 13%), quality of life, Qmax (33% versus 13%), and prostate volume (14% versus 7%). [bib_ref] A randomized double-blind placebo-controlled phase 2 dose-ranging study of onabotulinumtoxinA in men..., Marberger [/bib_ref] On post hoc analysis, a significant reduction in IPSS versus placebo was observed with onabotulinumtoxinA in prior alpha-blocker users. Adverse events were comparable across all groups. At present, more basic research is needed to understand the mechanism of action of this toxin.
NX-1207 is administered as a clinic-based procedure via ultrasound-guided transrectal intraprostatic injection and causes apoptotic cell death. Phases I and II clinical trials have demonstrated a significant treatment effect for LUTS/BPH. NX-1207 has not shown any of the bothersome and limiting sexual side effects of the oral therapies. Two large Phase III trials are ongoing to further confirm the efficacy, safety, and tolerability of this catheter on anesthetic-free procedure. [bib_ref] NX-1207: a novel investigational drug for the treatment of benign prostatic hyperplasia, Shore [/bib_ref] PRX302 is a modified form of proaerolysin; highly toxic bacterial pore-forming protons activated by PSA with the ability to effectively and safely ablate PSA-producing prostate tissue. PRX302 is injected transperineally under transrectal ultrasound guidance. Phase I and II studies have submit your manuscript | www.dovepress.com
## Dovepress
## Dovepress
## 134
Aoun et al also demonstrated improvement in IPSS, quality of life, and prostate volume with no deleterious effect on sexual function. [bib_ref] Phase 1 and 2 studies demonstrate the safety and efficacy of intraprostatic..., Denmeade [/bib_ref] A recent Phase IIb trial reported that this outpatient-based procedure was well tolerated with statistically significant improvement in patient subjective (IPSS) and quantitative objective (Qmax) measures sustained during 12 months of follow-up. [bib_ref] Prospective, randomized, double-blind, vehicle controlled, multicenter Phase IIb clinical trial of the..., Elhilali [/bib_ref] Adequately powered, randomized, placebocontrolled, blinded studies with long-term follow-up are still needed before intraprostatic injection can be considered a suitable therapeutic option in contemporary clinical practice. The concept of intraprostatic injections of different agents as a minimally invasive surgical therapy for LUTS /BPH is attractive, but thus far the results available are unclear.
# Conclusion
The concept of new minimally invasive procedures combined with technical advances and refinements is attractive as a method for alleviating some of the concerns regarding minimally invasive procedures. Recently, several new approaches have emerged. Some of these techniques had already shown safety and efficacy for the treatment of LUTS/BPH with the advantage of being able to be performed as an outpatient procedure under local anesthesia with no sexual adverse events. Other procedures are to date considered as experimental and under clinical investigation. More systematic laboratory research and currently ongoing clinical trials need to be completed to elucidate the potential role of these newer devices for the treatment of LUTS/BPH.
## Research and reports in urology
Publish your work in this journal Submit your manuscript here: http://www.dovepress.com/research-and-reports-in-urology-journal Research and Reports in Urology is an international, peer-reviewed, open access journal publishing original research, reports, editorials, reviews and commentaries on all aspects of adult and pediatric urology in the clinic and laboratory including the following topics: Pathology, pathophysiology of urological disease; Investigation and treatment of urological disease; Pharmacology of drugs used for the treatment of urological disease. The manuscript management system is completely online and includes a very quick and fair peer-review system, which is all easy to use. Visit http://www.dovepress.com/testimonials.php to read real quotes from published authors.
## Research and reports in
[table] Table 1: Characteristics of HiFU devices in use for the treatment of benign prostatic hyperplasia and localized prostate cancer [/table]
[table] Table 2: Characteristics of commonly used prostatic stents [/table]
|
Establishment of a Percutaneous Coronary Intervention Registry in Vietnam: Rationale and Methodology
Background:In lower-and middle-income countries across Asia there has been a rapid expansion and uptake of percutaneous coronary intervention (PCI). However, there has been limited routine collection of related data, particularly around quality, safety and cost. The aim of this study was to assess the viability of implementing routine collection of PCI data in a registry at a leading hospital in Hanoi, Vietnam. Method: A Vietnamese data collection form and collection strategy were developed in collaboration with the Vietnam National Heart Institute. Information on patient characteristics, treatments, and outcomes was collected through direct interviews using a standardised form and medical record abstraction, while PCI data was read and coded into paper forms by interventional cardiologists. Viability of the registry was determined by four main factors: 1) being able to collect a representative sample; 2) quality of data obtained; 3) costs and time taken for data collection by hospital staff; and 4) level of support from key stakeholders in the institute. Results: Between September 2017 and May 2018, 1,022 patients undergoing PCI were recruited from a total of 1,041 procedures conducted during that time frame. The estimated mean time to collect information from patients before discharge was 60 minutes. Of the collected data fields, 98% were successfully completed. Most hospital staff surveyed indicated support for the continuation of the activity following the implementation of the pilot study. Conclusions: The proposed methodology for establishing a PCI registry in a large hospital in Vietnam produced high quality data and was considered worthwhile by hospital staff. The model has the potential opportunity for replication in other cardiac catheterisation sites, leading to a national PCI registry in Vietnam.Coronary heart disease (CHD) is consistently the leading cause of death worldwide, responsible for approximately 16.6% of total deaths in 2016 and places a large economic burden on the population [1, 2]. Since its inception in 1977, percutaneous coronary intervention (PCI) has been recognised as a valuable procedure for treating CHD patients and has become a common part of routine practice worldwide[3,4]. The Asia-Pacific region is home to nearly 60% of the world's population, where CHD is now a leading cause of mortality[5,6]and the development of PCI registries is of growing interest [7-9]. As a clinical quality registry, a PCI of Sydney. Her research interests focus on the use of large epidemiological datasets and published data to study obesity, diabetes, and cardiovascular disease in adults. Her current projects include the analysis of primary care and registry datasets in Australia. Education in Therapeutics. He is a cardiovascular epidemiologist and clinical trialist with a specific interest in quality improvement and outcomes research. He holds a NHMRC Senior Research Fellowship and has had continual NHMRC research funding since 1997. He has published more than 250 papers with over 6785 citations (Google Scholar H-index 41). Over the past 5 years (from 2009), he has received 17 peer reviewed grants as a Chief Investigator from National and International funding agencies totalling over $70M.
database is an important mechanism for monitoring and benchmarking the performance of clinical care, improving safety and outcomes, contributing to reducing treatment cost and regulating guidelines. Nonetheless, there remains wide geographic variation in terms of the organisation, operation, management, sustainability and utilisation of data collected of PCI registries in Asia. Additionally, data are limited regarding the participation of less economically developed countries, particularly those in the South-East Asia region, including Vietnam.
As a nation undergoing rapid economic and epidemiological transition, Vietnam has experienced a high burden of CHD, causing more than 58,000 deaths (11.6% of all mortality) in 2017. Vietnam began to adopt PCI in 1995 at the Vietnam National Heart Institute (VNHI), and has to date introduced this procedure to approximately 70 cardiac centres nationwide. The annual number of PCI procedures is relatively large and increasing; for instance, there were 2,250 patients receiving this technique in 2013 in a single national centre and there is a 15% increase annually. Notwithstanding its widespread use, there has been no PCI registry in Vietnam.
This paper presents the rationale, design and conduct of a pilot PCI registry model in Vietnam. The viability of implementing routine collection of PCI data was also assessed and documented. If viable, this would be the initial step in developing a model for an expanded PCI registry in Vietnam. The major objectives of this pilot study are 1) to describe the implementation experience at a large cardiac centre in Vietnam; 2) to describe the methodology for developing the PCI registry in Vietnam; and 3) to report on the viability of the strategy as a model for the nation.
# Methods
## Study setting
This PCI pilot registry project was conducted at the VNHI. Located in Hanoi, the capital city of Vietnam, VNHI is the biggest cardiac and referral centre in the country, providing the highest quality of healthcare services for cardiovascular patients in the country. Numerous advanced catheter-based therapies including complex coronary stenting, aortic stent grafting, transcatheter cardiac structural interventions, have become integrated into routine clinical practice at VNHI. As a 450-bed medical institution, VNHI receives around 17,000 inpatients and 80,000 out-patients annually, from across Northern Vietnamese provinces. The number of cardiac interventional procedures undertaken at VNHI is increasing from approximately 2,800 in 2004 to 12,000 in 2018. In addition, VNHI is an education and training centre, from which cardiovascular technologies and therapies are transferred into practice at other lower level medical institutions. VNHI was therefore selected for implementing the pilot PCI registry to reflect the contemporary practice of PCI in Vietnam.
## Establishment of dataset
This study adapted the current versions of standardised data abstraction forms developed for the Victorian Cardiac Outcomes Registry (VCOR), Australia, including the standard case report form (CRF) and dataset definitions for all fields. The state-wide VCOR was built on the Melbourne Interventional Group registry, in which PCI data elements are in line with a number of current interventional registries worldwide, for instance, the American College of Cardiology -National Cardiovascular Data Registry. The standard dataset aimed to collect the minimum standard data and avoid cumbersome management:
(1) The three page baseline survey, administered at time of presentation for procedures, contains 13 sections: patient details, admission data, clinical symptoms, clinical presentation, pre-procedural left ventricular function, risk factors, renal status, medication, procedure details, postprocedural cardiac biomarkers, in-hospital complications, discharge details and medications. (2) The one page survey for 30-day and 12-month follow-up, including four sections: patient details, outcomes, medications and quality of life at 30 days and 12 months.
The VCOR data collection forms were translated into Vietnamese, and revised by two Vietnamese clinical cardiologists to reflect local practice. After discussion, consensus was reached around the addition of some new elements into the Vietnamese data collection forms, such as patient details (e.g. medical record number, ethnic group, poverty status, educational level, occupation, and income) and risk factors (e.g. smoking, dyslipidaemia). The Vietnamese data collection forms were designed in the TELEFORM software [21] and printed in paper records. The specific questions for all three data collection points are presented in the Appendix.
## Data collection
Following the finalisation of the data collection form, data recruitment commenced in September 2017. Potential participants were patients who underwent PCI at VNHI during the study period and met the following criteria: (1) Vietnamese residents aged 18 years and over;Had at least one active phone contact number; and (3) Able to communicate, understand the information sheet and did not opt-out of future follow-ups by the time of discharge. There were no exclusion criteria. Under the strict clinical audit and strategy for data collection, baseline data collection was conducted over a 9 month period (from September 2017 to May 2018), followed by the 30-day follow-up, while the 12-month follow-up is underway.
## Baseline survey
Baseline data were collected using a paper-based form through interviewing patients, visiting the catheterization laboratory where PCI was performed and extracting information from medical records. The data manager at VNHI and research assistants were responsible for conducting these activities. Compliance with the project protocol was supervised by staff trained in clinical audit processes. The patient interviews were largely conducted in wards, following the index PCI and prior to discharge when patients were medically well enough, as assessed by the responsible physician. Data on the index PCI (e.g. the PCI indication, entry location, adjunctive devices, lesion characteristics, in-stent restenosis, stent thrombosis, and stents used) were obtained from the catheterization laboratory. Images of coronary lesions were stored on protected disks, and printed, read and coded by a cardiologist. Other information was abstracted from medical records including time of admission, in-hospital management, medications used, clinical tests and pre-discharge complications (e.g. renal impairment, cardiogenic shock, bleeding [classified by the Bleeding Academic Research Consortium (BARC)]), stroke, new or current myocardial infarction, target vessel or lesion revascularisation after the procedure (PCI or coronary artery bypass grafting). It took approximately one hour to complete a CRF on average.
## Follow-up surveys
The 30-day and 12-month follow-up surveys were designed to capture data on the combined endpoint of major adverse cardiac and/or cerebrovascular events such as all cases of death, new or recurrent myocardial infarction or stent thrombosis, target vessel revascularisation or stroke; bleeding (BARC); rehospitalisation; medication use and health quality of life (mobility, personal care, usual activities, pain/discomfort, anxiety/depression, and own health state today) of participants after the index PCI.
At 30 days, a face-to-face interview was conducted by the data manager or research assistants if the participant was physically present at VNHI; otherwise, a phone interview was used. Supplementary information from a heart ultrasound and blood tests was recorded during the face-to-face interviews at VNHI, which lasted approximately 30 minutes.
The 12-month follow-up survey is beyond the scope of the current report, however the planned methods are as follows. A 15-minute phone interview is being undertaken by trained research assistants to obtain information from the patients directly. First-degree relatives are used as the proxy if the participant is not contactable. At least three attempts will be made to contact the participants. If patients have additional concerns regarding their health, then consultation will be available upon request following the interview.
## Perspectives from vnhi
We also conducted an online survey with qualitative open-ended questions to explore perspectives on the implementation of the registry and identify key factors associated with successful implementation of this new model at VNHI. We approached all clinical, nursing and leadership staff involved in coronary interventional activity. The survey that contained 10 questions was administered using Qualtrics Research Suite (Qualtrics, Provo, UT), a web-based tool that allows researchers to build, distribute, and analyse online surveys in real time. Analysis of qualitative data was guided by the principles of the conventional and summative content analysis. Briefly, responses obtained from participants were coded to identify and categorise different themes together with performing word counts. The interpretation focused on the several key factors that may influence the development and implementation of PCI registry in Vietnam.
## Registry viability
This is the first study focusing on developing a model to collect data on the contemporary practice of PCI at the largest cardiac institute in Vietnam. The viability of the pilot PCI registry as a model for a national registry in Vietnam was determined by the following elements: a) Being able to recruit a representative sample into the registry.
b) The quality of data collected, determined by data completeness and audit activities. c) Costs and time taken to collect the data by hospital staff. d) The level of support for the activity from patients, clinical staff and the cardiac institute.
## Ethics approval
Ethics for the study was approved by the Curtin University Human Research Ethics Committee (HRE 2017-0378). Every participant was provided with a Patient Information Sheet in which the purpose of the study, activities and rights of participants were described clearly. Participating in this study was voluntary and participants had the right to decline their participation or withdraw from the study at any time without any consequence via an ' opt-out' consent. A unique ID was assigned to each participant and linkable to the private information such as name, age, address, and phone numbers for follow-ups. All information that identifies participants was coded and stored confidentially.
# Results
## Implementation experience
During the study period (from September 2017 to May 2018), patients who underwent PCI at VNHI and met the inclusion criteria were approached and invited to take part in the study. Three strategies were used to enrol patients as the pilot study progressed, with changes applied to recruitment and the corresponding results. The modification of the data collection strategy was implemented to ensure that, in the absence of resources to capture all cases, we were able to capture a representative sample and minimise the potential for selection bias in registry enrolment when a single data manager was responsible.
For the first period of data collection (September and October 2017), 6-7 patients were randomly selected from the total number admitted of each day. However, to eliminate the potential bias of choosing only good prognosis patients (i.e., without complications prior to and after the procedure), we moved to second approach (from November 2017 to March 2018) which aimed to collect all cases in two weekdays. Two particular weekdays (Monday and Tuesday) were chosen to capture data from all operators including two teams of interventionists who performed PCI at VNHI on alternate days during the week. Notably, this second period of data collection coincided with Tet (Lunar New Year's holiday), the biggest cultural event in the year, which may explain a reduction in the number of patient visits. The actual number of patients recruited in the period one and three may be generalizable to the rest of the year (i.e., from June to August) because there were no similar events in the year. Finally, the last period (from April to mid-May 2018) included all patients undergoing PCI on two weekdays and the weekend, which was designed to capture both the normal practice (weekdays) and acute cases (weekend).
A total of 1,028 eligible coronary patients were approached and invited from a total of approximately 2,800 patients undergoing PCI at VNHI during the study period. Six patients refused to participate in the study by opt-out consent and thus 1,022 patients remained in the baseline study sample, which included 1,041 individual PCIs (as 19 patients underwent PCI twice at VNHI). There was an extremely high rate of data completeness, with only information of one lesion missed due to a lost disk (0.1%). Ninety-eight percent of fields were fully filled, with the exception of oral anti-coagulant therapy as patients had difficulty recognising the kinds of drug used and unknown information about their referral to cardiac rehabilitation. The successful follow-up rate at 30 days was high, with 993 patients followed-up (97.2%).
Of the 25 invited cardiovascular professionals, 12 consented to participate in the qualitative survey (response rate: 48%). These 12 respondents included 5 cardiologists with administrative leadership at VNHI, 3 clinical cardiologists and 4 nurses. Several key additional factors concerning the successful implementation of a PCI registry in Vietnam were raised. Nine respondents agreed with the importance of standardised data collection forms as used in the registry. Other key facilitating factors were also emphasized, including well-trained investigators, the use of professional clinical audit, and strong support from leaders of target cardiac institutions. They also raised concerns regarding the sustainability of such a study at VNHI, including lack of data storage systems, sufficient funding for infrastructure and human resources, and strong commitment from hospital leaders.
## Registry viability
The viability of the pilot PCI registry in VNHI was determined by the following elements: a) Being able to recruit a representative sample into the registry. After several amendments, the data collection strategy captured all patients undergoing PCI at VNHI four days per week (2 week days with routine practice and weekend with emergency cases only). Thus, the sample recruited into the registry could be considered to be representative of coronary patients treated with PCI at VNHI when there were not sufficient resources to collect all cases. b) Data quality. In total, we collected information on 99 data fields, and overall we had 98% data completeness. The reasons for this high completeness were the strong cooperation of patients and the availability of data resources (e.g., medical records, disks and machines in the catheterization laboratory). Additionally, the local clinical audit was performed monthly by well-trained staff, including case ascertainment (checking the collection of eligible data) and data quality assessment (reviewing source of data collection). Overall, 2% of cases were randomly checked and there were no significant errors in choosing patients and medical records. Thus, the quality and accuracy of data collected at VNHI was ensured. c) Costs for and time taken to collect the data by hospital staff. It took approximately an hour to complete a CRF, including 15 minutes for interviewing patients, abstracting data from medical records and reading procedure information in the secured disks with the cardiologist. In further studies, data collection will be the responsibility of hospital nurses. Therefore, the actual time for completing the data collection form would be around or less than one hour because of the familiarity with routine PCI practice. While hospital nurses will do data collection as part of routine clinical activity, we estimated the cost required for data collection. If we use an average monthly income of a nurse at VNHI, which is approximately 1066 USD 1, then the estimated time-cost of baseline data collection for one case, which is the income of hospital nurse in 1 hour, was equivalent to 1066/(30*8) = 4.4 USD.In follow-up survey, approximate 15-30 minutes (phone or direct interviews) will be required for each patient, which was roughly equivalent to 1.1-2.2 USD. d) The level of support from patients, clinical staff and the leader team. In the pilot registry, patients at VNHI had shown their strong engagement with the registry and there were no significant difficulties regarding the patients noted. The leader team and the hospital staff were well aware of the necessity to develop such a registry at the VNHI and in Vietnam and shown their universal support. They also belief that the establishment of the registry would be successful if there were sufficient data storing system, sufficient funding for human resources and improving the infrastructure, and strong commitment from hospital leaders.
# Discussion
In the context of the growing interest in developing clinical quality registries worldwide, this paper reports the development of the first PCI registry in Vietnam, using and adapting experiences from longstanding registries in Australia. The work to date has demonstrated a PCI registry to be feasible and suitable for Vietnamese circumstances, providing a significant opportunity to extend the approach to other cardiac centres looking to replicate the model. Importantly, implementing such a model not only provides crucial feedback on the performance of PCI for Vietnamese clinicians and cardiac care providers but also allowing a robust comparison with other regional registries such as those involved in the ASPECT collaborationas well as contributing to the literature on the use of PCI.
In comparison with other regional PCI registries in Thailand, China, India and Malaysia, the methodology in our pilot registry had some similarities, including using a standard abstraction form to collect consecutive patients undergoing PCI, providing sufficient training for investigators prior to data collection, performing clinical audit to ensure data quality and conducting follow-ups at 30 days and 12 months to investigate the outcomes of PCI. Nonetheless, due to resource constraint, we did not approach all the patients undergoing PCI in the study period. The CRF was also completed in a paper format only and data obtained were not transferred to a web based system as other PCI registries. Therefore, further studies might apply our methodology if there is limited resources or overcome our drawbacks if there is sufficient funding.
Although it is at an early stage, we are optimistic about the viability of the PCI model that we have implemented at the VNHI. The success of the PCI registry at the VNHI to date, is due to a variety of key factors such as the high number of patients undergoing PCI and their strong engagement with clinicians, the availability of data resources, and the supportive hospital staff team. We also faced several challenges in the first registry implement at VNHI, including high workload which might affect the time spent on data collection by the hospital staff and the lack of electronic record systems which made it difficult to collect comprehensive information when patients revisited the hospital. However, these obstacles can be minimised by providing sufficient training for clinical investigators, specifying sections of the data forms for investigators and conducting more detailed follow-up surveys.
One limitation is that the registry was conducted at VNHI, the leading cardiac centre in Vietnam where there is a highest patient throughput. Therefore, extending the methodology in establishing PCI registries in smaller settings with potentially less experienced staff may require a modified approach to that which we have done, but we believe this current work represents an important first step in doing so. Another potential concern may be bias in data collection, though attempt was made to enrol a representative sample. Thus, where sufficient resources are available, collecting all patients undergoing PCI in the study period would be recommended.
# Conclusion
This paper describes the methodology of establishing the first PCI registry at the leading cardiac centre in Vietnam and reports on the viability of this model. We hope that the successful implementation of a PCI registry at VNHI will encourage other cardiac intervention centres in Vietnam to adopt this model in their daily practice and by doing so, enables the opportunity to develop a nationwide PCI registry.
## Abbreviations
## Additional file
The additional file for this article can be found as follows:
-
## Richard norman
Richard Norman is an Associate Professor in Health Economics, based at the Curtin University School of Public Health. His research focuses on economic evaluation of healthcare, on the measurement and valuation of quality of life, and on patient and population preferences for different aspects of healthcare.
## Rachel r. huxley
Professor Rachel Huxley is Associate Pro-Vice Chancellor (Research) for the College of Science, Health and Engineering, and also Co-Director of the recently established Research Centre for Cardiovascular Biology and Disease at La Trobe University. Previously, she was the Head of School of Public Health at Curtin University, Australia. She completed her undergraduate, PhD and post-doctoral training in Oxford University. She has published more than 200 research articles, has a Google Scholar H-index of 64 and >20,000 citations. She currently holds several competitive research grants as Principal Investigator in areas related to obesity, diabetes and women's health.
## Ngoc m. pham
Ngoc Minh Pham is a medical doctor by training from Vietnam and received his Master of Public Health and PhD in Medical Science from the University of Melbourne in Australia and Kyushu University in Japan, respectively. He currently works as a visiting research associate at Curtin University, Australia and a lecturer in epidemiology at Thai Nguyen University of Medicine and Pharmacy, Vietnam. He has eight years of postdoctoral research experience in the epidemiology and prevention of lifestyle-related diseases and disorders, particularly type 2 diabetes, metabolic syndrome and cardiovascular risk factors. |
Joint association of physical activity and overweight with subsequent psychotropic medication: a register-linked follow-up study among employees
Background: Physical activity level and overweight have shown associations with mental health problems but it is not known whether the risk of mental health problems due to overweight varies by physical activity. We examined joint association of physical activity and overweight with subsequent psychotropic medication among 40-60-year-old employees. Methods: The questionnaire survey data were derived from Helsinki Health Study baseline postal questionnaires in 2000-02 among employees of the City of Helsinki aged 40-60 years (n = 8960, response rate 67 %). Baseline survey data were linked with prospective register data on prescribed psychotropic medication (ATC-codes N05 and N06, except N06D) among those with written consent (74 %) for such linkage. The analyses included 6169 responders (78 % women, corresponding to the target population). We divided participants into six groups according to their baseline self-reported body mass index and leisure-time physical activity using physically highly active normal-weight participants as a reference group. We used Cox regression analysis adjusted for age, gender, psychotropic medication prior to baseline, and socioeconomic position, marital status, working conditions, limiting long-standing illness, alcohol use, and smoking. Results: At baseline, 49 % were overweight and 23 % were physically inactive. After adjusting for age and gender, inactive normal-weight (hazard ratio (HR) 1.3, 95 % CI 1.1-1.5), moderately active overweight (HR 1.3, 95 % CI 1.1-1.5) and inactive overweight (HR 1.4, 95 % CI 1.2-1.6) had higher risk for any psychotropic medication compared with group of highly active normal-weight. After adjusting for prior medication, only the inactive overweight group had higher risk (HR 1.4, 95 % CI 1.2-1.6). Other covariates made but a minor contribution to the examined associations. For antidepressants the associations were somewhat stronger than for sedatives. Conclusions: Both normal-weight and physical activity help prevent psychotropic medication but physical activity dominates the association over normal-weight.
# Background
Mental health problems are increasing around the world and are topical issues from public health perspective [bib_ref] The global burden of mental disorders: an update from the WHO World..., Kessler [/bib_ref] and better treatment might be beneficial to decreasing costs of mental health problems. Psychotropic medication is widely used in general population of various countries [bib_ref] Prescription psychotropic medication use among the U.S. adult population: results from the..., Paulose-Ram [/bib_ref] [bib_ref] Psychotropic medication consumption patterns in the UK general population, Ohayon [/bib_ref] [bib_ref] Epidemiology of psychotropic drug use in Rio de Janeiro, Brazil: gaps in..., Quintana [/bib_ref] and there is evidence that prescribed psychotropic medication is more common among women than men [bib_ref] Psychotropic medication consumption patterns in the UK general population, Ohayon [/bib_ref] [bib_ref] Epidemiology of psychotropic drug use in Rio de Janeiro, Brazil: gaps in..., Quintana [/bib_ref]. Poor mental health further is a major cause for work disability and disability-retirement [bib_ref] Common mental disorders and subsequent work disability: a population-based health 2000 study, Ahola [/bib_ref].
Physical inactivity and overweight are associated with mental health problems [bib_ref] Leisure-time physical activity and psychotropic medication: a prospective cohort study, Lahti [/bib_ref] [bib_ref] Physical activity and likelihood of depression in adults: a review, Teychenne [/bib_ref] [bib_ref] The relation between body mass index, mental health, and functional disability: a..., Bruffaerts [/bib_ref] [bib_ref] Obesity and mental disorders in the adult general population, Scott [/bib_ref]. Additionally, physical inactivity alone is a public health risk factor and a number of studies have showed that leisure-time physical inactivity increases the risk of various diseases and mortality [bib_ref] The effects of physical activity and body mass index on cardiovascular, cancer..., Hu [/bib_ref] [bib_ref] Body mass index-independent effect of fitness and physical activity for all-cause mortality, Pedersen [/bib_ref] [bib_ref] Body mass index, physical activity, and the risk of decline in overall..., He [/bib_ref] [bib_ref] Association of leisure time physical activity with the risk of coronary heart..., Haapanen [/bib_ref]. Leisure-time physical activity may also be important in reducing the risk of mental health problems [bib_ref] Relationship between physical activity and general mental health, Kim [/bib_ref] and even small amounts of physical activity may protect against depression [bib_ref] Physical activity and likelihood of depression in adults: a review, Teychenne [/bib_ref]. Furthermore, some studies suggest that higher amounts of physical activity may show further benefits for mental health [bib_ref] Leisure-time physical activity and psychotropic medication: a prospective cohort study, Lahti [/bib_ref] [bib_ref] Association of physical activity with future mental health in older, mid-life and..., Griffiths [/bib_ref]. Physical activity has been successfully used to treat depression [bib_ref] Physical activity, exercise, depression and anxiety disorders, Strohle [/bib_ref] [bib_ref] Exercise and physical activity in mental disorders, Wolff [/bib_ref] [bib_ref] Exercise treatment for depression: efficacy and dose response, Dunn [/bib_ref] and some anxiety disorders [bib_ref] Physical activity, exercise, depression and anxiety disorders, Strohle [/bib_ref] although the evidence is still weak. A study of six European populations [bib_ref] The relation between body mass index, mental health, and functional disability: a..., Bruffaerts [/bib_ref] showed that obese people had more mood disorders and comorbid mental disorders compared to normal-weight population. The association between high BMI and mental health problems is seen clearest in post-traumatic stress disorder and in different anxiety disorders [bib_ref] Obesity and mental disorders in the adult general population, Scott [/bib_ref].
Earlier longitudinal study from the same data showed that physical activity is likely to be beneficial for selfreported mental health functioning among both normalweight and overweight adults after a follow-up of 5 to 7 years [bib_ref] Joint association of physical activity and body weight with subsequent physical and..., Lindholm [/bib_ref]. Furthermore, a cross-sectional study of physical activity and overweight with health-related quality of life showed that physical activity might be more important than normal-weight for health-related quality of life [bib_ref] Recommended levels of physical activity and health-related quality of life among overweight..., Heath [/bib_ref]. We still lack studies of joint-association of physical activity and overweight with subsequent mental health using objective register based measures such as psychotropic medication.
The aim of this study was to examine joint association of leisure-time physical activity and overweight with subsequent psychotropic medication. The contribution of prior medication and other key covariates such as smoking, drinking problems, limiting long-standing illness, socioeconomic position, employment status, and marital status to the association was also examined, because they have been associated with mental health in previous studies [bib_ref] Cigarette smoking, alcohol intoxication and major depressive episode in a representative population..., Hämäläinen [/bib_ref] [bib_ref] Social inequalities and the common mental disorders: a systematic review of the..., Fryers [/bib_ref] [bib_ref] Increased 12-month prevalence rates of mental disorders in patients with chronic somatic..., Härter [/bib_ref] [bib_ref] Revisiting the relationships among gender, marital status, and mental health, Simon [/bib_ref]. We expected that normal-weight and high physical activity is the best protection against mental health problems and psychotropic medication. Based on previous study [bib_ref] Leisure-time physical activity and psychotropic medication: a prospective cohort study, Lahti [/bib_ref] we further expected that high physical activity might protect against psychotropic medication also among the overweight.
# Methods
The questionnaire survey data were derived from the Helsinki Health Study baseline postal questionnaires in 2000-02. Participants were employees of the City of Helsinki, Finland, aged 40-60 years. There were 8960 responders (response rate 67 %) of whom 78 % were women. The response rate was somewhat lower among men, in younger age groups, lower occupational classes and those with long sickness absence. The data were linked to the register of prescribed psychotropic medication from the Social Insurance Institution of Finland. Seventy-four percent of participants gave permission for linkages and non-response analysis showed that nonresponse and non-consent unlikely biased results [bib_ref] Cohort profile: the Helsinki health study, Lahelma [/bib_ref] [bib_ref] Register-based study among employees showed small nonparticipation bias in health surveys and..., Laaksonen [/bib_ref]. The follow-up started at the day of returning the questionnaire and ended to first psychotropic medication purchase or death. The maximum follow-up time was 7.8 years and mean follow-up time was 6.4 years (SD 2.45). Those with psychotropic medication at baseline (n = 319) or with missing information on leisure time physical activity (n = 43), BMI (n = 47) or other covariates (n = 27) were excluded. The analyses included 6169 responders and 78 % on those were women. The ethics committees of the Department of Public Health, University of Helsinki and health authorities of the City of Helsinki have approved the Helsinki Health Study.
## Physical activity
The participants were asked about the average weekly time they spent in leisure-time physical activity (commuting included) within the last 12 months. They were first asked to estimate the intensity of their physical activity compared to common activities: walking, brisk walking, jogging, and running, or activities equivalent to these. After that they were asked to estimate on average how much time they spent in each intensity grade in 1 week. Based on responds of leisure-time physical activity, an approximate metabolic equivalent (MET) value was calculated [bib_ref] Compendium of physical activities: an update of activity codes and MET intensities, Ainsworth [/bib_ref]. The average MET hours of leisure-time physical activity in 1 week were calculated by multiplying time used weekly by the estimated MET value of the each intensity grade [bib_ref] Relationship of leisure-time physical activity and mortality: the Finnish twin cohort, Kujala [/bib_ref] and then summing all four MET values together. The participants were then grouped into inactive, moderately active, and highly active groups and as cut-off points we used the recommended levels of physical activity [bib_ref] Physical activity pie; a graphical presentation integrating recommendations for fitness and health, Fogelholm [/bib_ref]. The physically inactive group did under 14 MET hours weekly and it corresponds approximately to 1000 kcal per week or less. Moderately active group did 14-30 MET hours per week (approximately 1000-2000 kcal). Highly active group did over 30 MET hours (approximately over 2000 kcal) of physical activity per week, which is the recommended level for healthy weight maintenance [bib_ref] Physical activity pie; a graphical presentation integrating recommendations for fitness and health, Fogelholm [/bib_ref].
## Body mass index
Body mass index was calculated from self-reported weight (kg) and height (m). Participants whose BMI was under or equal to 25 kg/m 2 were grouped as normal-weight and whose BMI was over 25 kg/m 2 were grouped as overweight. Ca 1 % (n = 59) of participants were underweight (BMI under 18,5 kg/m 2 ) and those were excluded from the sample but no effect to the association was found and underweight participants were retained in the final sample.
## Psychotropic medication
Psychotropic medication data from Social Insurance Institution of Finland were classified according to the Anatomical Therapeutic Chemical (ATC) classification system. ATC codes N05 and N06 (except N06D which is medication for dementia) were included to any psychotropic medication. We also examined antidepressants (N06A) and anxiolytics and sedatives (N05B and N05C) separately but groups were not mutually exclusive. The follow up started at the day of returning the questionnaire and ended to first psychotropic medication purchase or death. Mortality data were derived from the registers of Statistics Finland.
## Covariates
Covariates derived from the baseline questionnaire included age, gender, marital status, physical and mental strenuousness of work, smoking, drinking problems, and limiting long-standing illness (LLI). Socioeconomic position was derived from the employer's registers and prior psychotropic medication was derived from the register of Social Insurance Institution of Finland. At baseline age was classified into five groups: 40, 45, 50, 55 and 60 years. Information on prior psychotropic medication 3 years preceding baseline were included. Socioeconomic position was measured by four occupational classes: managers/professionals, semi-professionals, routine nonmanual employees, and manual workers [bib_ref] Occupational class inequalities across key domains of health: results from the Helsinki..., Lahelma [/bib_ref]. Marital status was dichotomized to partnership or no partnership. Physical strenuousness of work was dichotomized to light or heavy. Similar procedure was followed for mental strenuousness of work. Smoking was divided to smokers and non-smokers. Alcohol use was measured by CAGE questionnaire, which describes drinking problems [bib_ref] Detecting alcoholism. The CAGE questionnaire, Ewing [/bib_ref]. LLI was dichotomized into those reporting any limiting long-standing illness and those not reporting LLI.
# Statistical methods
We divided participants into six groups according to their body mass index and leisure-time physical activity:
(1) inactive (under 14 METs) and normal-weight (BMI ≤ 25 kg/m 2 ), (2) moderately active [bib_ref] Relationship between physical activity and general mental health, Kim [/bib_ref] [bib_ref] Association of physical activity with future mental health in older, mid-life and..., Griffiths [/bib_ref] [bib_ref] Physical activity, exercise, depression and anxiety disorders, Strohle [/bib_ref] [bib_ref] Exercise and physical activity in mental disorders, Wolff [/bib_ref] [bib_ref] Exercise treatment for depression: efficacy and dose response, Dunn [/bib_ref] [bib_ref] Joint association of physical activity and body weight with subsequent physical and..., Lindholm [/bib_ref] [bib_ref] Recommended levels of physical activity and health-related quality of life among overweight..., Heath [/bib_ref] [bib_ref] Cigarette smoking, alcohol intoxication and major depressive episode in a representative population..., Hämäläinen [/bib_ref] [bib_ref] Social inequalities and the common mental disorders: a systematic review of the..., Fryers [/bib_ref] [bib_ref] Increased 12-month prevalence rates of mental disorders in patients with chronic somatic..., Härter [/bib_ref] [bib_ref] Revisiting the relationships among gender, marital status, and mental health, Simon [/bib_ref] [bib_ref] Cohort profile: the Helsinki health study, Lahelma [/bib_ref] [bib_ref] Register-based study among employees showed small nonparticipation bias in health surveys and..., Laaksonen [/bib_ref] [bib_ref] Compendium of physical activities: an update of activity codes and MET intensities, Ainsworth [/bib_ref] [bib_ref] Relationship of leisure-time physical activity and mortality: the Finnish twin cohort, Kujala [/bib_ref] [bib_ref] Physical activity pie; a graphical presentation integrating recommendations for fitness and health, Fogelholm [/bib_ref] normal-weight, (3) highly active (over 30 METs) normalweight, (4) inactive overweight (BMI > 25 kg/m 2 ), (5) moderately active overweight, and (6) highly active overweight. The group of active normal-weight participants was used as a reference group in all the analyses. Proportions and 95 % confidence intervals for subsequent psychotropic medication was first calculated. Cox regression analysis was then used to examine the effect of covariates and to calculate hazard ratios (HR) and their 95 % confidence intervals (95 % CI). Men and women were in the pooled data adjusting for gender because there was no significant difference between genders (p = 0.44 for interaction). In model 1 age and gender were adjusted for. In addition to covariates in model 1, in model 2 prior medication was adjusted for. In model 3 covariates in model 2 and socioeconomic position, marital status, and physical and mental strenuousness at work were adjusted for. In model 4 covariates in model 2 and smoking, drinking problems, and LLI were adjusted for. We used IBM SPSS Statistics version 22.0 for mac.
# Results
At baseline, 19 % were physically highly active and normal-weight and 18 % were physically inactive and overweight [fig_ref] Table 1: Baseline participant characteristics [/fig_ref]. During the follow-up, 30 % of participants had at least one reimbursed purchase of any psychotropic medication. Every fifth had antidepressant purchases and only slightly fewer had sedative purchases. Thirty-two percent of those who had any purchases of psychotropic medication purchased both sedatives and antidepressants. Exclusively antidepressant purchases were 35 % and exclusively sedatives 29 % of all psychotropic medication purchases. Nineteen percent of participants had had psychotropic medication within the last 3 years before the baseline survey.
Psychotropic medication tended to be more common in physically inactive overweight [fig_ref] Table 2: Prevalence of first psychotropic medication purchases [/fig_ref]. The patterns were similar for antidepressants and sedatives.
In age and gender adjusted model the inactive normalweight (HR 1.3, 95 % CI 1.1-1.5), moderately active overweight (HR 1.3, 95 % CI 1.1-1.5), and inactive overweight (HR 1.4, 95 % CI 1.2-1.6) had higher risk of psychotropic medication compared with the highly active normal-weight, however, the highly active overweight were not at increased risk (HR 1.0, 95 % CI 0.9-1.2) [fig_ref] Table 3: Hazard ratios for psychotropic medication purchases according to joint association of physical... [/fig_ref]. Adjusting for prior medication before baseline attenuated the associations after which only the inactive overweight group had a significantly higher risk (HR 1.4, 95 % CI 1.2-1.6) for any psychotropic medication. Adjusting for covariates in model 3 had only a small effect compared with model 2, but adjusting for LLI, smoking, and drinking problems (model 4) further attenuated the associations. However, the inactive overweight group still had a higher risk of psychotropic medication (HR 1.2, 95 % CI 1.0-1.4).
For antidepressant medication the associations were similar although slightly stronger than for the any psychotropic medication. After adjustment for prior medication the moderately active overweight, inactive normal-weight and the moderately active and inactive overweight groups had increased risk of antidepressant medication. After adjustment for baseline smoking, drinking problems and LLI (model 4), those physically inactive normal-weight (HR 1.3, 95 % CI 1.1-1.6) and overweight (HR 1.4, 95 % CI 1.1-1.6) had a higher risk of antidepressant medication.
For sedative medication the association was weaker than for antidepressants. Overweight was associated with increased risk of sedative medication among physically inactive (HR 1.3, 95 % CI 1.1-1.6) and moderately active (HR 1.2, 95 % CI 1.0-1.5) after adjusting for age and gender. After adjusting for prior medication the association remained only for the inactive overweight group. After full adjustments (model 4) no associations remained.
# Discussion
In this study we examined joint-association of physical activity and overweight with subsequent psychotropic medication over a follow-up of around 8 years.
Main findings were as follows:
(1)Both normal-weight and physical activity are associated with lower risk of psychotropic medication but physical activity dominates the association over normal-weight. (2)Joint association of physical activity and overweight were stronger with antidepressants than sedatives
The results of our study confirm results from previous longitudinal [bib_ref] Joint association of physical activity and body weight with subsequent physical and..., Lindholm [/bib_ref] [bib_ref] A prospective study of fitness, fatness, and depressive symptoms, Becofsky [/bib_ref] and cross-sectional [bib_ref] Recommended levels of physical activity and health-related quality of life among overweight..., Heath [/bib_ref] studies with self-reported measures of mental health. We conducted the analyses also for antidepressants and sedatives, which are the two main groups of psychotropic medication. Associations were quite similar to any psychotropic medication but antidepressants had a stronger association than sedatives. Unfortunately, we do not have information on which mental disorder the medication was prescribed for. There is comorbidity between depression and anxiety [bib_ref] Depression-anxiety relationships with chronic physical conditions: results from the world mental health..., Scott [/bib_ref] and antidepressants are also prescribed for anxiety disorders [bib_ref] Treatment of anxiety disorders in the Finnish general population, Sihvo [/bib_ref]. However, previous studies on physical activity and mental health problems have shown especially an association with depression [bib_ref] Exercise and physical activity in mental disorders, Wolff [/bib_ref]. We also conducted the analyses excluding those with both antidepressant and sedative purchases and it made a minor contribution to the examined associations. Overall, the associations in this study were relatively weak but still relevant from public health perspective. Mental health problems are commonly occurring [bib_ref] The global burden of mental disorders: an update from the WHO World..., Kessler [/bib_ref] and promoting physical activity among middle-aged employees may be useful for preventing mental health problems. Association between overweight or obesity and mental health problems are reciprocal [bib_ref] Overweight, obesity, and depression: a systematic review and meta-analysis of longitudinal studies, Luppino [/bib_ref] and in our study we only examined the obesity to mental health pathway. Although our study showed that overweight is unlikely to notably affect the risk of mental health problems, normal-weight has plenty of other considerable benefits for health such as better physical functioning.
A recent prospective study from United States [bib_ref] A prospective study of fitness, fatness, and depressive symptoms, Becofsky [/bib_ref] showed that cardiorespiratory fitness is more important than fatness in reducing depressive symptoms. Increasing physical activity is the only way to improve cardiorespiratory fitness thus those highly active in our study likely have better fitness than their less active counterparts which may also contribute to the differences found in psychotropic medication purchases. Previous review [bib_ref] Physical activity and likelihood of depression in adults: a review, Teychenne [/bib_ref] , however, concluded that even low doses of physical activity are beneficial in terms of depressive symptoms, supporting the view that there are positive effects of physical activity independent of fitness on mental health. Physical activity increases e.g. several neurotransmitters such as serotonin and dopamine which may protect especially against depression [bib_ref] Exercise and mental health: many reasons to move, Deslandes [/bib_ref]. Mental health problems may also contribute to physical inactivity, and in addition, some psychotropic medication tends to increase body weight [bib_ref] Associations between anxiety, depression, antidepressant medication, obesity and weight gain among canadian..., Grundy [/bib_ref]. We were interested in joint association between physical activity and overweight with psychotropic medication in a follow-up and therefore we excluded those with baseline psychotropic medication and adjusted for previous medication in the analyses. We conducted the analyses excluding those with prior psychotropic medication within the last three years before the baseline survey. The analyses showed that the association were similar to those with previous medication adjusted for (data not shown). Furthermore, those with drinking problems have more anxiety and depression symptoms [bib_ref] The CAGE questionnaire and psychological health, Schofield [/bib_ref] and economic difficulties and socioeconomic position are also associated with mental health problems [bib_ref] Multiple measures of socioeconomic circumstances and common mental disorders, Lahelma [/bib_ref]. We adjusted these as well as other key covariates, however, more detailed control analyses showed that LLI attenuated the association the most. Some LLI may restrict physical activity and also increase the body weight and further contribute to mental health problems.
Additional sensitivity analyses were made. We used different cut-off points for body mass index and physical activity. We examined obesity (30 kg/m 2 ) and also classified BMI groups as well as physical activity according to tertiles. Different cut-off points made a minor contribution to the examined associations suggesting that the associations are not sensitive to the cut-off points used (data not shown). In addition, we used defined daily dose (DDD) of 365 as outcome reflecting more severe problems in mental health. Thirty-six percent of those having any psychotropic medication had DDD over 365. We also divided psychotropic medication purchases to three categories: no purchases, 1 to 2 purchases, and 3 or more purchases. Sixty-three percent of those having any psychotropic medication had three or more purchases. However, these additional analyses with more severe outcomes showed similar associations with the first purchase as the outcome (data not shown).
The strengths of this study include a prospective design and a large sample of middle-aged women and men. Additionally, complete national register data of psychotropic medication purchases linked with the survey data, is strength. We excluded those with the baseline medication and adjusted for prior medication as well as several other covariates.
The limitations of this study include self-reported body mass index and leisure-time physical activity. People tend to overestimate physical activity and height and underestimate weight [bib_ref] How valid are self-reported height and weight? A comparison between CATI self-report..., Taylor [/bib_ref]. Despite of underestimation, self-reported BMI is an adequate predictor of associated health risks. BMI cannot distinguish whether the mass is from fat or from muscle but in middle-aged population dominated by women this is likely not a major concern. In leisure-time physical activity we used cut-off points based on recommendation. Self-estimated leisure-time physical activity was not validated, however, no single physical activity questionnaire has proved better than others for measurements of leisure-time physical activity [bib_ref] Physical activity questionnaires for adults: a systematic review of measurement properties, Van Poppel [/bib_ref]. The psychotropic medication purchases are not a direct measurement for mental health because they have also other indications such as epilepsy, pain, and sleep problems for several psychotropic medicines. Some people with mental health problems have no psychotropic medication and it is also unclear whether people use the medicines they have purchased. However, we were interested in the need of psychotropic medication rather than if the participants had used the medication purchased. The need of psychotropic medication is estimated when doctors decide to prescribe psychotropic medication. We used only baseline information of weight and leisure-time physical activity as well as covariates and all these variables could change during the follow-up. However, this is unlikely to result in overestimating of the associations. Because 78 % of participants were women and genders were combined in the analyses, the results are dominated by the women. However, the association were not significantly different between genders.
# Conclusions
Both normal-weight and physical activity may help prevent psychotropic medication but physical activity dominates the association. Higher amounts of physical activity may show further benefits for mental health. Promoting physical activity among middle-aged employees may prove useful for preventing mental health problems.
[table] Table 1: Baseline participant characteristics [/table]
[table] Table 2: Prevalence of first psychotropic medication purchases (%) during the follow-up [/table]
[table] Table 3: Hazard ratios for psychotropic medication purchases according to joint association of physical activity and BMI [/table]
|
Large-Scale Plantlet Conversion and Ex Vitro Transplantation Efficiency of Siberian Ginseng by Bioreactor Culture
To achieve large-scale low-cost ex vitro acclimatization of Siberian ginseng plants, heart-and torpedo-shaped secondary somatic embryos (SEs) induced from germinated SEs on agar medium were collected and then inoculated to 10-l bubble column bioreactor, respectively. For plantlet conversion, inoculation of torpedo-shaped secondary SEs was more effective than heart-shaped SEs. TS2 (culture of torpedo-shaped SEs in a bioreactor with a 2-week subculture interval) plantlets had a higher root number and leaf number and larger leaf area than did HS3 (culture of heart-shaped SEs in a bioreactor with a 3-week subculture interval) and HS2 (culture of heart-shaped SEs in a bioreactor with a 2-week subculture interval) plantlets. Of these converted plants, TS2 plantlets had higher survival rate (83.7%) and growth characteristics after transplantation in a simple shed covered with a 50% sunshade net only for 6 months. TS2 plantlets also showed significantly lower H 2 O 2 content and significantly increased superoxide dismutase (SOD), glutathione peroxidase (GPX), and glutathione transferase (GST) expression levels as compared to HS2 plants when exposure to ex vitro conditions.
# Introduction
Over the past decade, there has been a significant application of somatic embryogenesis in rapid multiplication of plants, which provided a possible from-laboratory-to-large-scale industrial production. However, the extensive use of somatic embryo (SE) is still restricted by its relative high production costs and low survival rate of in vitro plantlets during ex vitro transplanting [bib_ref] The morphological variability within a population of coffee somatic embryos produced in..., Barry-Etienne [/bib_ref]. During in vitro cultivation, plantlets grow in a unique aseptic microenvironment with special climatic conditions of high relative humidity and low light intensity to allow heterotrophic growth. These conditions may lead to changes in plant with abnormal morphology and physiology, which are often characterized by retardation in the deposition of epicuticular waxes and development of functional stomata associated with the low ability of the plantlets to regulate water loss, which can lead to losses during acclimatization [bib_ref] Acclimatization of tissue cultured plantlets: from laboratory to land, Chandra [/bib_ref]. Most of the management of environ-mental growth conditions has been focused on the ex vitro acclimatization stage; little attention has been placed in the management of environmental growth conditions of in vitro culture. Nevertheless, some studies demonstrated that partial desiccation treatment for early stage SEs could facilitate ex vitro establishment of SE-derived plantlets [bib_ref] The morphological variability within a population of coffee somatic embryos produced in..., Barry-Etienne [/bib_ref] [bib_ref] Improved growth and acclimatization of somatic embryoderived Oplopanax elatus plantlets by ventilated..., Park [/bib_ref].
Siberian ginseng (Eleutherococcus senticosus) is a medicinal woody plant species that is endangered because of overharvesting in its natural habitat. The cortical tissues of the roots, shoots, and leaves are used for various medicinal purposes and in subsidiary food production [bib_ref] New substances of plant origin which increase nonspecific resistance, Brekhman [/bib_ref]. Conventional propagation by seeds and stem cuttings is very difficult because of the long-term stratification needed to induce germination of the zygotic embryos, as well as the low rooting frequency of cuttings [bib_ref] Studies on the cultivation of Eleutherococcus senticosus Maxim. II. On the germination..., Isoda [/bib_ref]. Therefore, production of Siberian ginseng plantlets by somatic embryogenesis as an alternative method has been reported [bib_ref] High frequency of plant production via somatic embryogenesis from callus or cell..., Choi [/bib_ref] [bib_ref] Rapid propagation of Eleutherococcus senticosus via direct somatic embryogenesis from explants of..., Choi [/bib_ref]. The Scientific World Journal
The use of bioreactors is a technology suitable for largescale plant production of Siberian ginseng [bib_ref] Application of bioreactors for large scale micropropagation systems of plants, Paek [/bib_ref]. We previously used 10-l bubble column bioreactors to culture Siberian ginseng secondary SEs, which resulted in a higher productivity than in suspension flask cultures [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref]. Suspensionculture-propagated heart-shaped secondary SEs that were inoculated in bioreactors could germinate uniformly in plant growth regulators-free medium. Mass-produced germinated SEs attained 64.7% conversion frequency. About 90% of in vitro plants survived after acclimatization but only obtained under controlled temperature of 21 ∘ C, requireing very high production costs [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref]. Thus, a key problem of production of high-quality plants to decrease the cost at ex vitro acclimatization stage needs to be resolved before widespread commercial application. Moreover, most work are restricted to increase the induction, germination rate, and conversion into plantlets, with little focus on the underlying physiological and molecular processes during acclimatization [bib_ref] High frequency of plant production via somatic embryogenesis from callus or cell..., Choi [/bib_ref] [bib_ref] Rapid propagation of Eleutherococcus senticosus via direct somatic embryogenesis from explants of..., Choi [/bib_ref] [bib_ref] Cyclic secondary somatic embryogenesis and efficient plant regeneration in mountain ash, Yang [/bib_ref].
The aim of this study was to establish an efficient plant propagation and ex vitro acclimatization method for Siberian ginseng using bioreactor culture of agar-medium-developed secondary SEs and determine the effect of such conditions on the rate of plant conversion and subsequent ex vitro growth in a shed covered with a 50% sunshade net and the underlying physiological and molecular processes during acclimatization.
# Materials and methods
## Induction and development of secondary somatic
Embryogenesis on Agar Medium. Germinated SEs of Siberian ginseng (Eleutherococcus senticosus) developed in bioreactor culture from highly cyclic somatic embryogenic-competent cell lines (see [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref] were used as initial materials for the induction of secondary SEs. Bioreactor culture-germinated SEs were transferred to 1/3 MS [bib_ref] A revised medium for rapid growth and bioassays with tobacco tissue cultures, Murashige [/bib_ref] agar medium containing 1% (w/v) sugar and 2.5% (w/v) gelrite (DUCHEFA, The Netherlands) without plant growth regulators (PGRs) and were cultured at 21 ∘ C, 25 ∘ C, or 29 ∘ C. Ten plantlets were cultured in each 500-mL plastic culture vessel containing 70 mL medium (adjusted to pH 5.8, autoclaved at 121 ∘ C for 15 min) and were cultured with a 16-h photoperiod at 36 mol m −2 s −1 (cool white fluorescent tubes). After 8 weeks, the induction frequency of secondary SEs cultured at the different temperatures was calculated.
## Cultivation of secondary ses in bioreactors.
Clusters of heart-and torpedo-shaped secondary SEs were synchronously developed from germinated SEs cultured on 1/3 MS agar medium for 6-8 weeks as described (see [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref]. These secondary SEs were initially separated by gentle chopping with tweezers and inoculated into 250-mL Erlenmeyer flasks containing 50 mL 1/3 MS liquid medium and 1% (w/v) sugar. About 3000 heart-shaped or 1000 torpedo-shaped embryos were cultured in each flask. Cultures were agitated at 100 rpm on a gyrating shaker at 21 ∘ C.
After adapting to suspension culture for 1 week, SEs were inoculated in 10-l bubble column bioreactors (30 cm × 15 cm) containing 8-L 1/3 MS liquid medium (see [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref]. About 12000 heart-shaped or torpedo-shaped SEs were cultured in each bioreactor. The cultures were agitated at one volume of air per volume of medium per min at 21 ∘ C. The effect of medium subculture period (a 2-or 3-week interval between medium changes) on SE development was evaluated. After 4 weeks (for torpedo-shaped SEs), 6 weeks (for heartshaped SEs with 2-week subculture intervals), or 9 weeks (for heart-shaped SEs with 3-week subculture intervals) of bioreactor culture, fresh weight, hypocotyls diameter and length, and primary root number of germinated SEs were determined. Each set of bioreactor treatment conditions was replicated in three bioreactors, and each experiment was carried out twice. Germinated SEs were placed into 500-mL plastic culture vessels containing 70-mL 1/3 MS agar medium (∼200 SEs/vessel) and stored at 4 ∘ C before being transferred to conversion medium.
## Comparison of plantlet conversion and transplantation ability.
For plantlet conversion, germinated SEs were transferred to conversion medium, which was 1/3 MS agar medium containing 1% (w/v) sugar and 2.5% (w/v) gelrite. Ten germinated SEs were cultured in each vessel. To improve the conversion frequency, germinated SEs stored in the vessel were exposed to light conditions (16-h photoperiod at 36 mol m −2 s −1 ) for 10 days before transfer to conversion medium. After 6 weeks, plantlet conversion frequencies were determined.
Randomly selected in vitro plantlets were transferred to plastic boxes (35 × 55 × 15 cm) containing a mixture of autoclaved sand and potting soil (1 : 3, v/v). About 50 plantlets were cultured in a box. Boxes were covered with perforated polythene bags to maintain high humidity and removed after 4 weeks. Plants were grown in a shed covered with a 50% high-density polyethylene sunshade net. Each of about 1500, 1000, and 500 germinated TS2-, HS2-and HS3derived plantlets were used for transplantation experiment, respectively. The survival rate, area of leaf, length, and fresh weight of whole plants and root systems were compared after transplantation for 6 months.
## Comparisons of h 2 o 2 content of transplanted plants.
Differences between TS2-and HS2-derived in vitro plantlets after 0, 1, 3, and 5 days soil transplantation were investigated at the physiological level. H 2 O 2 content was measured by Zhang et al. [bib_ref] Increased abscisic acid levels in transgenic tobacco over-expressing 9 cisepoxycarotenoid dioxygenase influence..., Zhang [/bib_ref] with some modification. Fresh leaves (0.5 g) were ground to powder in liquid nitrogen and extracted with 5 mL of 5% trichloroacetic acid (TCA) and 0.15 g activated charcoal. The mixture was centrifuged at 10,000 g for 20 min at 4 ∘ C. The supernatant was adjusted to pH 8.4 with 17 M ammonia solution and then filtered. The filtrate was used for determination of H 2 O 2 .
## Quantitative reverse transcription pcr (rt-qpcr) assay of antioxidant gene expression. differences between ts2
and HS2-derived in vitro plantlets after 0, 1, and 3 days soil transplantation were investigated at the molecular level. Total RNA was extracted from leaf tissues using The Scientific World Journal and GST (GenBank: KC542393) were given in [fig_ref] Table 1: Primers used for quantification of antioxidant enzyme genes by real-time PCR [/fig_ref] and amplified under the following cycling conditions: 10 sec at 95 ∘ C followed by 40 cycles of 5 sec at 95 ∘ C, 30 sec at 60 ∘ C, and 1 s at 78 ∘ C for plate reading. RT-qPCR was repeated three times for each sample on a DNA Engine Opticon 2 (MJ Research) following the manufacturer's recommendations.
[formula] 3 (a) (b) (c) (d) (e) (f) (g) [/formula]
# Statistical analysis.
The data were analyzed with ANOVA using SPSS 16.0 for Windows (SPSS Inc., Chicago, IL, USA). Means that differed significantly were compared using Duncan's multiple range test at the 5% probability level.
# Results and discussion
## Induction and development of secondary ses on agar
Medium. After transformation of the germinated primary SEs from the bioreactors to solid medium, the frequency of secondary SE induction was significantly different among the three culture temperatures. At 21 ∘ C and 25 ∘ C, the induction frequency was 30% and 52%, respectively, whereas almost 90.6% of SEs induced secondary SEs at 29 ∘ C [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref] , demonstrating that the higher temperature was better for secondary SE induction. Most secondary SEs gradually developed to the heart-and torpedo-shaped stage about 6 and 8 weeks after culture with no medium exchange, respectively. Further development of torpedo-shaped SEs was arrested, and relative water content decreased to about 54% after 8 weeks. The number of secondary SEs per explants varied significantly. About 20% of germinated SEs produced 100 to 1000 SEs per explants at 29 ∘ C.
## Bioreactor culture of secondary ses.
To allow the SEs to adapt to the bioreactor culture environment, heartand torpedo-shaped SEs were first inoculated into 250-mL Erlenmeyer flasks with liquid medium. After adapting to suspension culture in flasks for one week, secondary SEs were inoculated into bioreactors. When cultured in 1/3 MS medium, the SEs inoculated in bioreactors grew quickly, but the growth rate decreased after ∼2 weeks of culture. The subculture period significantly affected SE growth. HS2 SEs grew normally and consistently with 2-week subculture intervals, but with a 3-week interval, the HS3 SEs grew slowly and then stopped until the medium was changed again. After subculture at 2-week intervals in bioreactors, heart-shaped SEs (referred to as HS2) germinated and needed 6 weeks for growth, whereas 9 weeks were needed with 3-week subculture intervals (referred to as HS3) [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref] (data not shown). Similarly, torpedo-shaped SEs germinated and grew fully after 4 weeks with a 2-week subculture interval (referred to as TS2; [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref]. The morphology of germinated SEs showed a significant difference depending on the different inoculation stages [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref]. The hypocotyl length of germinated TS2 SEs was half of germinated HS2 and HS3 SEs, but fresh weight, hypocotyl diameter, and root number in the germinated TS2 SEs were all significantly higher than corresponding values for germinated HS2 and HS3 SEs [fig_ref] Table 2: Effect of inoculation stage and medium exchange period on growth of SEs... [/fig_ref].
## Plantlet conversion.
Most bioreactor-cultured germinated SEs were pale yellow-green, possibly because of their high culture density, which would result in less absorbed light. Exposing the storage vessels with the SEs to light prior to culture on conversion medium resulted in quickly turning greened SEs produced. When exposing germinated SEs to light for 10 days, the conversion frequency extensively increased, and the conversion frequency of germinated TS2 SEs was significantly higher than that of HS3 and HS2 SEs [fig_ref] Figure 2: Effects of inoculation stage and medium exchange period on plantlet conversion of... [/fig_ref]. TS2 plantlets had a higher root number, leaf number, and larger leaf area than did HS3 and HS2 plantlets [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref] to 1(g)) (data not shown). The leaf shape of TS2 plantlets was also more similar to that of seedling plantlets [fig_ref] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos... [/fig_ref].
## Comparison of transplantation ability of hs2, hs3, and ts2 plants.
In vitro plantlets were transferred to soil and grown in a shed covered with a 50% sunshade net [fig_ref] Figure 3: Transplantation of bioreactor cultured SE-derived plantlets of Siberian ginseng in simple sunshade... [/fig_ref]. After transplantation, the leaves of HS2 and HS3 plants wilted or fell off within 2 weeks, followed by new shoots and leaves forming at the shoot apex within 4 weeks [fig_ref] Figure 3: Transplantation of bioreactor cultured SE-derived plantlets of Siberian ginseng in simple sunshade... [/fig_ref]. On the contrary, most leaves survived and maintained vigorous growth for TS2 plantlets [fig_ref] Figure 3: Transplantation of bioreactor cultured SE-derived plantlets of Siberian ginseng in simple sunshade... [/fig_ref].
Only for three days, TS2 plantlets had formed new roots, demonstrating that they can adapt quickly to the ex vitro environment. The results showed that survival rate of TS2 plantlets was significantly higher than that of HS2 and HS3 plantlets . TS2 plantlets also have higher growth characteristics than HS2 plantlets [fig_ref] Figure 3: Transplantation of bioreactor cultured SE-derived plantlets of Siberian ginseng in simple sunshade... [/fig_ref] and 3(f);. Although the survival rate in this study was similar to it in our previous report [bib_ref] High conversion frequency of germinated somatic embryos of Siberian ginseng (Eleutherococcus senticosus..., Yang [/bib_ref] , the cost fee of this shading shed method was relative lower, which would be of benefit to the commercial application of plantlet production, because the transplantation was completely performed in The Scientific World Journal The Scientific World Journal a simple sunshade shed. The previous reported method need an accurate temperature (21 ∘ C) environment for ex vitro acclimatization; therefore, it requires an instrument high demand of.
To achieve large-scale vegetative propagation by somatic embryogenesis, the biggest obstacle is the mass production of SEs with high plantlet conversion ability [bib_ref] Recent advances in conifer somatic embryogenesis: improving somatic embryo quality, Stasolla [/bib_ref]. Previous reports of somatic embryogenesis and plantlet production in Siberian ginseng showed that SEs germinate easily but have a relatively low conversion frequency and ex vitro acclimatization difficulty [bib_ref] High frequency of plant production via somatic embryogenesis from callus or cell..., Choi [/bib_ref] [bib_ref] Rapid propagation of Eleutherococcus senticosus via direct somatic embryogenesis from explants of..., Choi [/bib_ref] [bib_ref] Cyclic secondary somatic embryogenesis and efficient plant regeneration in mountain ash, Yang [/bib_ref]. In the present study, the main cause for the high conversion and acclimatization ability of SEs was probably due to their long partial desiccation process at earlier stages of development over 6 to 8 weeks. A partial desiccation treatment is a common method for improving the plantlet quality of fully developed SEs, which can be otherwise difficult to germinate [bib_ref] Synchronous and high frequency germination of interior spruce somatic embryos following partial..., Roberts [/bib_ref]. In coniferous species, SE desiccation treatment is the primary means for improving conversion rates [bib_ref] Recent advances in conifer somatic embryogenesis: improving somatic embryo quality, Stasolla [/bib_ref]. Roberts et al. [bib_ref] Synchronous and high frequency germination of interior spruce somatic embryos following partial..., Roberts [/bib_ref] improved the mature postembryonic development of SEs of Picea glauca × P. engelmannii by gradually lowering the high relative humidity to achieve partial desiccation, which decreased the germination time, synchronized stem and root development, and resulted in >90% conversion frequency. In addition, partial desiccation treatment also successfully applied on large-scale SE production of coffee, rubber, and citrus trees by RITA bioreactor with a temporary immersion system [bib_ref] Improvement of Citrus somatic embryo development by temporary immersion, Cabasson [/bib_ref] [bib_ref] Temporary immersion systems in plant micropropagation, Etienne [/bib_ref] [bib_ref] Direct sowing of Coffea arabica somatic embryos massproduced in a bioreactor and..., Etienne-Barry [/bib_ref].
## Decreased h 2 o 2 content and enhanced expression of antioxidant genes in ts2
Plants. Abiotic stress may increase the level of reactive oxygen species (ROS) which is toxic to plant cells [bib_ref] Direct sowing of Coffea arabica somatic embryos massproduced in a bioreactor and..., Etienne-Barry [/bib_ref]. During desiccation, free radical scavengers may provide additional protection because the development of desiccation tolerance coincides with an increase in antioxidant gene expression in seeds [bib_ref] The expression of a peroxiredoxin antioxidant gene, AtPer1, in Arabiodopsis thaliana is..., Haslekås [/bib_ref] [bib_ref] Partial desiccation augments plant regeneration from irradiated embryogenic cultures of sugarcane, Suprasanna [/bib_ref]. To investigate whether the decreased sensitivity to environmental stress in TS2 plants as compared to HS2 plants may be due to decreased ROS accumulation, the H 2 O 2 content of TS2 and HS2 plants leaves was compared at initiation period of ex vitro transplantation. Under these conditions, H 2 O 2 levels in TS2 plants were less than half of those in HS2 plants . Thus, TS2 plants may have more normal functional leaves and suffer less environmental stress as compared to HS2 plants. Our result strongly suggests that there is a correlation between the leaf damage and the imbalance of ROS production in early stage of transplantation.
It has been accepted that antioxidant defense systemsincluding nonenzymatic antioxidants such as ascorbate and reduced glutathione and enzymatic antioxidants such as SOD and GPX-play a crucial role in plant defense mechanisms against various stresses [bib_ref] Screening of genes induced by salt stress from Alfalfa, Jin [/bib_ref]. The expression levels of three stress inducible antioxidant enzyme genes SOD (GenBank: JQ365625.1), GPX (GenBank: KC542392), and GST (GenBank: KC542393) under ex vitro environment were compared between TS2 and HS2 plants using RT-qPCR. The results showed SOD, GPX, and GST expression levels in TS2 plants were increased significantly as compared to that in HS2 plants when exposed to ex vitro conditions [fig_ref] Figure 6: Comparisons of [/fig_ref]. We hypothesized that ex vitro transplanted plant cells can induce expression of various antioxidant enzymes under environmental stresses, and the increased expression of these The Scientific World Journal antioxidant enzyme genes in TS2 plants may contribute to the scavenging of ROS. Antioxidant enzyme genes' expression levels of HS2 plants also increased when exposed to environmental conditions in vitro, may be because the far more produced ROS could not protected leaves from oxidative damage.
# Conclusion
An efficient micropropagation protocol for Siberian ginseng has been established. Development of secondary SEs to torpedo-shaped stage on agar medium prior inoculated to bioreactor culture was found to be beneficial for plant conversion and thus low the cost acclimatization of Siberian ginseng. TS2 plants were found to have higher environmental stress adaption ability compared to the HS2 plants. The protocol in the present study can be successfully exploited to widespread commercial propagation of Siberian ginseng elite clones.
[fig] Figure 1: Bioreactor culture of agar medium induced heart-shaped and torpedo-shaped secondary somatic embryos of Siberian ginseng and plantlet conversion. (a) Numerous secondary SEs induced from germinated primary SEs on 1/3 MS medium without PGRs. Bar = 1 cm. (b) Germinated HS2 (left), HS3 (middle), and TS2 (right) SEs production in a 10 l bubble column bioreactor. Bar = 10 cm. (c) Germinated TS2 (left), HS2 (middle), and HS3 (right) SEs production in a 10 l bubble column bioreactor. Bar = 2 cm. (d) Plantlets derived from germinated HS3 (left), HS2 (middle), and TS2 (right) SEs. Bar = 4 cm. Plantlet conversion of germinated HS3 (e), HS2 (f), and TS2 (g) SEs. Bar = 2 cm.the CTAB method[13]. First-strand cDNA synthesized with 0.5 g purified RNA was reverse-transcribed with a reverse transcriptase kit (MBI Fermentas), and PCR was performed in a volume of 20 L, containing 10 L of 2 × SYBR premix ExTaq (TaKaRa Biotech, Dalian, China), 0.5 M of forward and reverse primers, and 2 L cDNA template (equivalent to 0.05 g of total RNA). The primer sequences of Siberian ginseng 18S rRNA (GenBank: AB080245.1),tubulin (GenBank: KC542394), and the previously defined stress inducible ROS-scavenging related genes (unpublished) SOD (GenBank: JQ365625.1), GPX (GenBank: KC542392), [/fig]
[fig] Figure 2: Effects of inoculation stage and medium exchange period on plantlet conversion of bioreactor cultured germinated SEs of Siberian ginseng after culturing on 1/3 MS medium for 2 months. Values with different letters in a column are significantly different according to Duncan's multiple range test at the 5% level. [/fig]
[fig] Figure 3: Transplantation of bioreactor cultured SE-derived plantlets of Siberian ginseng in simple sunshade shed. (a) Simple shed covered with a 50% sunshade net. Bar = 1.1 m. (b) In vitro plantlets were transferred to plastic boxes (35 × 55 × 15) containing a mixture of sand and soil. Bar = 20 cm. (c) HS2 plantlets 2 weeks after transplantation. Bar = 1 cm. (d) TS2 plantlets after transplantation for 2 weeks. Bar = 1 cm. (e) HS2 (left) and TS2 (right) plantlets after transplantation for 6 months. Bar = 16 cm. (f) TS2 (left) and HS2 (right) plantlets after transplantation for 6 months. Bar = 6 cm. [/fig]
[fig] Figure 4, Figure 5: Survival frequency of HS2, HS3, and TS2 plantlet after ex vitro transplantation for 6 months. Values with different letters in a column are significantly different according to Duncan's multiple range test at the 5% level. Comparison of H 2 O 2 content between HS2 and TS2 plantlets after transplantation in a simple shed covered with a 50% sunshade net. [/fig]
[fig] Figure 6: Comparisons of (a) SOD, (b) GPX, and (c) GST expression levels in HS2 and TS2 plantlet after transplantation in a simple shed covered with a 50% sunshade net by RT-qPCR. [/fig]
[table] Table 1: Primers used for quantification of antioxidant enzyme genes by real-time PCR. [/table]
[table] Table 2: Effect of inoculation stage and medium exchange period on growth of SEs in 10-l bioreactor. [/table]
|
LED Fluorescence Microscopy for the Diagnosis of Pulmonary Tuberculosis: A Multi-Country Cross-Sectional Evaluation
Background: The diagnosis of tuberculosis (TB) in resource-limited settings relies on Ziehl-Neelsen (ZN) smear microscopy. LED fluorescence microscopy (LED-FM) has many potential advantages over ZN smear microscopy, but requires evaluation in the field. The aim of this study was to assess the sensitivity/specificity of LED-FM for the diagnosis of pulmonary TB and whether its performance varies with the timing of specimen collection.Methods and Findings: Adults with cough $2 wk were enrolled consecutively in Ethiopia, Nepal, Nigeria, and Yemen. Sputum specimens were examined by ZN smear microscopy and LED-FM and compared with culture as the reference standard. Specimens were collected using a spot-morning-spot (SMS) or spot-spot-morning (SSM) scheme to explore whether the collection of the first two smears at the health care facility (i.e., ''on the spot'') the first day of consultation followed by a morning sample the next day (SSM) would identify similar numbers of smear-positive patients as smears collected via the SMS scheme (i.e., one on-the-spot-smear the first day, followed by a morning specimen collected at home and a second on-the-spot sample the second day). In total, 529 (21.6%) culture-positive and 1,826 (74.6%) culture-negative patients were enrolled, of which 1,156 (49%) submitted SSM specimens and 1,199 (51%) submitted SMS specimens. Single LED-FM smears had higher sensitivity but lower specificity than single ZN smears. Using two LED-FM or two ZN smears per patient was 72.8% (385/529, 95% CI 68.8%-76.5%) and 65.8% (348/529, 95% CI 61.6%-69.8%) sensitive (p,0.001) and 90.9% (1,660/1,826, 95% CI 89.5%-92.2%) and 98% (1,790/1,826, 95% CI 97.3%-98.6%) specific (p,0.001). Using three LED-FM or three ZN smears per patient was 77% (408/529, 95% CI 73.3%-80.6%) and 70.5% (373/529, 95% CI 66.4%-74.4%, p,0.001) sensitive and 88.1% (95% CI 86.5%-89.6%) and 96.5% (95% CI 96.8%-98.2%, p,0.001) specific. The sensitivity/specificity of ZN smear microscopy and LED-FM did not vary between SMS and SSM.Conclusions: LED-FM had higher sensitivity but, in this study, lower specificity than ZN smear microscopy for diagnosis of pulmonary TB. Performance was independent of the scheme used for collecting specimens. The introduction of LED-FM needs to be accompanied by appropriate training, quality management, and monitoring of performance in the field.
# Introduction
Tuberculosis (TB) is a major public health problem: there are 9.3 million new cases and 1.7 million deaths per year, with 90% of cases occurring in low-and middle-income countries (LMICs). Most patients suspected of having pulmonary TB (PTB) in LMICs are investigated by examining three sputum specimens collected over a minimum of 2 d. Sputum smears are stained with Ziehl-Neelsen (ZN), and each smear examination requires on average 5-10 min, creating considerable workloads for laboratories with limited resources. An alternative technique to ZN smear microscopy, fluorescence microscopy (FM), is reported to be 10% more sensitive than ZN smear microscopy [bib_ref] Use of light-emitting diode fluorescence microscopy to detect acid-fast bacilli in sputum, Marais [/bib_ref] and, since fluorescent acid fast bacilli (AFB) can be seen at lower magnification than ZN-stained AFB, FM smears can be examined in a fraction (about 25%) of the time needed for ZN smears [bib_ref] Use of light-emitting diode fluorescence microscopy to detect acid-fast bacilli in sputum, Marais [/bib_ref]. Although FM can reduce laboratory workloads [bib_ref] New policies, new technologies: modelling the potential for improved smear microscopy services..., Ramsay [/bib_ref] , it has been difficult to implement it widely in LMICs due to the high cost and complexity of the microscope and mercury vapour lamp lighting system, the need for a dark room, and perceived health risks associated with ultraviolet light exposure [bib_ref] The future looks bright: low-cost fluorescent microscopes for detection of Mycobacterium tuberculosis..., Hanscheid [/bib_ref].
Recent technical developments have the potential to improve some of the shortcomings of the smear diagnosis of TB. These include the development of illumination systems based on LEDs (light-emitting diodes), which resulted in LED fluorescence microscopy (LED-FM) becoming commercially available [bib_ref] Light-emitting diode technologies for TB diagnosis: what is on the market?, Minion [/bib_ref]. Furthermore, the World Health Organization (WHO) recently modified its guidelines for the diagnosis of PTBand reduced the number of AFB required to declare a smear as positive (from 10 to 1 AFB), the minimum number of specimens needed for diagnosis (from three to two), and the number of positive smears required to classify a patient as having smear-positive TB (from two to one smear). In a separate publication (also reported in PLoS Medicine [bib_ref] A multicountry non-inferiority cluster randomized trial of frontloaded smear microscopy for the..., Cuevas [/bib_ref] , we report that the ZN smear microscopy examination of two sputum specimens collected on the spot the first day of consultation, followed by the examination of a third specimen collected the following morning (i.e., spot-spot-morning [SSM]) identifies the same number of patients with smear-positive PTB as the frequently used spot-morning-spot (SMS) scheme (i.e., examination of one sputum specimen collected on the spot the first day of consultation, followed by the examination of a second specimen collected the following morning and a third specimen collected on the spot when the patient brings the morning specimen to the health centre) [bib_ref] Frontloading sputum microscopy services: an opportunity to optimise smear-based case detection of..., Ramsay [/bib_ref]. As most smear-positive PTB patients are identified by the first two specimens [bib_ref] Yield of serial sputum specimen examinations in the diagnosis of pulmonary tuberculosis:..., Mase [/bib_ref] , a SSM scheme could detect most smear-positive patients on the first day of consultation. It is thus conceivable that a service that examines SSM specimens using LED-FM and classifies patients using the revised WHO guidelines could ease workloads and speed up the diagnostic process in LMICs, with significant benefits for patients and health systems [bib_ref] New policies, new technologies: modelling the potential for improved smear microscopy services..., Ramsay [/bib_ref].
We therefore conducted a study, nested within a clinical trial, assessing the sensitivity and specificity of LED-FM using both SSM and SMS sputum collection schemes for the diagnosis of PTB.
# Methods
The protocol for the trial was registered (ISRCTN53339491), and ethical approval was obtained from the Liverpool School of Tropical Medicine and WHO ethics review committees, the institutional review boards, and respective national research ethics committees of the participating centres (see Text S1). Translated consent forms and patient information sheets were used for each site. Informed written or oral consent in front of a witness were obtained from all patients. Patients unable to provide informed consent and those who had received anti-TB treatment in the last month were excluded.
This was a prospective cross-sectional study carried out in Ethiopia, Nepal, Nigeria, and Yemen over a [fig_ref] Table 1: Characteristics of the patients on enrolment [/fig_ref]. The study was nested within a non-inferiority clinical trial assessing whether the sensitivity and specificity of ZN smear microscopy of sputum specimens collected under a SSM scheme were non-inferior to those of ZN smear microscopy following the standard SMS collection scheme for the diagnosis of PTB, which is also published in this issue [bib_ref] A multicountry non-inferiority cluster randomized trial of frontloaded smear microscopy for the..., Cuevas [/bib_ref]. The study sites were selected to obtain a wide geographical representation of LMIC settings and took advantage of a larger clinical trial on smear microscopy being undertaken at the same time. This facilitated training, as well as the use of protocols and standard operating procedures across locations.
All consecutive patients $18 y old with cough $2 wk duration presenting at participating health facilities were screened and invited to participate in the clinical trial. Of these, only the first 7-10 patients attending the health care centres each day were invited to participate in the LED-FM evaluation to avoid overloading the laboratories (see [fig_ref] Figure 1: Depiction of flow of participants [/fig_ref]. It is recognised that selecting the first patients may have introduced systematic biases (e.g., residents of nearby households with mild illnesses or those from remote locations with advanced disease stages). Random selection, however, was logistically difficult. The data obtained for smear microscopy (ZN and LED-FM) were paired and therefore relatively independent of the actual representativeness of the participants in relation to the reference population.
Patients enrolled in Ethiopia were attending Bushullo Major and Awassa Health Centres. These are the main health service providers for Awassa District in the Southern Region. Smear microscopy was conducted in the health centres' laboratories, and sputum specimens were cultured at the Armauer Hansen Institute, Addis Ababa. Patients in Abuja, Nigeria, were enrolled in Wuse District Hospital, and sputum specimens were processed in the Zankli Medical Centre laboratory, a private laboratory acting as a diagnostic centre for the National Tuberculosis Programme. In Nepal, patients were enrolled from the TB DOTS centre of Tribhuvan University Teaching Hospital and the Dirgh Jeevan Health Care and Research Centre, both in Kathmandu. Sputum specimens collected in both centres were processed in Tribhuvan University Health Research Laboratory. In Yemen, patients were enrolled at the Tuberculosis Institute, which is the reference centre and headquarters of the National Tuberculosis Programme and which provides diagnostic services to the surrounding population and referred patients.
All patients were requested to submit three sputum specimens following one of the two schemes. One scheme required patients to provide a sputum sample on first attendance at the health facility, a second sample (which was produced at home) the following morning, and a third sample collected on the spot when the patient brought the morning specimen to the laboratory (SMS). The second scheme required patients to provide a specimen on the spot upon first attendance at the health care facility, a second specimen one hour after the first specimen, and a third specimen that was produced at home the following morning (SSM). The schemes were allocated by week using a randomization list that assigned the scheme to be used in a given week. The scheme was disclosed at the start of the week by opening a sealed envelope. Patients were provided with pre-labelled sputum collection pots and instructed on how to produce a good quality sputum specimen following a standard operating procedure [bib_ref] Improvement of tuberculosis case detection and reduction of discrepancies between men and..., Khan [/bib_ref].
Specimens were assessed macroscopically, and smears were prepared in duplicate for staining by the hot ZN technique or with Auramine O, counterstained with potassium permanganate, for FM. Slides were labelled with study numbers. These labels were then covered with wrap-around stickers before mixing and reading to ensure laboratory technicians were unaware of the grades assigned to previous smears. Grading was conducted by laboratory technicians, and the stickers were removed when another technician entered the results into the study logbooks. LED-FM was conducted on standard microscopes fitted with adaptors consisting of a specialised objective with an attached LED light source (LUMIN and the QBC Paralens Fluorescence Microscopy System). ZN smears were examined at 1,0006 magnification, and LED-FM smears (in accordance with advice from the International Union Against Tuberculosis and Lung Disease Working Group on Smear Microscopy) at 2006, with confirmation of positive smears at 4006 magnification. Smears were classified as positive when $1 AFB was detected per 100 fields, and patients were considered smear-positive if they had $1 positive smear. Patients with missing results were classified according to the results of available smears (e.g., patients with two negative smears but missing the third specimen were classified as negative). The morning specimen, or, if not available, a spot specimen, was concentrated (Petroff's method) and cultured on solid medium. Identification of Mycobacterium tuberculosis complex was confirmed by standard biochemical tests.
The study was implemented at each site after training and a pilot phase, and was conducted in compliance with good clinical practice/good clinical laboratory practice. A lot quality assurance sampling scheme was used to determine the sample size for ZN smear microscopy external quality assessment (EQA) to assure a sensitivity of 90% relative to the controller, with a maximum of two errors. Sampling for ZN smear microscopy EQA was performed before, during, and at the end of the study and was conducted by WHO/International Union Against Tuberculosis and Lung Disease Supranational Reference Laboratories. A random selection of 20% LED-FM smears from all sites was also selected for EQA. In addition, all LED-FM smears graded as scanty (,10 AFB per 100 fields) were sent to Supranational Reference Laboratories and the German-Nepal Tuberculosis Project in Nepal at the end of the study, where smears were restained and re-checked using conventional FM. The main outcomes for comparison were the sensitivity and specificity of LED-FM compared with the sensitivity and specificity of ZN smear microscopy. Culture was used as the reference standard, and the analysis was stratified by scheme. As LED-FM smears with scanty AFB (,10 AFB) were often reported negative by conventional FM, both results are presented, stratified by culture to assign true-/false-positive results. It is however acknowledged that this partial verification could result in an inaccurate estimate of test performance and that this latter sub-analysis needs to be interpreted with caution.
Data are presented using summary statistics with 95% CI. Comparisons of the characteristics of the patients studied in the four participating countries were by one-way analyses of variance for continuous measures and Fisher exact tests for categorical measures. Comparisons of sensitivity and specificity values where both methods were applied to the same samples (i.e., the observations were paired) were done using the McNemar test; comparisons where samples were taken from two different (independent) cohorts were done using the Fisher exact test. The sample size calculations for the study assumed that ZN smear microscopy would identify 50% of culture-positive patients and that LED-FM would be 10% more sensitive than ZN smear microscopy [bib_ref] Use of light-emitting diode fluorescence microscopy to detect acid-fast bacilli in sputum, Marais [/bib_ref]. The target sample size of 1,605 was assumed to be an overestimate of the number required because the calculations presumed that specimens were not paired. Although studies evaluating paired samples have lower minimum sample sizes required than studies using independent samples, the overestimation was unavoidable without prior knowledge of the expected test discordance.
# Results
A total of 2,445 patients (37% of 6,628 patients enrolled for the larger trial [bib_ref] A multicountry non-inferiority cluster randomized trial of frontloaded smear microscopy for the..., Cuevas [/bib_ref] were enrolled for the LED-FM evaluation. Of these, 468 (19%) were enrolled in Ethiopia, 526 (22%) in Nepal, 685 (28%) in Nigeria, and 766 (31%) in Yemen. The characteristics of the participants are summarised in [fig_ref] Table 1: Characteristics of the patients on enrolment [/fig_ref] y, respectively, p,0.001). Patients in Ethiopia and Yemen were more likely to come from rural areas (58% and 51%) than patients in Nepal and Nigeria (7% and 13%, p,0.001), and patients in Nepal presented with longer cough duration than patients in the other three countries. The most frequent complaints were cough, chest pain, weight loss, anorexia, fever, and night sweating. Of the 2,445 patients enrolled, 529 (21.6%) were culture-positive, 1,826 (74.6%) were culture-negative, and 90 (3.7%) had invalid culture results (57 [2.3%] contaminated and 33 [1.4%] missing). Among patients with positive or negative cultures, 1,156 (48.9%) were screened using the SSM scheme and 1,199 (51.1%) the SMS scheme.
Individual ZN and LED-FM smear results by type of specimen are described in Text S3. Smears prepared from the morning specimens were more likely to be positive than smears prepared from spot specimens. Among culture-positive patients, ZN smears were less likely to be positive than LED-FM smears (937/ Single LED-FM smears therefore had higher sensitivity but lower specificity than single ZN smears. Out of 283 LED-FM smears with scanty grades that had paired conventional FM readings, 136 (48.1%) were culture-positive and 147 (51.9%) culture-negative according to conventional FM. Ninety-six (70.6%) of the culture-positive specimens were graded positive and 77 (52.3%) of the 147 culture-negative specimens were graded negative by conventional FM. LED-FM smears therefore were more sensitive but less specific than conventional FM smears for classifying smears with low AFB numbers. The sensitivity/ specificity obtained after incorporating the revised scanty smears in the analysis is shown in [fig_ref] Table 2: Sensitivity and specificity of using two or three ZN and LED-FM smears... [/fig_ref] for illustration.
Using two or three LED-FM smears per patient resulted in sensitivities (95% CI) of 72.8% (68.8%-76.5%) and 77.1% (73.3%-80.6%), compared to 65.8% (61.6%-69.8%) and 70.5% (66.4%-74.4%) when using two or three ZN smears (p,0.001 for both). Using two or three LED-FM smears per patient produced specificities (95% CI) of 90.9% (89.5%-92.2%) and 88.1% (86.5%-89.6%), respectively, compared to 98% (97.3%-98.6%) and 96.5% (96.8%-98.2%) when using two or three ZN smears (p,0.001 for both). The accuracy of using three LED-FM smear examinations per patient was 85% (2,017 of 2,355 patients correctly classified, 95% CI 84.2%-87%), which was lower than the 91.8% accuracy obtained when using three ZN smear examinations (2,136 of 2,355 patients, 95% CI 89.5%-91.8%, p,0.001). An illustration of the proportion of true-and falsepositive and -negative results obtained with ZN smear microscopy and LED-FM using two and three smears is included in [fig_ref] Table 3: Absolute difference the use of two-or three-smear ZN smear microscopy and LED-FM... [/fig_ref] for illustration. The use of two ZN smears would correctly classify the highest number of patients (91%), but would also miss the highest number of cases with TB, and, conversely, using three LED-FM smears would have the lowest number of cases classified correctly (85%), but would miss the lowest number of cases with TB, thus the increased sensitivity is at the expense of lower specificity. The sensitivity and specificity of the SSM and SMS schemes for both LED-FM and ZN smear microscopy are shown in Text S3. There were no statistical differences between the schemes.
# Discussion
The number of cases of PTB continues to increase worldwide despite major international control initiatives, and at least 50 million individuals with chronic cough are screened for TB each year. Although it is widely accepted that the early identification of cases through inexpensive point-of-care diagnostics is pivotal for effective control of TB, the diagnosis of PTB often relies on sputum smear microscopy, especially in resource-limited settings.
Sputum smear microscopy services are the most decentralized of the TB diagnostic services, allowing patients to be screened for tuberculosis at relatively basic health care facilities. In most LMICs, however, these centres are less numerous than treatment centres, and many patients still need to travel to submit sputum specimens for diagnosis, incurring considerable personal costs, and this often leads patients to abandon the process. The diagnosis of TB therefore is particularly difficult for poor and marginalised patients, who face multiple challenges to accessing a TB diagnosis [bib_ref] The Malawi National Tuberculosis Programme: an equity analysis, Simwaka [/bib_ref] , and the improvement of smear microscopy services remains necessary to increase patients' access to treatment [bib_ref] Reducing the global burden of tuberculosis: the contribution of improved diagnostics, Keeler [/bib_ref]. Furthermore, smear microscopy is known to be less sensitive in HIVassociated tuberculosis, and interventions which could increase this sensitivity are needed.
LED-FM holds multiple advantages over both the conventional mercury vapour lamp FM and standard light microscopy. LEDs are relatively inexpensive, can be powered by mains or batteries, and have an effective lifespan of thousands of hours [bib_ref] Light-emitting diode technologies for TB diagnosis: what is on the market?, Minion [/bib_ref]. The longer lifespan of the LED systems, as well as facilitating their implementation, may result in better quality FM than conventional systems since mercury vapour and halogen lamps are often used after their effective lifespan, with diminished power, to cause fluorescence. The lower capital and maintenance costs of LED-based systems, the faster grading of smears than with ZN smear microscopy, and FM's reputed higher sensitivity makes LED-FM devices potentially more suitable for LMIC laboratories. Despite this potential, there have been very few well-designed field evaluations using culture as the reference standard [bib_ref] Use of light-emitting diode fluorescence microscopy to detect acid-fast bacilli in sputum, Marais [/bib_ref] [bib_ref] Clinical evaluation of acid-fast smear examination with light emitting diode fluorescent microscopy, Mizuno [/bib_ref] to assess whether LED-FM performance varies in locations with no or limited experience with FM.
LED-FM in this study had higher sensitivity but lower specificity than ZN smear microscopy, which resulted in a significantly [bib_ref] Sensitivity and specificity of fluorescence microscopy for diagnosing pulmonary tuberculosis in a..., Cattamanchi [/bib_ref] , suggesting that the loss of specificity was more evident in patients who had scanty AFB in sputum. The investigators, however, had processed different specimens for ZN smear microscopy and FM; thus, the interpretation of this study is difficult [bib_ref] Misleading conclusion, Klatser [/bib_ref] , and further studies are needed. If these findings are corroborated by further studies, however, it is still important to consider that if a trade-off has to be made between sensitivity and specificity, it is preferable to err on the side of increased sensitivity and to treat a small number of patients who do not have TB. There are real consequences to a false-positive result: patients incorrectly diagnosed with TB have the cost and inconvenience of taking a six-month course of treatment, and troublesome side effects are common, but life-threatening treatment-related events (hepatitis and Stevens-Johnson syndrome) are infrequent (,5% incidence). It is also necessary to discuss whether solid culture, or liquid culture, can function as an adequate comparator, as neither is a perfect reference standard, and the underlying assumption that either culture system has higher specificity than smear microscopy might be debatable. The WHO manual for laboratory services in TB control states that the probability of obtaining a positive culture is related to the number of AFB in the specimen, with only about 50% of cultures of specimens with 1-2 AFB per 100 fields being identified as positive, increasing to 80% and 96.7% for specimens with ''scanty'' (1-9 AFB per 100 fields) and ''+'' AFB grades, respectively. The analysis of our smear-positive, culturenegative results indicated that more than 98% of these were scanty smears with a median of one AFB per smear, and 23.7% had a second positive smear by either ZN smear microscopy or LED-FM. It is thus highly likely that these patients either had paucibacillary tuberculosis in which M. tuberculosis failed to grow in culture, or were infected/colonised by non-tuberculous mycobacteria. This is unfortunately an inherent weakness of solid culture (and, to a lesser extent, liquid culture) as a diagnostic reference standard. Judgements on the performance of new diagnostic tests must be done with considerations of these limitations. The ''false positive'' LED-FM results are thus most likely due to technician error, but we cannot discount the possibility that these were cases with TB or non-tuberculous mycobacteria with a false-negative solid medium result that would have been detected using liquid medium. There might also be cases of TB that would have been missed by both culture technologies (because of delayed processing/destruction of marginally viable bacilli during decontamination or bacterial overgrowth). Unfortunately, HIV testing was not done systematically as part of this study. Overall, the vast majority of patients across the sites did not know their HIV status. The exception was the Nigerian site, where around half of patients reported knowing their HIV status. Only patients who volunteered this information were categorised as HIV positive or negative. Any sub-analysis of the results by HIV status would be underpowered, subject to self- selection bias, and confounded by the duration since last test and the methods used for testing. Clinical laboratories in our study had considerable experience with ZN smear microscopy but limited experience with mercury vapour lamp FM, and their conventional FM microscopes were neither functioning nor in recent use at the time of the study. Although technicians received 2-d refresher training at the bench, were supervised for a further day to monitor performance, and had photographic bench-aids available, ZN smear microscopy had consistently higher specificity.
Interestingly, the rechecking of scanty smears and comparison against culture suggests that LED-FM also had higher sensitivity but lower specificity than conventional FM. These differences could have been due to the devices' performance characteristics, as reference laboratories are equipped with high-quality microscopes with greatly improved optics, and to the novelty of LED-FM, with laboratory technicians conducting more careful examinations and spending more time reading the smears to identify paucibacillary smears, while lacking experience to distinguish fluorescing artefacts from scanty AFB. Furthermore, technicians in the reference laboratories routinely use dark rooms, which facilitate visualizing the AFB in conventional FM but are not required with LED-FM, and may have been more cautious to declare smears with one or two bacilli as positive. Also, the smears may have faded, and AFB could have been washed off during re-staining.
Although the comparison of LED-FM and conventional FM was based on the verification of a partial selection of smears and needs to be replicated, national TB control programmes considering introducing LED-FM devices may wish to bear in mind that their performance could vary with the service setting, staff experience, and the type of devices used. In contrast to the findings of this study, other studies conducted at established research centres have found similar specificity in conventional FM and LED-FM. Our lower specificity may thus be the upside of the trade-off implicit in the larger amount of operational data collected in this study, at the cost of a less-than-ideal reference standard. The LED-FM technology, however, is likely to be widely implemented, and the study still gives valuable insights into operational issues. The performance of LED-FM, and indeed any diagnostic test, depends critically on proficiency of test operators. Adequate and appropriate training is thus required. Current expert opinion is that only moderate training (3 d) is required to make laboratory technicians already proficient in ZN smear microscopy proficient in LED-FM. This training was provided to study staff, involved instruction in specimen processing and staining techniques, and included 2 d of supervised routine microscopy during which an expert helped resolve any doubtful results in order to instil confidence in workers. All staff reported familiarity and confidence in their ability to perform the technique. We believe that this kind of training is most likely sufficient for the induction of staff in laboratories where the technique is well established and there are experienced staff who can provide continued mentoring. However, as stand-alone training, after which laboratory staff are expected to maintain performance without access to a second opinion, it is most likely inadequate, and the development of structured and standardised training packages and supervisory schemes is needed to support scale up. The introduction of LED-FM therefore needs to be accompanied by careful training, appropriate quality control, and monitoring of the performance of LED-FM in the field. Implementers may need to consider whether trade offs between higher sensitivity and lower specificity are acceptable.
The study also indicates that the sensitivity and specificity of LED-FM is independent of whether the sputum specimens are collected as SSM or SMS. There is hardly any evidence that a particular sequence of sputum collection increases the probability of finding TB bacilli, and we have recently demonstrated that this is not so [bib_ref] A multicountry non-inferiority cluster randomized trial of frontloaded smear microscopy for the..., Cuevas [/bib_ref]. Although morning specimens have a higher yield (and single-smear sensitivity), examining two spot specimens identifies equivalent numbers of smear-positive patients as examining one spot specimen followed by one morning specimen. SMS requires that all patients return to provide the morning sample, while the SSM would identify most smear-positive cases in one day. Thus, what is crucial is the potential shortening of the process. Although many clinicians are reluctant to rely on spot specimens due to the higher sensitivity of the morning specimens, the 10% higher sensitivity of a single morning specimen becomes redundant once multiple specimens are used. The WHO has recently recommended using spot specimens (indeed spotspot) to allow the screening of patients in one day. However, the change in policy was based on the evaluation of the method using ZN smear microscopy. We now provide evidence that a frontloaded LED-FM approach that uses two spot specimens results in the same yield as approaches using a conventional SMS LED-FM scheme. This seemingly small change may hold the key to improving current microscopy services. If two smears are collected and examined on the first day and one of them is positive, as was the case in the majority of smear-positive cases in this study, patients would not have to return the next day to complete the sputum submission process. If health systems are modified to provide results on the same day, patients could save the costs of one more visit in the diagnostic process, the majority of smear-positive patients could be referred for treatment on the day of consultation, and dropout rates may be reduced.
Moreover, it is important to consider the role that LED-FM can play in the recent WHO endorsement of the new rapid, automated nucleic acid amplification test, Xpert. It is unlikely, in the short term, that Xpert can be scaled up and decentralized sufficiently to replace smear microscopy as the initial diagnostic test worldwide, even in areas with high rates of multi-drug-resistant TB or HIVassociated TB, where Xpert is recommended as the initial screening test. Furthermore, the WHO recommends continuing to use smear microscopy for treatment monitoring, even in areas served by new technologies such as TB culture, line probe assays, and the Xpert assay. In areas without high rates of multi-drugresistant TB or HIV-associated TB, cost-effectiveness considerations favour the use of smear microscopy as the initial diagnostic tool, and thus the improved performance of direct smear microscopy (and indeed a technique that facilitates faster examination of smears) is an important step for improving diagnostic services in LMICs, one that would reduce laboratory workloads and facilitate the faster identification of patients with negative smear microscopy who could undergo further tests.
Despite the reduced specificity of LED-FM, this approach still has operational advantages that make it an attractive tool for TB laboratory diagnosis. National control programmes introducing new LED-FM services should monitor the performance of the method under operational conditions, as training needs of staff may be greater than anticipated. This study has shown that LED-FM can play a key role in reaching WHO targets for TB detection, reducing laboratory workloads, and ensuring poor patients' access to TB diagnosis and prompt treatment.
## Supporting information
Text S1 Study protocol.
## (doc)
Text S2 STARD checklist.
## (doc)
Text S3 Supporting tables. [fig_ref] Table 1: Characteristics of the patients on enrolment [/fig_ref] shows single ZN and LED-FM smear results by culture; [fig_ref] Table 2: Sensitivity and specificity of using two or three ZN and LED-FM smears... [/fig_ref] shows sensitivity and specificity of using two and three ZN and LED-FM smears per patient, by specimen collection scheme. (DOC)
## Editors' summary
Background. Tuberculosis is a global public health problem. Every year, about 1.7 million people die from this contagious bacterial infection, and about 9 million new cases occur, mainly in low-and middle-income countries. Mycobacterium tuberculosis, which causes tuberculosis, is spread in airborne droplets when people with the disease cough or sneeze, and usually infects the lungs (pulmonary tuberculosis). Symptoms of tuberculosis include a persistent cough, weight loss, and night sweats. Because tuberculosis is easily transmitted and potentially deadly, it is important that it is diagnosed quickly and accurately and immediately treated. The ''gold standard'' diagnostic test for tuberculosis is mycobacterial culture (in liquid or solid medium), in which laboratory technicians try to grow M. tuberculosis from sputum (mucus brought up from the lungs by coughing). However, this test is expensive, so most patients suspected of having pulmonary tuberculosis in resource-limited countries are investigated using sputum smear microscopy. In this cheaper but less sensitive test, sputum samples are ''smeared'' onto microscope slides, stained with Ziehl-Neelsen (ZN) dye, and then examined with a microscope for the presence of M. tuberculosis.
Why Was This Study Done? With smear microscopy, multiple samples have to be examined to increase the test's sensitivity (the proportion of patients with culture-positive tuberculosis that the test detects). Because each smear examination takes up to 10 minutes, tuberculosis diagnosis with ZN smear microscopy creates a large laboratory workload. A variant form of smear microscopy-lightemitting-diode fluorescence microscopy (LED-FM)-could reduce this workload. With LED-FM, smears stained with a fluorescent dye can be examined in a quarter of the time it takes to examine ZN smears. In this study, the researchers evaluate the sensitivity and specificity (the proportion of people with a negative smear among people without tuberculosis; a high specificity indicates a low false-positive rate) of LED-FM using samples collected in a trial undertaken in four resource-limited countries (Ethiopia, Nepal, Nigeria, and Yemen) to investigate two schemes for sputum sample collection. In the spot-morning-spot (SMS) scheme, patients provide an on-the-spot specimen at their initial consultation, a specimen collected at home the next morning, and a second on-the-spot sample when they deliver their morning specimen. In the spot-spot-morning (SSM) scheme, patients provide two on-the-spot samples during their first clinic visit and a sample collected at home the next morning.
What Did the Researchers Do and Find? In the main trial, the researchers collected sputum samples using the SMS or SSM scheme from 6,627 patients with a cough lasting more than two weeks. For their investigation of LED-FM, they examined nearly 2,400 samples (half SSM and half SMS specimens, about a quarter of which were tuberculosis culture-positive) with both ZN smear microscopy and LED-FM and determined the sensitivity and specificity of both tests-with one, two, or three sputum samples per patientrelative to mycobacterial solid culture. Single LED-FM smears had higher sensitivity but lower specificity than single ZN smears. The sensitivities of two LED-FM and two ZN smears were 72.8% and 65.8%, respectively; the specificities of these tests were 90.9% and 98.0%. The sensitivities of three LED-FM and three ZN smears were 77% and 70.5%, respectively; the specificities of these tests were 88.1% and 96.5%. The sensitivity and specificity of both tests was similar for samples collected using the SMS and the SSM schemes.
What Do These Findings Mean? These findings show that in the resource-limited countries included in this trial, LED-FM has a higher sensitivity but lower specificity than ZN smear microscopy. The researchers calculate that in this study the accuracy of three LED-FM examinations was 85% (2,017 out of 2,355 patients were correctly classified as infected or uninfected), whereas the accuracy of three ZN smears was 91.8%. Thus, although LED-FM should identify more people with tuberculosis than ZN smear microscopy, because of its lower specificity, its use might also lead to more people without tuberculosis being needlessly treated for the disease. Nevertheless, provided that the introduction of LED-FM is accompanied by appropriate training and performance monitoring, LED-FM is an attractive potential tool for the laboratory diagnosis of tuberculosis that, together with a move towards the collection of two onthe-spot smears in a single clinic visit, could ensure that poor patients have access to timely tuberculosis diagnosis and prompt treatment.
[fig] Figure 1: Depiction of flow of participants. doi:10.1371/journal.pmed.1001057.g001 [/fig]
[fig] Additional: Information. Please access these Web sites via the online version of this summary at http://dx.doi.org/10. 1371/journal.pmed.1001057. N Details of the parent trial in which the samples used in this study were collected are available in a PLoS Medicine Research Article by Cuevas et al. N The World Health Organization provides information on all aspects of tuberculosis, including information on tuberculosis diagnostics; recent WHO policy statements on diagnosis of tuberculosis are available; the Stop TB Partnership provides information on global tuberculosis control (some information in several languages) N The US Centers for Disease Control and Prevention has information about tuberculosis, including information on the diagnosis of tuberculosis disease N The US National Institute of Allergy and Infectious Diseases also has detailed information on all aspects of tuberculosis N MedlinePlus has links to further information about tuberculosis (in English and Spanish) N A new Web site dedicated to the discussion and optimization of smear microscopy has recently been launched [/fig]
[table] Table 1: Characteristics of the patients on enrolment. Age in years, cough duration in weeks. p,0.01 for all variables. SD, standard deviation. doi:10.1371/journal.pmed.1001057.t001 [/table]
[table] Table 2: Sensitivity and specificity of using two or three ZN and LED-FM smears per patient. [/table]
[table] Table 3: Absolute difference the use of two-or three-smear ZN smear microscopy and LED-FM schemes would make per 1,000 persons tested in situations with similar TB prevalence. [/table]
|
Progranulin-derived granulin E and lysosome membrane protein CD68 interact to reciprocally regulate their protein homeostasis
Edited by Ursula JakobProgranulin (PGRN) is a glycoprotein implicated in several neurodegenerative diseases. It is highly expressed in microglia and macrophages and can be secreted or delivered to the lysosome compartment. PGRN comprises 7.5 granulin repeats and is processed into individual granulin peptides within the lysosome, but the functions of these peptides are largely unknown. Here, we identify CD68, a lysosome membrane protein mainly expressed in hematopoietic cells, as a binding partner of PGRN and PGRN-derived granulin E. Deletion analysis of CD68 showed that this interaction is mediated by the mucinproline-rich domain of CD68. While CD68 deficiency does not affect the lysosomal localization of PGRN, it results in a specific decrease in the levels of granulin E but no other granulin peptides. On the other hand, the deficiency of PGRN, and its derivative granulin peptides, leads to a significant shift in the molecular weight of CD68, without altering CD68 localization within the cell. Our results support that granulin E and CD68 reciprocally regulate each other's protein homeostasis.
Progranulin (PGRN) is a glycoprotein implicated in several neurodegenerative diseases. It is highly expressed in microglia and macrophages and can be secreted or delivered to the lysosome compartment. PGRN comprises 7.5 granulin repeats and is processed into individual granulin peptides within the lysosome, but the functions of these peptides are largely unknown. Here, we identify CD68, a lysosome membrane protein mainly expressed in hematopoietic cells, as a binding partner of PGRN and PGRN-derived granulin E. Deletion analysis of CD68 showed that this interaction is mediated by the mucinproline-rich domain of CD68. While CD68 deficiency does not affect the lysosomal localization of PGRN, it results in a specific decrease in the levels of granulin E but no other granulin peptides. On the other hand, the deficiency of PGRN, and its derivative granulin peptides, leads to a significant shift in the molecular weight of CD68, without altering CD68 localization within the cell. Our results support that granulin E and CD68 reciprocally regulate each other's protein homeostasis.
Frontotemporal lobar degeneration is a progressive neurodegenerative disease characterized by changes in personality and behavior as well as cognitive decline and language impairments (https://www.uptodate.com/contents/frontotemporaldementia-epidemiology-pathology-and-pathogenesis) [bib_ref] Frontotemporal lobar degeneration, Rabinovici [/bib_ref]. Mutation in the granulin (GRN) gene, resulting in haploinsufficiency of the progranulin (PGRN) protein, is one of the leading causes of frontotemporal lobar degeneration [bib_ref] Null mutations in progranulin cause ubiquitin-positive frontotemporal dementia linked to chromosome 17q21, Cruts [/bib_ref] [bib_ref] Mutations in progranulin are a major cause of ubiquitinpositive frontotemporal lobar degeneration, Gass [/bib_ref] , with over 70 disease-associated GRN mutations identified (https://www.uptodate.com/contents/frontotemporaldementia-epidemiology-pathology-and-pathogenesis). PGRN is a secreted glycoprotein that is involved in many cellular processes, including inflammation, wound healing, and tumorigenesis [bib_ref] Progranulin: a proteolytically processed protein at the crossroads of inflammation and neurodegeneration, Cenik [/bib_ref]. PGRN is widely considered to be antiinflammatory, aiding in the reduction of proinflammatory cytokines and reducing the activation of microglia and astrocytes [bib_ref] Progranulin, lysosomal regulation and neurodegenerative disease, Kao [/bib_ref] [bib_ref] Progranulin axis and recent developments in frontotemporal lobar degeneration, Nicholson [/bib_ref]. Structurally, PGRN is comprised of 7.5 granulin segments, denoted as granulin A, B, C, D, E, F, and G, and the half-granulin segment known as paragranulin [bib_ref] Progranulin: a proteolytically processed protein at the crossroads of inflammation and neurodegeneration, Cenik [/bib_ref]. The granulin peptides have been shown to possess functions that are independent of PGRN [bib_ref] Progranulin: a proteolytically processed protein at the crossroads of inflammation and neurodegeneration, Cenik [/bib_ref] [bib_ref] Progranulin, lysosomal regulation and neurodegenerative disease, Kao [/bib_ref].
In addition, accumulating evidence supports that PGRN is critical for proper lysosomal function [bib_ref] The lysosomal function of progranulin, a guardian against neurodegeneration, Paushter [/bib_ref]. Complete loss of PGRN, caused by homozygous mutations in the GRN gene, leads to neuronal ceroid lipofuscinosis, a lysosomal storage disorder that is characterized by degeneration of nerve cells and the accumulation of autofluorescent lipofuscin [bib_ref] Portuguese family with the co-occurrence of frontotemporal lobar degeneration and neuronal ceroid..., Almeida [/bib_ref] [bib_ref] Strikingly different clinicopathological phenotypes determined by progranulin-mutation dosage, Smith [/bib_ref]. PGRN is a resident lysosomal protein and trafficked to the lysosome via two independent pathways, directly by the sortilin receptor [bib_ref] Sortilin-mediated endocytosis determines levels of the fronto-temporal dementia protein, Hu [/bib_ref] or indirectly by binding to the soluble lysosomal protein, prosaposin (PSAP), which carries PGRN with it to the lysosome when it binds its own trafficking receptors, low-density lipoprotein receptor-related protein 1 or the cation-independent mannose-6-phosphate receptor [bib_ref] Prosaposin facilitates sortilin-independent lysosomal trafficking of progranulin, Zhou [/bib_ref]. Within the lysosome, PGRN is processed into individual granulin peptides by lysosomal proteases [bib_ref] The lysosomal protein cathepsin L is a progranulin protease, Lee [/bib_ref] [bib_ref] Lysosomal processing of progranulin, Zhou [/bib_ref] [bib_ref] Processing of progranulin into granulins involves multiple lysosomal proteases and is affected..., Mohan [/bib_ref] , and these granulin peptides have been proposed to be the functional units of PGRN to regulate lysosomal functions. In support of this, the activities of several lysosomal enzymes, including cathepsin D [bib_ref] Progranulin functions as a cathepsin D chaperone to stimulate axonal outgrowth in..., Beel [/bib_ref] [bib_ref] Progranulin stimulates the in vitro maturation of pro-cathepsin D at acidic pH, Butler [/bib_ref] [bib_ref] Progranulin-mediated deficiency of cathepsin D results in FTD and NCL-like phenotypes in..., Valdez [/bib_ref] [bib_ref] Regulation of cathepsin D activity by the FTLD protein progranulin, Zhou [/bib_ref] , glucocerebrosidase (GCase) [bib_ref] Progranulin deficiency leads to reduced glucocerebrosidase activity, Zhou [/bib_ref] [bib_ref] Impaired β-glucocerebrosidase activity and processing in frontotemporal dementia due to progranulin mutations, Arrant [/bib_ref] [bib_ref] Progranulin mutations result in impaired processing of prosaposin and reduced glucocerebrosidase activity, Valdez [/bib_ref] [bib_ref] Rescue of a lysosomal storage disorder caused by Grn loss of function..., Logan [/bib_ref] , and asparagine endopeptidase [bib_ref] Processing of progranulin into granulins involves multiple lysosomal proteases and is affected..., Mohan [/bib_ref] , have been shown to be affected by the loss of PGRN and granulin peptides. PGRN deficiency results in the accumulation of myelin debris in microglial lysosomes, which is further exacerbated by the reduction in cathepsin D levels [bib_ref] Microglial lysosome dysfunction contributes to white matter pathology and TDP-43 proteinopathy in..., Wu [/bib_ref]. In addition, PGRN was shown to interact with bis(monoacylglycero)phosphate, an endolysosomal phospholipid, and PGRN loss leads to a significant reduction in bis(monoacylglycero)phosphate levels [bib_ref] Rescue of a lysosomal storage disorder caused by Grn loss of function..., Logan [/bib_ref]. However, the function of each granulin peptide remains to be characterized. Interestingly, the levels of individual granulin peptides are differentially regulated despite being derived from the same precursor [bib_ref] Differential regulation of progranulin derived granulin peptides, Zhang [/bib_ref] , but the mechanisms involved in modulating the stability of individual granulins in the lysosome remain unknown.
Here, we report the identification of CD68 as a binding partner for full-length PGRN and granulin E. CD68 is a type I transmembrane (TM) protein that is abundantly expressed in microglia and macrophages and localized to the plasma membrane, late endosomal membrane, and lysosomal membrane [bib_ref] CD68/macrosialin: not just a histochemical marker, Chistiakov [/bib_ref]. Although the exact function of CD68 remains to be determined, our data support that CD68 and granulin E regulate each other's homeostasis in the lysosome.
# Results
## Pgrn and granulin e interact with cd68
To identify potential TM protein-binding partners of PGRN or granulin peptides, we cherry-picked 2815 complementary DNAs (cDNAs) encoding the TM proteins from the human ORFeome 8.1 library. The cDNAs were cloned into a mammalian vector containing a C-terminal GFP tag and transfected into COS-7 cells. The cells were incubated with alkaline phosphatase (AP)-tagged PGRN or granulin peptides. Positive interactions were visualized using AP substrates following a previously published protocol [bib_ref] Alkaline phosphatase fusions of ligands or receptors as in situ probes for..., Flanagan [/bib_ref]. From this screen, we identified the TM protein, CD68, as a binding partner for granulin E . To further confirm these interactions, we performed coimmunoprecipitation (co-IP) assays of CD68 with full-length PGRN or granulin peptides and found that CD68 binds to both full-length PGRN and granulin E but no other granulins . Deletion of the C-terminal fragment of PGRN, which contains the granulin E domain (PGRNΔE), abolishes the interaction with CD68 , further supporting that CD68 binds to PGRN via the granulin E domain.
CD68 contains three domains: the TM domain, the mucinproline-rich region (M+P), and the lysosomal-associated membrane protein (LAMP)-like domain (Lamp-D) [bib_ref] CD68/macrosialin: not just a histochemical marker, Chistiakov [/bib_ref]. To characterize which regions of CD68 mediate the interaction with granulin E, we generated deletion constructs of CD68. Deletion of the mucin-proline-rich domain, but not the LAMP-like domain, abolished CD68 binding to PGRN and granulin E [fig_ref] Figure 2: CD68 binds PGRN and granulin E via the mucin-proline-rich domain [/fig_ref] and C), supporting that the mucinproline-rich domain of CD68 is required for the interaction with PGRN and granulin E.
## Cd68 is not an essential pgrn trafficking receptor
Since CD68 is present in both the plasma membrane and the lysosomal membrane, we hypothesized that CD68 might function as a PGRN lysosomal trafficking receptor. To test this, we transfected CD68 into COS-7 cells and incubated the cells with the conditioned medium containing PGRN. Sortilintransfected cells were used as a positive control. Like sortilin, CD68 is able to mediate the uptake of PGRN but much less robustly . Next, we examined the lysosomal localization of PGRN in Cd68 −/− cells. We found that PGRN remains colocalized with lysosomal markers cathepsin D and LAMP1 of cultured Cd68 −/− macrophages [fig_ref] Figure 3: CD68 is not required for PGRN lysosomal trafficking [/fig_ref] and Cd68 −/− microglia in the mouse brain [fig_ref] Figure 3: CD68 is not required for PGRN lysosomal trafficking [/fig_ref]. Thus, although CD68 can potentially mediate PGRN uptake, ablation of CD68 does not have any obvious effect on PGRN lysosomal trafficking in vivo.
Defects in PGRN lysosomal trafficking often lead to increased levels of PGRN in the extracellular space, such as in . Physical interaction between CD68 and granulin E. A, COS7 cells transfected with CD68-GFP were incubated with conditioned media containing AP-tagged granulin peptides. Positive interaction between CD68 and the granulin E was observed after incubation with AP substrate. The scale bar represents 100 μm. B, Myc-His-tagged CD68 and GFP-tagged granulin A through G were transfected in HEK293T. Lysates were incubated with anti-GFP-conjugated beads. After washes, products were analyzed by Western blot using anti-GFP and anti-Myc antibodies as indicated. C, FLAG-tagged PGRN and GFP-tagged CD68 were transfected in HEK293T cells. Lysates were incubated with anti-GFP-conjugated beads. After washes, products were analyzed by Western blot using anti-GFP and anti-FLAG antibodies as indicated. D, conditioned media containing FLAG-tagged PGRN were incubated with beads bound to GFP or GFP-CD68. After washes, products were analyzed by Western blot using anti-GFP and anti-FLAG antibodies as indicated. E, HEK293T cells were transfected with GFP-tagged CD68 and FLAG-tagged PGRN or PGRN-ΔE. Lysates were incubated with anti-FLAG-conjugated beads. After washes, products were analyzed by Western blot using anti-GFP and anti-FLAG antibodies as indicated. AP, alkaline phosphatase; HEK293T, human embryonic kidney 293T cell line; PGRN, progranulin. the case of sortilin or PSAP deficiency [bib_ref] Sortilin-mediated endocytosis determines levels of the fronto-temporal dementia protein, Hu [/bib_ref] [bib_ref] Prosaposin facilitates sortilin-independent lysosomal trafficking of progranulin, Zhou [/bib_ref]. If CD68 were a PGRN lysosomal trafficking receptor, we may expect that CD68 deficiency would impair PGRN lysosomal trafficking and thus, cause an increase in the extracellular levels of PGRN. However, the levels of PGRN in the serum were not altered in Cd68 −/− mice [fig_ref] Figure 3: CD68 is not required for PGRN lysosomal trafficking [/fig_ref]. PGRN is processed to granulin peptides in the lysosome, and defects in PGRN trafficking lead to PGRN processing defects. To determine whether CD68 affects PGRN processing, we determined the levels of granulin peptides in the spleen, where CD68 is highly expressed. The overall levels of full-length PGRN and granulin peptides are not altered in CD68-deficient spleen lysates [fig_ref] Figure 3: CD68 is not required for PGRN lysosomal trafficking [/fig_ref] , further supporting that CD68 is not an essential PGRN lysosomal trafficking receptor.
## Cd68 deficiency decreases the levels of granulin e
Since CD68 specifically interacts with granulin E and the granulin E domain of PGRN, next we tested whether CD68 deficiency alters granulin E levels. To do this, we generated homemade antibodies against individual granulin peptides [bib_ref] Differential regulation of progranulin derived granulin peptides, Zhang [/bib_ref] and measured the levels of granulin E and other granulins in the lysate prepared from the spleen tissue and primary microglia, which have high levels of CD68 expression. Antibodies against granulin A, B, C, E, and F, but not D and G, can detect endogenous granulin peptides specifically [fig_ref] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E [/fig_ref] [bib_ref] Differential regulation of progranulin derived granulin peptides, Zhang [/bib_ref]. Notably, our results show that there is a significant reduction in the levels of granulin E, but no other granulins in Cd68 −/− spleen lysates [fig_ref] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E [/fig_ref] , A and B) and primary microglia cultured from Cd68 −/− pups [fig_ref] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E [/fig_ref] , C and D).
The specific reduction of granulin E in Cd68 −/− samples could be due to altered PGRN processing, reduced stability of granulin E in the lysosome, or changes in granulin E secretion or endocytosis. To determine whether CD68 regulates granulin E trafficking, we incubated CD68-expressing COS-7 cells with conditioned medium containing granulin E. We found that CD68 is able to mediate the endocytosis and lysosomal delivery of granulin E when overexpressed in COS-7 cells , indicating that granulin E trafficking could be affected by CD68. However, we failed to detect a clear signal for granulin E in the media of Cd68 −/− primary microglia, even after trichloroacetic acid precipitation or immunoprecipitation (data not shown), indicating that the majority of granulin E is intracellular and the decrease in granulin E levels in CD68 deficient cells is unlikely because of increased secretion or decreased endocytosis.
## Cd68 deficiency does not affect cathepsin d or gcase activities or psap processing
Since CD68 deficiency leads to reduced levels of granulin E and granulin E has been shown to regulate GCase (31) and cathepsin D activities [bib_ref] Progranulin functions as a cathepsin D chaperone to stimulate axonal outgrowth in..., Beel [/bib_ref] [bib_ref] Progranulin-mediated deficiency of cathepsin D results in FTD and NCL-like phenotypes in..., Valdez [/bib_ref] , we determined whether loss of CD68 would affect GCase and cathepsin D activities. No obvious difference was found between WT and Cd68 −/− spleen lysates [fig_ref] Figure 2: CD68 binds PGRN and granulin E via the mucin-proline-rich domain [/fig_ref]. In addition, we analyzed whether the levels of PSAP and saposins are altered by CD68 loss since PSAP interacts and traffics with PGRN to the lysosome [bib_ref] Impaired prosaposin lysosomal trafficking in frontotemporal lobar degeneration due to progranulin mutations, Zhou [/bib_ref] , and among all the granulins, granulins D and E have a strong interaction with PSAP [bib_ref] The interaction between progranulin and prosaposin is mediated by granulins and the..., Zhou [/bib_ref]. The levels of both full-length PSAP and saposins are not altered by CD68 ablation [fig_ref] Figure 3: CD68 is not required for PGRN lysosomal trafficking [/fig_ref] , indicating that CD68 does not affect PSAP expression or processing.
## Pgrn deficiency affects cd68 homeostasis
To explore how the deficiency of PGRN and granulin peptides affect CD68 protein homeostasis, we examined CD68 protein in Grn −/− cells and tissues. Western blot analysis of Grn −/− spleen lysates showed a decrease in the molecular weight (MW) of CD68 in addition to an increase in CD68 levels [fig_ref] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68 [/fig_ref]. MW changes of CD68 are also seen in PGRNdeficient primary microglia and bone marrow-derived macrophages (BMDMs) [fig_ref] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68 [/fig_ref]. Since CD68 is a resident lysosomal glycoprotein, the MW shift could be indicative of changes in lysosomal cleavage or glycosylation of CD68. To test the former, we treated the cells with bafilomycin and chloroquine, which are agents that alter the lysosomal pH and thus decrease lysosomal degradative capacity. CD68 MW in Grn −/− cells is not altered by lysosome inhibition [fig_ref] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68 [/fig_ref] , suggesting that CD68 MW changes under PGRN-deficient conditions are not because of lysosome-dependent cleavage.
To determine whether alterations in N-linked glycosylation are responsible for the changes in CD68 MW, PNGase F was used to remove N-linked glycosyl groups from CD68 immunoprecipitated from either WT or Grn −/− spleen lysates. The MW shift of CD68 is still present in Grn −/− samples upon the removal of all N-glycans [fig_ref] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68 [/fig_ref] , indicating that the MW change of CD68 in PGRN-deficient conditions is not due to alterations in N-glycosylation.
Another possibility is that CD68 protein localized on the plasma membrane could be processed by extracellular proteases under PGRN-deficient conditions, resulting in the changes in MW. To test this, we treated control and Grn −/− cells with various matrix metalloprotease inhibitors (GM6001 . B, brain sections from adult WT and Cd68 −/− mice were stained with anti-PGRN, LAMP1, and IBA1 antibodies. The scale bar represents 100 μm (inset: 25 μm). Representative images from three mice of each genotype were shown. C, serum PGRN levels in WT and Cd68 −/− mice were measured by ELISA. Six mice per genotype were analyzed (n = 6). D, Western blot analysis of PGRN and granulin peptides in spleen lysates of WT and Cd68 −/− mice. Six mice per genotype were analyzed (n = 6). PGRN and granulin intensities were normalized to GAPDH. LAMP1, lysosomal-associated membrane protein 1; ns, not significant; PGRN, progranulin. and TAPI-2), β-secretase (BACE1) inhibitor, or leupeptin. None of the inhibitors restored CD68 MW in Grn −/− cells [fig_ref] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E [/fig_ref] , indicating that the decrease in CD68 MW is not likely to be caused by proteolytic processing of the protein.
Next, we examined whether CD68 localization is altered by PGRN loss. In Grn −/− BMDMs and RAW264.7 cells, CD68 still localizes to the lysosome as indicated by the colocalization between CD68 and the lysosomal marker, cathepsin D [fig_ref] Figure 6: PGRN does not affect CD68 localization [/fig_ref]. Since CD68 is also present in the plasma membrane, we investigated whether the levels of CD68 at the cell surface are altered in Grn −/− cells. To test this, we measured cell surface levels of CD68 in nonpermeabilized conditions in control and Grn −/− RAW264.7 cells [fig_ref] Figure 6: PGRN does not affect CD68 localization [/fig_ref] and found that CD68 surface levels were not altered in Grn −/− RAW264.7 cells [fig_ref] Figure 6: PGRN does not affect CD68 localization [/fig_ref].
# Discussion
## Differential regulation of pgrn-derived granulin peptides
PGRN is processed into individual granulin peptides in the lysosome, and these granulin peptides are proposed to possess unique functions to regulate lysosomal activities [bib_ref] Structure dissection of human progranulin identifies wellfolded granulin/epithelin modules with unique functional..., Tolkatchev [/bib_ref]. Although the granulin peptides are derived from the same precursor, our recent studies have shown that the levels of individual granulin peptides differ from each other [bib_ref] Differential regulation of progranulin derived granulin peptides, Zhang [/bib_ref]. This is further supported by our observation that CD68 ablation specifically decreases the levels of granulin E but no other granulins tested [fig_ref] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E [/fig_ref]. How CD68 regulates the levels of granulin E remains to be determined. Since we failed to detect granulins in the extracellular space, it is unlikely that CD68 regulates the levels of granulin E via modulating its secretion or endocytosis, although CD68 can mediate the uptake of granulin E to the lysosome when overexpressed in COS-7 cells . Instead, we hypothesize that CD68 is required for the stability of granulin E in the lysosomal compartment. It is possible that each granulin peptide interacts with a different set of lysosomal proteins, which determine the stability and half-life of these peptides in the lysosome. Since CD68 is not expressed in neurons, there might exist another binding partner for granulin E in neurons and other cell types to stabilize granulin E. The levels of granulin peptides A, B, C, E, and F were quantified and normalized to GAPDH. Six mice per genotype were analyzed (n = 6). ns, not significant; *p < 0.05, ***p < 0.001, unpaired t-test. C and D, lysates from WT and Cd68 −/− microglia were probed with antibodies against individual granulin peptides as indicated. The levels of granulin peptides A and E relative to GAPDH were quantified and normalized to WT. Results were from three sets of microglia cultured independently from mouse pups of corresponding genotypes (n = 3). *p < 0.05, ns, not significant; unpaired t-test.
Since CD68 is critical for maintaining proper levels of granulin E, we expect that granulin E-dependent lysosomal activities should be altered in CD68-deficient cells. Unfortunately, we still do not have a full understanding of granulin E functions in the lysosome. In the literature, GCase (31) and cathepsin D [bib_ref] Progranulin functions as a cathepsin D chaperone to stimulate axonal outgrowth in..., Beel [/bib_ref] [bib_ref] Progranulin-mediated deficiency of cathepsin D results in FTD and NCL-like phenotypes in..., Valdez [/bib_ref] have been shown to be regulated by granulin E, but their activities are not altered by CD68 ablation [fig_ref] Figure 2: CD68 binds PGRN and granulin E via the mucin-proline-rich domain [/fig_ref]. Future work is needed to fully dissect the functions of granulin E in the lysosome and the consequence of CD68 ablation on granulin E-dependent lysosomal activities.
## Regulation of cd68 by pgrn and granulin e
Our data have shown that PGRN deficiency results in a shift in the MW of CD68 in addition to an increase in CD68 levels (Figs. 5, A-C and 6B). The MW shift is not caused by alterations in N-glycosylation or proteolytic cleavage. It is possible that loss of PGRN leads to alterations in the O-glycosylation pattern of CD68, since mouse CD68 contains 26 O-linked glycosylation sites [bib_ref] CD68/macrosialin: not just a histochemical marker, Chistiakov [/bib_ref]. Another important question is whether the function of CD68 is affected by PGRN loss. Unfortunately, although CD68 is frequently utilized as a marker for macrophages and activated microglia, little is known about its molecular functions [bib_ref] CD68/macrosialin: not just a histochemical marker, Chistiakov [/bib_ref]. Interactions between CD68 and phosphatidylserine and oxidized low-density lipoprotein on the plasma membrane have been reported, but the physiological importance of this interaction remains to be confirmed [bib_ref] CD68/macrosialin: not just a histochemical marker, Chistiakov [/bib_ref] [bib_ref] Cell surface expression of mouse macrosialin and human CD68 and their role..., Ramprasad [/bib_ref] [bib_ref] Evidence that the lipid moiety of oxidized low density lipoprotein plays a..., Terpstra [/bib_ref]. A better understanding of CD68 functions will help determine the significance of the observed shift in MW under PGRN-deficient conditions. Nevertheless, our work has identified CD68 as a specific binding partner for granulin E and indicated that regulation of CD68 might be one function of granulin E in the lysosome.
## Experimental procedures
## Primary antibodies and reagents
The following antibodies were used in this study: rat antimouse CD68 (Bio-Rad; catalog no.: MCA1957), rabbit anti-CD68 (Abcam; catalog no.: 125212), sheep antimouse PGRN (R&D Systems; catalog no.: AF2557), mouse anti-Myc (catalog no.: 9E10; Hybridoma Bank), mouse anti-GAPDH (Proteintech Group; catalog no.: 60004-1-Ig), rat antimouse LAMP1 (BD Biosciences; catalog no.: 553792), goat anti-CathD (R&D Systems; catalog no.: AF1029) and rabbit anti-IBA-1 (Wako; catalog no.: 01919741), rabbit anti-PGAM1 (Proteintech; catalog no.: 16126-1-AP), and GFP-Trap were from ChromoTek. Rabbit anti-GFP antibodies were a gift from Professor Anthony Bretscher and rabbit anti-cathepsin D antibodies were a gift from Dr. William Brown at Cornell University. Antibodies against each granulin peptide have been previously characterized [bib_ref] Differential regulation of progranulin derived granulin peptides, Zhang [/bib_ref]. Rabbit anti-mouse PSAPs were generated by Pocono Rabbit Farm and Laboratory and previously characterized [bib_ref] Prosaposin facilitates sortilin-independent lysosomal trafficking of progranulin, Zhou [/bib_ref]. Alexa Fluor-conjugated secondary antibodies (488/ 594/647 nm) were from Invitrogen. IRDye 680RD/800CW secondary antibodies were from Invitrogen and LI-COR Biosciences.
The following reagents were also used in the study: Dulbecco's modified Eagle's medium (DMEM) (Cellgro; catalog no.: 10-017-CV), 0.25% trypsin (Corning; catalog no.: 25-053-CI), Odyssey blocking buffer (LI-COR Biosciences; catalog no.: 927-40000), protease inhibitor (Roche; catalog no.:
## Plasmids
Human CD68 was cloned in the pEGFPs-N2, pcDNA3.1/ myc-His A, and the lentiviral vector pCDH-CMV-MCS-EF1-Puro vector (System Biosciences). The pAP5 (Gen Hunter; catalog no.: Q202) vector was used to clone AP-tagged PGRN and granulins A-G. PGRNΔE (amino acids 17-492) was cloned into the pSectag2A vector with an N-terminal FLAG tag.
About 2815 cDNAs encoding the TM proteins from the human ORFeome 8.1 library in the pDONR223 vector were [fig_ref] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68 [/fig_ref]. B, deletion of PGRN in RAW264.7 cells leads to the decreased molecular weight of CD68 and increased CD68 levels. CD68 levels were quantified and normalized to GAPDH. n = 3, *p < 0.05, unpaired t-test. C, control and Grn −/− RAW264.7 cells were fixed, permeabilized, and stained with anti-CD68, cathepsin D, and PGRN antibodies. The scale bar represents 10 μm (inset: 2.5 μm). D, live control and Grn −/− RAW264.7 cells were incubated with rat anti-mouse CD68 antibodies on ice followed by washing, fixation, blocking, and staining with secondary antibody and Hoechst. Representative images from three replicates are shown. The scale bar represents 10 μm. E, cell surface and total levels of CD68 were quantified by ImageJ for experiments in C and D. n = 3. BMDM, bone marrow-derived macrophage; ns, not significant; PGRN, progranulin.
obtained from DNASU and cloned into the pDEST47 vector (Invitrogen) via gateway cloning.
## Mouse studies
WT C57/BL6, Cd68 −/− (37), and Grn −/− (38) mice were purchased from Jax Laboratories. All animals (one to six adult mice per cage) were housed in a 12 h light/dark cycle. Mixed males and females are used in the study. All animal procedures have been approved by the Institutional Animal Care and Use Committee at Cornell University (2017-0056). For perfusion, mice were anesthetized with isoflurane at a concentration of 1 ml isoflurane for every 500 ml of the volume of the anesthesia chamber. The mice were monitored closely, and deep anesthesia was confirmed by toe pinching (1-2 min after induction). Mice were perfused with PBS afterward. To culture primary microglia, P0-P2 pups were decapitated using a razor blade and brains were harvested.
## Cell culture
Human embryonic kidney 293T and RAW264.7 cells were maintained in DMEM (Cellgro) supplemented with 10% fetal bovine serum (Sigma) in a humidified incubator at 37 C with 5% CO 2 . RAW264.7 cells with PGRN deletion or controls were generated by infecting the cells with lentivirus-expressing Cas9 and the PGRN-targeting guide RNAs (5 0 -caccgCGGACCCCG ACGCAGGTAGG-3 0 and 5 0 -aaacCCTACCTGCGTC GGGGTCCGc-3 0 ) or Cas9 only. Cells were selected with puromycin (8 μg/ml) 2 days after infection, and the knockout is confirmed by Western blot and immunostaining. For transient overexpression, cells were transfected with polyethyleneimine as previously described [bib_ref] Use of polyethyleneimine polymer in cell culture as attachment factor and lipofection..., Vancha [/bib_ref]. BMDMs from WT and Grn −/− mice were cultured according to published protocols [bib_ref] Bone marrow-derived macrophages (BMM): isolation and applications, Weischenfeldt [/bib_ref]. Briefly, the cells were flushed out from the femur, tibia, and fibula of mice using DMEM. Cells were spun, filtered, and resuspended in L929 media with 10% fetal bovine serum, and DMEM. Cells were counted with a hemocytometer (Lumictyte; catalog no.: 090001) and plated. Mouse primary cortical neurons and microglia were cultured from P0 to P2 pups according to published protocols [bib_ref] Impaired prosaposin lysosomal trafficking in frontotemporal lobar degeneration due to progranulin mutations, Zhou [/bib_ref] [bib_ref] A multifaceted role of progranulin in regulating amyloid-beta dynamics and responses, Du [/bib_ref].
## Ap-based cell surface binding screen
COS-7 cells grown in 96-well tissue culture plates (Corning) were transfected with 50 to 100 ng of expression construct for an individual TM protein in each well. Sortilin was used as a positive control for PGRN binding, and pEGFP-C1 vector was used as a negative control. After 2 days, the wells were washed 2× with HBH (Hank's balanced salt solution with calcium and magnesium with 20 mM Hepes, pH 6.4, and 0.1% bovine serum albumin) and incubated with conditioned medium containing AP-PGRN, AP-PGRN + AP-PSAP, or a mixture of AP-granulin A, B, D, E, and G at 50 nM. AP fusion protein conditioned media were diluted to working concentrations in HBH. About 100 μl was added to each well, and the plate was incubated for 3 h at room temperature. Following the incubation, wells were washed twice with cold HBH to remove traces of unbound AP-fused ligand, and then the cells were fixed with 3.7% formaldehyde at room temperature for 20 min. Wells were washed 1× with HBH, and then GFP expression was assessed by fluorescence microscopy using the ImageXpress system (Molecular Devices). Wells were then filled with HBH, and the plate was sealed tightly with adhesive film (VWR). The plate was incubated overnight at 65 C to inactivate endogenous APs. Wells were washed 1× with AP buffer (100 mM Tris-HCl, pH 9.5, 100 mM NaCl, and 5 mM MgCl 2 ) and incubated with 100 μl of developing solution composed of AP buffer with 350 μg/ml nitroblue tetrazolium chloride (NBT) and 175 μg/ml 5-bromo-4-chloro-3 0 -indolyl phosphate p-toluidine salt at room temperature for a minimum of 30 min or until signal developed. Positive interactions, as indicated by the presence of an insoluble black-purple precipitate, were confirmed by eye using brightfield microscopy.
## Immunoprecipitation and western blot analysis
Cells were lysed in a cold solution containing 150 mM NaCl, 50 mM Tris-HCl (pH 7.5), 1% Triton X-100, 0.1% deoxycholic acid, 1× protease inhibitors (Roche). After centrifugation at 14,000g, for 15 min, at 4 C, supernatants were transferred to clean tubes on ice, to which GFP-Trap or Myc-Trap beads (ChromoTek) were added and then rocked for 3 to 4 h at 4 C. Samples were run on 12% polyacrylamide gels and transferred to Immobilon-FL polyvinylidene fluoride membranes (Millipore Corporation). Membranes were blocked with either 5% nonfat milk in 1× PBS or Odyssey Blocking Buffer (LI-COR Biosciences) for 1 h at room temperature followed by incubation with primary antibodies and left rocking overnight at 4 C. Membranes were then washed with Tris-buffered saline with 0.1% Tween-20 for three times, 5 min each, and incubated with fluorescently tagged secondary antibodies (LI-COR Biosciences) for 1 h at room temperature, and then followed by three washes. Membranes were scanned using an Odyssey Infrared Imaging System (LI-COR Biosciences). Densitometry was performed using Image Studio (LI-COR Biosciences) and ImageJ (National Institutes of Health).
For granulin peptide detection, NuPAGE Bis-Tris gradient gels (Invitrogen; NP032C and NP032A) were used. Gels were run in NuPAGE Mes SDS cold running buffer (50 mM Mes, 50 mM Tris base, 0.1% SDS, 1 mM EDTA, pH 7.3) and transferred using cold NuPAGE Transfer buffer (25 mM bicine, 25 mM Bis-Tris [free base], 1 mM EDTA, pH 7.2, 20% methanol-final volume) onto a nitrocellulose membrane with 0.2 μm pore size (Millipore Corporation). Membranes were blocked using 5% nonfat milk in 1× PBS for 2 h at room temperature followed by incubation with primary antibody for 2 days. Washes and secondary antibody incubation were performed as aforementioned.
To prepare tissue lysates, mice were perfused with PBS, and tissues were dissected and snap-frozen with liquid nitrogen and kept at −80 C. On the day of the experiment, frozen tissues were thawed and homogenized on ice with a bead homogenizer (Moni International) in a cold solution of radioimmunoprecipitation assay buffer (150 mM NaCl, 50 mM Tris-HCl [pH 8.0], 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% SDS) with proteinase inhibitors. After centrifugation at 14,000g for 15 min at 4 C, supernatants were collected. Protein concentrations were determined via BCA assay and then standardized. Equal amounts of protein were analyzed by Western blotting using the indicated antibodies.
## Gcase activity assay
Spleen tissues from 10-month-old WT and Cd68 −/− mice were homogenized with 0.2% (w/v) sodium taurocholate and 0.1% (v/v) Triton X-100. Protein concentrations were determined via BCA assay and then standardized. Tissue lysates were then incubated with 100 nM MDW941 at 37 C for 30 min. Reactions were stopped by the addition of an equal volume of 2× Laemmli sample buffer with 10% β-mercaptoethanol before heating at 95 C for 5 min. An equal amount of each sample (50 μg total protein) was run on a 12% polyacrylamide gel, which was scanned at 532 nm excitation/ 580 nm emission with a Typhoon Imaging System (GE Healthcare), then Western blot and assessment were performed as described previously, with all values normalized to GAPDH.
## Cathepsin d activity assay
Spleen tissues from 6.5-month-old WT and Cd68 −/− mice were lysed in lysis buffer (0.2% [w/v] taurocholate and 0.2% [v/ v] Triton X-100 at pH 5.2 at a 1:10 ratio of tissue weight (g) to lysis buffer (ml). Equal amounts of tissue lysates were incubated at 37 C for 30 min in 100 μl assay buffer (50 mM sodium acetate [pH 5.5], 0.1 M NaCl, 1 mM EDTA, and 0.2% [v/v] Triton X-100) in the presence of the fluorogenic substrate (MOCAc-GKPILFFRLK(Dnp)-D-R-NH2) (Calbiochem; catalog no.: 219360). The fluorescence released as a result of cathepsin D proteolytic activity was read at 340 nm (excitation) and 420 nm (emission).
## Elisa
Blood samples were collected from four groups of 3-to 4-month-old mice and left at 4 C overnight. Blood was spun at 3500g for 15 min two times, and serum was collected and analyzed using mouse PGRN ELISA Kit (R&D Systems; catalog no.: MPGRN0).
## Immunofluorescence staining, image acquisition, and analysis
Cells were fixed, permeabilized with 0.05% saponin, and visualized using immunofluorescence microscopy as previously described [bib_ref] The frontotemporal lobar degeneration risk factor, TMEM106B, regulates lysosomal morphology and function, Brady [/bib_ref]. For cell surface staining, live cells were incubated on ice for 2 h with antibodies recognizing extracellular domain of LAMP1 or CD68 in 1× Hank's balanced salt solution with 20 mM Hepes, followed by fixation in 4% paraformaldehyde for 20 min at room temperature, then blocked with Odyssey blocking buffer, and followed with secondary antibody incubation with Hoechst for 1 h. The intensity of cell surface LAMP1 and the ratio between cell surface CD68 and total CD68 was quantified using ImageJ from three independent experiments (50-70 cells/experiment). The quantitative analysis of fluorescence images was performed using ImageJ. For the quantitative analysis of intracellular levels of LAMP1, the entire cell body was selected, and the fluorescence intensity was measured directly using ImageJ after a threshold application.
Mice were perfused and fixed with 4% paraformaldehyde in 1× PBS for 2 days, followed by daily incubations with 15% sucrose and then 30% sucrose in 1× PBS, respectively. About 18 μm brain sections were cut with a cryotome. The brain sections were blocked and permeabilized with 0.1% saponin in Odyssey buffer (Licor) followed by incubation with primary and secondary antibodies as mentioned previously.
## Deglycosylation assay
Spleens were harvested from WT, Cd68 −/− , and Grn −/− mice and lysed in cold IP lysis buffer (150 mM NaCl, 50 mM Tris [pH 8.0], 1% Triton X-100, 0.1% deoxycholic acid, and 1× protease inhibitors). The samples were centrifuged at 14,000g for 15 min at 4 C, and the supernatants were collected. After centrifugation, protein concentrations were quantified using the Pierce Protein Assay Kit (catalog no.: 23225). The samples were then incubated with 10 μl of protein G beads (Genscript; catalog no.: L00209) and placed in a nutator for 1 h at 4 C. The samples were then centrifuged at 3000g for 30 s at 4 C, and the supernatant was collected and placed in clean tubes, to which antibodies against CD68 were added and rocked for 3 to 4 h. About 15 μl of protein G-beads were then added and rocked for 2 h at 4 C. Samples were centrifuged at 3000g for 30 s at 4 C and then washed with 1 ml of a solution containing 150 mM NaCl, 50 mM Tris (pH 8.0), and 1% Triton X-100. This was repeated for a total of three washes. After the final centrifugation, 200 μl of IP wash buffer was added and divided into two clean tubes and centrifuged. One tube was eluted by the addition of 25 μl of Laemmli sample buffer with 5% β-mercaptoethanol, whereas the other tube was used to perform the deglycosylation assay (NEB; catalog no.: P0704S).
For the deglycosylation assay, 1 μl of glycoprotein denaturing buffer (10×) and 9 μl H 2 O were added to the tube. The sample was heated to 100 C for 10 min and then placed on ice for 10 s 2 μl of GlycoBuffer 2 (10×), 2 μl of 10% NP-40 and 6 μl of water was added, and the sample was lightly flicked. Finally, 1 μl of PNGase F was added, mixed gently, and incubated at 37 C for 1 h. Western blot was performed of the input, IP, and deglycosylated samples.
# Statistical analysis
All statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, Inc). All data are presented as mean ± SEM. Statistical significance was assessed by unpaired t-test (for two-group comparisons). p Values less than or equal to 0.05 were considered statistically significant. *p < 0.05; **p < 0.01; ***p < 0.001; and ****p < 0.0001.
## Data availability
The data supporting the findings of this study are available from the corresponding author on request.
Supporting information-This article contains supporting information.
Author contributions-T. Funding and additional information-This work is supported by the National Institute of Neurological Disorders and Stroke/National Institute on Aging (National Institutes of Health; grant nos.: R01NS088448 and R01NS095954) and the Bluefield project to cure frontotemporal dementia to F. H. The funders are not involved in the study. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Conflict of interest-The authors declare that they have no conflicts of interest with the contents of this article.
Ethical approval and consent to participate-All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. The work under animal protocol 2017-0056 is approved by the Institutional Animal Care and Use Committee at Cornell University.
Abbreviations-The abbreviations used are: AP, alkaline phosphatase; BCA, bicinchoninic acid; BMDM, bone marrow-derived macrophage; cDNA, complementary DNA; co-IP, coimmunoprecipitation; DMEM, Dulbecco's modified Eagle's medium; GCase, glucocerebrosidase; GRN, granulin; HBH, Hank's balanced salt solution with Hepes buffer; LAMP, lysosomal-associated membrane protein; MW, molecular weight; PGRN, progranulin; PSAP, prosaposin; TM, transmembrane.
[fig] Figure 2: CD68 binds PGRN and granulin E via the mucin-proline-rich domain. A, schematic of CD68 domain structure: mucin-praline-rich domain (M + P) or the LAMP-like domain (Lamp-D), and the transmembrane (TM) domain. B, Myc-tagged CD68 constructs containing the M + P or Lamp-D domains were expressed in HEK293T cells with GFP-tagged PGRN, and the lysates were coimmunoprecipitated using anti-Myc-conjugated beads. IP products were analyzed by Western blot and probed with Myc and GFP antibodies as indicated. Full length-CD68 (FL-CD68) is used as a control. C, M + P or the Lamp-D domains of CD68 were fused to the transmembrane domain of PDGFR (pDisplay vector; Invitrogen) and transfected into COS-7 cells. FL-CD68 is used as a control. Cells were incubated with AP-tagged PGRN and granulin E (100 nM). The scale bar represents 100 μm. AP, alkaline phosphatase; HEK293T, human embryonic kidney 293T cell line; IP, immunoprecipitation; LAMP, lysosomal-associated membrane protein; PDGFR, platelet-derived growth factor receptor; PGRN, progranulin. [/fig]
[fig] Figure 3: CD68 is not required for PGRN lysosomal trafficking. A, WT and Cd68 −/− bone marrow-derived macrophages (BMDMs) were stained with anticathepsin D (CathD), PGRN, and CD68 antibodies. The scale bar represents 10 μm (inset: 5 μm [/fig]
[fig] Figure 4: CD68 deficiency causes a specific reduction in the levels of granulin E. A and B, spleen lysates of 3-month-old WT and Cd68 −/− mice were analyzed via Western blot and probed with antibodies against individual granulin peptides as indicated. [/fig]
[fig] Figure 5: PGRN deficiency leads to decreased molecular weight of CD68. A, Western blot analysis of CD68 in WT and Grn −/− spleen lysates. CD68 levels were quantified and normalized to GAPDH. n = 4, **p < 0.01, unpaired t-test. B, WT and Grn −/− BMDM cells were treated with lysosomal inhibitors, bafilomycin (50 nM), and chloroquine (250 μM) for 8 h at 37 C. Representative blots from two replicates were shown. C, Western blot analysis of CD68 in primary microglia derived from WT and Grn −/− mice. Results were from three sets of microglia cultured independently from mouse pups of corresponding genotypes (n = 3). ns, not significant; unpaired t-test. D, spleen lysates from WT, Grn −/− , and Cd68 −/− mice were immunoprecipitated using rat antimouse CD68 antibodies. The beads were then incubated with or without PNGase F. Representative blots from three replicates are shown. Immunoglobulin G (IgG) bands are indicated by arrows. BMDM, bone marrow-derived macrophage; PGRN, progranulin.05056489001), Hochest 33342 (Invitrogen), bafilomycin (Sigma; catalog no.: 11707), chloroquine (Sigma; catalog no.: C6628), PNGaseF (New England Biolabs; catalog no.: P0704S), Pierce Bicinchoninic Acid (BCA) Protein Assay Kit (Thermo Fisher Scientific; catalog no.: 23225), OCT compound (Electron Microscopy Sciences, catalog no.: 62550-01), leupeptin (Sigma; catalog no.: L2884), TAPI-2 (EMD Millipore; catalog no.: 579052), GM6001 (EMD Millipore; catalog no.: CC1010), and β-secretase (BACE) IV inhibitor (EMD Millipore; catalog no.: 565788). [/fig]
[fig] Figure 6: PGRN does not affect CD68 localization. A, WT and Grn −/− BMDMs were stained with antibodies against CD68, cathepsin D, and PGRN. The scale bar represents 10 μm (inset: [/fig]
|
Cytoplasmic vacuolation with endoplasmic reticulum stress directs sorafenib induced non-apoptotic cell death in hepatic stellate cells
The activated hepatic stellate cells (HSCs) are the major cells that secrete the ECM proteins and drive the pathogenesis of fibrosis in chronic liver disease. Targeting of HSCs by modulating their activation and proliferation has emerged as a promising approach in the development of anti-fibrotic therapy. Sorafenib, a multi-kinase inhibitor has shown anti-fibrotic properties by inhibiting the survival and proliferation of HSCs. In present study we investigated sorafenib induced cytoplasmic vacuolation mediated decreased cell viability of HSCs in dose and time dependent manner. In this circumstance, sorafenib induces ROS and ER stress in HSCs without involvement of autophagic signals. The protein synthesis inhibitor cycloheximide treatment significantly decreased the sorafenib-induced cytoplasmic vacuolation with increasing cell viability. Antioxidant human serum albumin influences the viability of HSCs by reducing sorafenib induced vacuolation and cell death. However, neither caspase inhibitor Z-VAD-FMK nor autophagy inhibitor chloroquine could rescue the HSCs from sorafenib-induced cytoplasmic vacuolation and cell death. Using TEM and ER organelle tracker, we conclude that the cytoplasmic vacuoles are due to ER dilation. Sorafenib treatment induces calreticulin and GPR78, and activates IRE1α-XBP1s axis of UPR pathway, which eventually trigger the non-apoptotic cell death in HSCs. This study provides a notable mechanistic insight into the ER stress directed non-apoptotic cell death with future directions for the development of efficient anti-fibrotic therapeutic strategies. OPEN the mechanistic interconnections between apoptosis and autophagy 8 . It has been reported that in HCC, sorafenib induces the ER dilation, and activation of the unfolded protein response (UPR) pathway, and these events have direct connection with cytoplasmic vacuolation mediated non-apoptotic cell death6,[9][10][11][12][13]. In present study we found sorafenib induced cytoplasmic vacuolation in HSCs. However, the role of sorafenib induced ER stress, autophagy and their interaction have not been not well documented. While, the cross-talk of ER stress and autophagy are important cellular processes for fibrogenic activity of HSCs 14 , a high demand for extracellular matrix proteins and their folding disturb the ER homeostasis in activated HSCs 14,15 . To maintain ER homeostasis, the ER starts the UPR pathway through three major stress transducers localized at the ER membrane, including activating transcription factor 6 (ATF6), PKR-like ER kinase (PERK), and the inositol requiring enzyme 1α (IRE1α) 16 . When the adaptive UPR response fails to resolve the ER stress, it triggers cell death 17 . A recent study demonstrated that the UPR is an early non-critical event in the activation of HSCs and its prolonged induction triggers ER stress related cell death 18 .Here we have reported that sorafenib induces non-apoptotic cell death mediated by ER stress which subsequently activates the IRE1α-XBP1s axis of UPR pathway in activated HSCs. Sorafenib also increases reactive oxygen species (ROS) levels in treated cells along with cytoplasmic vacuolation due to ER luminal dilation that are independent of autophagy.
Hepatic fibrosis is a wound healing process characterized by the deposition of extracellular matrix (ECM) proteins such as collagen, around the inflamed or injured liver. Excessive deposition of ECM proteins disrupts the normal hepatic architecture and function, resulting in progression to cirrhosis, the major determinant of morbidity and mortality in chronic liver disease patients. The hepatic stellate cells (HSCs) are the principle cells responsible for hepatic fibrosis that become fibrogenic or activated in response to hepatic injury from a quiescent, non-fibrogenic state. Mechanistically, the quiescent state HSCs lose retinoid containing lipid droplets and become activated and transdifferentiate into myofibroblasts. Activated HSCs start to secrete and deposit ECM proteins, which results in fibrotic scar formation in the injured tissue 2,3 . Deactivation or apoptotic clearance of activated HSCs in the fibrotic liver is the key feature for successful fibrosis resolution after the cessation of tissue damage source. Deactivation of fibrogenic response, or clearance of activated HSCs by inducing cell death or apoptosis is a major therapeutic approach in the development of anti-fibrotic therapy. Some recent studies have highlighted a few promising anti-fibrotic drugs using apoptotic clearance as therapeutic approaches, sorafenib being one of them. Food and Drug Administration (FDA) approved the multikinase inhibitor Sorafenib as a frontline anti-cancer drug for the treatment of advanced human hepatocellular carcinoma (HCC). Sorafenib attenuates the liver fibrosis by reducing HSC proliferation and inducing cell death. Treatment with sorafenib also induces caspase mediated progressive apoptosis in activated HSCs having shrunken and crescent-shaped nuclear morphology. In another study, it was found that a low dose of sorafenib induces autophagic cell death in activated HSCs, whereas its higher dose inhibits autophagy and induces caspase mediated apoptosis, highlighting
# Results
Sorafenib induces dose and duration dependent suppression of viability with increased cytoplasmic vacuolation in hepatic stellate cells. The multikinase inhibitor sorafenib inhibits the fibrogenic activation of HSCs and affects their viability by blocking pro-fibrogenic platelet derived growth factor (PDGF) and transforming growth factor β1 (TGFβ1) receptor mediated signaling. To assess the cytotoxicity of sorafenib in activated human HSC cell line, we treated the LX2 cells with different concentrations of sorafenib , and 15 μM) for 24 h. Microscopic analysis showed that sorafenib treatment induces vacuole formation adjacent to the nucleus within the cytoplasm of LX2 cells. Small cytoplasmic vacuoles started to appear when LX2 cells were treated with 7.5 μM dose of sorafenib for 24 h and became bigger in size with increasing concentration of sorafenib. Interestingly, the nuclei of HSCs also appeared like a crescent-shape or spherical morphology at 7.5 μM and 10 μM dose of sorafenib as compared to intact nuclei in untreated control cells. Then we measured the viability of LX2 cells through flow cytometry analysis using propidium iodide (PI) staining and found that sorafenib decreased cell viability in a dose dependent manner. We further incubated the LX2 cells with a fixed 10 μM dose of sorafenib for various time intervals. The number and size of cytoplasmic vacuoles were increased with increased time duration of sorafenib treatments. These results suggested that the cytoplasmic vacuolation was associated with decreased cell viability and increased duration of treatment. Here we also compared our results using activated rat hepatic stellate cell line, HSC-T6 where 10 μM dose of sorafenib induced cytoplasmic vacuolation at 24 h similar to activated human HSCs (Suppl.. To eliminate any possibility of cytoplasmic lipid droplet accumulation we performed oil red staining in LX2 cells after treatment with 10 μM dose of sorafenib for 24 h. LX2 cells did not show any accumulated lipid droplets within the cytoplasmic vacuoles (Suppl. . All the above evidences suggest that sorafenib induced cytoplasmic vacuolation and cell death in activated HSCs is dose and time dependent.
## Sorafenib induced cytoplasmic vacuolation in lx2 cells coordinates with non-apoptotic cell death.
To investigate the relation between sorafenib induced cytoplasmic vacuolation and cell death in activated HSCs, we pre-treated the LX2 cells with 20 μM caspase inhibitor, Z-VAD-FMK [carbobenzoxy-valylalanyl-aspartyl-(O-methyl)-fluoromethylketone] 60 min prior to the treatment of 10 μM sorafenib. After 24 h of sorafenib treatment we found that the caspase inhibition was unable to rescue LX2 cells from cell death without alteration of cytoplasmic vacuolation . These results suggest the involvement of a caspase-independent non-apoptotic cell death in activated HSCs after sorafenib treatment. To confirm the non-apoptotic mode of cell death in sorafenib treated HSCs, we performed DNA fragmentation assay (DNA ladder assay) using agarose gel electrophoresis as the DNA breakdown is an unique feature of apoptotic cell death. In results, no DNA ladder formation was found as an indication of non-apoptotic cell death in 10 μM sorafenib-treated LX2 cells for 24 h (Suppl. .
To further investigate, whether sorafenib induces autophagy dependent cell death in activated HSCs, we pre-treated LX2 cells with a 25 μM dose of chloroquine (CQ) to inhibit autophagy, prior to treatment of 10 μM sorafenib for 24 h. CQ is an anti-malarial drug that inhibits autophagy by interfering with the fusion of autophagosomes and lysosomes within the cells. If sorafenib induced cell death in HSCs occurred via autophagy, the inhibition of autophagy would rescue the cell death and prolong cellular survival in HSCs. Interestingly, CQ unable to prevent cytoplasmic vacuolation and cellular death. In contrast, it enhanced PI + cells (~ 77% in comparison with ~ 54% with only sorafenib treated cells) after exposure of 10 μM sorafenib for 24 h . At a higher dose of 15 µM, sorafenib further enhanced the population of PI + LX2 cells pre-treated with CQ to ~ 87% compared to ~ 79% when no pre-treatment was done (Suppl. . These results suggest that autophagy inhibition further enhanced the non-apoptotic cell death in sorafenib treated HSCs without affecting the cytoplasmic vacuole formation.
Several anti-cancer compounds such as Gambogic acid (Xanthonoid), and Cyclosporine A stimulate the cytoplasmic vacuolation associated cell death, and display similar morphological features in the target cells as we observed in our study. These compounds triggered the cytoplasmic vacuolation associated cell death mediated with ROS generation and ER stress, which dilate the ER cisternae due to accumulation of misfolded protein in the ER lumen. For further investigation, we pre-treated LX2 cells with 25 μM cycloheximide (CHX), a protein synthesis inhibitor that could reduce the load of protein in ER lumen which may subsequently decrease www.nature.com/scientificreports/ the ER stress and cell death. Interestingly, here we found that the exposure of CHX reduced both cytoplasmic vacuolation as well as cellular death after treatment with either 10 µM or 15 µM sorafenib for 24 h. This suggests a coordination of protein synthesis regulation with sorafenib induced cytoplasmic vacuolation and caspase independent non-apoptotic cell death and Suppl. . The above results provide the clue of the www.nature.com/scientificreports/ cytoplasmic vacuoles that may emerge through misfolded protein accumulation and ER lumen dilation as a result of sorafenib induced ER stress. To investigate the role of ROS in cytoplasmic vacuolation and cell death, we pre-treated LX2 cells with 5 μM human serum albumin (ALB) or 5 μM N-acetylcysteine (NAC) that consist anti-oxidant properties. ALB pre-treatment completely inhibited the cytoplasmic vacuole formation in 10 μM sorafenib treated LX2 cells with reduced cell death to ~ 32%. The increased cell viability of sorafenib treated LX2 cells on pre-treatment with ALB was comparable to the increased viability of CHX pre-treated LX2 cells even at high dose of sorafenib and Suppl. . However, NAC was unable to protect the LX2 cells from sorafenib mediated cytoplasmic vacuolation and cell death. These observations suggest that sorafenib induced cytoplasmic vacuolation directed cell death in LX2 cells partially depends on ROS generation. Based on these results, we conclude that sorafenib induces cytoplasmic vacuolation with possible induction of ER stress along with caspase independent, non-apoptotic cell death in activated HSCs.
## Alterations of endoplasmic reticulum (er) are associated with sorafenib induced cytoplasmic vacuolation and non-apoptotic cell death in hscs.
To investigate whether the sorafenib-induced cytoplasmic vacuolation in activated HSCs through ER dilation, we examined the morphological changes in LX2 cells after the treatment with 10 μM dose of sorafenib for 24 h by Transmission Electron Microscopy (TEM). The untreated control cells showed intact nuclear morphology without dilation of the ER lumen. However, sorafenib treated LX2 cells showed large cytoplasmic vacuoles close to the nucleus . Sorafenib treated LX2 cells also displayed intact nuclear morphology without chromatin condensation, nuclear fragmentation or plasma membrane blebbing; all features being a hallmark of non-apoptotic cell death. In fact, the vacuoles were surrounded by membranes, some of which were decorated with ribosomes, indicating the chance of intracellular vacuolation from the rough ER (Suppl.. In addition, the vacuoles were surrounded by single layered membrane, and some LX2 cells showed bigger sized vacuoles close to the nucleus arising through ER lumen www.nature.com/scientificreports/ dilation and Suppl.. To further confirm, we stained the ER of both the untreated and 10 μM sorafenib-treated LX2 cells with the ER tracker dye that binds to sulphonylurea receptor of ATP-sensitive K + channel present on the surface of ER. We observed clear cytoplasmic vacuoles close to the nucleus in only sorafenib treated LX2 cells . The dilation of the lumen caused the dispersion of the dye throughout the ER. In this context, we also checked the expression of calcium-binding chaperon, calreticulin which is present in the lumen of the ER 26 . 10 μM sorafenib treatment enhanced calreticulin expression after 12 h in both cytoplasm and nucleus , suggesting the initiation of aggravated structural disorder of ER and nuclear translocation of calreticulin. Similarly in rat HSC-T6 cells, the calreticulin expression were also elevated in both cytoplasm and nucleus after 10 μM sorafenib treatment at 12 h (Suppl.. These findings confirmed that sorafenib induced cytoplasmic vacuolation mediated non-apoptotic cell death in HSCs are associated with ER dilation.
Dose dependent influence of sorafenib on LC3 signaling is not associated with cytoplasmic vacuole formation. Current literature and our findings based on cytopathological characteristics suggest that the cell death associated with cytoplasmic vacuolation is predominantly due to ER stress and lack of caspase activation. The induction of cytoplasmic vacuolation mediated non-apoptotic and non-autophagic death was reported in several cancers with a mechanism involving ER stress and LC3 (microtubule-associated protein 1 light chain 3). Alterations in the biochemical nature and subcellular localization of LC3s correlate with autophagy and are used as surrogate markers for its quantification. LC3s (MAP1-LC3A, B, and C) are structural proteins of autophagosomal membranes. While LC3A has been reported to show nuclear and perinuclear localization, LC3B was uniformly distributed throughout the cytoplasm. To investigate the role of autophagy in ER stress and cytoplasmic vacuole formation, we examined the LC3B localization in the cytoplasm of LX2 cells through immunofluorescence (IF) study along with the distribution of hepatic stellate cell activation marker (αSMA) after treatment with low or high dose of sorafenib. At low dose (5 µM) of sorafenib treatment for 12 h, we observed an increased expression of LC3B in LX2 cells with reduced expression of αSMA. Whereas, higher dose (10 µM) of sorafenib for 12 h suppressed the expression of both LC3B and αSMA. These results suggest that autophagic regulation was not involved with the ER stress mediated cytoplasmic vacuolation, and a higher dose of sorafenib bypasses the requirement of autophagy for inducing cell death in activated HSCs.
To further explain the results, we performed western blotting with low (5 µM) and high (10 µM) dose of sorafenib treated LX2 cells with or without inhibiting ATG5 by siRNA. ATG5 is a critical and indispensable protein for vesicle formation during autophagy. To inhibit autophagy, we inactivated ATG5 in LX2 cells by pre-incubating with 100 nM of ATG5 siRNA prior to sorafenib treatment. Then we examined the alteration of autophagic flux in 5 µM sorafenib treated LX2 cells after inactivation of ATG5 compared to wild type ATG5 sorafenib treated LX2 cells. We observed a similar ratio of LC3BI to LC3BII in control cells with respective to 5 µM sorafenib treated cells, possibly due to delayed autophagosome turn over that accumulated and enhanced LC3BII expression. Similarly we observed some autophagic vacuoles in the TEM images of untreated LX2 cells (Suppl., may be to maintain the cellular homeostasis because autophagy also plays a vital role in fibrogenic responses of activated HSCs 32 . In contrast, when LX2 cells were treated with 10 µM dose of sorafenib for 12 h, the LC3BI expression was reduced, which was comparable to the expression of LC3BI in LX2 cells treated with a lower dose of sorafenib (5 µM) following ATG5 inactivation. Together with the above findings, we conclude that a high dose of sorafenib (10 µM) inhibits autophagy and mediates non-autophagic cell death in activated HSCs. These results were also in concordance with our previous findings with autophagic inhibitors, where we showed that CQ was unable to suppress the cytoplasmic vacuolation and non-apoptotic cell death induced by 10 µM sorafenib in LX2 cells .
## Ros is critically involved in sorafenib induced er stress but cannot alone influence the cytoplasmic vacuolation mediated cell death.
Various studies have reported that ROS-mediated ER stress play a critical role in sorafenib induced cell death in various cancer types. Based on these findings, we predicted that ROS-mediated ER stress can play a critical role in sorafenib induced cytoplasmic vacuolation and cell death in activated HSCs. Therefore, we analysed the intracellular ROS that labeled with 2′,7′-dichlorodihydrofluorescein diacetate (H 2 DCFDA) fluorescent signals and quantified by flow cytometry analysis. We observed that the production of ROS was significantly increased in LX2 cells after treatment with 10 µM sorafenib for 24 h. To determine whether the increased intracellular ROS levels mediate the sorafenib induced ER stress in LX2 cells, we inhibited ROS production by pre-treating LX2 cells with anti-oxidants NAC and ALB prior to 10 μM dose of sorafenib treatment. ALB and NAC inhibited the sorafenib induced ROS generation as indicated by a decrease in H 2 DCFDA positive cell population to ~ 11% and ~ 6% respectively compared to ~ 36% H2DCFDA positive cells in LX2 cells treated with sorafenib alone. As mentioned above the NAC pre-treated LX2 cells showed no change in cytoplasmic vacuolation and cell death after sorafenib treatment . On the other hand, ALB pre-treatment reduced both the sorafenib induced ROS production, cytoplasmic vacuolation, and cell death in LX2 cells after 10 µM of sorafenib treatment for 24 h (Figs. 2 and 5). These observations suggest the indirect involvement of ROS in sorafenib induced cytoplasmic vacuolation mediated cell death. Interestingly, CHX pre-treatment also resulted a similar ROS suppression effect as observed with ALB, however CHX was unable to completely rescue vacuole formation as seen with ALB pre-treated HSCs. LX2 cells pre-treated with caspase inhibitor vZAD-FMK showed no alteration in sorafenib induced ROS production, cytoplasmic vacuolation as well as cell death (Figs. 2 and 5). Here, we conclude that sorafenib mediated ROS generation induced ER dilation that subsequently results in cytoplasmic vacuolation and triggers non-apoptotic cell death in activated HSCs. www.nature.com/scientificreports/ Sorafenib induces non-apoptotic cell death in activated HSCs through ROS, ER stress, and UPR pathway. Excessive ROS production can generate oxidative stress which further triggers the ER stress. It may lead to accumulation of large amounts of unfolded or misfolded proteins in the ER lumen and initiate the cellular ER stress response known as un-coupled protein response (UPR) pathway. In this context, we evaluated the gene expression of ROS generating enzymes such as nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 1 (NOX1), NADP oxidase 4 (NOX4), NADPH Oxidase Activator 1 (NOXA1), Cytochrome B-245 Alpha Chain (CYBA), and Flavin-containing monooxygenase 2 (FMO2). We found that sorafenib induced oxidative stress in LX2 cells by upregulating the mRNA expression of NOX1, NOX4, NOXA1, CYBA, and FMO2 in a dose dependent manner. LX2 cells showed maximum expression of these genes when treated with 10 μM sorafenib dose for 24 h. Pre-treatment with CHX, NAC, and ALB suppressed the www.nature.com/scientificreports/ increased mRNA expression of ROS generating enzymes in LX2 cells after 10 µM sorafenib treatment for 24 h. These results suggest the involvement of ROS signals in sorafenib induced cytoplasmic vacuolation and cell death. However, the anti-oxidant NAC was unable to supress the sorafenib induced cytoplasmic vacuolation mediated non-apoptotic cell death.
## Scientific reports
To investigate the possible ER stress mediated signaling, we quantified the mRNA expression of ER stress or UPR pathway markers such as Binding immunoglobulin protein/(GPR78/BiP), inositol-requiring enzyme 1 (IRE1α), PKR-like ER kinase (PERK), X-box-binding protein 1 (XBP1), and C/EBP Homologous Protein (CHOP) in LX2 cells on treatment with different concentration of sorafenib and different time durations with 10 μM sorafenib. We found that sorafenib enhanced the mRNA expression of UPR markers in a dose and time dependent manner. LX2 cells treated with the 10 μM sorafenib dose for 24 h showed the highest expression of UPR genes. However, the sorafenib induced upregulation of UPR genes in LX2 cells was attenuated on pre-treatment with CHX, NAC and ALB.
Upon ER stress, IRE1/endoRNAse activity regulates the expression of the transcription factor cleaved XBP1 (XBP1s). Here the mRNA level of XBP1 showed a steady increase significantly from 12 h time point after 10 μM sorafenib treatment, determining a possible involvement of IRE1α-XBP1s axis to induce ER stress. www.nature.com/scientificreports/ Dose dependent sorafenib induced UPR is associated with functional activation of the IRE1α-XBP1s axis. To further confirm the role of the IRE1α-XBP1s axis of the UPR to ER stress, we analysed the protein expression patterns of related genes through western blot. It was found that the IRE1α were significantly overexpressed in LX2 cells on treatment with 10 μM sorafenib at both the 12 h and 24 h time points. During ER stress IRE1α activates the endoribonuclease domain, which primarily acts through XBP1. Here the protein levels of XBP1s were also enhanced in LX2 cells after treatment with 10 μM sorafenib for 12 h. Calreticulin and GRP78 (BiP) chaperon proteins that bind to misfolded or un-folded proteins were also upregulated after sorafenib treatment for 12 h. As per literature IRE1α autophosphorylation enhanced further oligomerization of the protein to stimulate RNase activity. In contrast, sometimes IRE1α can bypass its autophosphorylation to cleave XBP1 for activation of the UPR pathway. In our results we found that phosphorylation of IRE1α (pIRE1α) were unaltered in LX2 cells without or with 10 μM sorafenib at both 12 and 24 h of treatment.
To further explain the effect of ER stress on the cytoplasmic vacuolation-mediated cell death, we blocked the ER stress signalling using 3-Ethoxy 5, 6-dibromosalicylaldehyde (EDBS). EDBS is a non-competitive reversible inhibitor that binds specifically to the IRE1α protein to inactivate its endoribonuclease activity. To inhibit the IRE1α protein we pre-treated the LX2 cells with 10 μM EDBS for 60 min prior to exposure of sorafenib for 12 h and 24 h. EDBS pre-treatment decreased the expression of the GRP78, IRE1α, XBP1s and calreticulin after 12 h of 10 μM sorafenib treatment. There was a very marginal reduction of basal level pIRE1α in EDBS pre-treated LX2 cells after sorafenib treatment for 12 h. We again confirmed the overexpression of IRE1α and its suppression through pre-treated EDBS after 10 μM sorafenib treatment in rat HSC-T6 cells for both 12 h and 24 h (Suppl., suggesting the involvement of IRE1α mediated UPR pathway in sorafenib induced ER stress.
Most strikingly, when we pre-treated human HSCs, LX2 cells and rat HSC-T6 with EDBS there was significant reduction in cytoplasmic vacuolation after 12 h of 10 μM sorafenib treatment, whereas when sorafenib Relative protein ratios (normalized with loading control) were shown in a plot graph. Data represent mean ± s.d. from three independent experiments (ns > 0.05, *P < 0.05; **P < 0.01, ***P < 0.001 One-way analysis of variance). www.nature.com/scientificreports/ treatment was extended to 24 h, the vacuolation was delayedand Suppl.. Here, we also found that EDBS not only effectively suppressed the cytoplasmic vacuolation but also drastically reduced the population of PI + LX2 cells after sorafenib treatment at both the 12 h and 24 h time pointsThese results demonstrate the induction of ER stress with the involvement of IRE1α-XBP1s axis during sorafenib induced cytoplasmic vacuolation-mediated non-apoptotic cell death in activated HSCs.
# Discussion
As the central effector of liver fibrosis, activated HSCs have been the focus of many studies examining mechanisms underlying the disease. The conception of activated HSCs as a target for the treatment of liver fibrosis has stimulated the investigation of pathways that promote HSC apoptosis, as a means to facilitate disease www.nature.com/scientificreports/ regression. Various anti-fibrotic agents have been identified such as gliotoxin, sulfasalazine, and tectorigenin that target the activated HSC cell survival and proliferation, ultimately inducing cell death to limit the fibrogenic activity of HSCs. Sorafenib has also been reported to have anti-fibrotic effects by limiting cell proliferation and inducing apoptosis in activated HSCs that leads to fibrosis regression. Originally, sorafenib is a frontline anti-cancer drug that is used for treatment of advance HCC. Mechanistically, sorafenib blocks vascular endothelial growth factor (VEGF) and platelet derived growth factor (PDGF) receptors to suppress tumour angiogenesis or inhibit MAP Kinase pathways to suppress tumour cell proliferation. The suppression of proliferation and induction of apoptosis are accompanied by a down regulation of cyclins and cyclin dependent kinases (Cdks). Jiang et al. showed that HSCs treated with sorafenib exhibited shrunken chromatin that was aggregated and condensed inside the nuclear membrane, with crescent-shaped or spherical nuclear morphology. Other studies have reported that sorafenib reduced proliferation, induced autophagy and apoptosis in HSCs 8 . Sorafenib induced autophagy and apoptosis in HSCs have been shown to interlink through mechanisms of cross-talk.
In our study, we highlighted a new mechanism of ER stress induced autophagy independent non-apoptotic cell death in activated HSCs after the treatment with sorafenib. We found that sorafenib induces cytoplasmic vacuolation adjacent to the nucleus in activated HSCs and subsequently cell death. Sorafenib induced vacuolations become bigger with increased dose and time duration of treatment. However, pre-treatment with CHX, rescued the activated HSCs from the sorafenib induced effects. CHX treatment halts the synthesis of proteins and their subsequent accumulation in the ER lumen eventually rescues the HSCs from ER stress. Based on these results we reasoned that cytoplasmic vacuolation after the sorafenib treatment are dilated ER cisternae. TEM analysis and confocal analysis of sorafenib treated HSCs with ER tracker dye and calreticulin expression further confirmed the results. Interestingly, we found a basal autophagic flux in activated human HSCs and LX2 cells by TEM analysis and confirmed with LC3B protein expression through immunoblotting and confocal microscopy. As autophagy is essential for cellular homeostasis, some basal autophagy is present in activated HSCs. Moreover autophagy is a critical event for the induction of fibrogenic response. It is rapidly up-regulated as an adaptive response under a variety of cellular stress conditions including nutrient deprivation, oxidative stress, and infections.
Friedman et al. showed the induction of autophagy in hepatic stellate cells in Carbon tetrachloride (CCl 4 ) and Thioacetamide (TAA) induced liver injury model. They suggested that autophagy fulfills the high energy demand required to initiate and maintain the stellate cell activation by liberation of free fatty acid (FFA), lipid droplet (LD) mobilization, and mitochondrial-oxidation. Autophagy is also involved in a cell death process called as autophagic cell death that differs from apoptosis in the presence of characteristic autophagosomes and autophagolysosomes within the dying cells. Consistent with previous studies, we found that a low concentration of sorafenib for short duration of treatment induces autophagy, however at high concentrations and longer durations of sorafenib treatment inhibits autophagy. Interestingly, CQ treatment did not rescue the HSCs from the cytoplasmic vacuolation mediated cell death. The vacuoles start to appear only after treating activated HSCs with a higher dose of sorafenib i.e. 7.5 µM for 24 h. This indicated that the cytoplasmic vacuoles are not autophagic vacuoles, and a higher dose of sorafenib treatment bypasses autophagic cell death to cytoplasmic vacuolation mediated cell death. Studies also reported that cell death processes switched from autophagic to apoptotic and non-apoptotic depending upon whether the exposure to stimuli was extended to longer durations or the drug concentration was increased. Many studies have already demonstrated that sorafenib inhibits the proliferative activity of activated HSCs through caspase mediated apoptosis. Surprisingly, when we pre-treated LX2 cells with caspase inhibitor before the sorafenib treatment, we found that caspase inhibition was unable to rescue from their cell death without effecting vacuole formation. During flow cytometry analysis with annexin-PI, we found lesser number of early apoptotic and late apoptotic LX2 cells at 12 h or 24 h of 10 μM sorafenib treatment. Thus, previous findings together with our current results confirm that the viability of LX2 cells decreases due to cytoplasmic vacuolation mediated non apoptotic cell death upon sorafenib treatment depending on the dose and time . The cytoplasmic vacuole formation in HSCs at higher dose of sorafenib treatment is caspase and autophagy independent.
The role of ER stress has been studied in a variety of diseases including liver fibrosis 52 . Kim's group demonstrated the effects of ER stress on the activation of HSCs 53 . Similar to other reports we have highlighted the sorafenib induced ROS mediated ER stress and the accumulation of misfolded proteins in the ER lumen resulting in the ER dilation 12 . To overcome this stress or to restore the normal ER function, the ER starts the UPR pathway to avert the ER stress or induce cell death if stimuli persists. In our study, we have shown that dose dependent influence of sorafenib induces the gene expression markers of ER stress such as XBP1, CHOP, GPR78 (BiP), IRE1α, PERK, and oxidative stress markers such as NOX1, NOX4 NOXA1, CYBA, and FMO2. However, CHX, anti-oxidant NAC, and ALB pre-treated HSCs showed decreased expression of both ER stress and oxidative stress markers after sorafenib treatment. We observed that ALB completely abolished the cytoplasmic vacuole formation and rescued the HSCs from sorafenib induced cell death but we could not find similar results with using NAC. XBP1 showed a steady increase in mRNA expression with sorafenib treatment, further indicating the involvement of the IRE1α-XBP1s axis of the UPR. We evaluated also the expression of pIRE1α in sorafenib treated HSCs and compared with IRE1α through immunoblot analysis. The expression of pIRE1α was not significantly enhanced after either treatment of sorafenib as compared with control. EDBS showed no effect and was unable to decrease the pIRE1α expression either with or without sorafenib treatment in HSCs. On the other hand, EDBS inhibited the overexpressed IRE1α, GRP78 and XBP1s which eventually enhanced the cell viability in sorafenib treated HSCs indicating sorafenib induced cytoplasmic vacuolation through the IRE1α-XBP1s UPR axis.
Thus, the present study delivers unique insights into the anti-fibrotic effects of sorafenib treatment in activated HSCs, and highlights the complex interplay between ER stress and cell death pathways. This study provides evidence for a new mechanism of sorafenib action in activated HSCs within the liver fibrosis microenvironment. Further investigation of the molecular mechanisms underlying sorafenib induced cytoplasmic vacuolation www.nature.com/scientificreports/ mediated non-apoptotic cell death may lead to the development of a novel therapeutic approach for the more effective management of liver fibrosis. For immunofluorescence, cells were fixed with 4% paraformaldehyde (PFA) after the sorafenib treatment for 10 min and washed with 1 × PBS three times. Next, cells were incubated with blocking reagent (5% bovine serum albumin, 0.3% Tritron X-100 in PBS) for 60 min for blocking. Antibodies were diluted in antibody dilution buffer (1% bovine serum albumin, 0.15% Tritron X-100 in PBS). Next, after the blocking, to determine the LC3B and α-SMA expression in LX2 cells, we incubated with primary antibody LC3B and α-SMA along with respective fluorescence-tagged secondary antibody mentioned in Supplementary informations. Then, cover clip was mounted with VECTASHIELD antifade mounting medium with DAPI (#H-1200). Confocal images were taken using Advanced Nikon A1 confocal microscope at Amity University Uttar Pradesh, India and at Advanced Technology Platform Centre (ATPC), Regional Centre for Biotechnology, Haryana, India.
# Methods
Small-interfering RNA (siRNA) transfection. Small-interfering RNA (siRNA) against ATG5 and nonspecific scrambled siRNA were purchased from Dharmacon. LX2 cells were cultured in 6 well plates. Lipofectamine 2000 (11668-027; Invitrogen, San Diego, CA, USA) was mixed with serum free DMEM containing 100 nM siRNA or scrabbled siRNA final concentration and incubated for 20 min at room temperature. Transfection mixture were incubated on cells at 37 °C in 5% CO2 for 6 h in serum free routine cell culture media. Experiments were performed after the 3 days of siRNA transfection.
## Measurement of reactive oxygen species (ros).
To measure intracellular ROS production, 0.5 × 10 6 cells were seeded in 60 mm cell culture dish. We performed the experiments after over-night attachment of cells.
We detached the cells with trypsin/EDTA and washed with 1 × PBS buffer to remove cellular debris. We incubated cells with 5 µM 2′,7′-dichlorodihydrofluorescein diacetate (H 2 DCFDA) for 30 min in the dark, washed with 1 × PBS buffer and further processed for flow cytometry analysis using BD FACS Calibur flow cytometer (BD Biosciences). All experiments were performed in triplicate. Data were analysed using FlowJo software (BD Biosciences). www.nature.com/scientificreports/ Real-time PCR was performed using Powerup SYBR green master mix (A25742 Applied Biosystems, Waltham, MA, USA). The copy number of the target mRNA in each sample was normalized as a ratio using the copy number for 18S rRNA in the denominator.
## Rna isolation and quantitative polymerase chain reaction (qpcr
## Western blot analyses. treated or untreated cells were washed with pbs and homogenized in ripa lysis
buffer in presence of 1 × protease inhibitor (11697498001; Roche, St. Louis, Missouri, USA) and 1 × phosphatase inhibitor (4906845001; Roche, St. Louis, Missouri, USA). Protein concentrations were determined using Bradford protein assay (20279; Thermo Fisher Scientific, Waltham, MA, USA). 30 µg protein lysates were separated by 12% (w/v) SDS-PAGE, and proteins were transferred to PVDF membrane (1620177; BioRad, CA, USA). Membrane were incubated with primary antibody for overnight at 4 °C with gently shaking. Secondary ant-rabbit or anti-mouse was incubated for 2 h and visualized using an Enhanced Chemiluminescence (ECL) detection kit (34094; Thermo Fisher Scientific, Waltham, MA, USA). Details of primary and secondary antibodies mentioned in . For statistical analyses and densitometry analyses was measured using prism and ImageJ software.
Statistical analysis. All data were presented as mean ± SD (standard deviation) from at least three separate experiments. Student's t test was applied to evaluate the differences between treated and control groups. Data from multiple groups were analyzed by one-way or two way ANOVA using Prism-GraphPad. For all the tests, the level of significance was values of P < 0.05.
Received: 14 April 2020; Accepted: 19 January 2021 |
Neural Pattern Classification Tracks Transfer-Appropriate Processing in Episodic Memory
The transfer-appropriate processing (TAP) account holds that episodic memory depends on the overlap between encoding and retrieval processing. In the current study, we employed multivariate pattern analysis (MVPA) of electroencephalography to examine the relevance of spontaneously engaged visual processing during encoding for later retrieval. Human participants encoded word-picture associations, where the picture could be a famous face, a landmark, or an object. At test, we manipulated the retrieval demands by asking participants to retrieve either visual or verbal information about the pictures. MVPA revealed classification between picture categories during early perceptual stages of encoding (ϳ170 ms). Importantly, these visual category-specific neural patterns were predictive of later episodic remembering, but the direction of the relationship was contingent on the particular retrieval demand of the memory task: a benefit for the visual and a cost for the verbal. A reinstatement of the category-specific neural patterns established during encoding was observed during retrieval, and again the relationship with behavior varied with retrieval demands. Reactivation of visual representations during retrieval was associated with better memory in the visual task, but with lower performance in the verbal task. Our findings support and extend the TAP account by demonstrating that processing of particular aspects during memory formation can also have detrimental effects on later episodic remembering when other aspects of the event are called-for and shed new light on encoding and retrieval interactions in episodic memory.
# Introduction
Episodic memory allows us to revisit the past and to re-experience previous events from a first-person perspective . Such re-experiencing is consid-ered to be mediated by the reinstatement of the brain activity that was present during processing of the original event [bib_ref] Simple memory: a theory for archicortex, Marr [/bib_ref]. The contents of memory depend on what type of information was at the focus of attention during encoding, which explains how different people may remember the same event differently and how the relevance of a memory for guiding future behavior depends on whether the encoded information is applicable in times to come. The current study used a novel approach to investigate the relationship between processing engaged during encoding and later retrieval demands. We employed multivariate pattern analysis (MVPA) of oscillatory brain activity to examine the relevance of spontaneously engaged category-specific visual processing during encoding for later memory retrieval. By manipulating retrieval demands over two memory tasks, we here demonstrate that the particular attentional focus adopted during the encoding of an event can both facilitate and hinder performance on subsequent tests of memory for that event.
The transfer-appropriate processing (TAP) account holds that the likelihood of successful episodic memory depends on the extent to which the processing engaged by a retrieval cue overlaps with that engaged at encoding [bib_ref] Modeling hippocampal and neocortical contributions to recognition memory: a complementarylearning-systems approach, Mccarthy [/bib_ref]. A successful retrieval cue is thought to trigger the reinstatement of the cortical patterns active at the time of encoding [bib_ref] Simple memory: a theory for archicortex, Marr [/bib_ref]. A growing body of research has consistently provided evidence for this notion by showing that successful retrieval co-varies with the replay of the encoding-related brain patterns at the time of retrieval [bib_ref] Modeling hippocampal and neocortical contributions to recognition memory: a complementarylearning-systems approach, Mccarthy [/bib_ref] [bib_ref] Neurophysiological evidence for the time course of activation of global shape, part,..., Rossion [/bib_ref]. Furthermore, previous studies have shown that encoding-related brain activity (i.e., predictive of subsequent memory) is shaped both by the type of processing occurring at encoding and by the overlap between encoding and retrieval processes [bib_ref] Brain oscillatory subsequent memory effects differ in power and long-range synchronization between..., Bauch [/bib_ref] [bib_ref] Goal-directed mechanisms that constrain retrieval predict subsequent memory for new ЉfoilЉ information, Tulving [/bib_ref]. The typical procedure in these previous studies has been to investigate encoding-related brain activity when explicitly instructing participants to attend to particular attributes during the encoding and retrieval. In the current study, participants were not directed to process specific aspects of the stimuli, instead we used measures of brain activity to track their spontaneously adopted attentional focus during encoding and examine the extent to which processing during encoding matched later retrieval demands.
Participants learned paired-associates formed by a word and a picture from one of three different categories (famous faces, landmarks, and objects). Retrieval demand was manipulated in two memory tasks in which participants were cued by the word and instructed to recall either visual (orientation) or verbal (name) information about the associated picture. Previous work involving electroencephalography-based MVPA decoding of picture categories has shown that classification relies on early brain activity related to low-level visual processing involving category-specialized brain regions . This allowed us to use classification accuracy as a proxy for spontaneously engaged visual processing during picture encoding. Thus, high classification accuracy in our MVPA approach indicates attention directed toward visual features of the pictures. To examine how categoryspecific visual processing during encoding transfers to and matches with later retrieval demands, the accuracy of the pattern classifiers was related to performance in the two memory tasks. Moreover, the classifiers established at encoding were used to decode the oscillatory brain activity during retrieval, when only the word cue was presented, which allowed us to examine cortical reinstatement and its functional significance depending on retrieval demand. In this novel paradigm, we expected TAP to be revealed in a positive relationship between classification accuracy and episodic remembering in the visual rather than in the verbal memory task, and further that the replay of encoding brain activity would occur and co-vary with performance in the visual memory task exclusively.
# Materials and methods
## Participants
Thirty-six participants took part in the study in exchange for a movie ticket. Eighteen participants took part in the visual memory task (six males, average 23 years old, range 20 -28) and the remaining 18 took part in the verbal memory task (six males, average 24 years old, range 21-40). An independent-sample t test confirmed that the two groups of participants were matched in age (t (34) ϭ 1.13, p Ͼ 0.25). All participants were right-handed, native Swedish speakers, and reported no history of neurologic diseases. The study was conducted in accordance with the Swedish Act concerning the Ethical Review of . Participants gave their written informed consent, and the study followed the local ethical guidelines at Lund University. axe). All photographs were converted to a black-andwhite format with 600 ϫ 600 pixels and with a resolution of 72 pixels/inch. Each word was assigned to one picture from each category, with no obvious pre-experimental association between word and pictures. Three stimulus sets were created that counterbalanced the word-picture selection across participants.
## Experimental design and procedures
The memory task comprised eight blocks, each including a study phase, a distractor task and a test phase. In each block, participants were asked to memorize 24 paired-associates composed of a cue word and a target picture that could be either a face, a landmark or an object. In the visual memory task, participants were instructed to retrieve the semantic category of the picture, followed by a forced choice about two versions (original vs mirrored) of the picture that was studied with the word. Participants indicated their response by pressing a response button. In the verbal memory task, participants were instructed to verbally retrieve the semantic category and the name of the picture studied with each word. Except for the retrieval demand differences in the test phase, the encoding phase was identical in the two tasks [fig_ref] Figure 1: A, Trial structure during study [/fig_ref]. Each encoding trial started with a 1-s fixation cross followed by the presentation of the word cue in red for 2.5 s. Next, the target image was presented for 2.5 s, preceded and followed by a 1-s fixation period. Immediately thereafter, the target picture appeared with the red word cue presented on top and remained in the screen for another 2.5 s. Each trial ended with participants rating how easy it was to associate the word cue with the target picture (1 ϭ very difficult, 2 ϭ OK, 3 ϭ very easy). The trials were randomly presented with the constraint that consecutive trials were from different categories.
To eliminate active rehearsal of the last associate pair, an arithmetic task separated the study from the test phase. At test, each trial started with a fixation period of 1 s followed by the display of the word cue for 3 s [fig_ref] Figure 1: A, Trial structure during study [/fig_ref]. In the visual memory task, participants selected the category of the target picture associated with the word cue (1 ϭ face, 2 ϭ landmark, 3 ϭ object, or 4 ϭ don't know). If they were correct, participants were asked to make a forced-choice decision about which version of the picture was presented during encoding (1 ϭ left and 2 ϭ right). The assignment of the buttons to category and target picture selection was counterbalanced across participants. In the verbal memory task, participants were asked to verbally retrieve the semantic category and the name of the exemplar in the target picture (e.g., Big Ben) associated with the word cue. To avoid muscle artifacts in the electroencephalogram (EEG) recordings, participants were instructed to withhold their response during the presentation of the word cue. If they did not remember the exact name, participants were encouraged to tell everything they remembered about the picture (e.g., "the clock in London"). In both tasks, each trial was separated by an interstimulus interval of 0.5 s.
To ensure that participants were familiar with all the stimulus material, previous to the EEG recordings, participants were familiarized with each picture and respective name. Each image was displayed on screen for 2.5 s, followed by its name for 1 s and participants were asked to rate how familiar they were with each of the identities represented by the picture using a five-point scale (1 ϭ not familiar and 5 ϭ very familiar). Notice that study phase was identical in the visual and the verbal memory tasks. The classifiers were trained and tested for decoding the picture category (face, landmark or object) based on the EEG TFRs (4 -45 Hz) at different time bins when the picture was presented alone (outlined in green). B, Trial structure during retrieval for the visual and the verbal memory tasks. Replay of the encoding category-specific neural patterns was examined during the presentation of the word cue (outlined in green). C, Exemplars of the stimuli used in the paradigm.
## Eeg recording
The EEG was recorded continuously using a Neuroscan (Compumedics) Grael amplifier (2048-Hz sampling rate; left mastoid reference) from 31 Ag/AgCl scalp electrodes mounted in an elastic cap and positioned according to the extended 10 -20 system. The montage included five midline electrode sites and 13 sites over each hemisphere (FP1/FP2, F7/F8, F3/F4, FC5/FC6, FC1/ FC2, T7/T8, C3/C4, CP5/CP6, CP1/CP2, P7/P8, P3/P4, PO9/PO10, and O1/O 2 ). Additional electrodes were used as ground (AFz), reference sites (mastoids), and for recording the electrooculogram (EOG). EOG electrodes were placed below the left eye and at the left and right outer canthi.
## Eeg preprocessing and data preparation
The EEG data were preprocessed using FieldTrip . Offline, the data were downsampled to 512 Hz, and divided into epochs of 4 s, from -1.5 to 2.5 s relative to both the onset of the word cue and target image in the study phase and relative to the word cue in the test phase. The data were transformed to a linkedmastoid reference, and a baseline correction was applied (subtraction by the average amplitude of the epoch; as in . Bipolar EOG was computed using the FP1 and the electrodes placed vertically and horizontally around the eyes. EEG epochs were physically inspected and those containing muscle or other artifacts, not related to blinks and horizontal eye movements, were manually removed. Independent components analysis was conducted and components representing blink and other oculomotor artifacts clearly distinct from EEG were removed and bad channels (if any) were interpolated. The data were thereafter again visually inspected and trials with residual artifacts were manually excluded.
The signals from individual trials were transformed into time-frequency representations (TFRs). Brain oscillatory activity from the low range to the high ␥ frequency represents core mechanisms of episodic memory [bib_ref] Brain oscillatory subsequent memory effects differ in power and long-range synchronization between..., Bauch [/bib_ref] and have been previously used for training pattern classifiers to distinguish early visual encoding brain representations correspondent to faces and visual scenes . TFRs were obtained for frequencies ranging from 4 to 45 Hz, with a frequency step of 1 Hz, a time step of 0.01 s, and a wavelet width of five cycles, using the complex Morlet wavelet transform as implemented in Field-Trip.
## Multivariate pattern classification and statistical analysis
Classification analysis was used to build pattern classifiers that could distinguish the encoding brain activity related to the three picture categories (faces, landmarks, and objects). MVPA was performed using support vector machine (SVM), with a linear kernel, and a one-against-all strategy, as implemented in the MATLAB bioinformatics toolbox. The pattern classifiers were trained on the TFR elicited when the target picture was shown alone during the encoding phase [fig_ref] Figure 1: A, Trial structure during study [/fig_ref]. Twenty different classifiers were trained separately for each participant. The 20 classifiers were trained using 20 different time bins that cov-ered a time period from -45 to 920 ms after picture onset. The 20 time bins, each with a duration of 39 ms (spanning over five time points), were centered at - , and 902 relative to stimulus onset, to cover the whole epoch. Each of the classifiers used the TFR in the 31 EEG channels and in five time points within each time bin. Thus, for each classifier, there were 6510 possible features (31 channels ϫ 42 frequencies ϫ five time points). No additional baseline correction was performed on the TFR and instead the power at each timepoint, frequency, and channel was normalized across trials (as in . Each classification used a ten-fold cross validation, that is, the data were randomized and partitioned into ten rough equal-sized subsets, over which ten training-test iterations were performed. Each partition was used as the test set exactly once, with the remaining nine partitions used for training the classifier in that fold. In the visual memory task, classification was performed using the TFR signal of an average of 62 trials corresponding to faces (range 60 -64), 62 corresponding to landmarks (range 59 -64), and 62 corresponding to objects (range 58 -64). In the verbal memory task, analysis was performed on an average of 63 trials corresponding to faces (range 60 -64), 62 corresponding to landmarks (range 57-64), and 62 corresponding to objects (range 57-64). Before the classification, in each cross-validation iteration, a feature-selection step was performed by calculating a univariate statistical test across the training subset (excluding the test subset) on spectral power at each frequency, time point, and channel that constituted the features of the classifier. The features that were found to be significantly different between categories using a one-way ANOVA (p Ͻ 0.05) were selected for classifier training and z-transformed. In each cross-validation iteration, the model was used to predict the category of the left-out trials (i.e., test subset). Thus, the classification accuracy here reported represents the performance of the classifier averaged over categories (faces, landmarks, and scenes), cross-validation iterations, and participants. Classification performance for target picture was contrasted against classification performance for word cue at study. During word cue presentation, the pattern classifiers do not carry information about the specific stimulus category, and the word cue therefore offers a perfect baseline control condition. The data for this baseline classification analysis was preprocessed and treated in the same way as described above for the target picture classification. To account for multi-comparisons problem the significance level of each test was Bonferroni corrected (corrected p value: 0.05/20 classifiers ϭ 0.0025).
The pattern classifiers built during encoding were used to predict retrieval without any tuning to optimize crossvalidation performance. The testing was performed at 20 separate time bins centered at -25, , and 902 relative to word cue onset (the same time bins as for the encoding phase). The classification accuracy was calculated in relation to the category of the target picture (i.e., the to-be-retrieved picture). We looked at replay in trials for which participants successfully re-trieved the target. To generate the statistical significance of the replay, a permutation test procedure, with 500 iterations, was used to generate the null distribution for the significance thresholds. In each iteration, the labels for the stimulus category were shuffled. Thus, each iteration yielded a distribution that contained no true information about the category of the picture but preserved overall smoothness and other statistical properties. At each iteration a one-sample t test comparing classification performance against chance (33.3%) was conducted. The distribution of the t tests formed the nonparametric empirical null distribution and the 95th percentile of this distribution is reported as the significance threshold. All study (target picture onset) ϫ retrieval (word cue onset) maps of classification performance were smoothed by a 2-D Gaussian kernel spatial filter with a of 1 for display and for calculating the t test of the non-parametric distribution.
To investigate the contribution of each channel for classification accuracy, we re-run the classification training at study and the category prediction at test using one channel and its neighbors at a time and storing the classification performance at the center channel. to compare the classification topography between the two tasks, the normalized mean classification at each channel was compared between the visual and the verbal task. Bonferroni correction was performed for correction of multiple comparisons (corrected p value: 0.05/31 channels ϭ 0.0016).
## Relationship between pattern classifier accuracy and retrieval demands
The relationship between classifier accuracy during encoding and episodic remembering was tested using the Robust Correlation Toolbox in MATLAB (http://sourceforge.net/projects/ robustcorrtool/). First, we correlated the highest classifier accuracy during encoding with memory performance in the verbal and visual memory tasks. We calculated the Person correlation and the robust 95% confidence intervals (CIs) which are computed by bootstrapping the data after removing outliers to prevent them from exerting disproportionate leverage . Correlations were considered significant if the 95% CI did not include zero.
We also correlated the accuracy obtained in each classifier during encoding with memory performance to evaluate if the correlation was also present in the neighboring time bins. To correct for multiple comparisons, a permutation test procedure (50,000 iterations) was used to generate a null distribution for the significance thresholds. In each iteration of the test, we randomly scrambled the order of the subjects (thereby eliminating any inherent correlation) and the correlations were recomputed. Next, the resulting Pearson r values corresponding to the 95th and 5th percentiles were calculated and used for the significance threshold. For display, the resulting Pearson r values were smoothed by a 2-D Gaussian kernel spatial filter with a of 1.
Classification accuracy during retrieval was also correlated with memory performance. For this, we used Person r correlations with a two-tailed level of significance.
# Encoding-related time-frequency analysis
We investigated the subsequent memory effect (SME) in the time window where a significant relationship between encoding pattern classification and episodic remembering was observed. The data from encoding were preprocessed in the same way as described for the pattern classification. The TFR was averaged separately for successful and non-successful memory retrieval and the power estimates for each time point were logtransformed and baseline corrected by the average power in a -0.5-to 0-s time window relative to onset of the target picture.
The statistical significance of these effects was conducted using a nonparametric cluster-based permutation test implemented in . In a first step, a dependent-samples t test is performed to compare the conditions (successful and non-successful target retrieval) and identify statistically significant data samples (␣ ϭ 0.05). All adjacent data samples (either spatial or temporal neighbors) are then grouped into clusters and the t values within each cluster are summed and used to generate a cluster-level t value. The type-1 error rate for the complete data matrix is controlled by evaluating the cluster-level test statistic under the randomization null distribution of the maximum cluster-level test statistic. This was obtained by randomizing the data between conditions for each participant. By creating a reference distribution from 6000 random draws, the p value was estimated according to the proportion of the randomization null distribution exceeding the observed maximum cluster-level test statistic (the so-called Monte Carlo p value). In this way, significant clusters extending over time, frequency, and electrodes can be identified.
Furthermore, to investigate whether the underlying neural mechanisms supporting these effects in the visual and the verbal memory task were different, a topographical analysis of the observed effects was conducted. The TFR of the effects (successful vs unsuccessful target retrieval) on a set of representative sites (F7/F3/Fz/F4/F8, T7/C3/ Cz/C4/T8, and P7/P3/Pz/P4/P8) was subjected to a repeated-measures ANOVA, including the factors memory task (visual vs verbal), region (frontal vs central vs posterior), and hemisphere (left peripheral vs left central vs midline vs right central vs right peripheral). The data were vector scaled [bib_ref] Modeling hippocampal and neocortical contributions to recognition memory: a complementarylearning-systems approach, Mccarthy [/bib_ref] , and Greenhouse-Geisser corrections were applied when appropriate (Greenhouse and Geisser, 1959).
# Results
# Behavioral results
For each task and participant, we calculated the percentage of exemplar and category hits. In the visual memory task, a given trial was considered a category hit if participants correctly identified the picture category and a target hit if they additionally identified the original picture. Overall, participants were successful identifying the category of the target 69 Ϯ 13% (mean Ϯ SD of category hits). Of these trials, 74 Ϯ 6% (mean Ϯ SD) corresponded to exemplar hits. A one-sample t test confirmed that participants were able to identify the original target picture significantly above the 50% chance level (t (17) ϭ 16.7; p Ͻ 0.001).
For the verbal memory task, a trial was considered an exemplar hit if participants correctly named the exemplar in the target picture or alternatively, provided rich details of it (e.g., "the clock that is in London" or "the actress in the Mamma Mia movie"). In general, participants successfully recalled the name of the target picture 53 Ϯ 16% (mean Ϯ SD) of the times. A category hit was a trial for which participants provided evidence of knowing the category of the associated picture. Thus, trials in which participants answer with the category name but did not retrieve any detail of the specific identify represented by the picture were accounted as a category hit (e.g., a "place in a city" or a "face of female"). Overall, participants provided evidence of knowing the category of the target 62 Ϯ 18% (mean Ϯ SD).
Pairwise comparisons showed that the percentage of exemplar hits was higher in the visual compared with the verbal memory task (t (34) ϭ 5.35; p Ͻ 0.001), revealing that it is easier to retrieve the picture than the name of the identity represented by the picture. However, there was no differences in the percentage of category hits between the two memory tasks (t (34) ϭ 1.33; p ϭ 0.194). This indicates similar behavioral performance between the tasks when the retrieval requirement is the semantic category.
We also investigated whether participants were better at retrieving information from one specific category [fig_ref] Table 1: Mean ؎ Standard Deviations for Exemplar and Category Hits, shown for each... [/fig_ref]. A repeated-measures ANOVA including the factors task (visual vs verbal) and category (faces vs landmarks vs objects) was run separately for exemplar and category hits. The analysis of the exemplar hits revealed significant main effects of task (F (1,34) ϭ 31.03; p Ͻ 0.001; 2 ϭ 0.48) and category (F (2,68) ϭ 8.24; p ϭ 0.001; ; 2 ϭ 0.20) and also a significant two-way interaction between the two factors (F (2,68) ϭ 19.43; p Ͻ 0.001; ; 2 ϭ 0.36).
Pairwise comparisons showed that for the visual memory task participants were worse at selecting the correct associated picture when the picture was a face compared with both landmarks tasks (t (17) ϭ -5.1; p Ͻ 0.001) and objects tasks (t (17) ϭ -7.7; p Ͻ 0.001). For the verbal memory task, participants were worse at recalling the names of the landmarks compared with both the faces (t (17) ϭ -3.2; p ϭ 0.005) and objects (t (17) ϭ -3.1; p ϭ 0.006). In terms of category hits, the analysis also revealed a main effect of category (F (2,68) ϭ 38.1; p Ͻ 0.001; ; p ϭ 0.53). Participants were better at retrieving the category if the image was a face compared with both landmarks (t (35) ϭ 5.3; p Ͻ 0.001) and objects (t (35) ϭ 9.0; p Ͻ 0.001), and they were better at retrieving landmarks compared with objects (t (35) ϭ 3.2; p ϭ 0.002). However, no significant effect of task (F (1,34) ϭ 1.8; p ϭ 0.18; ; p ϭ 0.05) nor interaction between task and category (F (2,68) ϭ 1.7; p ϭ 0.18; ; p ϭ 0.05) was observed. In sum, although we identified differences in memory performance as a function of category between the two tasks when the exemplar hits were analyzed, these differences disappeared when the nature of the response between the two tasks is matched, that is, when category hits are analyzed.
A final analysis revealed that participants' rated associative strength was comparable between the two tasks, in total (visual ϭ 1.7 Ϯ 0.3; verbal ϭ 1.6 Ϯ 0.3; t (34) ϭ 0.94; p ϭ 0.352) and also for each stimulus category (faces: t (34) ϭ 1.59; p ϭ 0.120; landmarks: t (34) ϭ 0.76; p ϭ 0.455; and objects: t (34) ϭ 0.37; p ϭ 0.714).
## Eeg-based decoding tracks early visual processing during encoding
To quantify the visual processing during encoding, we trained pattern classifiers to decode the brain activity related to picture encoding. Twenty pattern classifiers were trained at different time bins from -45 to 920 ms after picture onset, and their performance was compared with a baseline classification after word cue onset. Compared with baseline, classifiers trained on picture-related activity showed significant classification accuracy in both memory tasks (all ps Ͻ 0.05). The classifiers that survived correction for multiple comparisons are highlighted )ء( in separately for the visual and verbal memory tasks. An early peak between 120 and 320 ms after picture onset emerged from the classification in both tasks. The classifier with highest accuracy was centered at ϳ170 ms after picture onset for both the visual (mean Ϯ SD ϭ 51 Ϯ 11%) and the verbal (mean Ϯ SD ϭ 57 Ϯ 12%) memory tasks. A direct comparison of the classification accuracy between the two tasks indicated comparable classification accuracy in the two memory tasks.
MVPA classification of brain activity in this early time window has previously been linked to visual processing necessary to distinguish different categories of stimuli . to substantiate this and provide further evidence that classification accuracy in the present study can be used as proxy for visual processing occurring during encoding, we investigated the channel contribution to classification performance of the 20 pattern classifiers trained during picture encoding. Confirming that classification accuracy is based on visual processing, the analysis showed that posterior channels contribute the most to classification accuracy in both tasks [fig_ref] Figure 3: Contribution of each channel to the performance accuracy of the 20 pattern... [/fig_ref]. Interestingly, even in later time windows the posterior channels show the highest contribution to classification performance, indicating that classification performance is based on continued visual processing throughout all the epoch. The topographies were statistically indistinguishable in the two memory tests. In summary, the pattern of the MVPA classification is, in terms of both time and topography, similar in the visual and verbal memory tasks, indicating that classification accuracy reflects visual processing operations in both tasks.
## Early visual processing during encoding interacts with later retrieval demands
To test the prediction that visual processing at study would only be beneficial when the memory task requires the retrieval of visual information, we investigated the relationship between classification accuracy at study and later episodic memory performance. We selected the pattern classifier for which performance accuracy was highest (ϳ170 ms) and correlated its accuracy with successful memory performance (exemplar hits %) for both tasks. As expected, we observed a significant positive correlation for the visual memory task (r ϭ 0.534 [0.07, 0,81]; p ϭ 0.022; [fig_ref] Figure 4: Relationship between the accuracy of the pattern classifier trained at 170 ms... [/fig_ref]. In accordance with our predictions, participants who spontaneously directed attention to percep-tual features of the stimuli, as indicated by pattern classification accuracy, were more likely to correctly identify the original target picture from encoding. This finding offers novel support for the TAP account in a paradigm that assessed the modulatory role of attention during encoding without explicitly directing participants to particular stimulus attributes. Interestingly, we also observed a negative association between visual processing at study and memory performance in the verbal memory task (r ϭ -0.503 ; p ϭ 0.038; [fig_ref] Figure 4: Relationship between the accuracy of the pattern classifier trained at 170 ms... [/fig_ref] , indicating that individuals who adopted a visual attentional focus during study were less likely to retrieve the correct name of the exemplar target picture. This negative correlation suggests that the allocation of resources to stimulus aspects that are not relevant for later retrieval demands (here, visual focus in a verbal memory task) is not only nonbeneficial, as predicted, but can even be detrimental when other aspects are goal relevant (i.e., verbal, lexical information).
## Figure 2.
Averaged accuracy of the 20 cross-validated pattern classifiers trained to discriminate between faces, landmarks, and objects during the encoding phase of the visual and of the verbal memory task. The colored line represents the accuracy of the classifiers trained during the presentation of the target picture. The black line shows the baseline accuracy of the classifiers trained during word cue presentation. Highlighted )ء( are the times bins for which classification performance survived multiple comparison correction. In gray is highlighted the time bins for which classification accuracy was highest.
## Temporal profile of the association between processing during study and later retrieval demands
We next explored the temporal dynamics of the association between visual processing tracked by pattern classification at study and retrieval demands in the ensuing visual and verbal memory tasks. We predicted that the association between the visual processing tapped by the EEG-based decoding and memory retrieval should be seen not only for the classifier with the highest accuracy (ϳ170 ms) but also for its temporal neighbors. To do so, we correlated memory performance in terms of both successful (i.e., exemplar hit) and unsuccessful target retrieval (i.e., when participants completely failed to retrieve any information about the target picture; thus, trials for which participants only correctly retrieved the semantic category are excluded from this analysis) with the accuracy of all the 20 pattern classifiers (for details, see Materials and Methods). shows the temporal profile of these associations in the two memory tasks. As predicted, the associations are not sporadic but extend to neighboring pattern classifiers. Interestingly, the accuracy of two pattern classifiers trained at later time bins (ϳ707 and ϳ805 ms) also show a significant association with memory performance in the visual memory task. [fig_ref] Figure 3: Contribution of each channel to the performance accuracy of the 20 pattern... [/fig_ref] shows that these later pattern classifiers are also most likely based on visual processing, indicating that sustained visual processing is associated with a benefit in retrieving the originally encoded target picture in the visual memory task.
Moreover, for the verbal memory task we observed a positive relationship between the accuracy of early pattern classifiers and unsuccessful memory performance, indicating that participants with high levels of visual processing, tracked by classification accuracy, were not only less likely to correctly retrieve the name of the image but also more likely to fail at retrieval . This is a novel and interesting finding, corroborating the idea that focusing on task-irrelevant aspects of the stimuli during encoding may have a negative impact when later tests of memory requires retrieval of other aspects of the same event.
Next, the same analysis was conducted separately for each picture category. to evaluate if the profile of the observed association between classifier accuracy and memory performance is general or driven by one specific semantic category, we correlated classifier accuracy with memory performance for each of the individual categories. shows the temporal profile of this association. In general, the pattern of results remains when overall classifier accuracy is correlated with memory performance for each of the three categories of stimuli, indicating that the reported associations due not depend on a particular semantic category. However, no significant association for landmarks was observed for the visual memory task. One possible explanation for this result is that participants used some kind of verbal coding to memorize the landscape pictures (e.g., "the tower is in the right side of the image"), which was more difficult to adopt for faces and objects.
## Task-relevant memory encoding differs as a function of retrieval demands
The analyses above demonstrate that early visual processing during encoding is predictive of later memory performance depending on the particular retrieval demands of the test. A subsequent memory analysis was used to further investigate whether processing captured by the pattern classifiers is relevant for successful encoding. Additionally, because previous studies have shown that focusing on different aspects of the stimulus at encoding affects the resulting memory representation , we predicted that the memory representation necessary to perform the upcoming retrieval task will differ as a function of the retrieval demands. An SME analysis was used to test the prediction that early memory formation (120 -320 ms), when pattern classifier showed significant relationship with retrieval demands, would be different across the two memory tasks. A cluster permutation test was used to investigate the SME in the time period between 120 and 320 ms on the range of frequencies (4 -45 Hz) used for pattern classifier training. This analysis identified no significant effects. However, when the analysis was constrained to the frequency range (5-7 Hz) we observed significant SMEs in both tasks (visual memory task: p ϭ 0.04; verbal memory task: p ϭ 0.05; . SMEs characterized by power increases are well documented in previous literature [bib_ref] Brain oscillatory subsequent memory effects differ in power and long-range synchronization between..., Bauch [/bib_ref]. Additional analyses were run in the ␣ (8 -12 Hz),  (13-30 Hz), and ␥ (30 -45 Hz) bands, and no significant additional effects were observed (all ps Ͼ 0.19).
## Figure 5.
Pearson r values showing the temporal profile of the association between visual processing (tracked by EEG-based decoding) and memory performance in the visual and verbal memory task. Highlighted (- ) are time bins for which the relationship with behavioral performance survived correction for multiple comparisons (for details, see Materials and Methods). A, The green line shows the association for retrieval success (i.e., exemplar hits) and the red line shows the association for retrieval failure (i.e., when participants fail to retrieve any information about the target picture). B, Association for retrieval success (green lines) separated for faces (solid green line), landmarks (dashed green line) and objects (dotted green line), and for retrieval failure (red lines) separated for faces (solid red line), landmarks (dashed red line), and objects (dotted red line).
Crucially, a topographical analysis revealed a significant two-way interaction between the factors memory task and hemisphere (F (4,136) ϭ 2.78; p ϭ 0.05; ; 2 ϭ 0.075), showing that the SME is supported by different neural generators in the two memory tasks. While the SME for visual memory task was left lateralized, the SME for the verbal memory task showed a more bilateral distribution , confirming that the task-relevant memory representation formed already at this very early encoding stage varies as a function of the retrieval demands.
## Replay of encoding brain activity at retrieval differs as a function of retrieval demands
Next, we investigated whether the early categoryspecific neural patterns from encoding were replayed at any time during retrieval, as predicted from cortical reinstatement theory. Considering the observed association between classifier performance and retrieval demands, we expected the functional significance of the replay to be different in the two memory tasks. In the visual memory task, we predicted replay to be beneficial. Conversely, in the verbal memory task we predicted replay to be unrelated to performance or even detrimental given the observed negative association between pattern classification at study and retrieval performance in this task. The replay was tested during the word cue presentation at retrieval and for exemplar hit trials, that is, when participants correctly selected the original target picture in the visual memory task or remembered the name of the exemplar depicted target picture in the verbal memory task.
The analysis revealed significant classification of the category of the images associated with the word cue elicited by the onset of the word at retrieval. For the visual memory task, we observed that the early categoryspecific activity during encoding (ϳ120 -320) was replayed at 463-512 ms after word cue onset during memory retrieval . Consistent with our prediction, this neural replay was associated with an increase in memory performance (r ϭ 0.60 [0.18, 0.87]; p ϭ 0.009; [fig_ref] Figure 8: Relationship between the replay of the early visual pattern classifier at retrieval... [/fig_ref]. Additionally, the early pattern classifier trained at encoding at around -25 ms was replayed at 170 -220 ms after word cue onset during retrieval. Note that the reemergence of this classifier, trained in a pre-stimulus interval, is most likely due to poststimulus effects being temporally smeared into the prestimulus interval due to wavelet filtering. However, no significant functional relationship (r ϭ 0.06 between this re-emergency effect and memory performance was observed which clouds interpretation of the effect.
We also observed significant replay in the verbal memory task. The visual classifiers identified during encoding (ϳ350 -420 ms) were replayed later during retrieval (ϳ707 ms; . Following our prediction, the re-emergence of these visual pattern classifiers at retrieval in the verbal memory task was negatively associated with memory performance (r ϭ -0,51 [-0.80, -0.07]; p ϭ 0.033; [fig_ref] Figure 8: Relationship between the replay of the early visual pattern classifier at retrieval... [/fig_ref]. . A, Topography of the SME (5-7 Hz) observed between 120 and 320 ms for the visual and the verbal memory task. Electrodes that reached significance in the cluster-based permutation test are highlighted .)ء( B, TFRs from a representative channel for the visual (FC5) and verbal (F3) memory task.
To further understand the nature of the replay of the encoding brain patterns at retrieval, a final analysis investigated the contribution of each channel to classification (using the replaying pattern classifiers that correlated with memory performance). The classification procedure was repeated for one channel and its neighbors at a time. Classification performance was recalculated and allo-cated to the central channel. Interestingly, the channels that showed the highest contribution to classification were not the posterior ones, and they were clearly different for the visual and the verbal memory task . Recent research has also provided evidence that the reinstated information at retrieval may be a transformed representation of the encoded information rather than a . For the visual memory task, the classifiers trained between 120 -320 ms after picture onset during study and tested around 450 ms after word cue onset at retrieval decode the stimuli category previously associated with the cue. For the verbal memory task, the classifiers trained between 350 -450 ms after picture onset during study and tested around 700 ms after word cue onset at retrieval decode the stimuli category previously associated with the cue. A, Accuracy of the decoding at retrieval for both tasks. The black outlines show p ϭ 0.05 significance thresholds generated by a permutation test. B, Results of comparing the accuracy of the replay against chance (33%). ء Denotes critical t-value. C, Contribution of each channel for the accuracy of the significant replay of the early visual pattern classifiers at retrieval for the visual and verbal memory task.
week replica of the encoded representation [bib_ref] Transformed neural pattern reinstatement during episodic memory retrieval, Xiao [/bib_ref]. The results observed in the present study align well with these recent findings.
# Discussion
A prominent idea in the memory literature is that episodic remembering depends on the extent to which cognitive operations engaged during encoding match those engaged during retrieval [bib_ref] Modeling hippocampal and neocortical contributions to recognition memory: a complementarylearning-systems approach, Mccarthy [/bib_ref]. The present study employed a novel experimental approach to investigate this principle of TAP, by capitalizing on MVPA of brain activity to track spontaneously engaged processing during encoding and to assess transfer depending on later retrieval demands. Our approach thus allowed participants to freely allocate and adjust their processing resources and attentional focus to whatever attribute they considered relevant. Categoryspecific neural patterns observed at encoding and replayed at retrieval were indeed predictive of later episodic remembering. These findings provide novel support for the TAP account and shed new light on the dynamics of encoding and retrieval.
In two memory tasks, with different retrieval demands, we used MVPA to decode, from the oscillatory brain data, category-specific representations of faces, landmarks, and objects. Our EEG-based decoding captured spontaneously engaged visual processing necessary to distinguish the different stimulus categories, presumably driven by category-specific brain regions along the inferior temporal cortex. We trained consecutive classifiers during picture presentation at encoding, which allowed us to track the development of these category-specific brain oscillatory representations over time. The early onset (ϳ170 ms) and the marked posterior topography of these representations indicate that they reflect lowlevel visual processing necessary for the selection of the object models corresponding to faces, landmarks, and objects . Interestingly, these early visual category-specific patterns were not only associated with the predicted benefit when retrieving visual information about an event, but with a cost, when instead verbal information, the picture name, was demanded at retrieval . Specifically, early classification of visual processing predicted successful memory in the visual task and memory failure in the verbal task. These results indicate that participants who spontaneously focused their attention on the visual attributes of the pictures during encoding, were more likely to succeed in the visual memory task but, conversely, fail in the verbal task.
In line with previous research [bib_ref] Neurophysiological evidence for the time course of activation of global shape, part,..., Rossion [/bib_ref] , SMEs were observed in the band. Our results add weight to previous claims underscoring the important role of oscillations in the binding of contextual information supporting episodic memories, and the role of power in encoding-retrieval overlap (for review, see . Interestingly, while the SME in the visual memory task was left lateralized, the SME observed in the verbal memory task was more widespread . The different topographies indicate that encoding was supported by nonoverlapping neural generators, providing evidence that different memory representations formed at encoding are applicable in the two upcoming memory tasks. This finding underscores the modulatory role of attention during encoding and is consistent with predictions from the TAP account underscoring the processing overlap between encoding and retrieval. Notably, the encoding-related category-specific representations were evident already during very early encoding, between 120 and 300 ms after picture onset, when pattern classification performance was highest. Speculatively, it is conceivable that participants may spontaneously adjust their encoding to fit later retrieval demands. An interesting objective for future work is to examine how processing during encoding is modified over time, across multiple study-test blocks.
To examine cortical reinstatement during episodic remembering, the classifiers established during picture encoding were used to decode the oscillatory brain activity during retrieval, when only the word cue was presented. Our results demonstrate that visual category-specific processing was replayed during memory retrieval, and further that the timing and consequence for performance differed as a function of retrieval demands. In the visual memory task, the replay occurred relatively early (ϳ463 ms) after word cue onset and was predictive of successful retrieval, whereas the replay in the verbal memory task occurred later (ϳ707 ms) and was conversely associated with lowered memory performance [fig_ref] Figure 8: Relationship between the replay of the early visual pattern classifier at retrieval... [/fig_ref]. These replay re- sults thus mirror those observed for the category-specific processing during encoding.
Long-term remembering is dependent on the processing occurring during both encoding and retrieval, and the greater the overlap between the cognitive operations that took place during encoding and retrieval the greater the likelihood of successful retrieval [bib_ref] Goal-directed mechanisms that constrain retrieval predict subsequent memory for new ЉfoilЉ information, Tulving [/bib_ref] [bib_ref] Modeling hippocampal and neocortical contributions to recognition memory: a complementarylearning-systems approach, Mccarthy [/bib_ref]. This core prediction of the TAP account has received support from imaging studies [bib_ref] Brain oscillatory subsequent memory effects differ in power and long-range synchronization between..., Bauch [/bib_ref] [bib_ref] Goal-directed mechanisms that constrain retrieval predict subsequent memory for new ЉfoilЉ information, Tulving [/bib_ref]. Here, we replicate and extend the results from these previous studies by showing that the cortical reinstatement of the encoding brain patterns is only associated with beneficial effects for remembering if the replayed patterns are task relevant. Conversely, when the encoding brain patterns were task irrelevant, the cortical reinstatement was associated with detrimental effects on memory performance. Recent work has reported episodic remembering costs, despite a perfect overlap between the encoding and the retrieval contexts, in situations where context does not only overlap with the target episode but also with additional currently irrelevant memory traces . The results of the present study are consistent with this previous research by showing that the reinstatement of the encoding brain patterns at retrieval, under certain conditions, can be associated with detrimental effects on memory.
Face-selective cortical processing is known to be reflected in the N170 event-related potential (ERP) component [bib_ref] Neurophysiological evidence for the time course of activation of global shape, part,..., Rossion [/bib_ref] , which raises the question whether the early visual category-specific representations here observed are reduce to this face-sensitive mechanism. Our data, notably, speak against this explanation as the predictive value of the early categoryspecific brain representations was observed for the episodic retrieval of all three picture categories. There was one exception however. In the visual memory task, the early category-specific representations were not predictive of visual retrieval of the landmarks. It is conceivable that discrimination of the target and distracter landmarks might have been influenced by verbal cues to guide selection (e.g., "the tower is to the right"). Such strategies would be harder to implement for faces and objects.
The memory tasks used in this study are hippocampally dependent, and the cortical reinstatement at retrieval of the encoded brain patterns depends on pattern completion operations [bib_ref] Simple memory: a theory for archicortex, Marr [/bib_ref]. Previous studies have shown that hippocampal pattern completion and ensuing cortical reinstatement are accomplished ϳ500 ms after stimulus onset [bib_ref] Brain oscillatory subsequent memory effects differ in power and long-range synchronization between..., Bauch [/bib_ref]. Others have shown that cortical reinstatement may be more sustained in time for as long as 2000 ms after retrieval cue onset [bib_ref] Neurophysiological evidence for the time course of activation of global shape, part,..., Rossion [/bib_ref]. Our data show that pattern completion operations leading to the cortical reinstatement of encoding brain patterns may occur within ϳ463 ms after cue onset if the encoding brain patterns carry task-relevant representations. On the other hand, cortical reinstatement of task-irrelevant encoding patterns may occur later during retrieval (ϳ707 ms after word cue onset). Also, it should be noted that the reinstated activity in the verbal task was formed at later stages of encoding (ϳ350 -420 ms) compared with the visual task (ϳ120 -320 ms). Although the timing of these encoding patterns is different, the consistent topography throughout the epoch suggests that the classification accuracy was based on sustained visual processing. The MVPA may not have captured category-specific representations reflecting lexical processing relevant for the verbal memory task, which may explain the lack of a positive association between classification accuracy and performance in the verbal memory task. Although we cannot offer a conclusive account for the timing differences of replay in the two memory tasks, our results indicate an important role of task requirement and demands in, at least, the timing of the cortical reinstatement.
The topographical distribution of the replay of the encoding brain patterns differs from the topography of the category-specific representations observed at encoding [fig_ref] Figure 3: Contribution of each channel to the performance accuracy of the 20 pattern... [/fig_ref]. While this finding aligns well with recent research demonstrating that retrieval may involve the reinstatement of a transformed representation of the encoded information [bib_ref] Transformed neural pattern reinstatement during episodic memory retrieval, Xiao [/bib_ref] , further research is needed to fully understand the functional significance of these transformations. One interesting possibility is that they reflect memory reconstruction mechanisms (i.e., "recontextualization") leading to a neural representation that varies in the degree of overlap with details of the original episode [bib_ref] Competitive trace theory: a role for the hippocampus in contextual interference during..., Yassa [/bib_ref]. Similarly, it is conceivable that cortical reinstatement captured with EEG during episodic remembering represents to a lesser degree the low-level sensory activation evoked by the external stimulus input during encoding. In any case, we may conclude that cortical reinstatement during retrieval not only involves literal replay of the processing that occurred in the previous event.
In summary, this is the first study to examine TAP in a paradigm that allowed participants to freely allocate their attention during encoding to whatever attribute they considered relevant. MVPA revealed encoding-related category-specific neural patterns that were replayed at retrieval, and that predicted episodic remembering. Extending the TAP account, we show that the processing engaged during encoding may be associated with both retrieval success and failure depending on the match with later retrieval requirements, thus highlighting also transfer "inappropriate" processing. The present results inform current cognitive neuroscience theories of memory by shedding new light on encoding and retrieval interactions in episodic memory.
[fig] Figure 1: A, Trial structure during study. [/fig]
[fig] Figure 3: Contribution of each channel to the performance accuracy of the 20 pattern classifiers during the encoding phase of the visual and the verbal memory task. The gray frame highlights the time bins for which classification accuracy was highest. [/fig]
[fig] Figure 4: Relationship between the accuracy of the pattern classifier trained at 170 ms after picture onset during study and the proportion of exemplar hits in the visual and in the verbal memory task. [/fig]
[fig] Figure 8: Relationship between the replay of the early visual pattern classifier at retrieval and the proportion of exemplar hits in the visual and in the verbal memory task. [/fig]
[table] Table 1: Mean ؎ Standard Deviations for Exemplar and Category Hits, shown for each memory task and stimulus category. [/table]
|
Association of Human Leukocyte Antigen Alleles and Nevirapine Hypersensitivity in a Malawian HIV-Infected Population
Background. The nonnucleoside reverse transcriptase inhibitor nevirapine is the cornerstone of treatment for human immunodeficiency virus (HIV) in many sub-Saharan African countries. However, nevirapine is associated with a 6%-10% risk of developing a hypersensitivity reaction, with different phenotypes, including the blistering conditions Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN). Our aim was to identify predictive human leukocyte antigen (HLA) markers that are associated with nevirapine hypersensitivity.Methods. We identified 117 HIV-infected Malawian adults with nevirapine hypersensitivity (15 drug-induced liver injury [DILI], 33 SJS/TEN, 20 hypersensitivity syndrome, and 46 nevirapine-induced rash plus 3 with both DILI and SJS phenotype) and 155 age-, sex-and ethnicity-matched nevirapine-exposed controls. HLA typing for 5 loci (A, B, C, DRB1, and DQB1) was undertaken using a sequence-based high-resolution protocol. Logistic regression analysis included CD4 + cell count as a covariate.Results. HLA-C*04:01 was found to markedly increase the risk for SJS (odds ratio [OR] = 17.52; 95% confidence interval, 3.31-92.80) and all hypersensitivity phenotypes (OR = 2.64; 95% CI, 1.13-6.18) when compared to the baseline rare allele group in a binary logistic regression model. The OR for absolute risk of SJS/TEN associated with carriage of HLA-C*04:01 was 5.17 (95% CI, 2.39-11.18). Positive predictive value was 2.6% and negative predictive value was 99.2%. In addition, a number of alleles within the HLA-DQB1 loci protected against nevirapine-induced hypersensitivity phenotypes.Conclusions. Our study has identified HLA-C*04:01 carriage as a risk factor for nevirapine-induced SJS/TEN in a Malawian HIV cohort. Validation of these findings in a larger cohort of patients and mechanistic investigation of the pathogenesis are required.
The nonnucleoside reverse-transcriptase inhibitor nevirapine is widely used as a first-line treatment of human immunodeficiency virus (HIV) [bib_ref] Comparison of firstline antiretroviral therapy with regimens including nevirapine, efavirenz, or both..., Van Leth [/bib_ref] infection in developing countries because of its low cost. Nevirapine is usually given in combination with 2 nucleoside reverse-transcriptase inhibitors (stavudine or zidovudine and lamivudine).
Though effective [bib_ref] Stavudine, lamivudine and nevirapine combination therapy for treatment of HIV infection and..., Siegfried [/bib_ref] , nevirapine causes hypersensitivity in 6%-10% of patients [bib_ref] Determinants of nevirapine hypersensitivity and its effect on the association between hepatitis..., Phillips [/bib_ref] [bib_ref] Discontinuation of nevirapine because of hypersensitivity reactions in patients with prior treatment..., Wit [/bib_ref] , which can manifest clinically in various ways from nevirapine-induced rash (without any systemic manifestations) to severe blistering skin reactions such as Stevens-Johnson syndrome and toxic epidermal necrolysis [bib_ref] Nevirapine and the risk of Stevens-Johnson syndrome or toxic epidermal necrolysis, Fagot [/bib_ref] (1-2 per 1000 exposed individuals [bib_ref] Incidence of toxic epidermal necrolysis and Stevens-Johnson syndrome in an HIV cohort:..., Mittmann [/bib_ref]. Extracutaneous involvement typically manifests as hepatotoxicity [bib_ref] Hepatotoxicity following nevirapine-containing regimens in HIV-1-infected individuals, De Maat [/bib_ref]. Reactions most commonly manifest within the first 6 weeks of starting therapy.
Immunogenetic factors, including a number of human leukocyte antigen (HLA) alleles, have been previously identified as risk factors for hypersensitivity reactions to the antiretroviral abacavir [bib_ref] Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1..., Mallal [/bib_ref] and many other classes of drugs, including the antiepileptic drug carbamazepine [bib_ref] Medical genetics: a marker for Stevens-Johnson syndrome, Chung [/bib_ref] [bib_ref] HLA-A*3101 and carbamazepine-induced hypersensitivity reactions in Europeans, Mccormack [/bib_ref] and the antigout drug allopurinol [bib_ref] HLA-B*5801 allele as a genetic marker for severe cutaneous adverse reactions caused..., Hung [/bib_ref]. Research has also focused on immunogenetic risk factors for nevirapine hypersensitivity, identifying the HLA-DRB1*01:01 (whites [bib_ref] Predisposition to nevirapine hypersensitivity associated with HLA-DRB1*0101 and abrogated by low CD4..., Martin [/bib_ref] [bib_ref] HLA-DRB1*01 associated with cutaneous hypersensitivity induced by nevirapine and efavirenz, Vitezica [/bib_ref] [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] , HLA-C*04 (Thai [bib_ref] HLA-Cw*04 allele associated with nevirapine-induced rash in HIV-infected Thai patients, Likanonsakul [/bib_ref] , Chinese [bib_ref] HLA-dependent hypersensitivity reaction to nevirapine in Chinese Han HIV-infected patients, Gao [/bib_ref] , and blacks [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] , HLA-C*08 ( Japanese [bib_ref] HLA-Cw8 primarily associated with hypersensitivity to nevirapine, Gatanaga [/bib_ref] and HLA-B*35:05 (Thai [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] [bib_ref] HLA-B*3505 allele is a strong predictor for nevirapine-induced skin adverse drug reactions..., Chantarangsu [/bib_ref] as risk alleles [fig_ref] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/fig_ref].
To date, little is known regarding genetic risk factors for nevirapine-induced hypersensitivity in sub-Saharan African HIV-infected populations. Using a cohort of carefully phenotyped Malawian patients, we have undertaken high-resolution sequence-based genotyping to determine whether alleles in 5 loci in the class I and II major histocompatibility complex (MHC) regions on chromosome 6 (HLA-DRB1, DQB1, A, B, or C) are predisposing factors for nevirapine hypersensitivity.
# Patients and methods
## Patients
Between March 2007 and December 2008, we prospectively recruited 1117 antiretroviral-naive adult patients from the outpatient clinic at the Queen Elizabeth Central Hospital, Blantyre, Malawi. At time of recruitment, this clinic had approximately 10 000 patients registered as having started on antiretroviral therapy since 2004. Patients were self-reported black African; were older than 16 years; and gave informed consent approved by the research ethics committees at the College of Medicine Research and Ethics Committee, Malawi, and Liverpool School of Tropical Medicine. Patients presenting with jaundice at baseline were excluded.
Patients commenced antiretroviral therapy as recommended by the World Health Organization eligibility criteria at the time of recruitment. All patients were diagnosed as clinical stage 3/4 or had a CD4 + count <250 cells/µL; commenced preparations, which contained a fixed dose of nevirapine, lamivudine, and stavudine; and were followed up for 26 weeks. Clinical and laboratory parameters including CD4 + count and liver function tests were monitored at 0, 6, 14, and 18 weeks.
The study was a nested case-control study; however, because of the low incidence of hypersensitivity in the prospectively recruited cohort, an additional 177 patients attending the same outpatient clinic who developed nevirapine hypersensitivity were also recruited, either prospectively (n = 149) or identified from patient records retrospectively (n = 28). Of 177 patients, 65 were excluded owing to insufficient DNA quality and quantity.
Careful clinical assessment of all patients was undertaken to identify and characterize the hypersensitivity reactions, using the Naranjo causality assessment tool [bib_ref] Evaluation of Naranjo adverse drug reactions probability scale in causality assessment of..., Garcia-Cortes [/bib_ref] [bib_ref] Clinical detection and assessment of drug induced neurotoxicity, Naranjo [/bib_ref]. Phenotypes were retrospectively reviewed independently by a dermatologist using both clinical data and photographs. These were defined as:
- Nevirapine-induced rash (NIR): widespread maculopapular rash without systemic manifestations and getting worse on treatment continuation.
- Hypersensitivity syndrome (HSS; also known as drug reaction with eosinophilia and systemic symptoms or drug-induced hypersensitivity syndrome): widespread rash and systemic manifestations such as fever, cough, or abnormal liver function tests.
- Stevens-Johnson syndrome (SJS): extensive rash with the involvement of at least 2 mucous membranes or blistering eruptions affecting <10% of body surface area.
- Toxic epidermal necrolysis (TEN) [bib_ref] Nevirapine and the risk of Stevens-Johnson syndrome or toxic epidermal necrolysis, Fagot [/bib_ref] ; blistering rash affecting >30% of body surface area and mucous membrane involvement as per SJS [bib_ref] Toxic epidermal necrolysis and Stevens Johnson syndrome: our current understanding, French [/bib_ref]. Blistering between 10% and 30% of body surface area was termed overlap syndrome.
- Drug-induced liver injury (DILI) [bib_ref] Hepatotoxicity following nevirapine-containing regimens in HIV-1-infected individuals, De Maat [/bib_ref] : jaundice and abnormal alanine aminotransferase level.
Patients meeting criteria for drug-induced reactions had nevirapine withdrawn in accordance with international guidelines. It is important to note that as part of the Malawian treatment guidelines, liver function tests are not routine, and therefore, abnormal tests without clinical jaundice did not fulfill criteria for treatment cessation and were not included as cases. Furthermore, some patients who developed transient nonsevere rash without systemic symptoms underwent close observation, were treated continuously with rash resolution, and again were not classified as cases.
Control patients (n = 155) were identified from the prospective cohort and followed up for at least 6 months while taking nevirapine without developing any signs of hypersensitivity. Cases and controls were matched by age and sex, and were also from the same region of Malawi.
## Dna extraction and high-resolution sequence-based hla typing
DNA was extracted from whole blood using a salt precipitation protocol. High-resolution, sequencing-based HLA typing of 5 loci (HLA-A, B, C, DRB1, and DQB) was undertaken by Histogenetics (Ossining, New York). Sequencing data files were analyzed using Histogenetics' proprietary analysis software (Histomatcher and HistoMagic) for HLA genotype calling. Allele assignments are based on IMGT/HLA Database release version 2.21.0, dated April 2008 (http://www.ebi.ac.uk/ imgt/hla/). Where odds ratios and 95% CIs are not reported, a 2 × 2 χ 2 test was performed on the data available.
Abbreviations: CI, confidence interval; DILI, drug-induced liver injury; HLA, human leukocyte antigen; HSR, hypersensitivity reaction; OR, odds ratio; SJS, Stevens-Johnson syndrome. a Denotes reported associations which withstood correction for multiple testing (P corrected < 0.05).
# Statistical analysis
Sample size calculations were performed assuming that a 10% background frequency of an HLA allele would provide 80% power (α = .05) to detect an odds ratio (OR) of 3.0 (and 90% power to detect an OR of 3.4). We included all patients with hypersensitivity in the analysis. Subgroup analyses were performed for all phenotypes, (DILI, SJS/TEN, HSS, NIR) where we compared the frequency of HLA alleles in patients with nevirapine-induced adverse reaction with the frequency in tolerant individuals.
Nongenetic factors identified a priori as being of importance, such as CD4 + count, were first tested univariately for association with hypersensitivity reaction (all cases) using the Student t test. The distribution of CD4 + count was skewed, and a square-root transformation was used to achieve normality. CD4 + count for 20 cases was missing, and these observations were substituted by the mean-transformed CD4 + count for all cases where CD4 + count was observed. Differences in frequencies of alleles in individual HLA locus between tolerant patients and each of the hypersensitivity phenotypes were determined from 2 × N contingency tables using a χ 2 test within the CLUMP software package (http://www.smd.qmul.ac.uk/ statgen/dcurtis/lc/clump.html). To determine association with specific alleles within hypersensitivity-linked HLA loci, 2 logistic regression models were fitted. The first included covariates representing the nongenetic factors identified from univariate analysis (P < .05). The second included a covariate to represent HLA alleles assuming a dominant mode of inheritance. Rare alleles were grouped into a single allele category and, because this represented the largest category, it was assumed to be the baseline allele category for the purpose of regression modeling. To assess for significance of the genetic locus, a likelihood-ratio test was undertaken comparing the models and the P value was recorded. Analyses were undertaken in R version 2.13.0. To account for multiple comparisons (5 phenotypes and 5 loci), we used the false-discovery rate method within the "p.adjust" function of R. The HLA multiple locus haplotypes were generated using PyPop 0.7.0 software [bib_ref] PyPop: a software framework for population genomics: analyzing large-scale multi-locus genotype data, Lancaster [/bib_ref].
A random-effects OR meta-analysis of pooled data from our study and previously published data was undertaken using StatsDirect version 2.6.8 (StatsDirect Ltd, Atrincham, UK).
# Results
From the prospective cohort (n = 1117), 57 patients developed hypersensitivity (5.1%), of whom 31 were successfully HLAtyped. Of the 149 supplementary hypersensitive patients, 86 were HLA-typed, giving a total of 117 HLA-typed hypersensitive patients (15 DILI, 33 SJS/TEN, 20 HSS, and 46 NIR, plus 3 individuals with the DILI and SJS/TEN phenotype). One control sample failed HLA typing, leaving 154 HLA-typed drug-tolerant controls. The overall HLA-allele call rates were 182 of 271 (67%) for DRB1*; 241 of 271 (89%) for DQB1*; and 296 of 271 (99%) for A, B, and C.
A summary of the HLA allele frequencies for each of the phenotypes and controls is provided in [fig_ref] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/fig_ref].
Median CD4 + cell count at the start of antiretroviral therapy was 235 cells/µL (interquartile range [IQR], 128-424 cells/µL) in cases and 166 cells/µL (IQR, 83-250 cells/µL) in controls. This represented a statistically significant difference; thus, CD4 + cell count was adjusted for in the analyses of association with genetic loci.
We undertook χ 2 analyses in CLUMP focusing on the association of each locus with the different phenotypic manifestations [fig_ref] Table 2: Logistic Regression Analysis for 5 Human Leukocyte Antigen Loci in All Patients... [/fig_ref]. After correction for multiple comparisons, we identified HLA-DQB1 as the only significant (P corrected < .05) HLA locus for nevirapine-induced hypersensitivity, when all phenotypes were combined, and with SJS/TEN specifically. The locus-specific analysis provided an indication that the HLA-DQB1 region protected against nevirapine hypersensitivity. Given the high degree of linkage disequilibrium in the MHC, and the multiple alleles present within each locus, we then undertook an analysis of the individual HLA alleles. Consistent with the locus-specific data, a number of HLA-DQB1 alleles were found to protect against nevirapine hypersensitivity when compared to the "rare allele" group. Our analysis showed that HLA-C*04:01 predisposed to nevirapine hypersensitivity. Individuals who carry HLA-C*04:01 were at higher risk of developing hypersensitivity reactions in general (OR = 2.64 [95% CI, 1.13-2.64]), and, specifically SJS/ TEN (OR = 17.52 [95% CI, 3.31-92.80]) when exposed to nevirapine than were carriers of the "rare alleles," the most common group of HLA alleles in C locus. This association was not observed with any other phenotype.
Multilocus haplotypes for class I and II HLA loci were generated to determine the structure of haplotypes across multiple loci in our cohort from Malawi [fig_ref] Table 4: Linkage Disequilibrium Analysis of Class I and II Human Leukocyte Antigen Loci... [/fig_ref]. The data suggest high linkage disequilibrium between the HLA-B and C loci in both the nevirapine hypersensitive and tolerant patients (D′ = 0.946 and 0.924, respectively). Significant linkage disequilibrium was observed in the hypersensitive and tolerant groups between the DQB1 and DRB1 loci (D′ = 0.890 and 0.898 respectively). Haplotype frequencies were calculated for 5-loci haplotypes and for combinations of HLA-B, C, DRB1, and DQB1 loci haplotypes containing the HLA-C*04:01 allele [fig_ref] Table 2: Logistic Regression Analysis for 5 Human Leukocyte Antigen Loci in All Patients... [/fig_ref]. The frequency of the HLA B53:01:01/C*04:01 haplotype was significantly higher in the hypersensitive cohort (0.121) than the tolerant group Only loci/phenotype associations determined as significant in [fig_ref] Table 2: Logistic Regression Analysis for 5 Human Leukocyte Antigen Loci in All Patients... [/fig_ref] are included. Results in bold are those allele/phenotypes associations where the 95% confidence interval for the odds ratio excludes 1.
Abbreviations: CI, confidence interval; DILI, drug-induced liver injury; HLA, human leukocyte antigen; HSR, hypersensitivity reaction; HSS, hypersensitivity syndrome; n/a, odds ratios were not possible to calculate due to the low frequency of the allele in that phenotype; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis.
(0.039). There was no difference in DRB1/DQB1 haplotype frequencies in haplotypes containing the DQB1*05:01:01 allele [fig_ref] Table 5: Frequency of Human Leukocyte Antigen Haplotype Frequencies in the Nevirapine-Hypersensitive and -Tolerant... [/fig_ref]. Subsequent analysis was undertaken incorporating all HLAtyped individuals to determine the absolute risk of SJS/TEN and predictive values of HLA-C*04:01, the HLA-B*53:01:01/ C*04:01 haplotype, and DQB1*05:0101 carriage [fig_ref] Table 6: Predictive Value for Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis Risk of HLA-C [/fig_ref]. The OR for overall risk of SJS/TEN associated with HLA-C*04:01 was 5.17 (95% CI, 2.39-11.18; P < .0001). Positive predictive values (PPVs) and negative predictive values (NPVs), based on a SJS/TEN prevalence of 1.07%, were 2.6% and 99.2%, respectively. The OR for overall risk of SJS/TEN associated with carriage of the HLA-B*53:01:01/C*04:01 haplotype (5.17 [95% CI, 1.83-14.28]) was comparable to HLA-C*04:01 alone, although the NPV was lower (91.6%). For DQB*05:0101 the OR was 0.17 (95% CI, .05-.60), and the PPV and NPV were 0.3% and 98.6%, respectively. Other HLA allele/hypersensitivity phenotype associations noted indemonstrated PPVs between 0.08% and 3.1% and NPVs between 22.3% and 36.2% (data not shown).
# Discussion
Nevirapine-induced hypersensitivity reactions have shown an association with a number of HLA alleles [fig_ref] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/fig_ref] , which vary according to ethnicity and the phenotype of the reaction [bib_ref] HLA and pharmacogenetics of drug hypersensitivity, Pavlos [/bib_ref]. The main finding of the present study is that HLA-C*04:01 predisposes to nevirapine-cutaneous reactions with the greatest risk observed with SJS/TEN (OR = 5.17 [95% CI, 2.39-11.18];, the severest form of hypersensitivity in terms of mortality [bib_ref] The incidence of erythema multiforme, Stevens-Johnson syndrome, and toxic epidermal necrolysis. A..., Chan [/bib_ref] [bib_ref] Toxic epidermal necrolysis (Lyell syndrome), Roujeau [/bib_ref] [bib_ref] Toxic epidermal necrolysis and Stevens-Johnson syndrome. An epidemiologic study from West Germany, Schopf [/bib_ref]. The risk associated with HLA-B*53:01:01/C*04:01 haplotype carriage was comparable. Its sensitivity as a biomarker for SJS/TEN was 31.4%, compared to 63.9% for HLA-C*04:01 alone [fig_ref] Table 6: Predictive Value for Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis Risk of HLA-C [/fig_ref] , suggesting that the association is driven by carriage of 1 allele at a single HLA locus. This is supported by the haplotype analysis of the HLA loci in this particular Malawian population [fig_ref] Table 4: Linkage Disequilibrium Analysis of Class I and II Human Leukocyte Antigen Loci... [/fig_ref]. Although HLA-C*04 (along with B*35) has been associated with the development of AIDS in whites [bib_ref] HLA and HIV-1: heterozygote advantage and B*35-Cw*04 disadvantage, Carrington [/bib_ref] , no association between HLA-C*04 and HIV has been reported in African populations or any other ethnic group. This is the first report of an association between nevirapine-induced SJS/TEN and HLA-C*04:01, but is consistent with previous studies in black African (OR = 5.17) [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] , Thai (OR = 3.79) [bib_ref] HLA-Cw*04 allele associated with nevirapine-induced rash in HIV-infected Thai patients, Likanonsakul [/bib_ref] , and Chinese (OR = 3.23) [bib_ref] HLA-dependent hypersensitivity reaction to nevirapine in Chinese Han HIV-infected patients, Gao [/bib_ref] populations that have reported an association with HLA-C*04 nevirapine-cutaneous reactions. A metaanalysis of our data, related to HLA-C*04:01 carriage and cutaneous nevirapine hypersensitivity reactions (n = 102), but excluding patients who had DILI only (n = 15), with the only eligible previous study in a black American population (OR = 5.17 [95% CI, 1.81-14.78]) [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] gave a combined OR of 3.34 (95% CI, 1.60-4.98) (Supplementary .
Although it would be useful to replicate in other populations, the data available to date strongly suggest that Positive and negative predictive values as well as sensitivity and specificity are shown. Prevalence of SJS/TEN is assumed at 1.07% based on an incidence of 12 of 1117 observed in the prospective study. Odds ratio with 95% confidence and P value are determined using a 2 × 2 χ 2 test.
Abbreviations: CI, confidence interval; HLA, human leukocyte antigen; NPV, negative predictive value; OR, odds ratio; PPV, positive predictive value; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis.
HLA-C*04:01 predisposes to nevirapine-induced cutaneous reactions of different severities (including SJS/TEN) in several ethnic groups. The predictive value of HLA-C*04:01 as a biomarker of nevirapine-induced SJS/TEN is limited [fig_ref] Table 4: Linkage Disequilibrium Analysis of Class I and II Human Leukocyte Antigen Loci... [/fig_ref]. The incidence of nevirapine-induced SJS/TEN in our prospective study was 12 of 1117 patients (1.07%). This gives a PPV of 2.6%, which is of no diagnostic value. Although the NPV is 99.2%, it does not reach 100%, which has been recommended [bib_ref] HLA and pharmacogenetics of drug hypersensitivity, Pavlos [/bib_ref]. It is important to note that in our population, HLA-C*04:01 was associated with SJS/TEN, which is almost always drug-related, severe, and perhaps more easily recognizable than other drug-induced adverse phenotypes. This contrasts with abacavir hypersensitivity, which varies in severity and can be more difficult to differentiate from other causes. This is reflected in the fact that the NPV of HLA-B*57:01 was 95.5% for clinically diagnosed abacavir-induced hypersensitivity and 100% for immunologically confirmed abacavir hypersensitivity [bib_ref] HLA-B*5701 screening for hypersensitivity to abacavir, Mallal [/bib_ref]. We did not replicate a previous associations between HLA-DRB1*01:01 and nevirapine-induced hypersensitivity [bib_ref] Predisposition to nevirapine hypersensitivity associated with HLA-DRB1*0101 and abrogated by low CD4..., Martin [/bib_ref] [bib_ref] HLA-DRB1*01 associated with cutaneous hypersensitivity induced by nevirapine and efavirenz, Vitezica [/bib_ref] [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] in whites. This is possibly due to ethnic differences in the frequency of HLA alleles. HLA-DRB1*01:01 was observed at a low frequency in our Malawian cohort (0.008) with 1 tolerant and 2 hypersensitive carriers out of 182 individuals genotyped.
The observation of an apparent protective effect of 6 different HLA-DQB1across a number of different phenotypes is interesting. A number studies have identified protective HLA alleles [fig_ref] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/fig_ref] , but there is no common pattern. These may not represent the actual protective alleles given the high degree of linkage disequilibrium across the MHC. Further work in larger populations is needed to elucidate the interaction between risk and protective HLA alleles in predisposing to different forms of nevirapine hypersensitivity. Of note here is that the pathogenesis of nevirapine hypersensitivity is immune-mediated, as shown by a positive lymphocyte transformation test in a patient with DILI [bib_ref] Drug-specific T cells in an HIV-positive patient with nevirapine-induced hepatitis, Drummond [/bib_ref]. However the mechanism by which this occurs is unknown. Three possible hypotheses have been suggested, including the hapten hypothesis [bib_ref] Idiosyncratic drug reactions: current understanding, Uetrecht [/bib_ref] , pharmacological interaction hypothesis [bib_ref] Pharmacological interaction of drugs with immune receptors: the p-i concept, Pichler [/bib_ref] , and altered peptide binding profile [bib_ref] Drug hypersensitivity caused by alteration of the MHC-presented self-peptide repertoire, Ostrov [/bib_ref]. It is possible that based on the HLA profile of an individual, the interaction between nevirapine (and its antigen) and the HLA molecules leads to either a protective or a predisposing effect. Such an allele-competing effect has been postulated for the HLA-associated disease narcolepsy [bib_ref] HLA-DQ association and allele competition in Chinese narcolepsy, Han [/bib_ref].
Studies have also evaluated the role of CYP2B6, which metabolizes nevirapine, in predisposing to hypersensitivity. CYPB6 shows wide interindividual variability in expression and activity in human livers [bib_ref] Human cytochrome P4502B6: interindividual hepatic expression, substrate specificity, and role in procarcinogen..., Code [/bib_ref]. It contains a functional exonic variant (c.516G>T), which causes loss of enzymatic function [bib_ref] Extensive genetic polymorphism in the human CYP2B6 gene with impact on expression..., Lang [/bib_ref] ; is associated with higher plasma concentrations in black and white populations [bib_ref] Influence of CYP2B6 polymorphism on plasma and intracellular concentrations and toxicity of..., Rotger [/bib_ref] [bib_ref] Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients..., Penzak [/bib_ref] ; and has been associated with nevirapine-induced cutaneous adverse reactions [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] and neuropsychological toxicity [bib_ref] Influence of CYP2B6 polymorphism on plasma and intracellular concentrations and toxicity of..., Rotger [/bib_ref] , though not hepatotoxicity [bib_ref] Pharmacogenetics of nevirapine-associated hepatotoxicity: an adult AIDS Clinical Trials Group collaboration, Haas [/bib_ref]. The combination of c.516G>T and HLA-Cw*04 alleles showed a stronger association in black, white, and Asian populations than c.516 G>T alone [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref]. In our cohort, however, the CYP2B6 c.516G>T polymorphism was not a significant risk factor for any of the hypersensitivity phenotypes (data not shown); therefore, a combination analysis has not been undertaken.
Our study has several strengths: (1) we investigated a Malawian HIV cohort originating from a highly homogeneous population from a small geographic area; thus, the effect of ethnicity admixture is likely to be minimal; (2) sex-and agematching of our tolerant controls also minimized the effect of these nongenetic factors; (3) when compared to previous studies, our sample size was larger; (4) we used strict phenotypic characterization with independent adjudication by a dermatologist; (5) the majority of patients were recruited prospectively where detailed phenotypic data could be gathered, although we did include 28 retrospectively identified patients; (6) close monitoring of patients with nonsevere rash, treated through and excluded as cases, further strengthened the phenotype by omitting potential false positives; and (7) we used sequencebased HLA typing to at least 4 digits, which is particularly important in this population because of the presence of some rare alleles. However, there are also some limitations. First, despite a large sample size, when subdividing the groups into phenotypes, the numbers within categories fell, limiting our power to detect true associations. This is a recognized drawback of studying nevirapine hypersensitivity, where the phenotypic manifestations not only vary, but have different allelic associations [fig_ref] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/fig_ref]. Second, we could not genotype all patients, particularly for the class II HLA alleles, which would have strengthened haplotype analysis, nevertheless, our study was larger than previous studies despite the missing data. Third, given the homogeneity of the Malawian population, it is possible that although the HLA association identified here is relevant, it may not be applicable to other ethnicities including other African populations.
In conclusion, we have identified an association between the HLA-C*04:01 allele nevirapine-induced hypersensitivity phenotypes, including the first report of an association between HLA-C*04:01 and the most severe phenotype, SJS/TEN. Our study appears to replicate previous observations [bib_ref] Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African,..., Yuan [/bib_ref] of an association between HLA-C*04:01 and risk of nevirapine cutaneous adverse drug reactions in a black population. Further work is required to replicate the association identified here, and to evaluate in more detail the effects of risk and competing HLA alleles. Additionally, functional in vitro or in silico models are needed to clarify the mechanisms of the immunemediated response to nevirapine and its metabolites [bib_ref] Quantifying the metabolic activation of nevirapine in patients by integrated applications of..., Srivastava [/bib_ref].
[table] Table 1: Previously Reported Human Leukocyte Antigen Allele Associations With Nevirapine Hypersensitivity [/table]
[table] Table 2: Logistic Regression Analysis for 5 Human Leukocyte Antigen Loci in All Patients With Nevirapine-Induced Hypersensitivity [/table]
[table] Table 4: Linkage Disequilibrium Analysis of Class I and II Human Leukocyte Antigen Loci in Nevirapine-Hypersensitive and -Tolerant Malawian Cohorts [/table]
[table] Table 5: Frequency of Human Leukocyte Antigen Haplotype Frequencies in the Nevirapine-Hypersensitive and -Tolerant Cohorts [/table]
[table] Table 6: Predictive Value for Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis Risk of HLA-C [/table]
|
A multicenter prospective, randomized, placebo-controlled phase II/III trial for preemptive acute graft-versus-host disease therapy
Acute graft-versus-host disease (aGvHD) contributes to about 50% of transplant-related mortality (non-relapse mortality) after allogeneic hematopoietic stem cell transplantation (HSCT). Here the predictive value of a urinary proteomic profile (aGvHD_MS17) was tested together with preemptive prednisolone therapy. Two-hundred and fifty-nine of 267 patients were eligible for analysis. Ninety-two patients were randomized upon aGvHD_MS17 classification factor above 0.1 to receive either prednisolone (2-2.5 mg/kg, N = 44) or placebo (N = 47; N = 1 randomization failure) for 5 days followed by tapering. The remaining 167 patients formed the observation group. The primary endpoint of the randomized trial was incidence of aGvHD grade II between randomization and day +100 post HSCT. Analysis of the short-term preemptive prednisolone therapy in the randomized patients showed no significant difference in incidence or severity of acute GvHD (HR: 1.69, 95% CI: 0.66-4.32, P = 0.27). Prednisolone as preemptive treatment did not lead to an increase in relapse (20.2% in the placebo and 14.0% in the prednisolone group (P = 0.46)). The frequency of adverse events was slightly higher in the placebo group (64.4% versus 50%, respectively). Taken together, the results of the Pre-GvHD trial demonstrated the feasibility and safety of preemptive prednisolone treatment in the randomized patients.
# Introduction
Acute graft-versus-host disease (aGvHD) is a severe complication of allogeneic hematopoietic stem cell transplantation (HSCT) and is diagnosed by clinical features, such as skin rash, diarrhea, or elevation of liver enzymes followed by biopsies and histopathological examination if appropriate. Between 30 and 80% of patients develop aGvHD, depending on primary disease, patient age, conditioning regimen, and GvHD prophylaxis. Initial standard therapy for aGvHD is prednisolone at a dose of 2-2.5 mg/kg body weight per day, resulting in a response rate of about 70% for patients with aGvHD grades I-II.
Proteome analysis of urine using capillary electrophoresis coupled online to mass spectrometry (CE-MS) to define differentially excreted urinary peptides is a broadly applicable and powerful diagnostic tool in a variety of diseases. The proteomic classifier "aGvHD_MS17" was based on 17 differentially excreted peptides identified by CE-MS analysis in the urine of patients after allogeneic HSCT. Upon application of support vector machine (SVM)-based analyses tools, the dimensionless classification factor (CF) for aGvHD_MS17 was calculated. Receiver-operated characteristic curves were used to determine that a CF above 0.1 could be used to separate patients with pending aGvHD grades II-IV from those who never developed aGvHD or had aGvHD grade I up to 21 days prior to clinical manifestation of aGvHD. The aGvHD_MS17 classifier has been tested on more than 700 patients transplanted at four different transplant centers. Apart from the aGvHD_MS17 profile, others have described the use of ELISA to detect plasma biomarkers, to predict aGvHD outcome and non-relapse mortality (NRM). Bacigalupo et al.published the first preemptive treatment trial for aGvHD, using ATG in all patients stratified into three risk groups and resulting in a reduction of acute and chronic GvHD and better overall survival (OS) in the high-risk subgroup. The same group studied the influence of steroid treatment in patients with aGvHD grade I randomized to receive 1 mg prednisolone or placebo to prevent severe GvHD. Based on our previous studies, we initiated a German multicenter trial in 2008, to evaluate the effect of preemptive prednisolone treatment of imminent aGvHD grades II-IV upon aGvHD_MS17 CF positivity. The trial was an investigator initiated, prospective, multicenter, placebo controlled, double-blind clinical trial, registered under www.isrctn.com: ISRCTN03911524 and www.clinicaltrialsregister.eu: EudraCT number: 2008-005862-30.
# Methods
The PRE-GvHD trial
The "Pre-emptive therapy of aGvHD according to specific proteomic patterns after allogeneic hematopoietic stem cell transplantation (Pre-GvHD) trial" was an investigator initiated, prospective multicenter, double-blind, placebo controlled, randomized phase II/III trial conducted in 11 German centers. The trial was designed and overseen by the authors. The trial protocol was approved by the leading ethics committee of Hannover Medical School (MHH), those of the contributing sites and the federal regulatory agency (BfArM). The trial was performed in accordance with the principles of the Declaration of Helsinki.
## Patients
Adult patients (from 18 years on) were included after informed consent. Inclusion was on day +3 (range: 2-5) after the first allogeneic HSCT. Detailed inclusion and exclusion criteria are provided in Supplementary material. The first patient entered the trial on December 14, 2009. Eight patients were excluded from analysis, since no urine samples were collected and analyzed, due to intensive care treatment, dialysis or death prior day +7. Clinical and demographic data for all eligible patients are summarized in. Urine was collected at regular predefined intervals up to 80 days post HSCT and analyzed centrally. Ninetytwo patients were randomized upon aGvHD_MS17 CF positivity to receive either prednisolone (2-2.5 mg/kg, N = 44) or placebo (N = 48) for 5 days followed by a taper of 19 days in the absence of aGvHD. One patient of the placebo group was randomized by mistake despite only negative samples in the aGvHD_MS17 test. The remaining 167 patients formed the observation group. Clinically manifest aGvHD was graded according to the modified Glucksberg criteriaand verified by biopsieswhere appropriate. Visits were made weekly until day +35 and on days +50, +80 (all ±3 days), +100, and +130 (±10 days) for clinical evaluation and development of aGvHD. The follow-up was 1 year after HSCT.shows the trial flow chart and the disposition of patients.
## Urine sample collection
Samples were collected from all patients enrolled weekly between days +7 and +35, on days +50 and +80 (all: ±3 days) after HSCT, frozen at −20°C and shipped to the central laboratory (Mosaiques, Hannover, Germany) for further analysis. The protocol required that the time between sample arrival at Mosaiques and the transmission of the aGvHD_MS17 CF score for randomization had to be within 72 h. Sample preparation and CE-MS analyses were described previously and summarized in Supplementary material.
## Blinding and randomization procedures
All patients included in the screening process received a consecutive patient number at their study site. The patient identifier was the combination of the study site number and the patient number. For patients not eligible for the trial, only demographic data and inclusion/exclusion criteria were recorded. Two independent teams at each study site ensured double-blinding. Team A (clinical ward team) received only blinded medication and did not know the treatment code. Team B (pharmacy) knew the treatment code prepared by the independent trial statistician and prepared the medication accordingly. Randomization lists were prepared by the statistician at Prometris (Mannheim, Germany) who was not involved in other tasks of the trial. Randomization envelopes containing the treatment code were deposited at each study site, accessible only for team B. Additionally, sealed emergency envelopes containing the randomization number and the treatment group was prepared for each patient to be opened in case of signs of aGvHD grades II or higher or another medical emergency.
## Primary and secondary endpoints of the clinical trial
The primary endpoint was the occurrence of aGvHD grades II-IV between randomization and 100 days after HSCT. Death occurring in patients without aGvHD (grade II or higher) between randomization and 100 days after HSCT was considered as treatment failure, equivalent to aGvHD grade II. The secondary efficacy endpoints were (1) the severity of aGvHD between time of randomization and 100 days after allo-HSCT, (2) occurrence of aGvHD grades II-IV, (3) severity of all aGvHD, (4) transplant-related mortality (NRM), (5) OS, (6) occurrence of leukemic relapses, and (7) infectious complications.
## Safety
The exposure of study treatment was characterized by the number of administrations and the cumulative dose for both treatment groups. Adverse events (AEs) were coded using the Medical Dictionary for Regulatory Activities, version 19.1. Annual Safety Reports were submitted to National Competent Authority and leading ethic committee as required by ICH guidelines and national regulations.
# Statistical analysis
Primary endpoint of the Pre-GvHD trial was incidence of aGvHD grades II-IV within 100 days post HSCT. Sample size calculation was based on previous data. We expected that aGvHD_MS17-positive samples would predict 80% probability of developing aGvHD grades II-IV and a reduction to 41% by preemptive therapy (pilot study,;. Seventy-eight patients (2 × 39) were required to detect a reduction of incidence or severity of aGvHD grades II-IV or death from 80% in the placebo group to 50% (odds ratio: 0.3) in the prednisolone group with a type I error of 5% and a power of 80% using a two-sided Cochran-Mantel-Haenszel (CMH) test. The number of patients for randomization was increased to 2 × 45 to account for dropout. A detailed description of the statistical plan is provided in the Supplementary materials. The CMH test and competing risk analysis were used to compare the two treatment arms for incidence of aGvHD II-IV within 100 days post HSCT. OS in both treatment arms was compared using Cox proportional hazard models with left truncation at time of randomization and right censoring in case of lost to follow-up. Time dependent Cox models were used to compare OS between all randomized patients with samples positive for the aGvHD_MS17 test versus all patients in the observation group with continuously negative aGvHD_MS17 test results.
# Results
## Patient characteristics
Between 2009 and 2015, 267 patients were enrolled into the pre-randomization phase of the Pre-GvHD trial after the first HSCT; 259 patients were eligible for assessment of outcome. The disposition of patients and trial flow diagram are shown in(Supplementary. Urine samples were monitored at indicated time points and the aGvHD_MS17 CF was calculated by application of an SVM-based softwareand considered positive when the CF was above 0.1and Supplementary material exclusion and inclusion criteria). All patients in the observation group were treated upon clinical signs of aGvHD with the firstline standard treatment of aGvHD using corticosteroids and/or second-line treatment, if necessary, at the discretion of the project leaders in each center.
## Randomized study population
The intent-to-treat (ITT) population of the randomized phase of the Pre-GvHD trial consisted of 92 patients, 44 receiving prednisolone and 48 receiving placebo for 5 days followed by tapering for 19 days in the absence of clinical signs of aGvHD. One patient in the placebo group was randomized by error despite only negative aGvHD_MS17 test results. In the randomized patients, the median time of aGvHD_MS17 CF sample positivity prior to aGvHD I-IV manifestation was 25 days (mean: 14 days, range: 7-90), in randomization to onset aGvHD: 24 days (0-89). Preemptive therapy was started 2 days after randomization (median 2; range: 0-6 days) in both groups. The as-per-protocol population (PP group; N = 84; 42 placebo, 42 prednisolone arm) excluded patients not receiving the study drug for at least 3 days. The data generated in the ITT and the PP groups were similar in all statistical analyses, thus most analysis shown here are on the ITT population. Reasons for premature withdrawal were withdrawal of consent (placebo N = 1; prednisolone N = 1; observation N = 3) and protocol violation (placebo: N = 1). One patient in the prednisolone group died prior to randomization (Supplementary. The reasons for early discontinuation of the study medication in 17 patients (placebo N = 11; prednisolone N = 6) are summarized in . Unblinding due to signs of clinical manifest aGvHD II or higher occurred in 14 patients (placebo N = 9; prednisolone N = 5).
## Outcome of the randomized phase of the pre-gvhd trial
The primary endpoint was the incidence of aGvHD II-IV between randomization and day +100 after HSCT. The incidence of aGvHD grades II-IV was 25% in the randomized patients. Death during the first 100 days was considered an event also in patients without aGvHD II-IV.shows that the cumulative incidence of aGvHD II-IV, with death as competing event, was similar in the prednisolone and placebo groups (HR: 1.69, 95% CI: 0.66-4.32, P = 0.27).summarizes the incidence and severity of aGvHD, NRM, and OS in the placebo, prednisolone, and the observation groups.shows the OS in the PP group. The probability of OS was similar in both arms (HR: 1.06; 95% CI: 0.52-2.14; P = 0.88; Supplementary time from HSCT to NRM was not analyzed, due to the low number of events in both randomized treatment arms. Patients with only aGvHD_MS17 CF negative test results in the observation group (N = 102) had a very low risk to die compared to the randomized patients with aGvHD_MS17 positive samples (N = 91), who had a 2.5-fold (P = 0.002) increased risk of death within 1 year post HSCT. The cumulative incidence of leukemic relapse or progression (including death due to relapse/progression) at 1 year after HSCT was slightly higher in the placebo group (20%) than in the prednisolone group (14%), but the difference was not statistically significant (P = 0.46). In this analysis, death unrelated to relapse or progression was considered as competing risk and patients lost to follow-up were considered as censored cases (Supplementary . The frequency of infections during the screening period prior to randomization was already higher in the prednisolone group (23%) than in the placebo group (8%). After the screening period, infections were reported in 91% of the patients in the prednisolone group and in 83% of the patients in the placebo group (Supplementary . The slightly higher frequency of infections was observed at all time points during the study, with the exception of the final visit at 1 year after allo-HSCT.
## Safety data
The mean duration of treatment (including taper) was 16 days in the placebo group and 18 days in the prednisolone group (Supplementary . The shorter treatment duration in the placebo arm is explained by the higher frequency of premature discontinuation of the study medication in the placebo group (27%) compared to 17% in the prednisolone group. AEs observed in the two treatment groups are summarized in . The frequency of AEs was higher in the placebo when compared to the prednisolone group (64% and 50%, respectively). The most common AEs were diarrhea (18% and 10%, respectively), nausea (11 and 5%), vomiting (4 and 5%), and acute kidney injury (7 against 2%). Serious AEs were more frequent in the placebo group (13 and 7%). The frequency of (S)AEs attributed to the study medication by the investigators was lower in the placebo (13%) than in the prednisolone group (21%). AEs leading to withdrawal from study drug were only observed in the placebo group (five patients with gastrointestinal complications, one GvHD of the skin, one patient with insomnia, and one with acute kidney injury). In addition, one placebo patient died from pulmonary hemorrhage .
# Discussion
This is the first prospective randomized multicenter trial using proteomic peptide profiling of patients to assess the effect of preemptive GvHD-directed therapy post-HSCT.
With sufficient numbers of patients randomized and long follow-up, the results are disillusioning and reflect the outcome of previous attempts of preemptive therapy. The preemptive treatment of imminent aGvHD with 2-2.5 mg prednisolone/kg BW for 5 consecutive days after randomization did not reduce the clinical development of severe aGvHD. Other drugs, like etanercept or ruxolitinib, may be more effective than corticosteroids, but this would have to be tested in another clinical trial. Several reasons may account for this negative outcome. The short treatment period of only 5 days followed by tapering for 19 days (mean treatment time: 16 days placebo; 18 days prednisolone; Supplementary may not be sufficient to inhibit the proliferation of allogeneic T cells that induce aGvHD. Lower overall aGvHD rates in the current trial could be due to differences in the patient population as compared to our previous studies. The pilot study leading to the design of the present trial (Supplementary had more patients with no CR and included mismatch transplantation. In the Pre-GvHD trial, only goodrisk patients in CR, chronic phase of chronic myeloid leukemia, very good PR, or patients with untreated MDS were included until 2013. Only when recruitment was still not completed by 2013, the inclusion criteria were expanded to include patients transplanted in relapse or refractory patients with more than 10% leukemic blasts (AML, ALL, and MDS/MPN). In the previous pilot, the majority of the patients were transplanted after relapse or treatment failure. Furthermore, although conditioning regimens did not influence the data significantly in former analyses, some RIC protocols are designed to allow for more allogeneic reactions. For example, in the previous analyses the intensive fludarabine-cytosine arabinoside-amsacrine (FLAMSA)either with TBI or busulfan and ATG protocol was the most commonly applied RIC, while in the current trial protocols without particular influence on immune reactions such as fludarabine/melphalan or fludarabine/busulfan made up more than 50% of the RIC protocols (Supplementary. All these factors together may explain the lower aGvHD rate in the Pre-GvHD trial compared to previous studies. Considering safety of preemptive therapy with prednisolone, neither infections including virus reactivations , nor disease relapses (Supplementary differed significantly between the prednisolone and the placebo arms. In addition, the frequency of other AEs was not increased in the prednisolone group when compared to the placebo group . Thus, no specific safety risk of the preemptive therapy with prednisolone was identified, although this was a relatively small patient number to detect rare events.
There was a slight difference in aGvHD development in the randomized patient groups. As shown in, the incidence of aGvHD grades II-IV was slightly higher in the prednisolone group (N = 11) compared to the placebo group (N = 8). This may be explained by more HLA-mismatched donors (N = 5 (10%) placebo; N = 8 (18%) prednisolone) and higher rate of gender mismatch HSCT in the prednisolone group (male recipients transplanted from female donors in the placebo N = 6 (12%); prednisolone N = 9 (20%);.
We can draw further conclusions from the follow-up of the patients who entered this large multicenter trial but were not randomized. aGvHD_MS17 CF positivity predicted NRM and lower OS within the 1st year after HSCT. Patients with only aGvHD_MS17 CF negative samples (observation group; N = 102) had a lower risk for NRM compared to patients with at least one aGvHD_MS17-positive test (randomized patients; N = 91; HR: 2.49; P = 0.002;. Preemptive therapy based on laboratory biomarkers was investigated by Bacigalupo et al., using ATG treatment in patients after HSCT. They found that the highest risk group benefitted from this treatment. Later a prospective randomized trial of prednisolone treatment (1 mg/kg) or placebo included patients with aGvHD grade I. The outcome was that aGvHD grade II development was reduced, but infectious complications were higher and the incidence of aGvHD III-IV was not reduced. In our trial, the incidence of infectious complications post preemptive prednisolone treatment was not significantly different to those in the placebo group. The most promising plasma diagnostic biomarkers appear to be regenerating islet-derived protein 3 alpha (Reg-3a;, suppression of tumorigenicity 2 (ST2;, and soluble TNF-receptor 1 (sTNFR1;as diagnostic markers for GvHD and prediction of performance after HSCT. Reg-3a was tested in samples from 1014 HSCT patients from three transplantation centers. Recently, the so-called Ann Arbor Scoreof aGvHD, which relies on three biomarkers (Reg-3a, ST2, and sTNFR1), was implemented in the analysis of patients after HSCT. It predicts treatment response by day 28 post GvHD therapy and 6-month NRM irrespective of center-specific strategies. The international Mount Sinai Acute GvHD International Consortium has been recently founded and a prospective monitoring study for correct diagnosis and NRM prediction is currently ongoing in an international setting. Our findings add to the current literature where a cluster of plasma proteins detected by ELISA were used to predict outcome after HSCT. Two to six differentially secreted plasma biomarkers after clinical diagnosis of aGvHD grade II or higher indicated decreased OS and increased NRM. Prediction of OS and NRM was also studied by Luft et al., who used an "Endothelial Activation and Stress Index" (EASIX, an algorithm using lactate dehydrogenase (U/L) multiplied by creatinine (mg/dL) and divided by platelets (10 9 cells per L)) to predict NRM and death in patients after HSCT. However, only aGvHD_MS17 can be used to predict overall aGvHD and/or to guide preemptive therapy in patients after HSCT without prior clinical aGvHD diagnosis. In addition, aGvHD_MS17 monitoring has the potential to characterize new biomarkers/ pathways involved in the development of aGvHD and may be useful to detect new alternative and much needed therapeutic targets for treatment of severe aGvHD. In addition, combinations with the other biomarkers, like the EASIX score, used for predicting outcome prior to HSCTmay be used to pre-select high-risk patients for further analysis with aGvHD_MS17. Application of aGvHD_MS17 monitoring between days +7 and +21 led to a higher predictive value of aGvHD than at later time points. An early analysis (e.g., day +14 (±7) post HSCT) using aGvHD_MS17 in the clinical practice could be helpful to recognize patients at high risk to develop aGvHD and/or is prone to NRM within the 1st year after HSCT. In our study, we observed that the previously established aGvHD_MS17 urine peptide marker pattern allowed the prediction of 2.5-fold higher NRM. Thus, a positive aGvHD_MS17 test result by day +14 may help to guide intensified clinical monitoring of patients with high risk of developing aGvHD or NRM to administer effective other immunosuppressive therapy or intensified care early on.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. |
The unequal impact of the coronavirus pandemic: Evidence from seventeen developing countries
We recruited participants for the survey using paid advertisement posts on Facebook and Instagram. To be eligible to participate in the survey, participants had to be at least 18 years old. In each country, we targeted the social media campaign to people ages 19 and up. We utilized keywords that have broad appeal in the country, such as futbol (soccer) and the names of futbol stars or local celebrities, and are unrelated to COVID-19 in order to avoid selecting respondents based on their knowledge or experience regarding the current pandemic. In countries in which the average age of our sample was high or expected to be high, we utilized a second, simultaneous social media campaign with the same images and keywords, but specifically targeted to millennials aged 24-35.In each country, we used three posts, an image of a male wearing a face mask, an image of a female wearing a face mask, and an image of an iconic landmark in that country. The posts were adapted to each country by changing the image of the landmark, changing the image of the male and female if appropriate, changing the country name in the text, and the changing the language (Spanish, English, or Dutch) of the posts. Facebook algorithms select which post to show each user.
## Questionnaire
The questionnaire collects data on monthly total household income in January 2020 and expected monthly total household income in April 2020 in ranges constructed as multiples of the minimum wage in that country. The questionnaire asked if total household income was reduced during the past week. The questionnaire asked whether a household member lost their job or closed their business. The recall period for these questions was randomized between 1 week, 2 weeks, and 1 month. For respondents that remain in the labor market, the questionnaire asks whether the respondent worked outside the home or from home during the past week and we code a variable to represent teleworking if the respondent reported working from home. Using the past week as the recall period, the questionnaire asks if any member of the household went hungry due to lack of food and asks respondents to report the strength of their agreement with a statement "I am eating less healthy than normal" on a scale from 1 (complete disagreement) to 5 (complete agreement). The questionnaire asks if any member of the household received (gave) a transfer or a loan from (to) another household during the past week. The questionnaire also asks respondents to report the strength of their agreement with a statement that COVID-19 should be the top priority of the national government on a scale from 1 (complete disagreement) to 5 (complete agreement) and asked whether respondents think that non-essential businesses should close.
Sample questionnaires in English, Spanish and Dutch can be downloaded here: https: //www.dropbox.com/sh/uuv17cfaz94kw4h/AAAWvgJUNYEBHhNxZPcQVv3Za?dl=0.
## Data construction
The questionnaire was implemented online using Qualtrics. Column [bib_ref] How Much Should We Trust Differences-In-Differences Estimates?, Bertrand [/bib_ref] of [fig_ref] Table S1: Date of Launch and Number of Observations by Country [/fig_ref] displays the launch date of the survey in each country. We construct the data set in several steps and the resulting number of observations that comprise the sample for each country is shown in column [bib_ref] Insuring Consumption Against Illness, Gertler [/bib_ref] of [fig_ref] Table S1: Date of Launch and Number of Observations by Country [/fig_ref]. First, we restrict the sample to completed surveys. Overall, approximately 59% of surveys that are started are completed. Second, we restrict the sample to surveys associated with IP addresses within the borders of the country for which the respondent is completing the survey. Across all countries in the sample, over 99% of completed surveys comply with this criteria. Third, Qualtrics flags surveys that are completed on the same device and likely to be repeat surveys completed by the same individual or household based on cookies. This is an imperfect filter. For example, it will not recognize repeated surveys by the same individual or household that are completed on different devices. Surveys flagged as repeats comprise less than 2.3% of completed surveys and we drop these surveys from the sample.
## Re-weighting and external validity
## Estimation of weights
For each country, we modeled the probability of being in the nationally representative sample as a function of demographic, time-invariant characteristics such as gender, age and education of the respondent, indicators for the presence of children of 5 years old or younger in the household and elderly (60 years old or older) in the households, as well as the number of household members and the number of children enrolled in school. Finally, we also include region-specific indicators.
We conducted this process country by country. For each country, we used the most recent nationally representative survey available in the Inter-American Development Bank harmonized survey data repository. We estimated the model by fitting a logistic function (logit) and computed predicted probabilities of being in the nationally representative data set (p i,c ) for each respondent i in country c (see [fig_ref] Table S2: Parameters of logit models for the probability of being in the nationally... [/fig_ref]. We then used inverse probability weights (ipw i,c = 1/(1 −p i,c )) to, at least in terms of observable characteristics, resemble those from nationally representative surveys.
To prevent differences in response rates from driving the results, we re-scale the within-country weights ipw i,c by the inverse share of re-weighted number of responses per country, relative to the country's population size (P opulation c /( Nc i ipw i,c )).
## Out-of-sample validation
We validate the country-specific weights by looking at their out-of-sample performance. For each country we randomly selected 60% of the observations in the online survey and the household (field) surveys. We used this sample to estimate logit models. We then use the resulting models to predict the probability of being observed in the household survey over the observations in the remaining 40% of the sample, which we call the testing sample. We then re-weight the observations in the testing sample and compare the adjusted means of demographic characteristics to those obtained from household (field) survey data. [fig_ref] Table S3: Differences between Online Survey Data and Household [/fig_ref] shows this exercise in Columns 3 and 4. We were able to reduce the differences in respondent characteristics between the online and household surveys. This out-of-sample exercise suggests that the improvement in balance is not driven by overfitting. We also show that our online data does not differ substantially from the household survey data in non-targeted moments such as the proportion of respondents by income category.
## Countries by type of mobility-restriction policies
The following list details the type of mobility restrictions implemented by each country. Data was collected from official government websites and press articles. The information was collected on April 23rd 2020.
No mandatory policies: Bahamas, Suriname, Trinidad and Tobago, Uruguay. Curfews: Dominican Republic, Guyana. Local Quarantines: Chile, Jamaica, Mexico. National Quarantines Barbados, Bolivia, Colombia, Costa Rica, Ecuador, El Salvador, Panama, Peru. Job loss is a dummy variable that equals one if the respondent indicates that they lost their job within the last week, two weeks or month (these horizons were randomly assigned). Similarly, the variable for business closures equals one if subjects report that their business was forced to close by the government or by lack of demand within the same randomly assigned time horizon.
# Statistical analysis details
To explore differences in job loss and business closure across the income distribution, responses are first aggregated at the country and income bracket level using within-country weights as described in the Supplementary Information Appendix. Country averaged outcomes are then regressed by OLS on indicators for income bin and control for country fixed effects, weighting by country population. Because there are only have 17 countries in the sample, using standard clustered standard errors would result in incorrect inference. Therefore, the analysis follow [bib_ref] How Much Should We Trust Differences-In-Differences Estimates?, Bertrand [/bib_ref] by aggregating the data at the country level by income category.the point estimates and 95 percent confidence intervals for each income bin average. As robustness, results are shown for the raw pooled data, without applying weights in. For this, respondents that reported that at least one household member lost their job during or had to close their businesses during the survey reference period were identified, and calculated shares using weights to correct for sampling issues (See Supplementary Information section). Data corresponding to the share of self-employed workers was obtained from the World Bank's World Development Indicators.
## Figure 3
The share of household income in each income bin for incomes reported by respondents corresponding to January 2020 and April 2020 are presented in The sample is restricted to responses collected between April 13th to May 1st 2020 to include all countries. Shares are calculated by re-weighting counts by within-and across-country weights as detailed in Supplementary Material section Estimation of weights. As robustness, results are shown for the raw pooled data, without applying weights in .
## Figure 4
Fig 4 shows the share of respondents supporting extending the lockdown policies by at least one month over time since the first COVID-19 case in the country. One challenge to explore how support changes over time is that the timing of responses varies between countries. Over time as new countries enter the sample, the change in sample composition could bias our results. We address this issue in two steps. First, we construct new weights and re-weight respondents after day 31 to match those responding at the beginning of the series (days 30 and 31) on observable demographic characteristics.
A model was estimated by fitting a logistic function (logit) and computed predicted probabilities of having responded to the survey on day 30 or 31 relative to first case. Explanatory variables included were: demographic characteristics such as household size, age indicators for: presence of children or elderly, having felt hunger, woman, education primary or less, college education or higher, household income categories (<1, , and for day of week. Then, inverse probability weights were estimated based on the propensity score obtained.
Second, OLS regressions were separately estimated for respondents whose household member lost job or closed business and for those who did not, where the dependent variable is an indicator variable that equals one if they support extending the lockdown by at least a month, and zero otherwise. The independent variables are indicators for days since first COVID-19 case, day of week indicators, and country fixed effects. The regression is weighted by the inverse probability weight described in step one and employs robust standard errors. the point estimates and 95 percent confidence intervals for these estimates. Similar estimates when using the pooled raw data without re-weighting are presented in . Table ?? reports comparisons of differences in the relevant outcomes between respondents whose households experienced a loss of livelihood during the pandemic period and those that did not. To control for time-varying characteristics, we perform comparisons focusing on respondents in the same locality and who completed the survey during the same day. We exploit granular data from over 3000 localities in 17 countries. Thus, we focus our analysis on the subset of 18,000 locality-date-of-response cells that include more than one observation. This approach allows us to isolate time-varying locality shocks, and thus purge regional confounding factors. In addition we control for industry sector fixed effects, to prevent differences in the sectors related to the household's main economic activities from driving the results, as exposure to the effects of the pandemic may vary across sectors.
We operationalize our approach by estimating the following specification, which is similar to that used in reference [bib_ref] Insuring Consumption Against Illness, Gertler [/bib_ref] :
[formula] Y i,l,c,t = βLostlivelihood i,l,c,t + X i,c,l,t Σ + δ l,c,t + θ s + i,l,c,t(1) [/formula]
Here, Y i,l,c,t denotes the outcome of interest corresponding to respondent i in locality l from country c collected in date t. Lost livelihood i,l,c,t is an indicator of whether any member of the respondent's household lost her/his job or closed her business during the the past week, two-weeks, or month. X i,c,l,t is vector of demographic characteristics of the respondent (age, education level and gender) and of household characteristics (household size, presence of children younger than 5, presence of school-age children, and people 60 years old or older). δ l,c,t denotes locality-date fixed effects, and θ s denotes industry fixed effects based on the main pre-pandemic source of income of the respondent's household. i,c,l,t denotes unobserved shocks. To account for possible serial correlation of outcomes within localities, we cluster the standard errors at the locality-country level.
The parameter of interest, β, is reported in and captures within-locality differences in outcomes between respondents whose households experienced a loss of livelihood during the pandemic and those who didn't. The reported models are estimated using weights to achieve country-level representativeness, to correct for differences in sample size across countries, and to provide higher based on country population size. All results are robust to excluding these weights (See Robustness Section in Supplementary Material).
Dependent variables: Went hungry is an indicator of whether any household member went hungry during the past week due to lack of food. Eats less healthy takes the value of one if the respondent somewhat or totally agrees with the statement "I eat more unhealthy foods than normal". Gifts/Loans is an indicator of whether the respondent's household received a gift or transfer from either friends or relatives during the preceding week. Gov. Priority is an indicator of whether the respondent somewhat agrees or totally agrees with the statement "The government's priority should be to stop the spread of the pandemic". Lockdown (>= month) is an indicator of whether the respondent reports agreeing with closing non-essential business for one month or longer. As this question was asked only to people that reported agreeing with policies that require non-essential businesses to close, Lockdown(>= month) takes the value of zero when the respondent reported not supporting measures of keeping non-essential businesses closed at all, regardless of the time.
## Figure 2
To investigate whether the correlates of job loss or business closures and household outcomes are stronger in countries with high levels of informality, we also estimate the following specification: (2) +X i,c,l,t Σ + δ l,c,t + θ s + i,l,c,t
[formula] Y i,l,c,t = β 1 Lostlivelihood i,l,c,t + β 2 Lostlivelihood i,l,c,t × Self − employment c [/formula]
Where Self-employment denotes the share of self-employed workers in country c and was obtained from the World Bank's World Development Indicators, using the most-recent observations for each country. We also report results using the share of informal workers (as a share of non-agricultural workers) in each country. This information was not available for Suriname, Jamaica, Trinidad and Tobago, Barbados, and Bahamas. See results in [fig_ref] Table S4: Loss of livelihoods and changes in well-being, and policy support [/fig_ref]. Date of launch is the date on which the social media posts began. The date that the survey was rolled out in each country was largely determined by bureaucratic processes and approvals. With the except of Costa Rica, data collection in each country continued until April 30, 2020. The number of observations for each country reflected the number in the sample after data cleaning. The percent of localities is the percent of localities of each country for which we have one or more observations in the sample. Household size -0.002 -0.575*** -0.206*** -0.123*** -0.073*** -0.061*** -0.084*** -0.047*** -0.195*** -0.326*** -0.048*** -0.085*** -0.097*** -0.197*** -0.067*** -0.125*** (0 -5.185*** -1.623*** -6.316*** -6.035*** -5.904*** -5.698*** -6.427*** -6.900*** -6.614*** -6.726*** 0.000 -6.797*** -6.041*** -8.058*** -10.927*** -5.377*** ( The table presents estimates of coefficients from a logit model of the probability of being the nationally representative household survey as a function of demographic characteristics. The models were estimated separately for each country. All models also included region-specific indicators, except in the case of Ecuador in which data regarding regions was not available. Empty cells in the table imply that the relevant variable was not available in the household surveys. Robust standard errors are presented in parenthesis. The table presents means of household and survey respondent demographic characteristics using data from the online survey and nationally representative surveys, and pooling observations from all study countries (weighting by country size). Column (1) reports raw means using all the observations from the online surveys. Column (2) reports means using all available observations in the household (field) surveys using sampling weights. Column (3) reports means from the online survey data using only data from the testing sample (i.e., the sample not used for the estimation of the inverse probability weights). Column (4) reports means using data from the household (field) surveys corresponding to the testing sample. The testing sample corresponds to a randomly selected subsample corresponding to 40% of all the observations in the online and household (field) surveys. MW stands for national minimum wage. The inverse probability weights are computed based on logit models of the probability of being observed in the household survey which are estimated country by country. The models include including age, gender, and education categories of the respondent as well as household-level demographic characteristics such as the presence of children younger than 5 years old, the presence of elderly children in the household, # of children enrolled in school, household size, as well as region fixed effects. Panel A reports unweighted regression coefficients capturing the relationship between livelihood losses during the pandemic and outcomes. Each column reports results of a regression of the dependent variable on an indicator of whether any household member either lost a job or closed a business and a vector of covariates. In addition, all regressions control for locality × day of survey completion fixed effects (18,764), as well as economic-sector fixed effects. Standard errors are clustered at the locality level (3,165). Panels B and C replicate the results using country-date fixed effects (300) and standard errors clustered at the country level (17). See the Empirical Methods section in the main text for more details. Panel A reports regression coefficients capturing the relationship between livelihood losses during the pandemic and outcomes, as a function of the share of self-employed workers in each country. Each column reports results of a regression of the dependent variable on an indicator of whether any household member either lost a job or closed a business, and an interaction term of job loss or business closure with the share of self-employed workers in each country, and a vector of covariates. In addition, all regressions control for locality × day of survey completion fixed effects (18,764), as well as economic-sector fixed effects. Standard errors are clustered at the locality level (3,165). Panels B replicates the results using the percentage of informal workers as in each country instead of the share of self-employed workers. See the Empirical Methods section in the main text for more details.
# Supplementary figures
# Supplementary tables
[fig] Fig 1: shows how job loss and business closures vary throughout the income distribution. [/fig]
[fig] Fig 2: presents a cross-country comparison of the shares of households who lost their livelihoods as a function of the share of self-employed workers in the economy. [/fig]
[fig] Fig S1 Fig S2 Fig S3: Sample has broad geographic coverage at the sub-national level. The sub-national regions of each country in the sample are shaded according to number of observations as a share of population (in %). Sources of population data for each country are shown in the Supplementary Information section. Administrative boundary shapefiles were obtained from the GADM database (www.gadm.org). Higher rates of job loss and business closure among households in the lowest income group. Point estimates and 95% confidence intervals from a regression of the outcome on income bin indicators and country fixed effects. Pooled data (no weights), robust standard errors. See the Empirical Methods section in the Supplementary Information section for more details. The share of households in the bottom part of the income distribution is expected to increase. Shares of households in each income bin for incomes reported for January 2020 and April 2020. Pooled data (no weights). See the Empirical Methods section in the Supplementary Information section for more details. [/fig]
[table] Table S1: Date of Launch and Number of Observations by Country [/table]
[table] Table S2: Parameters of logit models for the probability of being in the nationally representative survey [/table]
[table] Table S3: Differences between Online Survey Data and Household (field) Survey Data [/table]
[table] Table S4: Loss of livelihoods and changes in well-being, and policy support (unweighted results) Decreased income Went hungry Eats less healthy Gift/Loan Gov. Priority Lockdown (>= month) Decreased income Went hungry Eats less healthy Gift/Loan Gov. Priority Lockdown (>= month) Decreased income Went hungry Eats less healthy Gift/Loan Gov. Priority Lockdown (>= month) * p < 0.1, * * p < 0.05, * * * p < 0.01 [/table]
[table] Table S5: Impacts of livelihood loss on income and food security by labor market characteristics Panel A: Impacts on food security by country-level rates of self-employment Decreased income Went hungry Eats less healthy Lost job or closed business × % Self-Employed Panel B: Impacts on food security by country-level rates of informality Decreased income Went hungry Eats less healthy Lost job or closed business × % Informal workers * p < 0.1, * * p < 0.05, * * * p < 0.01 [/table]
|
Streptovertimycins A–H, new fasamycin-type antibiotics produced by a soil-derived Streptomyces morookaense strain
Eight new fasamycin-type polyketides, streptovertimycins A-H (1−8), were isolated from soil-derived Streptomyces morookaense SC1169 cultivated on wheat grains. Their structures were established by extensive spectroscopic analysis and theoretical computations of ECD spectra. Compounds 1−8 have a fasamycin-type pentacyclic structure featuring a 15-Omethyl group. They exhibited potent activity against methicillin-resistant Staphylococcus aureus (MRSA) and vancomycinresistant Enterococcus faecium (VRE) with MIC values in the range of 0.63-5.0 μg/ml. The activity profile provided new insights into the structure-activity relationships of fasamycin-type antibiotics.
# Introduction
Fasamycins and congeners are a class of rare aromatic polyketides possessing a partially reduced 1-phenyltetracene pentacyclic core. The first disclosed members of this class of natural products are the Streptomyces sp. KB-3346-5 derived naphthacemycins, including three (A-C) different series of 17 compounds [bib_ref] Naphthacemycins, novel circumventors of β-lactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] Naphthacemycins, novel circumventors of βlactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] KB-3346-5 substances, their fermentative manufacture, and antibacterial agents containing them, Omura [/bib_ref] , which were originally termed as "KB-3346-5 substances" in a Japanese patent [bib_ref] KB-3346-5 substances, their fermentative manufacture, and antibacterial agents containing them, Omura [/bib_ref].
Fasamycins A and B were obtained by the heterologous expression of a clone derived from environmental DNA isolated from soil [bib_ref] Functional analysis of environmental DNA-derived type II polyketides synthase reveals structurally diverse..., Feng [/bib_ref] [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref]. Other congeners so far reported are all produced by Streptomyces formicae isolated from an African fungus-growing plant-ant species, which include fasamycins C-F, formicamycins A-M [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref] [bib_ref] A role for antibiotic biosynthesis monooxygenase domain proteins in fidelity control during..., Qin [/bib_ref] , and formicapyridines A-I [bib_ref] A role for antibiotic biosynthesis monooxygenase domain proteins in fidelity control during..., Qin [/bib_ref]. Most of this class of compounds, except the pyridine-containing formicapyridines, were shown to exhibit potent activity against Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus (MRSA) [bib_ref] Naphthacemycins, novel circumventors of β-lactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] Naphthacemycins, novel circumventors of βlactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] KB-3346-5 substances, their fermentative manufacture, and antibacterial agents containing them, Omura [/bib_ref] [bib_ref] Functional analysis of environmental DNA-derived type II polyketides synthase reveals structurally diverse..., Feng [/bib_ref] [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref] [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref] and vancomycin-resistant Enterococcus faecalis (VRE) [bib_ref] Functional analysis of environmental DNA-derived type II polyketides synthase reveals structurally diverse..., Feng [/bib_ref] [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref] [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref]. Fasamycins A and B were found to inhibit the FabF enzyme associated with the biosynthesis of type II fatty acids (FASII) in bacteria [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref]. This mode of action differs from those of antibiotics currently in clinical use [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref]. In addition, some naphthacemycins were reported as novel inhibitors of Poly (ADP-ribose) polymerase-1 (PARP1) [bib_ref] Discovery of naphthacemycins as a novel class of PARP1 inhibitors, Shen [/bib_ref]. The fascinating biological profile of fasamycins and congeners has attracted organic synthetic and medicinal chemists to embark on their synthesis. The total synthesis of fasamycin A and naphthacemycin A [bib_ref] Total synthesis of (±)-naphthacemycin A 9 , possessing both antibacterial activity against..., Hirose [/bib_ref] have been achieved and the structure-activity relationship (SAR) of the analogues were discussed [bib_ref] Total synthesis of (±)-naphthacemycin A 9 , possessing both antibacterial activity against..., Hirose [/bib_ref] [bib_ref] Total synthesis of tetarimycin A, (±)-naphthacemycin A 9 , and (±)-fasamycin A:..., Huang [/bib_ref].
Streptomyces morookaense (formerly Streptoverticillium morookaense) SC1169 is an actinomycete isolated from soil. Its culture extract was previously found to be active against the economically important fungal phytopathogen Peronophythora litchii and two new active alkaloids were isolated from the extract [bib_ref] Two new antifungal alkaloids produced by Streptoverticillium morookaense, Feng [/bib_ref]. During our ongoing search for antibacterial natural products from microorganisms [bib_ref] Cytotoxic and antibacterial quinone sesquiterpenes from a Myrothecium fungus, Fu [/bib_ref] [bib_ref] Penicillitone, a potent in vitro antiinflammatory and cytotoxic rearranged sterol with an..., Xue [/bib_ref] [bib_ref] Bisacremines A-D, dimeric acremines produced by a soil-derived Acremonium persicinum Strain, Wu [/bib_ref] [bib_ref] Bisacremines E-G, three polycyclic dimeric acremines produced by Acremonium persicinum SC0105, Wu [/bib_ref] [bib_ref] Acremotins A-D, peptaibiotics produced by the soil-derived fungus Acremonium persicinum SC0105, Wang [/bib_ref] , the culture extract of this strain was found to be also active against S. aureus and MRSA. Thus, the strain SC1169 was reinvestigated by focusing on metabolites responsible for the antibacterial activity, leading to the isolation of eight new fasamycin congeners, streptovertimycins A-H (1-8), from wheat-grown cultures. We report here the isolation, structure elucidation, and antibacterial activity of these compounds.
# Results and discussion
S. morookaense SC1169 was cultivated on wheat grains. The culture was extracted with 95% EtOH and the resulting crude extract was successively partitioned with petroleum ether, EtOAc, and n-BuOH. The EtOAc-soluble fraction was subjected to silica gel column chromatography (CC) followed by ODS CC, Sephadex LH-20 CC, and preparative HPLC, to afford the compounds 1−8 [fig_ref] Figure 1: Structures of streptovertimycins A-H [/fig_ref].
Streptovertimycin A (1) was obtained as a yellow powder. Its molecular formula was determined as C 29 H 25 ClO 7 based on the HR-ESIMS and 13 C NMR data. The 1 H NMR spectrumshowed signals for 11 hydrogen-bearing carbons corresponding to the above non-exchangeable protons as indicated by the HSQC spectrum. Other 13 C NMR resonances included 16 aromatic quaternary carbons, a conjugated carbonyl carbon [δ C 190.2 (C-11)], and an aliphatic quaternary carbon [δ C 38.8 (C-18)]. The connectivities were determined from extensive HMBC correlations [fig_ref] Figure 2: Key HMBC correlations [/fig_ref]. H-C long range correlations from H-2 to C-1 (δ C 137.6), C-3 (δ C 159.6), C-4 (δ C 98.3), and C-6 (δ C 122.5), from H-4 to C-5 (δ C 152.9), C-2 (δ C 107.7), C-6, and C-3, from H 3 -25 (δ H 1.96) to C-1, C-2 and C-6, and from 3-OCH 3 protons to C-3 established a 3,4-disubstituted 1-methoxy-3methylbenzene ring (A); correlations from H-22 to C-21 (δ C 141.8), C-23 (δ C 157.6), C-24 (δ C 121.9), and C-8 (δ C 117.3) and from H-24 to C-7 (δ C 137.6), C-22 (δ C 110.7), C-23, and C-8 constructed a 1,2,3,5-tetrasubstituted benzene ring (B); and the H-24/C-6 correlation revealed a biaryl linkage between C-6 of ring A and C-7 of ring B. The presence of a 1,8-dihydroxy-10,10-dimethylanthracenone fused with ring B via C-8/C-21 was indicated by HMBC correlations from H-20 to C-10 (δ C 107.4), C-18 (δ C 38.8), C-11 (δ C 190.2), C-8, and C-22, from H-16 to C-14 (δ C 109.5), C-15 (δ C 161.4), C-12 (δ C 107.9), C-17 (δ C 151.8), C-11, and C-18, from H 3 -26 (δ H 1.72) and H 3 -27 (δ H 1.76) to C-18, C-19, and C-17, from 9-OH to C-9 (δ C 166.0), C-10, and C-8, and from 13-OH to C-13 (δ C 160.0), C-12, and C-14 [fig_ref] Figure 2: Key HMBC correlations [/fig_ref]. The additional HMBC correlation of the OCH 3 protons at δ H 4.04 with C-15 located the OCH 3 group at C-15. The low field shifts of C-5 (δ C 152.9) and C-23 (δ C 157.6) and the high field shift of C-14 (δ C 109.5), in combination with the molecular formula, suggested the presence of hydroxyl groups at C-5 and C-23 and the chlorine at C-14. Thus, compound 1 was characterized to be a fasamycin-type metabolite with the structure as shown in [fig_ref] Figure 1: Structures of streptovertimycins A-H [/fig_ref].
Once the structure of 1 was established, analysis of UV absorptions (λ max 251, 289, 351, and 425 nm), ESIMS data, and NMR data (Tables 1 and 2) readily found that compounds 2−8 are also fasamycin-type polyketides with structures containing more chlorine atoms than that of 1. Structural variations among these compounds are the position and number of O-methyl groups and chlorine atoms in rings A and B.
Streptovertimycins B (2) and C (3) were both determined to have a molecular formula of C 29 H 24 Cl 2 O 7 , with an additional chlorine atom relative to that of 1. The 1 H NMR data of 2 and 3were closely similar to those of 1 except that one of two sets meta-coupled aromatic protons appeared as a proton singlet, suggesting that one of aromatic protons in ring A or B of 1 is replaced by the second chlorine in 2 and 3. Comparison of 13 C NMR spectrum of 2 with that of 1 readily found that 2 contains an additional aromatic quaternary carbon [δ C 113.9 (C-22)], which was assigned to the second chlorinated carbon on the basis of significantly upfield (by 4.0−5.0 ppm) shifts of C-20, C-21, and C-23 relative to those in 1 as well as the presence of an HMBC correlation of this carbon with the proton singlet [δ H 7.10 (H-24)], which further correlated with C-8 (δ C 117.9) and C-6 (δ C 122.0). The second chlorinated carbon in 3 was located at C-2 (δ C 114.5) as indicated by HMBC correlations of this carbon with H-4 (δ H 5.52) and H 3 -25 (δ H 2.04) together with the upfield 13 C NMR shifts of C-1, C-3, C-5, and C-25 relative to those in 1 and 2. Thus, 4) is also a dichlorinated metabolite. It has a molecular formula containing a CH 2 unit less than that of 2 and 3. Its 1D NMR spectra were almost identical to those of 3. Significant differences found were that the resonances for 3-OMe (δ H 3.94, δ C 56.3) in 3 was absent in 4 and chemical shifts of some carbons in ring A were upfield (C-2, C-3) or downfield (C-4) relative to those in 3. Accordingly, 4 was found to be a 3-O-demethylated analogue of 3.
The molecular formulae of streptovertimycins E (5) and F (6) were both determined as C 28 H 21 Cl 3 O 7 , requiring an additional chlorine atom compared with that of 4. The 1D NMR spectra of 5 and 6 were both closely similar to those of 4. Obvious differences observed between the spectra of 5 and 4 were in the signals attributable to ring B and vicinal positions, including H-20, H-24, and C-19−C-24, which instead, were almost identical with those in 2 (Tables 1 and 2), indicating that C-22 is also chlorinated in 5. The 1D NMR data of 6 differed from those of 4 in the absence of the signals for C-4 methine and the presence of the carbon signal for an additional quaternary carbon (δ C 105.1), indicating that C-4 is the third chlorinated position in 6. The molecular formula of streptovertimycin G (7) was determined to be C 29 H 23 Cl 3 O 7 , with one more CH 2 unit than that of 6. Analysis of the 1D NMR data of 7 revealed a structure similar to that of 6, except for the presence of an additional OMe group (δ H 4.00, δ C 60.8). The attachment of this group at C-5 was indicated by the HMBC correlation of the OMe protons with C-5 (δ C 151.2). Therefore, 7 was a 5-O-methylated analogue of 6.
Streptovertimycin H (8) gave a molecular formula of C 28 H 20 Cl 4 O 7 , indicating the presence of an additional chlorine atom compared with that of 5 and 6. The 1D NMR data of 8 were comparable to those of both 5 and 6, in which the resonances for the carbons of ring A of 8 were almost identical to those of 6, whereas the proton and carbon resonances for ring B and vicinal positions (C-19 and CH-20) were closely similar to those of 5 (Tables 1 and 2), indicating that 8 bears the same ring A as that in 6 and ring B as that in 5. Thus, 8 was characterized as a 22-chlorinated analogue of 6.
Compounds 1−8 exhibited optical activity with various [α] D values due to the presence of biaryl axial chirality [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref]. The configurations of 1, 3, and 6 were solved by theoretical computations of electronic circular dichroim (ECD) spectra. As can be seen in , the calculated ECD spectra for (S)atropisomers of 1, 3, and 6 were all in good fit with the measured spectra, whereas the theoretical spectra of (R)isomers were obvious mirror images of the experimental spectra, which designated these compounds as S. Other compounds were also assigned to this configuration based on the similarity of their measured ECD spectra with those of 1, 3, and 6 . It is interesting that compounds 1−8 and the previously reported congeners of known stereochemistry [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref] [bib_ref] A role for antibiotic biosynthesis monooxygenase domain proteins in fidelity control during..., Qin [/bib_ref] are all (S)-atropisomers, though they are from different microbial sources. This configuration is presumable to be a common characteristic in stereochemistry of this class of natural products.
The isolated metabolites were evaluated for their antibacterial activity against a panel of pathogenic bacteria including four Gram-positive bacteria, S. aureus ATCC6548 (MSSA), methicillin-resistant S. aureus 11646 (MRSA), E. faecium (VSE), and vancomycin-resistance E. faecium (VRE), and two Gram negative bacteria, Escherichia coli ATCC8739 and Shigella dysenteriae CMCC51252. Compounds 1−8 exhibited remarkable antibacterial activity toward Gram-positive bacteria MSSA, MRSA, VSE, and VRE with minimal inhibitory concentration (MIC) values in the range of 0.63−5.0 μg ml −1 , but were inactive (MIC > 50 μg ml −1 ) against Gram negative bacteria E. coli and S. dysenteriae. The results were consistent with previous findings for fasamycins, nephthacemycins, and formicamycins [bib_ref] Naphthacemycins, novel circumventors of β-lactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] Naphthacemycins, novel circumventors of βlactam resistance in MRSA, produced by Streptomyces sp...., Fukumoto [/bib_ref] [bib_ref] KB-3346-5 substances, their fermentative manufacture, and antibacterial agents containing them, Omura [/bib_ref] [bib_ref] Functional analysis of environmental DNA-derived type II polyketides synthase reveals structurally diverse..., Feng [/bib_ref] [bib_ref] Environmental DNA-encoded antibiotics fasamycins A and B inhibit FabF in type II..., Feng [/bib_ref] [bib_ref] Formicamycins, antibacterial polyketides produced by Streptomyces formicae isolated from African Tetraponera plant-ants, Qin [/bib_ref]. As can be seen in [fig_ref] Table 3: Antibacterial activity [/fig_ref] , compounds 4, 5/6, and 8, containing two, three, and four chlorine atoms, respectively, showed a successive drop in the activity, suggesting that more chlorine atoms could not enhance their antibacterial potency. Compounds 1−3 in comparison with compound 4 supported that a free OH group at C-3 is a favoured structural feature for the activity. Furthermore, the 5-O-methyl bearing 7 displayed a 2−4 fold activity superior to the 5-OH containing 6, indicating that methylation of 5-OH could enhance the potency.
In conclusion, the present study discovered eight new fasamycin-type polyketides, streptovertimycins A−H , and demonstrated that S. morookaense SC1169 is a new producer of this class of promising antibacterial substances. Compounds 1−8, with structures distinct from those of reported congeners, also showed strong activity against MRSA and VRE. Their activity profile provided new insights into SARs of fasamycins and congeners, which would be useful clues to structure optimization of this type of antibiotics.
# Material and methods
## General experimental procedure
Optical rotations were measured in MeOH on a Perkin-Elmer 343 spectropolarimeter. UV spectra and CD spectra were obtained simultaneously on a Chirascan CD spectrometer (Applied Photophysics Ltd., England) with MeOH as solvent. 1D NMR, and 2D NMR experiments were performed on a Bruker Avance III 500 MHz spectrometer. HR-ESIMS data were collected on a Bruker maXis Q-TOF mass spectrometer. Preparative and semipreparative HPLC were performed on a Shimadzu LC-6AD pump and a Shimadzu RID-10A refractive index detector with a Shimadzu Shim-packed Pro-ODS column (20 × 250 mm) and a Waters Nova-Pak ® HR C-18 column (7.8 × 300 mm). Silica gel 60 (100-200 mesh, Qingdao Marine Chemical Ltd., Qingdao, China), YMC ODS (75 μm, YMC Co. Ltd., Kyoto, Japan) and Sephadex LH-20 (GE Healthcare, Uppsala, Sweden) were used in CC. Analytical TLC was performed on HSGF254 silica gel plates (0.2 mm, Yantai Jiangyou Silica Gel Development Co. Ltd., Yantai, China); spots were visualized by spraying with 10% H 2 SO 4 solution followed by heating.
## Producing actinomycete and fermentation
The actinomycete, Streptomyces morookaense (formerly Streptoverticillium morookaense) SC1169 as previously described [bib_ref] Two new antifungal alkaloids produced by Streptoverticillium morookaense, Feng [/bib_ref] , was isolated from a soil sample collected in the Dinghu Mountain Biosphere Reserve, Guangdong, People's Republic of China. The culture plugs of SC1169 grown on PDA medium for 10 days were inoculated into twenty five 500 ml Erlenmeyer flasks containing 100 ml of YMG medium (glucose 0.4%, malt extract 1.0%, yeast extract 0.4%, pH 5.5) and shaken on a rotatory (150 rpm) at 28°C for 72 h. Ten millilitres of culture broth was transferred into each of two hundred 500 ml flasks containing 60 ml of YMG medium and 60 g of wheat grains, and the incubation was performed in the dark at 28°C for 30 days.
## Extraction and isolation
The solid fermentation cultures were extracted with 95% EtOH three times. The resultant crude extract was dissolved into water and partitioned sequentially with petroleum ether, EtOAc, and n-BuOH. The EtOAc-soluble fraction (67.5 g) was separated by a silica gel column using CHCl [bib_ref] KB-3346-5 substances, their fermentative manufacture, and antibacterial agents containing them, Omura [/bib_ref] Comparison between the M06/TZVP/PCM calculated and measured ECD spectra of compounds 1, 3 and 6 and comparison of the measured ECD spectra of 1−8 C NMR data, see
## Evaluation of antibacterial activity
Four Gram-positive bacterial strains, S. aureus (ATCC 6538), MRSA (no. 11646), VSE, and VRE, and two Gram negative bacterial strain, E. coli (ATCC 8739) and S. dysenteriae (CMCC 51252), were used for antibacterial evaluation. The reference strains, MSSA, E. coli, and S. dysenteriae, were purchased from Microbial Culture Collection Center of Guangdong Institute of Microbiology (Guangzhou, China). VRE (no. 151458137) and VSE (no. 160119481), the clinical isolates from abdomen of patients, were obtained from The First Affiliate Hospital, Sun Yat-sen University. MRSA (no. 11646) was kindly provided by the State Key Laboratory of Respiratory Disease, First Affiliated Hospital of Guangzhou Medical University (Guangzhou, China). The activity was tested by microplate Alamar Blue assay (MABA) as previously described [bib_ref] Cytotoxic and antibacterial quinone sesquiterpenes from a Myrothecium fungus, Fu [/bib_ref] with brain heart infusion medium for VRE and VSE, and Mueller-Hinton broth medium for other strains. The strains were incubated in medium on a rotary shaker at 150 rpm in 28°C for 12 h. The bacterial suspensions of the strain were quantified with a hemocytometer and adjusted with medium to 1 × 10 5 CFU ml −1 for the MABA. Tested compounds were twofold diluted with the DMSO to give serial concentrations. One hundred microlitres of bacterial suspension of each strain containing Alamar Blue (8%, v/v) and the compound solution (4%, v/v) was added into a 96-well microtiter plate in triplicate. The final concentrations of tested compounds were 10, 5, 2.5, 1.25, 0.63, and 0.31 μg ml −1 . Vancomycin and kanamycin were used as positive control. Negative control wells were added with DMSO instead of the test compound, and blank control wells contained Alamar Blue but without bacterial suspension. The plates were incubated in the dark at 28°C for 6−8 h. The final concentration of the well, which was closest to the red one and remained blue, was deemed as the MIC.
[fig] Figure 1: Structures of streptovertimycins A-H (1-8) compounds 2 and 3 were characterized as the 22-and 2-chlorinated derivatives of streptovertimycin A, respectively. Streptovertimycin D ( [/fig]
[fig] Figure 2: Key HMBC correlations (arrows) of 1 [/fig]
[fig] 13: C NMR data, see Tables 1 and 2; HR-ESIMS m/z 553.0841 [M − H] − (calcd for C 29 H 23 Cl 2 O 7 , 553.0826). Streptovertimycin C (3): yellow powder; [α] 20 D +4.0 (c 0.1, MeOH); UV (MeOH) λ max (log ε) 204 (3.81), 250 (3.35), 291 (3.18), 355 (3.07), 426 (3.11) nm; CD (MeOH) Δε 225 (−10.8), 246 (+4.8), 416 (+0.4); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 553.0837 [M − H] − (calcd for C 29 H 23 Cl 2 O 7 , 553.0826). Streptovertimycin D (4): yellow powder; [α] 20 D −11.0 (c 0.1, MeOH); UV (MeOH) λ max (log ε) 203 (3.87), 250 (3.43), 291 (3.25), 353 (3.17), 427 (3.22) nm; CD (MeOH) Δε 214 (−15.0), 245 (+5.1), 312 (−0.8), 411 (+0.4); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 539.0691 [M − H] − (calcd for C 28 H 21 Cl 2 O 7 , 539.0670). Streptovertimycin E (5): yellow powder; [α] 20 D +43.5 (c 0.2, MeOH); UV (MeOH) λ max (log ε) 203 (3.85), 252 (3.49), 292 (3.29), 351 (3.18), 425 (3.23) nm; CD (MeOH) Δε 214 (−12.4), 252 (+5.0), 316 (−1.6), 440 (+1.0); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 573.0293 [M − H] − (calcd for C 28 H 20 Cl 3 O 7 , 573.0280). Streptovertimycin F (6): yellow powder; [α] 20 D +43.0 (c 0.1, MeOH); UV (MeOH) λ max (log ε) 205 (3.86), 250 (3.37), 291 (3.23), 353 (3.16), 425 (3.19) nm; CD (MeOH) Δε 210 (−8.3), 240 (+3.7), 312 (−1.3); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 573.0302 [M − H] − (calcd for C 28 H 20 Cl 3 O 7 , 573.0280). Streptovertimycin G (7): yellow powder; [α] 20 D +63.5 (c 0.2, MeOH); UV (MeOH) λ max (log ε) 205 (3.90), 252 (3.41), 292 (3.21), 354 (3.15), 426 (3.20) nm; CD (MeOH) Δε 218 (−6.2), 240 (+1.5), 281 (+2.6), 423 (+1.1); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 587.0445 [M − H] − (calcd for C 29 H 22 Cl 3 O 7 , 587.0437). Streptovertimycin H (8): yellow powder; [α] 20 D +18.0 (c 0.1, MeOH); UV (MeOH) λ max (log ε) 204 (3.62), 251 (3.27), 290 (3.07), 351 (2.94), 423 (2.94) nm; CD (MeOH) Δε 210 (−3.7), 250 (+2.6), 293 (+1.3); 1 H and 13 C NMR data, see Tables 1 and 2; HR-ESIMS m/z 606.9901 [M − H] − (calcd for C 28 H 19 Cl 4 O 7 , 606.9890). [/fig]
[table] Table 3: Antibacterial activity (MIC, μg ml −1 ) of compounds 1-8 [/table]
|
Smoking Mechanics and Impact on Smoking Cessation: Two Cases of Smoking Lapse Status Post Lung Transplantation
BACKGROUND:Smoking behavior includes mechanisms taken on by persons to adjust for certain characteristic changes of cigarettes. However, as lung function declines due to lung-specific diseases, it is unclear how mechanical smoking behavior changes affect persons who smoke. We review two cases of patients who stopped smoking prior to and then subsequently resumed smoking after lung transplantation.
# Introduction
Smoking behavior is distinct, highly complex, and dynamic. [bib_ref] Tobacco use across the formative years: a road map to developmental vulnerabilities, Jamner [/bib_ref] Smoking topography attempts to characterize an individual's smoking behavior holistically, addressing cigarette properties, smoking mechanics, and triggers. [bib_ref] Doses of nicotine and lung carcinogens delivered to cigarette smokers, Djordjevic [/bib_ref] [bib_ref] Smoking topography predicts abstinence following treatment with nicotine replacement therapy, Strasser [/bib_ref] [bib_ref] Smoking history, cigarette yield and smoking behavior as determinants of smoke exposure, Bridges [/bib_ref] Topography parameters have been utilized to exhibit the compensatory smoking behavior resulting from cigarette brand-switching. Participants regulated their nicotine intake by smoking more of the low-yield cigarettes in comparison to the high-yield cigarettes. [bib_ref] Puffing behavior during the smoking of a single cigarette in a naturalistic..., Kolonen [/bib_ref] [bib_ref] Influence of smoking fewer cigarettes on exposure to tar, nicotine, and carbon..., Benowitz [/bib_ref] [bib_ref] Reduced tar, nicotine, and carbon monoxide exposure while smoking ultralow-but not low-yield..., Benowitz [/bib_ref] It remains unclear, though, what part of smoking topography is a significant driver in consumption rate and/or abstinence for certain populations of active smokers.
Smoking mechanics is a part of smoking topography involving parameters for puffing, inhalation, and exhalation. [bib_ref] Smoking topography predicts abstinence following treatment with nicotine replacement therapy, Strasser [/bib_ref] [bib_ref] Smoking history, cigarette yield and smoking behavior as determinants of smoke exposure, Bridges [/bib_ref] The inhalation and exhalation parameters of smoking mechanics depend, in part, on the condition of airways and lung tissue. Airway strength can be measured with forced expiratory volume (FEV1) and forced vital capacity (FVC). While smoking has been shown to reduce variables such as FEV1, [bib_ref] Smoking and lung function of Lung Health Study participants after 11 years, Anthonisen [/bib_ref] [bib_ref] Persistence of the effect of the lung health study (LHS) smoking intervention..., Murray [/bib_ref] it is unclear if such changes impact smoking continuation and/or cessation or overall smoking topography. It is possible that patients with diminished lung function may smoke less of a cigarette due to mechanical limitations, suggesting an interplay between smoking mechanics and lung health.
Lung transplantation is currently a treatment option for those with end-stage lung diseases. [bib_ref] Selection of candidates for lung transplantation, Kreider [/bib_ref] For lung transplant candidates, active smoking is a contraindication to transplant, and many centers recommend a period of abstinence from smoking prior to waitlisting. [bib_ref] ISHLT/APM/AST/ ICCAC/STSW recommendations for the psychosocial evaluation of adult cardiothoracic transplant candidates..., Dew [/bib_ref] Around 40% are active smokers before approaching lung transplantation candidacy. [bib_ref] Smoking resumption after lung transplantation: a sobering truth, Vos [/bib_ref] Unfortunately, smoking cessation therapy compliance of transplant recipients is low, with 10-40% relapsing post-lung transplantation. [bib_ref] Smoking resumption after lung transplantation: a sobering truth, Vos [/bib_ref] [bib_ref] Tobacco smoking and solid organ transplantation, Corbett [/bib_ref] The role of smoking mechanics in successful smoking cessation and smoking relapse in post-lung transplant patients has yet to be determined.
In this study, we present the cases of two patients with tobacco dependence who received lung transplants after maintaining smoking abstinence, though relapsed to tobacco usage after transplantation. We discuss in detail how changes in their lung function affected their smoking mechanics and may have impacted their relapse to smoking.
# Methods
For both patients, we collected several variables to better understand their smoking phenotype. This included brand of cigarette, smoking topography (variables of smoking mechanics such as puff duration, inhalation duration, and puffs per cigarette), [bib_ref] Smoking topography predicts abstinence following treatment with nicotine replacement therapy, Strasser [/bib_ref] and triggers to smoking. Further, we identified the age that they began to smoke consistently (defined as daily use of cigarettes) as well prior quit attempt strategies and duration of abstinence from smoking. The study was approved by Institutional Review Board at Johns Hopkins School of Medicine (IRB00282725) and all actions undertaken by the authors were in accordance with the Declaration of Helsinki, along with written consent for both patients.
Screening for smoking cessation at our institution takes on both self-report and laboratory data. Specifically, we screen for urine nicotine and cotinine. If the patient is using nicotine replacement therapies (NRT) to assist in smoking cessation, then we screen for anabasine to assure the nicotine and cotinine are from an NRT source and not tobacco products. [bib_ref] Anabasine and anatabine as biomarkers for tobacco use during nicotine replacement therapy, Jacob P 3rd [/bib_ref] Pretransplantation, the aforementioned laboratory data is monitored until the patient achieves a status of 6-months smoke-free. For patients who are post-lung transplant, since there are no active international guidelines suggesting how to monitor for smoking relapse, [bib_ref] Smoking resumption after heart or lung transplantation: a systematic review and suggestions..., Hofmann [/bib_ref] at our institution we continue to assess smoking status with self-reports and with the aforementioned smoking-related biomarkers. These biomarkers are ordered at the discretion of the transplant team in conjunction with clinical assessments of the patient.
We collected data on their pulmonary-specific morbidity, as well as any other morbidity that would influence the patient's tobacco dependence (e.g., major depressive disorder and general anxiety disorder). In regards to the patient's pulmonary-specific morbidity, we documented management strategies prior to pre-transplantation, pulmonary function testing, and management of tobacco dependence. Post-transplantation, we collected data on ongoing management of the patient's tobacco dependence and pulmonary function testing. Finally, we selected patients who relapsed into smoking post-lung transplant and, after confirmation of their smoking resumption, were enrolled into the Tobacco Treatment Clinic at the Johns Hopkins University School of Medicine.
# Results
## Patient a
Patient A is a 54-year-old woman status post a double lung transplant secondary to chronic obstructive pulmonary disease (COPD). The date of transplant was October 2017. Her COPD was considered a GOLD Class D given her exacerbation history, and was managed with an inhaled long-acting muscarinic antagonist, inhaled corticosteroid, inhaled long-acting beta2 agonists, as well as chronic macrolide (azithromycin), and an as needed short-acting beta2 agonist inhalers and nebulizers. She also underwent pulmonary rehabilitation for her COPD management as well. Given the inability to improve her symptoms as well as her overall prognosis, the patient began to pursue candidacy for lung transplantation.
Patient A quit smoking in 2015, with an FEV1 at the time of .56 L (21% predicted). She used varenicline to help with her smoking cessation strategy in 2015. She discussed that she has been smoking since the age of 15, and smoked consistently a pack of cigarettes a day, if not more, up until her early 40s. When she was diagnosed with COPD in 2004, she found herself frequenting hospitalizations and outpatient clinics due to exacerbations, often making it difficult to smoke given her "breathlessness." She discussed her smoking mechanics as long puffs, deep inhalations, as well as taking more than 5 puffs per cigarette [fig_ref] Table 1: Patient characteristics and smoking topography [/fig_ref]. Her choice of cigarette usage is Marlboro Reds, smoking a pack or more a day. She identified triggers to smoking as stress and anxiety, specific places (e.g., bathrooms of her home and her car), and after food. She never attempted to quit in the past, but did have periods without smoking, often due to COPD-related health issues. For the purpose of obtain a lung transplant, she was started on varenicline, using the medication for 6-weeks in total, stopping in July 2015 and receiving a transplant in 2017.
Patient A discussed a lapse in smoking at 5-months posttransplantation after identification of a positive urine nicotine and cotinine level, sent at the discretion of the transplant team given a decline of lung function on pulmonary function testing. The patient highlighted that stress at that time influenced significantly the need to re-engage with smoking. The ability to smoke in the manner she was accustomed to (long puffs, long inhalation) reaffirmed as well a "satisfaction" gain from smoking. At the time of relapse, her FEV1 had increased to 2.12 L, an over 200% increase in FEV1 from the patient's prior baseline when she began to abstain from cigarettes.
## Patient b
Patient B is a 40-year-old man status post a double lung transplant secondary to sarcoidosis. The date of transplant was January 2015. The impact of sarcoidosis on the patient's lungs was severe, with pre-transplant imaging identifying extensive central bronchiectasis, interstitial lung disease present with predominance near hila and upper lobes. Further, there was a cavitary lesion in the right upper lobe as well. Patient B was managed mainly with prednisone and short-acting beta agonist inhalers, though he had frequent hospitalizations for his sarcoidosis and resulting pulmonary symptoms of dyspnea. As his disease progressed, he discussed his ability to breathe was a challenge, even at rest, coupled with excessive sputum production. Both of these symptoms impaired his ability to smoke, ultimately resulting in his abstinence transition. Patient B began to smoke at the age of 12. His preferred cigarette choice was Newport Slims, smoking up to a pack of cigarettes a day. He described his smoking mechanics as long puffs with a long inhalation, and would smoke the cigarette completely [fig_ref] Table 1: Patient characteristics and smoking topography [/fig_ref]. He identified his triggers to smoke as stress and anxiety, after eating, and certain areas (e.g., in his car). He eventually stopped smoking, identifying while he wanted a lung transplant, he found it difficult to smoke in the manner he preferred. His abstinence was through avoidance and without pharmacological assistance. His last cigarette was in February 2012, with an FEV1 1.15 L at the time.
Patient B resumed smoking in April 2017, and discussed this at a lung transplant office visit. While it was self-reported, it was confirmed with urine nicotine and cotinine levels being positive. He identified stress as the main variable resulting in his lapse to smoking. With his improved lung function, he was able to resume smoking with the lung mechanics that were consistent with his prior usage of cigarettes. He lapsed at 26-months postlung transplantation. His FEV1 at the time of resuming smoking cigarettes was 2.37 L, more than double of FEV1 at the time of abstinence from smoking.
# Discussion
We highlight two patients status post-lung transplantation who relapsed to smoking, with both patients identifying the ability to smoke with improved lung function mechanics contributing to resuming smoking of cigarettes. Other similarities these patients shared was initiation of smoking at a young age (under the age of 18) and variables that result in cravings for cigarettes, specifically that of anxiety and stress. Such findings warrant further discussion in reaffirming tobacco dependence as a chronic disease whose management should continue in recipients of lung transplant.
The dynamics of smoking mechanics is associated with many unique variables, for both the individual smoking and in regards to the cigarette itself. For instance, the nicotine yield of a cigarette may dictate the puff duration, puff volume per cigarette, and number of puffs per cigarette. [bib_ref] Puffing behavior during the smoking of a single cigarette in a naturalistic..., Kolonen [/bib_ref] Further, the nicotine itself from cigarette smoke may provide a positive consequence for a person actively smoking. [bib_ref] Smoking with reason, Warburton [/bib_ref] For instance, in anxiety disorders may increase smoking behavior and the risk for nicotine dependence, as cigarettes may be seen as an anxiolytic selftreatment. [bib_ref] Depression, anxiety, and smoking initiation: a prospective study over 3 years, Patton [/bib_ref] [bib_ref] Distal and proximal family predictors of adolescents' smoking initiation and development: a..., Tjora [/bib_ref] Both patients' description of their smoking mechanics, specifically their puff duration and inhalation hold (puff volume), appeared to be impacted by specific external influences (stress and anxiety). When faced with these triggers to smoke in the past, during low FEV1 variables, their inability to smoke in the manner they desired resulted in the ability to pursue abstinence. However, while faced with the same triggers post-transplant, the ability to match their desired smoking mechanics due to the improved FEV1 to the trigger that is driving their craving for nicotine (anxiety and stress) resulted in a lapse to cigarette consumption.
A significant insight into the lapse of smoking by both patients' status post post-lung transplantation is in how it reaffirms tobacco dependence and nicotine addiction as a chronic disease. For instance, proteins impacting tolerance and craving in nicotine addiction, specifically cAMP response element-binding protein (CREB) and delta fos b, remain elevated after an individual's last cigarette for a significant amount of time (if not permanent). [bib_ref] Effects of protracted nicotine exposure and withdrawal on the expression and phosphorylation..., Pandey [/bib_ref] [bib_ref] DeltaFosB: a sustained molecular switch for addiction, Nestler [/bib_ref] Both proteins play a significant role in nicotine's ability to impact the brain's reward system as well as the conditioning that occurs in regards to person's compulsion towards smoking.Further, if nicotine exposure occurs at a young age, at a time when the prefrontal cortex continues to evolve, epigenetic changes may sensitize the brain to nicotine craving and future and continued use of cigarettes. This is the case for both patients, who began smoking while they were adolescents. Therefore, insight into both active smoking topography as well as past usage of cigarettes (e.g., age of initiation of smoking) may result in an understanding of a smoking phenotype that warrants management both before and after lung transplantation.
Smoking resumption post solid organ transplant is not an uncommon diagnosis in transplant centers. In regards to postlung transplantation, smoking resumption has been reported from 0 to 10% in the United States to up to 15% at a European transplant center. [bib_ref] Cigarette smoking following lung transplantation: effects on allograft function and recipient functional..., Bauldoff [/bib_ref] [bib_ref] Continued smoking in lung transplant patients: a cross sectional survey, Zmeskal [/bib_ref] Smoking cessation and smoking relapse are both complex, with multiple factors contributing to cessation and relapse. Improved lung function is an exceptional outcome for patients after lung transplantation, one that aims to improve a person's quality of life. However, the challenge with improved lung function is it may also contribute to smoking resumption, a plausible parallel since the opposite, a decreased lung function resulting in smoking cessation, was confirmed in our case series. Therefore, as Hofmann et al. [bib_ref] Smoking resumption after heart or lung transplantation: a systematic review and suggestions..., Hofmann [/bib_ref] of smoking resumption in post-transplant patients is likely underestimated, [bib_ref] Smoking resumption after heart or lung transplantation: a systematic review and suggestions..., Hofmann [/bib_ref] we believe our case series adds to the necessity that monitoring of smoking resumption should be standard of care for post-lung transplantation evaluation. Tobacco dependence and nicotine addiction are chronic diseases with many factors, both intrinsic and extrinsic, affecting the ability to quit and ultimately become tobacco independent. For patients undergoing lung transplant evaluation, a need to abstain from cigarettes prior to transplant should be met with similar priorities post-transplant. The management, like other chronic diseases, warrants consistent reassessments. We believe the aforementioned cases emphasize the need for international guidelines to offer recommendations on screening for smoking resumption post-lung transplantation and availability of immediate resources to help patients if they should relapse in an effort to preserve transplanted lungs.
## Orcid id
Panagis Galiatsatos https://orcid.org/0000-0002-7752-3841
[table] Table 1: Patient characteristics and smoking topography. [/table]
|
Low Serum Vitamin D in COVID-19 Patients Is Not Related to Inflammatory Markers and Patients’ Outcomes—A Single-Center Experience and a Brief Review of the Literature
Citation: Hut , anu, A.; Georgescu, A.M.; Voidȃzan, S.; Andrejkovits, A.V.; Negrea, V.; Dobreanu, M. Low
# Introduction
Vitamin D is a fat-soluble vitamin with an important role in multiple metabolic and immune processes. The body has the capacity to synthesize a large proportion of D3-cholecalciferol when exposed to UVB radiations; it needs further activation by two hydroxylation steps-hepatic and renal [bib_ref] The immunological implication of the new vitamin D metabolism. Cent, Bivona [/bib_ref]. There are many controversies about classifying vitamin D as a hormone. Unlike other vitamins, it is the only one produced by endogenous synthesis from a cholesterol precursor [bib_ref] Vitamin D: Vitamin or Hormone?, Ellison [/bib_ref]. Vitamin D has multisystemic effects mediated through the vitamin D receptor (VDR) present in various tissues [bib_ref] Steroid Hormone Vitamin D, Demer [/bib_ref] , including immune cells [bib_ref] Immunologic Effects of Vitamin D on Human Health and Disease, Charoenngam [/bib_ref] , and more specifically on leucocytes [bib_ref] 1,25-dihydroxyvitamin D3 receptors in human leukocytes, Provvedini [/bib_ref] , monocytes [bib_ref] Increased vitamin D receptor expression from macrophages after stimulation with M. tuberculosis..., Fiske [/bib_ref] , and dendritic cells [bib_ref] Regulation of Dendritic Cell Function by Vitamin D, Barragan [/bib_ref]. Activation through T cell receptors (TCRs) induces upregulation of VDR in naïve human T cells and enhances the vitamin D-VDR signaling, important in T cell differentiation and Mures , County Clinical Hospital, Târgu Mures , , Romania. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Hospital Ethics Committee . Before initiation of any study procedures and recruitment, the signed informed consent was obtained from the enrolled patients.
Additionally, for comparison purposes, we included a control group derived from the national program for vitamin D testing. For this group (considered here as a control group, COVID-19 negative), we analyzed only the vitamin D serum levels in relation to age and gender; we did not use all personal data, comorbidities, or vitamin D supplementation.
## Inclusion criteria
Adult patients (>18 years old) hospitalized with COVID-19 (Coronavirus Disease 2019) were included. The infection was confirmed with RT-PCR and subsequently, the patients were admitted to the infectious disease department for isolation and treatment; the cohort of patients was recruited between January 2021 and September 2021, when the alfa (B.1.1.7), beta (B.1.351), as well as the delta (B.1.617.2) variant were in circulation.
For the control group, the serum vitamin D levels were retrospectively retrieved from the laboratory information system (LIS), including ambulatory care patients between January 2020 and March 2022. The median for each age group was used.
## Exclusion criteria
Patients ≤ 18 years old or diagnosed with possible causes of immunodeficiency (malignancies, HIV infection, hemopathies, or pregnancy) were excluded. Patients unable to read and sign the written informed consent were also excluded.
## Clinical evaluation for disease severity
Patients were classified into three severity categories according to the updated treatment guidelines from the National Institutes of Health (https://www.covid19treatmentguidelines. nih.gov/overview/clinical-spectrum/, 19 October 2021, accessed on 22 March 2022):
## 1.
Mild infection: patients who had any of the various signs and symptoms of COVID-19, but without shortness of breath, dyspnea, or abnormal chest imaging.
## 2.
Moderate infection: patients with evidence of lower respiratory disease during clinical assessment or imaging and with an oxygen saturation (SpO2) ≥ 94% on room air.
## 3.
Severe infection: SpO2 < 94% on room air, a respiratory rate > 30 breaths/min, or lung infiltrates > 50%.
The patients with severe respiratory failure or acute single or multiple organ dysfunctions were transferred to the intensive care unit (ICU).
## Sampling procedure and data collection
Demographic data, duration of hospitalization, clinical characteristics (chronic comorbidities, admission to the ICU, survival/clinical outcome, and treatment administrated), and laboratory parameters were routinely collected during the hospital admission. Whole blood collected on EDTA was used for complete blood count (CBC) and analyzed using a fully automated hematological analyzer Sysmex XS-800i (Sysmex Corp, Kobe, Japan). Serum samples collected on clot activator tubes were used for inflammatory serum biomarker evaluation, C-reactive protein (CRP), and ferritin, assessed on Architect 4000c (Abbott Laboratory). Blood samples were collected during the first day after admission; the serum levels of 25-OH vitamin D (total vitamin D) were evaluated from serum aliquots stored at −80 - C.
For 25(OH) vitamin D serum levels, a fully automated assay was used based on a capture electrochemiluminescence (ECLIA) method on a Cobas e411 instrument (Roche Diagnostics). The test allows the quantification of both D3 (25-OH) and D2 (25-OH) vitamins for a total vitamin D estimation. The test protocol uses ruthenium-labeled vitamin D binding protein (VDBP) for analyte capture, and streptavidin-coated microparticles and biotin-labeled 25-OH vitamin D for competition (www.roche.com, accessed on 22 March 2022). According to the American Endocrine Society, the recommended levels for vitamin D range between 30-50 ng/mL, insufficiency is defined as 21-29 ng/mL, while vitamin D deficiency is defined as a 25(OH) vitamin D ≤ 20 ng/mL [bib_ref] Evaluation, treatment, and prevention of vitamin D deficiency: An endocrine society clinical..., Holick [/bib_ref].
The inflammatory indices were calculated from the routine blood cell count and used in the analyses to fully characterize the relationship between inflammation and vitamin D status in patients with COVID-19.
Neutrophil to lymphocyte ratio (NLR) was defined as: NLR = # neutrophils/# lymphocytes, where # denotes the absolute values of neutrophils and lymphocytes Platelet to lymphocyte ratio (PLR) was calculated as: PLR = platelets/# lymphocytes Systemic inflammatory index (SII) was also defined as a ratio and expressed as ×10 3 /µL: SII = (# neutrophils X platelets)/# lymphocytes The SII was identified as a strong predictor for severe disease and invasive ventilation, as well as a worse clinical outcome.
# Statistical analysis
For the statistical analysis, the SPSS software version 22 and Graph Pad 3.6 State Software, San Diego, CA, USA, were used; p < 0.05 was considered to be the threshold for significance. For the normal data distribution estimated with the Kolmogorov-Smirnov test, the results were expressed as mean ± SD, while for nonparametric distribution, median and min.-max. values or interquartile range (IQR) were used. The categorical variables were expressed as number (n) and percentage (%) of patients from the cohort. A chi-square test was used in order to compare the frequencies of nominal variables. Differences in the mean or median between groups were analyzed using the t-test, Mann-Whitney test, and Kruskal-Wallis test with Dunn's test for multiple comparison groups. Multivariate regression of predictive factors used regression of in-hospital mortality as the dependent variable and demographic parameters, vitamin D status, and inflammatory biomarkers as independent predictive variables in a single model. Results were expressed as hazard ratios (Expβ) with 95% CIs.
# Results
## Characterization of the patients with covid-19
A total of 203 patients infected with SARS-CoV-2 were analyzed for the vitamin D status in relation to disease severity. The demographic characteristics of the studied population are detailed in [fig_ref] Table 1: Distribution of patients in relation to demographic data and comorbidities [/fig_ref] ; the age classification was carried out according to the WHO criteria: young age, from 25 to 44 years old; middle age, from 44 to 60 years old; elderly age, between 60 and 75 years old; and senior age, 75 to 90 years old [bib_ref] Age Periods of Human Life, Dyussenbayev [/bib_ref]. One very young person, 21 years old, was included in the study and classified accordingly. The median age of the studied group was 64 (21-87) years, the majority of the patients (89.7%) were older than 45 years, with 58.1% male and 41.9% female. The severe (50.7%) and moderate (42.4%) cases were more frequent than the mild (6.9%) forms of COVID-19. The majority of patients (84.3%) presented at least one comorbidity; the most prevalent was hypertension (62.6%).
## Comparative analysis of the vitamin d serum levels for covid-19 patients and the control group
The retrospective analysis of vitamin D serum levels in outpatients, addressed through the national vitamin D program, from the laboratory information system (LIS) was performed. After the exclusion of the outliers, a very high number of results remained (n = 1830), which is nine times higher than the patients' group. Out of the total of 1830 subjects, 1593 (87%) were women and only 237 (13%) were male. The median vitamin D serum level for the control group was slightly lower 27.12 (IQR 21.14-36.62) ng/mL for women compared to 29.08 (IQR 21.81-35.89) ng/mL for males, without reaching the statistical significance (p = 0.678). The median value for the overall control group was 28.13 (IQR 21.22-36.61) ng/mL, a value close to the vitamin D levels in mild COVID-19 patients.
After the analysis of the vitamin D serum levels in relation to age in the control group, the corresponding vitamin D values were identified: for the [bib_ref] Vitamin D and COVID-19 severity and related mortality: A prospective study in..., Campi [/bib_ref] [bib_ref] Vitamin D Serum Levels in Subjects Tested for SARS-CoV-2: What Are the..., Gallelli [/bib_ref].08) ng/mL was found; p = 0.001 when comparing the extreme groups of youngest and oldest patients and also p = 0.0001 for multiple group comparisons.
The comparative analysis of vitamin D serum levels between COVID-19 patients and non-COVID patients is detailed in [fig_ref] Table 2: The comparative analysis of vitamin D serum levels between COVID-19 and non-COVID-19... [/fig_ref]. Except for the age group 21-44 years, there was a significant difference in vitamin D serum levels, with higher values in non-COVID-19 subjects compared to COVID-19 patients. However, due to the difference between the number of subjects for the two compared groups (with and without COVID-19), the results must be interpreted with caution. Additionally, while in the studied COVID-19 patients, there was a relative balance regarding the gender (58.1% male and 41.9% female), among the subjects included in the control group (COVID-19 negative) for the retrospective analysis of the vitamin D levels, women were prevalent (87%).
## Characterization of the covid-19 patients in relation to vitamin d status and inflammatory biomarkers
In our study, vitamin D deficiency was observed in 118 patients (57.20%) with a median of 11.85 (IQR 8.95-16.38) ng/mL; it was more prevalent in significantly older patients.
The main demographic, clinical and biological characteristics of the COVID-19 patients after dichotomization of the study population according to the vitamin D status (vitamin D deficiency ≤ 20 ng/mL) are presented in detail in [fig_ref] Table 2: The comparative analysis of vitamin D serum levels between COVID-19 and non-COVID-19... [/fig_ref]. The median time for hospitalization was 13 days (min. 1-max. 36 days), n = 39 (18.75%) patients were admitted or transferred to ICU, and n = 25 (12.3%) of the subjects died during hospitalization. Except for age and heart rate, there was no difference between the two groups [fig_ref] Table 3: The demographic and clinical data after patient dichotomization according to vitamin D... [/fig_ref]. After Spearman correlation analysis, we did not find a significant relationship between vitamin D serum levels and disease severity stratified by four different age groups, possibly due to the relatively small sample size/group.
After the analysis of the vitamin D serum levels in relation to age, the smallest values were identified at the extreme ages (76-90 years) with a median of 13.0 (IQR 8.83-21.55) ng/mL, while for the 21-44 age group, the median value of vitamin D was 19.07 (IQR 16.77-29.51) ng/mL, p = 0.005 when comparing the extreme groups of youngest and oldest patients and p = 0.0001 for multiple group comparisons . The analysis of the vitamin D status in relation to disease severity showed the highest value in mild COVID-19 patients, with a median of 29.39 (IQR 12.12-44.02) ng/mL, while for moderate and severe forms, the serum levels were significantly lower, with a median of 15.10 (IQR 9.56-24.11) ng/mL for moderate forms, and 18.86 (IQR 12.50-27.88) ng/mL for severe forms, and p = 0.009 for all cases [fig_ref] Figure 2: Serum levels of vitamin D in different disease severity [/fig_ref]. The highest prevalence of moderate and severe COVID-19 forms was found among vitamin D deficient patients. A significant difference in serum levels of vitamin D was found in hy tients, with a median of 16.0 (IQR 9.85-25.28) ng/mL compared to those w tension, with a median of 19.71 (IQR 12.88-28.41) ng/mL, p = 0.01. Similarly were found in patients with diabetes mellitus, with a median of 11.03 (IQ ng/mL compared to patients without diabetes mellitus, where the median w 11.78-27.60) ng/mL, p = 0.008. For other comorbidities, we did not find a s nificant difference in vitamin D serum levels (p > 0.05).
The univariate analysis among survivors (n = 178) and non-survivors tailed in [fig_ref] Table 4: The characteristics of the surviving vs [/fig_ref]. Except for the inflammatory biomarkers, the vitamin D significantly different between groups. When vitamin D level was analyzed in relation to comorbidities, the patients with no personal history of illness showed a significantly higher level of vitamin D, with a median of 24.72 (IQR 16.05-31.52) ng/mL compared to subjects with at least one comorbidity, with a median of 16.02 (IQR 9.81-25.22) ng/mL, p = 0.004. A significant difference in serum levels of vitamin D was found in hypertensive patients, with a median of 16.0 (IQR 9.85-25.28) ng/mL compared to those without hypertension, with a median of 19.71 (IQR 12.88-28.41) ng/mL, p = 0.01. Similarly, lower values were found in patients with diabetes mellitus, with a median of 11.03 (IQR 7.45-20.66) ng/mL compared to patients without diabetes mellitus, where the median was 18.75 (IQR 11.78-27.60) ng/mL, p = 0.008. For other comorbidities, we did not find a statistically significant difference in vitamin D serum levels (p > 0.05).
The univariate analysis among survivors (n = 178) and non-survivors (n = 25) is detailed in [fig_ref] Table 4: The characteristics of the surviving vs [/fig_ref]. Except for the inflammatory biomarkers, the vitamin D status was not significantly different between groups.
None of the analyzed laboratory inflammatory parameters showed significant differences between the groups with and without vitamin D deficiency (p > 0.05). Demographic and laboratory parameters (age, leucocytes, NLR, PLR, SII, ferritin, and leucocytes) were each significantly related to in-hospital mortality while vitamin D serum levels were not [fig_ref] Table 5: The relationship for the in-hospital mortality with demographic parameters, vitamin D status,... [/fig_ref].
# Discussion
In this study, we assessed the status of vitamin D in patients infected with SARS-CoV-2. Our patient cohort was graded as mild to moderate/severe COVID-19 based on clinical findings, imagistic parameters and SpO 2 on room air, the necessity for mechanical ventilation, and the presence of pulmonary infiltration. The extensive pulmonary and multiorgan lesions after the activation of the cytokine storm ultimately lead to acute respiratory distress syndrome and multiorgan failure, respectively. There are several studies that underline the effect of vitamin D in modulating the inflammatory response. Interestingly, a recent paper by Kongsbak-Wismann et al. showed a normal T cell and B cell response during a SARS-CoV-2 infection in one homozygous and two heterozygous patients for mutations in VDR. These mutations totally abolish transcriptional activity after the vitamin D binds the receptor. In the above-mentioned patients, the specific T cell response characterized by IFN gamma, IL-2, and TNF secretion as well as B cell response characterized by anti-RBD IgG antibodies were preserved. This demonstrates that an efficient immune response during SARS-CoV-2 infection does not require vitamin D signaling; however, vitamin D may influence the course of the disease by modulating other antiviral mechanisms, such as antibacterial peptides. An experimental study on vitamin D supplementation in mice with Pneumocystis murina pneumonia revealed its anti-inflammatory actions due to the cytokine synthesis modulation, increased expression of toll-like receptors (TLRs) 2 and 4, and glutathione reductase, thus reducing the oxidative stress [bib_ref] Mechanisms of action of vitamin D as supplemental therapy for Pneumocystis pneumonia, Lei [/bib_ref].
The reduction in the vitamin D levels with age and the seasonal variation, with the lowest values during the cold season, seems to be in relation to higher case-fatality rates in older patients [bib_ref] Evidence that vitamin d supplementation could reduce risk of influenza and covid-19..., Grant [/bib_ref] and similarly with an increased seasonal incidence of the viral pathology. In a recent study, Dror et al. found that hypovitaminosis D before SARS-CoV-2 infection was associated with more severe/critical illness, compared to patients with moderate/mild form, and the mortality rate was higher among patients with vitamin D insufficiency [bib_ref] Pre-infection 25-hydroxyvitamin D3 levels and association with severity of COVID-19 illness, Dror [/bib_ref]. Similar to the previously mentioned study, we found that vitamin D deficiency is prevalent in moderate/severe disease forms and advanced age, compared to mild disease forms and younger subjects. A recent article published by Gallelli et al. on a group of 117 individuals studied with and without COVID-19 revealed a significant difference in vitamin D status between groups, with lower levels found in acute COVID-19 compared to non-COVID-19 patients. Similarly, when we analyzed the vitamin D status in COVID-19 patients versus the control group for the same age categories, we found a significant difference, with lower values in infected patients compared to controls (p < 0.0001), except for the age group 20-44 years (p > 0.05). Additionally, in patients with severe hypovitaminosis D, Gallelli et al. found that IL-6 serum levels were not significantly increased; however, the supplementation with active vitamin D improved the clinical status and IL-6 serum levels in acute disease [bib_ref] Vitamin D Serum Levels in Subjects Tested for SARS-CoV-2: What Are the..., Gallelli [/bib_ref]. Another study found a significant relationship between low levels of vitamin D and increased inflammatory biomarkers, as well as the disease severity. A report by Jain et al. revealed that inflammatory biomarkers are significantly increased in the hypovitaminosis D group and associated with a higher fatality rate [bib_ref] Analysis of vitamin D level among asymptomatic and critically ill COVID-19 patients..., Jain [/bib_ref]. Although we did not find a significant correlation between inflammatory biomarkers and vitamin D levels, the proportion of transfers to ICU was higher among vitamin D deficient patients.
De Smet et al. found a strong association between low vitamin D levels on admission and mortality; however, the study included only patients with severe respiratory syndrome and the endpoint of the survival rate [bib_ref] Serum 25(OH)D Level on Hospital Admission Associated With COVID-19 Stage and Mortality, Smet [/bib_ref]. In our study, half of the patients were with mild/moderate forms of COVID-19, which might be a reason for the lack of predictive capacity for vitamin D levels related to patients' outcomes. In a retrospective case-control study, authors found lower levels of vitamin D in hospitalized patients compared to the control group; however, when considering the COVID-19 group, there was no correlation between vitamin D levels and disease severity [bib_ref] Vitamin D Status in Hospitalized Patients with SARS-CoV-2 Infection, Hernández [/bib_ref]. In the same study, no other inflammatory biomarkers were found to be significantly different between patients with and without vitamin D deficiency [bib_ref] Vitamin D Status in Hospitalized Patients with SARS-CoV-2 Infection, Hernández [/bib_ref].
The supplementation strategy and the frequency of serum vitamin D level determination for both COVID-19 prevention and severity were extensively reviewed by Grant et al. These authors stated that most randomized controlled trials have not reported a significantly reduced risk of infection after vitamin D supplementation [bib_ref] Evidence that vitamin d supplementation could reduce risk of influenza and covid-19..., Grant [/bib_ref].
Since there are some reports on the lack of relation between low levels of vitamin D and inflammation biomarkers, it is possible that the mechanisms for positive effects on viral infections are modulated by means of antibacterial peptides, cytokine, and chemokine response in immune cells exposed to the pathogens, or by the effects on viral cell entry or viral replication, and modulatory effect on IFN signaling pathway. Indeed, a recent review highlighted the important role of vitamin D in the innate antiviral immune response [bib_ref] Emerging Roles of Vitamin D-Induced Antimicrobial Peptides in Antiviral Innate Immunity, White [/bib_ref]. There is increasing evidence that calcitriol is involved in the production of antimicrobial peptides (AMP), enhancement of TLR signaling, and more importantly, the induction of LL-37 synthesis [bib_ref] Emerging Roles of Vitamin D-Induced Antimicrobial Peptides in Antiviral Innate Immunity, White [/bib_ref]. The review underlined the LL-37 in antiviral defense by multiple mechanisms, one of the important ones being the blockade of the interaction between spike SARS-CoV-2 protein and ACE2 receptor, thus blocking the viral entry. Additionally, LL-37 facilitates the recognition of the nucleic viral material by the endosomal TLRs and activates the IFN response through TRL3, TLR7, and TLR9 signaling [bib_ref] Emerging Roles of Vitamin D-Induced Antimicrobial Peptides in Antiviral Innate Immunity, White [/bib_ref]. Moreover, human betadefensin-2 (HBD-2) acts as a calcitriol-induced antiviral protein with a potential protective role in SARS-CoV-2 infection [bib_ref] Emerging Roles of Vitamin D-Induced Antimicrobial Peptides in Antiviral Innate Immunity, White [/bib_ref].
In addition to activation of vitamin D by renal alfa 1 hydroxylase, the active vitamin D forms could be produced under the collective action of VDR (considered a ligandactivated transcription factor) and CYP27B1 extensively expressed on various cells, including macrophages and dendritic cells. This extrarenal production of active vitamin D is regulated by cytokine Th1 signaling (IFN gamma) and TLR1 and TLR2. A sufficient level of vitamin D is required for a normal IFN gamma-induced antimicrobial peptide expression in macrophages [bib_ref] Vitamin D Is Required for IFN-γ-Mediated Antimicrobial Activity of Human Macrophages, Fabri [/bib_ref] , yet a dysregulated or delayed IFN gamma response will allow viral immune evasion [bib_ref] Innate immunology in COVID-19-A living review. Part I: Viral entry, sensing and..., Coveney [/bib_ref] and severe forms of COVID-19 after IFN gamma upregulation during the later phase with enhanced cytokine production and hyper-inflammation, thus leading to severe disease. However, a randomized clinical trial evaluating the effect of a single high dose of vitamin D3 in patients with COVID-19, found no significant difference for the in-hospital mortality between group and placebo, although after a single dose, the vitamin D levels increased significantly from 19.8 ng/mL to 44.4 ng/mL [bib_ref] Effect of a Single High Dose of Vitamin D3 on Hospital Length..., Murai [/bib_ref].
A study by Mardani et al. revealed lower levels of vitamin D in positive SARS-CoV-2 patients compared to the negative group, and that patients with lower vitamin D levels exhibit a higher expression of angiotensin-converting enzyme (ACE) [bib_ref] Association of vitamin D with the modulation of the disease severity in..., Mardani [/bib_ref] , thus interfering with the rennin-angiotensin-aldosterone system RAAS and proper functionality of the immune system through VDR. It is worth mentioning that the synthesis of active vitamin D is maintained even at low/deficient substrate (precursor) levels for up to several months.
Our results are similar to other recent publications revealing a lack of association between vitamin D levels and severity, or disease outcome [bib_ref] Lack of association between vitamin D insufficiency and clinical outcomes of patients..., Davoudi [/bib_ref]. On the other hand, in a systematic review of 54 studies regarding the outcome of SARS-CoV-2 infection and vitamin D status, the authors found a strong relationship between low vitamin D levels and admission to ICU and mortality due to COVID-19, or susceptibility to SARS-CoV-2 infection [bib_ref] Vitamin D and COVID-19 severity and related mortality: A prospective study in..., Campi [/bib_ref]. However, a large vitamin D Mendelian randomization study assessed the evidence for causality between vitamin D levels and severity and susceptibility for SARS-CoV-2 infection; after all the potential confounders were excluded, no association was found [bib_ref] Vitamin D and COVID-19 susceptibility and severity in the COVID-19 host genetics..., Butler-Laporte [/bib_ref]. Another large study conducted on more than 300,000 participants failed to prove the beneficial role of vitamin D in protection against severe disease or mortality due to SARS-CoV-2 infection, after adjustment for confounders [bib_ref] Vitamin D and COVID-19 infection and mortality in UK Biobank, Hastie [/bib_ref]. Another large-scale Mendelian randomization study for the identification of many traits for severe COVID-19 was performed by Sun et al. using a COVID-19 genome-wide association study (GWAS): host genetics initiative (HGI) for comparison between severe COVID-19 patients with respiratory failure and controls, a HGI study for hospitalized COVID-19 patients with control comparison, and a NEJM study that compared COVID-19 patients with controls in Spain and Italy [bib_ref] Extensive Mendelian randomization study identifies potential causal risk factors for severe COVID-19, Sun [/bib_ref]. Among studied risk factors, the authors analyzed the genetically predicted vitamin D traits (ukb-d-30890_raw) and found neither association with the risk of COVID-19 severity by comparing COVID-19 patients with confirmed severe respiratory symptoms to a control population, nor for hospitalized COVID-19 patients compared to the population control, and for COVID-19 patients with respiratory failure compared to healthy controls from Italy and Spain [bib_ref] Extensive Mendelian randomization study identifies potential causal risk factors for severe COVID-19, Sun [/bib_ref].
Mendelian randomization allows the creation of a study design that is not susceptible to environmental or confounding factors, an important aspect especially in studies for vitamin D. In this regard, genetically predicted vitamin D analysis is a more promising parameter for drawing pertinent conclusions. Another Mendelian randomization study performed by Liu et al. did not find evidence supporting genetically predicted vitamin D concentration to be significantly associated with COVID-19 severity or disease susceptibility [bib_ref] Association between 25 Hydroxyvitamin D Concentrations and the Risk of COVID-19: A..., Liu [/bib_ref].
Our study has several limitations. First, the serum level of interleukin-6, a potent pro-inflammatory cytokine, was not assessed. Second, the number of the included subjects was relatively low, with a potential impact on results; hence, an additional study would be useful for clarification of the vitamin D involvement in infections in general, and especially in COVID-19 patients. Another potential limitation is the unbalanced proportion between the healthy control group (vitamin D retrieved from a retrospective analysis) and COVID-19 patients. Nevertheless, although few studies have been performed to assess the prevalence of vitamin D deficiency in the Romanian population, they all indicate a high prevalence of vitamin D deficiency, especially in extreme age groups and women [bib_ref] Vitamin D Status: A Different Story in the Very Young versus the..., Chirita-Emandi [/bib_ref]. In our study, retrospective analysis of vitamin D serum levels categorized by age in subjects from outpatient settings (otherwise not fully analyzed, unpublished data) revealed that except for the age interval 21-44 years where we found vitamin D insufficiency, in the rest of the age groups, the medians of the vitamin D serum levels were at the lower threshold of the recommended interval, around 30 ng/mL. The vitamin D supplementation before the study adherence was not assessed as a part of the study, only the baseline vitamin D status was assessed shortly after hospital admission.
# Conclusions
Except for significantly lower vitamin D levels in moderate and severe COVID-19 forms compared to mild disease forms, we did not find an association between vitamin D levels and patients' outcomes or with the inflammatory biomarkers. Compared to the non-COVID-19 group, SARS-COV-2 infected patients had lower vitamin D serum levels, except for the age group 21-44 years, where the difference was without significance. Future studies on a larger cohort, which also include an appropriate control group, are needed for a better characterization of the relationship between vitamin D status and viral infections, in general, and SARS-CoV-2 infection, in particular. Informed Consent Statement: Informed consent was obtained from all participants prospectively recruited in the study.
# Data availability statement:
The data presented in this study are available on request from the corresponding author.
[fig] Figure 1, Figure 1: Serum levels of vitamin D were stratified by age groups. Results were analyzed Kruskal-Wallis test and Dunn's for multiple comparisons groups. Box plot (median and in tile range): the horizontal lines depict the 75% percentile, and the top of the box plot is the The analysis of the vitamin D status in relation to disease severity showed the value in mild COVID-19 patients, with a median of 29.39 (IQR 12.12-44.02) ng/m for moderate and severe forms, the serum levels were significantly lower, Serum levels of vitamin D were stratified by age groups. Results were analyzed with the Kruskal-Wallis test and Dunn's for multiple comparisons groups. Box plot (median and interquartile range): the horizontal lines depict the 75% percentile, and the top of the box plot is the median. [/fig]
[fig] Figure 2: Serum levels of vitamin D in different disease severity. Results were an Kruskal-Wallis test and Dunn's for multiple comparisons groups. Box plot (media tile range), the horizontal lines represent the 75% percentile, and the top of the bo dian.When vitamin D level was analyzed in relation to comorbidities, the p personal history of illness showed a significantly higher level of vitamin D, of 24.72 (IQR 16.05-31.52) ng/mL compared to subjects with at least one com a median of 16.02 (IQR 9.81-25.22) ng/mL, p = 0.004. [/fig]
[fig] Author: Contributions: Conceptualization, A.H. and A.M.G.; methodology, A.H.; software, S.V.; validation, A.H., A.M.G. and M.D.; formal analysis, A.H. and A.M.G.; investigation, A.V.A., V.N. and A.H.; data curation, A.H., S.V. and A.V.A.; writing-original draft preparation, A.H. and A.M.G.; writing-review and editing, A.M.G. and M.D.; supervision, M.D. and A.M.G.; project administration, A.H.; funding acquisition, A.H. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research was funded by George Emil Palade University of Medicine, Pharmacy, Science, and Technology of Targu Mures, Romania, grant number 10126/1/17.12.2020 and the publication fee is supported by George Emil Palade University of Medicine, Pharmacy, Science, and Technology of Targu Mures, Romania. Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of Clinical County Hospital of Targu Mures (No. 19038/21.12.2020 and Ad. No. 10096/29.04.2022) and by the Ethics Committee of George Emil Palade University of Medicine, Pharmacy, Science, and Technology of Targu Mures, Romania (No. 1237/08.01.2021). [/fig]
[table] Table 1: Distribution of patients in relation to demographic data and comorbidities. [/table]
[table] Table 2: The comparative analysis of vitamin D serum levels between COVID-19 and non-COVID-19 patients. [/table]
[table] Table 3: The demographic and clinical data after patient dichotomization according to vitamin D deficit. [/table]
[table] Table 4: The characteristics of the surviving vs. non-surviving patients as related t tory parameters and vitamin D status. [/table]
[table] Table 5: The relationship for the in-hospital mortality with demographic parameters, vitamin D status, and inflammatory biomarkers. [/table]
|
Flexible transparent graphene/polymer multilayers for efficient electromagnetic field absorption
Thanks to its high electrical conductivity, a graphene plane presents a good shielding efficiency against GHz electromagnetic radiations. Several graphene planes separated by thin polymer spacers add their conductivities arithmetically, because each of them conserves the intrinsic properties of isolated graphene. Maximum absorption of radiations for frequency around 30 GHz is achieved with six separated graphene planes, which is the optimum number. This remarkable result is demonstrated experimentally from electromagnetic measurements performed in the K a band on a series of multilayers obtained by piling 1, 2, 3 … graphene/PMMA units on a silica substrate. Theoretical calculations convincingly explain the observed absorption and transmission data in the GHz domain. It is concluded that graphene/PMMA multilayers can be used as an efficient optically transparent and flexible shielding media.Microwave is the ultimate range for open space communication. The rapid growth of satellite data transmission in last decades, reinforced by the Internet traffic, makes inevitable the overcrowding of the spectral bands allocated to the different communication channels. Along with the increasing density of emitters in the environment, this has made the Electromagnetic Compatibility (EMC) an important issue 1 . EMC rules require that each new equipment have adequate immunity in order to function consistently and reliably, be resilient to major electromagnetic (EM) disturbances, and coexist with other equipments. In order to meet these requirements, there is a real need for new materials to be used as coating layers, shields and filters in future nanoelectronic devices. These materials must have high electrical conductivity to achieve a good EM shielding effectiveness, defined as the reciprocal of the transmission factor. It is worth noting, however, that EM radiations reflected from the protective coatings do contribute to the microwave pollution. Consequantly, absorption rather than reflection losses should dominate the EM attenuation in the new materials. In addition, protection of the future nanodevices require submicron thick shields, making useless conventional conductive coatings capable to produce an EM attenuation of up to 3 dB/mm.The challenging problem of the electromagnetic interference (EMI) protection at the nanoscale may be resolved if one exploits the unusual electronic properties of graphene. Very recently we have proposed microwave shielding based on nanometrically thin pyrolytic carbon (PyC) films 2 composed of intertwined graphene ribbons. Specifically, 30-35 nm thick PyC films secure about 60% attenuation at 28 GHz mainly due to absorption of the EM radiations. The absorption efficiency of the EM protective coating can be increased even further by replacing PyC with few-layer graphite (FLG), which is capable to absorb 40% of microwave radiation at a thickness of only 5 nm 3 . Interlayer interactions affect the band structure of FLGs near the Fermi level in a qualitative way: compared to graphene, the dispersion of the p bands is no longer linear around the Dirac points 4 . This modification has direct consequences on properties such as the electrical conductivity that crucially depend on the topology of the Fermi surface 5 . When the distance between the graphene layers exceeds that realized in graphite by a fraction of a nanometer and more, the interlayer interaction plays no role anymore. In these conditions, the measured sheet conductivity increases in proportion to the number of graphene layers 6 . Therefore, one may expect that by separating graphene layers from one another in a controlled way by dielectric spacers, one can create a nano-metamaterial with enhanced shielding efficiency. This approach has been recently theoretically discussed in terms of an ultra-broadband THz absorber 7 , in which high-order surface plasmon resonances can be introduced by patterning graphene layers with micrometer periodicity.
In this paper we demonstrate both experimentally and theoretically that a stack of PMMA thin films sandwiching graphene sheets between them has remarkable microwave EMI shielding effectiveness. Specifically, in the K a band (frequency range between 26.5 and 40 GHz), strong absorption of the EM radiation can be achieved without pre-patterning of the graphene whose overall thickness of two nanometers is enough to absorb 50% of the incident EM power.
# Results
Experiment. [fig_ref] Figure 1 |: Schematic representation of graphene sandwich fabrication, consisting of a number of repeating... [/fig_ref] shows schematically the graphene/PMMA multilayers produced for EM shielding analysis. The thickness of the PMMA spacers is in the range 600-800 nm. These sandwich structures were produced as follows. The first graphene layer was synthesized by chemical vapor deposition (CVD) at 1000uC in methane atmosphere on a copper foil. It was next covered by a PMMA layer obtained by spin coating, after what Cu was wet etched in ferric chloride. The obtained PMMA film with its graphene layer was washed in distilled water, placed on a quartz (fused silica) substrate and dried. The very same procedure was repeated several times, each newly-produced graphene/PMMA unit being deposited on top of the stack obtained at the previous step. This procedure allowed us to fabricate sandwich multilayers containing several graphene sheets.
The SEM image reproduced indemonstrates that the PMMA-based transfer technique 8 allowed us to produce a continuous graphene sheet with a tiny number of holes. In addition, the optical transmittance of a graphene layer without the carrier PMMA film, taking out the response of the bare substrate (see), was 97.8% at 800 nm, which is only 0.3% below that expected for a single graphene layer [bib_ref] Counting graphene layers on glass via optical reflection microscopy, Gaskell [/bib_ref]. This measurement indicates that areas covered by bi-and multi-layer graphene are small. The Raman spectrum shown inpresents sharp G and 2D lines. More specifically, a sharp G peak (G FWHM , 20 cm 21 ) and the absence of D peak point to a high crystallinity of the graphene layer [bib_ref] Raman spectroscopy in graphene, Malard [/bib_ref]. Furthermore, the 2D line peaks at 2688 cm 21 is typical for graphene monolayer [bib_ref] Raman spectroscopy in graphene, Malard [/bib_ref]. However, the 2D peak is slightly broader (,40 cm 21 ) than expected for a graphene monolayer (30 cm 21 ), indicating some presence of bilayer graphene.
Microwave measurements were made with a scalar network analyzer R2-408R (ELMIKA, Vilnius, Lithuania). The setup includes a sweep generator, a waveguide reflectometer and an indicator unit. The IEC 62431:2008(E) standard for reflectivity measurements of EM materials was used. The frequency stability of the oscillator was controlled at the 10 26 level by a frequency meter. The power used was 7.0 mW 6 10 mW. The error on the measured EM attenuation was 7% in the range 0 to 240 dB. The ratios of transmitted/ input (S 21 ) and reflected/input (S 11 ) signals were measured at 30 GHz before and after insertion of the specimen into the waveguide. The lateral dimensions of the samples fitted precisely the waveguide 7.2 3 3.4 mm 2 cross section.
The transmission (T) and reflection (R) coefficients of graphene/ PMMA sandwich structure consisting of one to four graphene layers were deduced from the measured S parameters. The measurements were performed for the two possible normal orientations of the samples: in the first orientation, the guided EM wave enters the sample through the silica substrate and leaves it from the graphene/PMMA side (upward), the second orientation (downward) being just the reverse. [fig_ref] Figure 3 |: Absorption [/fig_ref] displays the variation of the absorption coefficient A 5 1 2 T 2 R versus the number N of graphene/PMMA units for the upward geometry: the absorbance varies between 25% (one graphene layer) to 48% (four graphene layers). The silica substrate and PMMA being non-absorbing in the K a band, all the incident EM power is absorbed by the graphene layers. The absorption coefficient is approximately a factor of 1.2 smaller for the downward geometry [fig_ref] Figure 3 |: Absorption [/fig_ref] compared to the upward geometry: A varies between 21% (one graphene layer) to 40% (four graphene layers). The transmission coefficient [fig_ref] Figure 3 |: Absorption [/fig_ref] ) is the same for both orientations of the sample. It decreases monotonously between 58% and 25%, the silica substrate itself reflects 21% of the incident radiation.
Theory. Before proceeding to a quantitative analysis of these measurements, it is worth understanding why graphene is so good in absorbing EM radiations in the K a band. Indeed, a single graphene layer was shown to absorb 21 to 25% of the incident radiation 3 , which is a lot for a one-atom thick material. For a qualitative explanation, it is sufficient to refer to the Fresnel formulas for the transmission and reflection of a plane wave at the interface between two dielectric media when the interface sustains an infinitely thin -at the scale of the incident wavelength-conducting layer. In normal incidence, the electric and magnetic fields are parallel to the interface. Their amplitudes must satisfy the boundary conditions
[formula] E i zE r~Et and H i zH r~Ht zsE tð1Þ [/formula]
where the subscripts i, r, and t refer to the incident, reflected, and transmitted waves, respectively. The last term in the second relation is the surface current density induced by the electric field in the interfacial layer with sheet conductance s. Starting from the boundary conditions (eq. 1), the reflectance R, transmittance T and absorbance A of the interfacial layer between media with refractive indices n 1 and n 2 are readily derived [bib_ref] Unusual reflection of electromagnetic radiation from a stack of graphene layers at..., Bludov [/bib_ref]. In particular,
[formula] T~4 n 1 n 2 n 1 zn 2 z s 0c 2 and A~4 n 1 Re s 0 n 2 c n 1 zn 2 z s 0 c 2ð2Þ [/formula]
where 0 is the vacuum permittivity and c the speed of light in vacuum. Both T and A are functions of the real and imaginary parts of the interfacial layer conductance. The maximum of absorbance is reached when Im s 5 0 and Re s~n 1 zn 2 ð Þ 0 c. The maximum value of A is n 1 /(n 1 1 n 2 ). When EM radiations propagating in air (n 1 5 1) enter a dielectric medium coated by a thin conductive layer (analogue of the downward sample orientation, [fig_ref] Figure 3 |: Absorption [/fig_ref] , the bigger the refractive index n 2 of the dielectric, the lower the absorbance. However if EM radiations propagate in the opposite direction, i.e. from the dielectric medium to air (n 2 5 1) (analogue of the upward sample orientation, [fig_ref] Figure 3 |: Absorption [/fig_ref] , the bigger the refractive index n 1 of the dielectric, the higher the absorbance. In particular the absorbance can exceed 50% when n 1 . n 2. Reaching maximum absorption requires Re s= 0 c<2, typically.
As shown below, s= 0 c for a single graphene layer already reaches 0.32 in the K a band. Such a large value of the sheet conductance comes from intraband transitions almost exclusively and is determined by Drude type conductivity. Indeed, even a very small doping of graphene blocks the interband transitions for frequencies in the GHz domain and below [bib_ref] Optical properties of graphene, Falkovsky [/bib_ref] [bib_ref] Unusual microwave response of Dirac quasi-particles in graphene, Gusynin [/bib_ref]. This is not the case in the optical range, where the intraband transitions are much less efficient than the interband transitions, the latter yielding the well-known result [bib_ref] Optical properties of graphene, Falkovsky [/bib_ref] [bib_ref] Measurement of the optical conductivity of graphene, Mak [/bib_ref] s= 0 c~0:023.
For a quantitative interpretation of the experimental data plotted in [fig_ref] Figure 3 |: Absorption [/fig_ref] , plane waves must be replaced by guided waves 2 . The electromagnetic field propagating in a plane-stratified medium put inside a waveguide can be expressed in terms of the longitudinal electric and magnetic components
[formula] E + x,j~i +K j LE + z,j . Lxzvm 0 LH + z,j . Ly . k 2 j {K 2 j H + y,j~{ i v j 0 LE + z,j . Lx+K j LH + z,j . Ly . k 2 j {K 2 j E + y,j~i +K j LE + z,j . Lyzvm 0 LH + z,j . Lx . k 2 j {K 2 j H + x,j~{ i v j 0 LE + z,j . Ly+K j LH + z,j . Lx . k 2 j {K 2 jð3Þ [/formula]
where the 6 signs label the incident and reflected modes counterpropagating along the waveguide axis (z direction) and subscript j 5 1, 2 and 3 label air, PMMA and silica, respectively. In eqs.(3), j , k j~ffi ffiffi j p v c and K j are the relative dielectric permittivity, wavevector and the mode propagation constant in the jth medium, respectively. For the transverse electric TE 01 mode of a rectangular waveguide, E z 5 0 while the longitudinal magnetic component H z can be presented in the following form:
[formula] H + z,j~i C + j cos px=a ð Þexp +iK j z À Áð4Þ [/formula]
where a is the waveguide dimension in x direction,
[formula] K j~j v=c ð Þ 2 { p=a ð Þ 2 Â Ã 1=2 . [/formula]
In Eq. 4, C + j are the mode amplitudes in the jth medium that are related to each other through the boundary conditions (eq. 1) at the successive interfaces.
The PMMA layers, being transparent and having a very small optical thickness in the GHz domain, can be ignored. Thus a multilayer containing N graphene/PMMA units on its silica substrate behaves like a bilayer composed of the silica substrate slab and N non-interacting graphene sheets squeezed at one of its two interfaces with air. This simplified approach allows us to arrive at the following equations for the reflectance and transmittance of the graphene/ PMMA sandwich placed in the waveguide for the upward (silica 1 graphene, R 1 and T 1 ) and downward (graphene 1 silica, R 2 and T 2 ) geometries, respectively:
[formula] R 1~1 {a ð Þ 1zazB ð Þ e { { 1za ð Þ 1{azB ð Þ e z 1za ð Þ 1zazB ð Þ e { { 1{a ð Þ 1{azB ð Þ e z 2 R 2~1 za ð Þ 1{a{B ð Þ e { { 1{a ð Þ 1za{B ð Þ e z 1za ð Þ 1zazB ð Þ e { { 1{a ð Þ 1{azB ð Þ e z 2 T 1~T2~4 a 1za ð Þ 1zazB ð Þ e { 1{a ð Þ 1{azB ð Þ e z 2ð5Þ [/formula]
Here a 5 K substrate /K air , B~Ns 1 = 0 c with s 1 the sheet conductance of a single graphene layer, and e 6 5 exp(6iK substrate d) with d the substrate thickness.
# Discussion
The above formalism yields values of the absorption and transmission coefficients plotted in [fig_ref] Figure 3 |: Absorption [/fig_ref]. The agreement with the experimental measurements is excellent. The dielectric constant of the silica substrate (d 5 0.5 mm) used in the calculations was 3.7 [bib_ref] Miscellaneous data on materials for millimetre and submillimetre optics, Lamb [/bib_ref]. The only adjustable parameter that remained in the theory was s 1 = 0 c, which was taken at 0.32. By comparison, the sheet conductance of graphene grown by CVD on Ni, measured between 1 GHz and 1 THz 17 , at the frequency of the experiment, n 5 30 GHz is very close (0.37) to our adjusted value. For six layer of graphene, the dimensionless sheet conductance is B~6s 1 = 0 c<2. It is worth mentioning that a graphite slab having a thickness equivalent to 6 graphene layers (2 nm) would give s= 0 c~0:1, 20 time smaller than here if we take a typical value 125000 S/m for the in-plane conductance of graphite. This simple argument demonstrates the tremendous superiority of separated graphene planes over graphite in the present context. The calculated dependence of the absorbance on the number N of graphene layers for both upward and downward geometries is plotted in [fig_ref] Figure 4 |: Computed dependence of the absorbance on the number N of graphene monolayers... [/fig_ref]. There is a point with maximum absorption equal to 54% for the upward geometry and 41% for the downward geometry. In each case, the maximum corresponds to six graphene layers. The transmission computed for N 5 6 is 21% for both sample orientations.
Such an interesting electromagnetic response of the sandwich structure can be captured by eq. 2. Forgetting the PMMA spacers, which are very thin compared to wavelength, the interfacial conducting layer stands for the number N of graphene layers contained in the graphene/polymer multilayer, with total conductivity s 5 Ns 1 . The two dielectric media are the air and the silica substrate. It follows from eq. 2 that the transmission coefficient decreases monotonously with N. The absorption A increases, reaches a maximum value at some N opt as mentioned above and then decreases asymptotically like 1/N. The asymmetry of the absorption for both sample orientations can be qualitatively understood from the presence of the prefactor n 1 in the expression (Eq. 2) of A. For a self-supported film (n 1 5 n 2 5 1), the maximum absorption would be 50%.
The overall thickness of a graphene/PMMA sandwich sample containing 6 graphene layers is about 4.2 mm, not including the silica substrate. The overall thickness could be reduced to 2 mm by using twice as thin PMMA films. A metallic film of 2-mm thickness would give better shielding efficiency than graphene/PMMA multilayers of equivalent thickness. However, the low transmittance of a metallic film is the result of high reflectivity more than than high absorption. For applications where reflecting microwaves back is not desirable, absorption must be favoured over reflection, which requires decreasing the film thickness drastically. For ultrathin aluminium films on glass, for instance, the maximum of absorption at 37.5 GHz is realized for a film thickness of 2.5 nm 18 and reaches 50% in the upward geometry. Graphene makes it possible to obtain similar and even slightly better characteristics, not mentioning the big advantages of graphene over any artificial metallic layer that are its high flexibility and mechanical strength.
# Conclusions
To summarize, we have demonstrated both experimentally and theoretically a remarkable shielding efficiency in the K a band of graphene layers separated by thin PMMA films. The transmission coefficient of the multilayers decreases monotonously with the number of graphene/PMMA units. At 30 GHz and around, the absorption coefficient reaches a maximum for 6 graphene layers, which is the optimum number. The largest contribution to the shielding then comes from radiation absorption by the separated graphene layers. Increasing the number of layers above the optimum value lowers the absorption coefficient and increases the reflection contribution to the shielding. A strong radiation absorption by extremely small thickness of graphene sandwiches is very appealing for nanodevice shielding applications and for the detection of low-power electromagnetic signals. It is worth mentioning that the idea of using graphene/dielectric multilayers for EM wave absorption is not new. An ultra-broadband absorbing multilayer operating in the THz has been theoretically proposed [bib_ref] An ultra-broadband multilayered graphene absorber, Amin [/bib_ref]. In this device, the graphene layers are pat- terned at the micrometer scale in order to generate high-order surface plasmon resonances. Nothing similar is required with our devices, which are much simplest, but work at smaller frequencies.
Still, experiments performed by time domain spectroscopy on one samples containing three graphene/PMMA units was shown to shield (by absorption and reflection) 75% of the incident power at 1 THz, which remains appreciable.
[fig] Figure 1 |: Schematic representation of graphene sandwich fabrication, consisting of a number of repeating steps, and final graphene/PMMA multilayer structure containing here four graphene sheets (the thickness of the different layers were not drawn in scale). The lateral dimensions of the samples are 7.2 mm 3 3.4 mm. [/fig]
[fig] Figure 2 |: (a) SEM picture of a graphene layer transferred from copper foil to a silica substrate. (b) Wavelength dependence of the visible-UV transmittance of a graphene layer on silica substrate corrected from that of the bare substrate). (c) Raman spectrum of the graphene layer. [/fig]
[fig] Figure 3 |: Absorption (a,b) and transmission (c) coefficients at 30 GHz of graphene/PMMA multilayers on silica containing N~0 Á Á Á 4 graphene/ PMMA units. In panel (a), the incident EM waves come from the side of the silica substrate (upward orientation of the graphene/PMMA sandwiches), in (b) they come from the other side of the samples (downward). The upward or downward orientation of the samples has no influence on the transmission. Both measured and calculated (eq. 5) values are represented. [/fig]
[fig] Figure 4 |: Computed dependence of the absorbance on the number N of graphene monolayers for the upward (silica 1 graphene) and downward (graphene 1 silica) orientations of the multilayer sample. www.nature.com/scientificreports SCIENTIFIC REPORTS | 4 : 7191 | DOI: 10.1038/srep07191 [/fig]
|
Successful atrial fibrillation ablation without pulmonary vein isolation utilizing focal impulse and rotor mapping in an atriopulmonary Fontan
# Introduction
Fontan palliation is designed to separate systemic and pulmonary circulations to ameliorate systemic hypoxemia and relieve ventricular volume overload in patients with a functional univentricular heart. Since its introduction in 1971, the surgical techniques have evolved considerably. [bib_ref] The Fontan procedure for tricuspid atresia: early and late results of a..., Mair [/bib_ref] Early atriopulmonary (AP) versions of the Fontan incorporated the right atrium (RA) into the systemic venous circuit and are associated with considerable burden of early and late atrial dysrhythmias. [bib_ref] The Fontan procedure for tricuspid atresia: early and late results of a..., Mair [/bib_ref] [bib_ref] Arrhythmia management in the Fontan patient, Deal [/bib_ref] [bib_ref] Chamberlain Memorial Paper for congenital heart surgery. 111 Fontan conversions with arrhythmia..., Mavroudis [/bib_ref] [bib_ref] Effect of age and surgical technique on symptomatic arrhythmias after the Fontan..., Cecchin [/bib_ref] [bib_ref] Education in heart management of the failing Fontan circulation, Deal [/bib_ref] Although the majority of atrial arrhythmias are macroreentrant RA circuits, such as intra-atrial reentrant tachycardia (IART), the prevalence of late atrial fibrillation (AF) in Fontan patients is increasing. [bib_ref] Arrhythmia management in the Fontan patient, Deal [/bib_ref] [bib_ref] Effect of age and surgical technique on symptomatic arrhythmias after the Fontan..., Cecchin [/bib_ref] [bib_ref] Education in heart management of the failing Fontan circulation, Deal [/bib_ref] [bib_ref] Successful radiofrequency catheter ablation of atrial fibrillation late after modified Fontan operation, Takahashi [/bib_ref] [bib_ref] Ablation of focal atrial arrhythmia in patients with congenital heart defects after..., De Groot [/bib_ref] The mechanisms responsible for AF initiation and maintenance in this population and the role and techniques of catheter ablation for treatment of AF are not well defined. [bib_ref] Successful radiofrequency catheter ablation of atrial fibrillation late after modified Fontan operation, Takahashi [/bib_ref] [bib_ref] Rotors as drivers of atrial fibrillation and targets for ablation, Schricker [/bib_ref] [bib_ref] Mechanisms Underlying AF: triggers, rotors, other? Curr Treat Options Cardio, Krummen [/bib_ref] [bib_ref] Ablation of arrhythmias in adult patients after Fontan operation, Tomkiewicz-Pajak [/bib_ref] [bib_ref] Atrial fibrillation ablation without pulmonary vein isolation in a patient with Fontan..., Ruckdeschel [/bib_ref] We present a case of successful AF ablation without pulmonary vein (PV) isolation by utilizing focal impulse and rotor mapping to identify and target mechanisms of AF maintenance in a patient with atrial arrhythmias and medication intolerance in the setting of an AP Fontan.
## Case report
A 25-year-old woman with complex congenital heart disease presented with recurrent palpitations and exertional syncope. She had a history of tricuspid atresia, initially palliated with a left modified Blalock-Taussig shunt and a subsequent shunt ligation and AP Fontan. Her childhood and adolescence were relatively unremarkable with the exception of an embolic stroke at age 20 while taking oral contraception, from which she had a complete neurologic recovery and for which she is maintained on anticoagulation. Because of progressive cyanosis and arrhythmias, her adult cardiologist referred her to a children's hospital, where catheterization revealed significant collateralization without focal hemodynamic derangement. A Fontan conversion was recommended, but she declined and shortly afterward relocated and was referred to our program.
On the initial visit, the patient reported significant functional impairment due to fatigue and daily symptomatic palpitations, lasting minutes to hours. Examination was notable for cyanosis and clubbing, with resting O 2 saturation of 83%. Echocardiography demonstrated tricuspid atresia with an unobstructed AP Fontan, severe RA dilation, and normal systemic ventricular function. Baseline electrocardiography demonstrated sinus rhythm with a prolonged PR interval of 216 ms, evidence of RA enlargement, and left-axis deviation. Exercise testing was notable for desaturation to 72%, poor exercise capacity (peak VO 2 17.2 mL/kg/min), and salvos of a regular tachyarrhythmia with ventricular rate of 166 bpm . Event monitoring revealed symptomatic paroxysmal AF . Catheterization revealed unobstructed Fontan flow with mean resting Fontan pressure of 5-6 mm Hg. A large venoatrial collateral from the superior vena cava to the left atrium was identified and the coil embolized, with improvement in systemic saturation to 89%. Medical management with beta-blocker was not tolerated. Given her overall normal hemodynamics and her aversion to Fontan conversion or antiarrhythmics, the decision was made to attempt catheter ablation. Preprocedural imaging was performed with cardiac magnetic resonance imaging to evaluate the atrial anatomy and identify pulmonary venous structures.
## Description of the procedure
The patient was brought to the electrophysiology laboratory in sinus rhythm. Intracardiac echocardiography was used to KEYWORDS Adult congenital heart disease; Atrial fibrillation; Catheter ablation; Congenital heart disease; Focal triggers; Fontan procedure; Mechanisms (Heart Rhythm Case Reports 2018;4:241-246) Drs Sauer and Nguyen receive significant research and consulting grants from Biosense Webster and CardioNXT and educational grants from St. Jude Medical, Boston Scientific, and Medtronic. Drs Nguyen and Sauer have non-public equity interests/stock options in CardioNXT. No specific grant from funding agencies in the public, commercial, or not-for-profit sectors was utilized for the preparation of this manuscript. There are no disclosures related to the topic of this manuscript. Address reprint requests and correspondence: Dr Duy Thai Nguyen, Cardiac Electrophysiology, guide catheter placement, to create an anatomic geometry of the Fontan pathway to merge with her preprocedural cardiac magnetic resonance imaging, and to monitor hemodynamics and ablation lesions. Therapeutic anticoagulation was maintained during the case. Various catheters, including a decapolar catheter, PentaRay multielectrode catheter (Biosense Webster, Diamond Bar, CA), and a ThermoCool Smart-Touch ablation catheter (Biosense Webster), were positioned throughout the Fontan pathway . A dense voltage map of the Fontan showed mostly normal voltage, except for a dense strip of scar within the inferolateral limb of the Fontan, contiguous with the inferior vena cava .
Electrophysiological testing was performed once the voltage map was acquired. Triggers initiating an IART with a tachycardia cycle length of 280-310 ms were noted during isoproterenol infusion . Activation mapping demonstrated a microreentrant circuit with early activation within the inferolateral Fontan, at the border zone between normal tissue and scar. A series of ablation lesions delivered within the border zone resulted in slowing of tachycardia before termination and transition into AF. Given the persistence of AF, substrate modification was then performed, extending the ablation lesions from the region of IART termination superiorly and inferiorly through the border zone to transect channels within scar. Furthermore, in order to identify possible drivers of AF maintenance, the decision was made to map potential rotors and focal impulses. Using a focal impulse and rotor mapping multielectrode basket catheter and RhythmView mapping system (Topera Inc, Abbott EP, Abbott Park, IL), rotors and focal impulses were identified. A focal impulse and an adjacent fixed rotor were identified in the lateral aspect of the Fontan pathway near the previously identified border zone of the Fontan scar . The region containing the rotor and focal impulse was ablated to noncapture, and further ablation was performed to anchor this region inferiorly to the previously identified area of scar. AF selfterminated. Bidirectional block was then confirmed across ablation lines. Several other rotors noted to be within areas of normal voltage or along the atrioventricular node/His region were not targeted for ablation given the potential risks of complete heart block and proarrhythmia.
An additional macroreentrant cavo-tricuspid isthmus (CTI)-dependent atrial flutter (AFL) circuit with a tachycardia cycle length of 230 ms was induced with rapid atrial pacing and isoproterenol infusion. CTI ablation was performed with AFL termination, and bidirectional block was confirmed with differential pacing. Repeat pacing failed to induce sustained atrial arrhythmias. The patient has remained free of significant arrhythmias without medications for over 18 months, with marked improvement in her functional capacity and quality of life.
# Discussion
Atrial arrhythmias are highly prevalent in patients with Fontan palliation for functional univentricular heart and are associated with significant morbidity and mortality. [bib_ref] Arrhythmia management in the Fontan patient, Deal [/bib_ref] [bib_ref] Education in heart management of the failing Fontan circulation, Deal [/bib_ref] Arrhythmias may arise from obstructive hemodynamic perturbations, progressive atrial dilation and remodeling, or scar. There is a high failure rate of antiarrhythmic drugs used to treat atrial arrhythmias in these patients. [bib_ref] Management of late atrial tachyarrhythmia long after Fontan operation, Fujita [/bib_ref] Catheter ablation of atrial arrhythmias in patients with Fontan palliation, the majority of which are right-sided, has historically been associated with lower success than in other forms of congenital heart disease, with recurrence rates of 30%-60% reported within 1 year. [bib_ref] Education in heart management of the failing Fontan circulation, Deal [/bib_ref] Ablation of atrial arrhythmias is challenging in Fontan patients for myriad reasons, including distorted anatomy resulting in anatomic and electrical barriers to native atrial tissue, restricted catheter access, inability to deliver lesions of sufficient penetrance, and hemodynamic instability during protracted cases.
AF ablation in Fontan patients is uncommon. Whether the optimal ablative strategy is to target all possible triggers utilizing empiric lines for IART prophylaxis and PV isolation or to selectively identify and target clinically relevant triggers is unclear. We and others have demonstrated durable AF ablation in Fontan patients by solely targeting right-sided triggers. 7,11,14 AF recurrence rates are higher in patients with non-PV triggers for AF,and non-PV triggers for AF, including right-sided triggers, are more common in patients with a normal-sized left atrium, [bib_ref] Atrial fibrillation ablation without pulmonary vein isolation in a patient with Fontan..., Ruckdeschel [/bib_ref] such as many
## Key teaching points
Atrial tachyarrhythmias are a common source or morbidity in patients who undergo classic atriopulmonary (AP) Fontan. The prevalence of atrial fibrillation (AF) is increasing in this population.
Catheter ablation of AF in AP Fontan patients can be feasible, safe, and efficacious.
Factors driving AF in Fontan patients are ill defined, and nonpulmonary vein triggers should be considered. Pulmonary vein isolation may not be necessary to maintain sinus rhythm in AP Fontan patients with nonpulmonary vein triggers for AF.
Ablative strategies targeted toward mechanisms of AF maintenance, such as rotors and focal impulses, should be considered in AP Fontan patients with AF.
Meticulous identification of triggers and judicious, focused ablative approaches applied by electrophysiologists skilled in catheter ablation and familiar with complex congenital heart disease should be considered an option to treat drug refractory atrial arrhythmias in Fontan patients. A sustained reduction in arrhythmia burden and improvement in quality of life can be achieved in select patients.
patients with AP Fontan. In this case, the patient did not have significant hemodynamic derangements and had not previously undergone ablative attempts. Multiple arrhythmias were demonstrated, as is often the case in Fontan patients. [bib_ref] Time course of atrial fibrillation in patients with congenital heart defects, Teuwen [/bib_ref] We targeted all relevant mechanisms of initiation and maintenance of AF that we uncovered. A microreentrant IART circuit in the inferolateral Fontan, a macroreentrant CTI-dependent AFL, and focal impulses and rotors within zones of abnormal voltage were targeted [fig_ref] Figure 3: Representative electroanatomic map of the right atrium/Fontan pathway demonstrating the various identified... [/fig_ref]. Empiric PV isolation was not performed. In this case, the ablation was successful and durable.
We propose that catheter ablation is a viable treatment strategy to address arrhythmia burden in AP Fontan patients who have drug intolerance or drug-refractory atrial arrhythmias without apparent hemodynamic abnormalities, especially those who are at high surgical risk. [bib_ref] Arrhythmia management in the Fontan patient, Deal [/bib_ref] Surgical conversion of AP Fontan to an extracardiac total cavopulmonary connection with biatrial surgical maze procedure is A: Regular narrow-complex long RP tachycardia elicited during exercise study. B: Symptomatic irregularly irregular tachycardia without P waves consistent with atrial fibrillation was noted on event monitor. C: During isoproterenol infusion, triggers were noted with runs of intra-atrial reentrant tachycardia with earliest signals at the ablation catheter (yellow arrow). D: An array of catheters was utilized to appropriately investigate and ablate tachycardias. The ablation catheter is positioned at the border zone. Embolization coils are seen within the collateral vessel (red arrows). E: Voltage map of the right atrium/Fontan pathway revealed a dense strip of scar (red) within the inferolateral limb of the Fontan, continuous with the inferior vena cava, with a border zone between normal tissue and scar. ICE 5 intracardiac echocardiography; IVC 5 inferior vena cava; PA 5 posteroanterior; RAO 5 right anterior oblique; SVC 5 superior vena cava.
## Figure 2
A: Fontan angiography revealed a markedly dilated IVC and RA, with unobstructed Fontan pathway to branch pulmonary arteries. A large venoatrial collateral from the SVC to left atrium was seen (green arrow) and subsequently embolized. B: A 64-pole FIRM multipolar basket (blue arrow) was introduced into the Fontan. C, D: Electroanatomic mapping was performed to define any underlying electrophysiological abnormalities contributing to atrial fibrillation maintenance. The FIRM basket catheter is visualized within 3-dimensional high-density electroanatomic maps. A rotor (yellow area) was identified within the Fontan near the border zone. E: Raw unipolar electrograms collected from the FIRM basket catheter, which are then exported to a proprietary mapping system that determines physiologically plausible activation paths. F: FIRM-derived computational phase map depicting electrical propagation during atrial fibrillation, revealing a rotor (yellow arrows) and adjacent focal impulse (asterisks). AP 5 anteroposterior; FIRM 5 focal impulse and rotor mapping; IVC 5 inferior vena cava; PA 5 posteroanterior; RA 5 right atrium; RAO 5 right anterior oblique; SVC 5 superior vena cava. associated with decreased burden of recurrent atrial arrhythmias, including AF, and has been increasingly proposed as the definitive therapy for significant atrial arrhythmias in AP Fontan patients. 2,3,5 However, conversion is associated with an early surgical mortality of 1%-6% and a prolonged recovery, and it may not be the optimal choice in patients without significant hemodynamic derangements who have not undergone attempts at catheter ablation.
# Conclusion
Radiofrequency catheter ablation is an alternative for reduction of arrhythmia burden and improvement in quality of life in carefully selected Fontan patients with drug-refractory or drug-intolerant arrhythmias. Ablation of AF is challenging in Fontan patients. Novel mapping techniques, remote magnetic navigation catheters, and ablative strategies targeting potential triggers and drivers of AF are powerful tools to aid in the identification of electrophysiological triggers for AF in patients with complex univentricular anatomy. We demonstrate successful and durable AF ablation using methodical identification of potential triggers without targeting PV triggers. Such approaches may be widely applicable. Meticulous individualized ablative strategies applied by electrophysiologists skilled in catheter ablation and familiar with complex congenital heart disease are of paramount importance in this patient population.
[fig] Figure 3: Representative electroanatomic map of the right atrium/Fontan pathway demonstrating the various identified pathologic triggers and mechanisms of maintenance of arrhythmia and subsequent sites of ablation. CTI 5 cavo-tricuspid isthmus; IART 5 intra-atrial reentrant tachycardia; IVC 5 inferior vena cava; RA 5 right atrium; RAO 5 right anterior oblique; SVC 5 superior vena cava. [/fig]
|
Current status and recent advances in resection cavity irradiation of brain metastases
Despite complete surgical resection brain metastases are at significant risk of local recurrence without additional radiation therapy. Traditionally, the addition of postoperative whole brain radiotherapy (WBRT) has been considered the standard of care on the basis of randomized studies demonstrating its efficacy in reducing the risk of recurrence in the surgical bed as well as the incidence of new distant metastases. More recently, postoperative stereotactic radiosurgery (SRS) to the surgical bed has emerged as an effective and safe treatment option for resected brain metastases. Published randomized trials have demonstrated that postoperative SRS to the resection cavity provides superior local control compared to surgery alone, and significantly decreases the risk of neurocognitive decline compared to WBRT, without detrimental effects on survival. While studies support the use of postoperative SRS to the resection cavity as the standard of care after surgery, there are several issues that need to be investigated further with the aim of improving local control and reducing the risk of leptomeningeal disease and radiation necrosis, including the optimal dose prescription/fractionation, the timing of postoperative SRS treatment, and surgical cavity target delineation. We provide a clinical overview on current status and recent advances in resection cavity irradiation of brain metastases, focusing on relevant strategies that can improve local control and minimize the risk of radiation-induced toxicity.
# Introduction
Brain metastases are a common and devastating complication of cancer. Surgical resection remains an effective treatment for brain metastases, especially for larger lesions causing mass effect and consequentially serious neurological symptoms. Postoperative whole brain radiation therapy (WBRT) has been traditionally employed in patients with resected brain metastases owing to its efficacy in reducing the risk of local recurrence in the surgical bed and the incidence of new distant metastases [bib_ref] Postoperative radiotherapy in the treatment of single metastases to the brain: a..., Patchell [/bib_ref]. Stereotactic radiosurgery (SRS), which is the recommended treatment for patients with a limited number of brain metastases [bib_ref] Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one..., Kocher [/bib_ref] [bib_ref] Effect of radiosurgery alone vs radiosurgery with whole brain radiation therapy on..., Brown [/bib_ref] , has been increasingly employed to target the postoperative resection cavity as an alternative to WBRT [bib_ref] Stereotactic radiosurgery versus whole-brain radiotherapy after intracranial metastasis resection: a systematic review..., Lamba [/bib_ref] [bib_ref] Post-operative stereotactic radiosurgery following excision of brain metastases: A systematic review and..., Akanda [/bib_ref] [bib_ref] Multidisciplinary patient-centered management of brain metastases and future directions, Palmer [/bib_ref]. Several retrospective series of stereotactic irradiation given as single fraction, referred to as SRS, or delivered in few fractions, typically named hypofractionated stereotactic radiotherapy (HSRT) or fractionated SRS, have shown local control rates from 70 to 90% at one year with low incidence of radiation-induced toxicity [bib_ref] Stereotactic radiosurgery versus whole-brain radiotherapy after intracranial metastasis resection: a systematic review..., Lamba [/bib_ref] [bib_ref] Post-operative stereotactic radiosurgery following excision of brain metastases: A systematic review and..., Akanda [/bib_ref]. Data from two randomized trials [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] have demonstrated that (1) SRS to the resection cavity significantly reduces bed recurrence rates compared with observation alone, and (2) decreases the risk of cognitive decline in patients with brain metastases as compared to WBRT, without diminishing survival [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref].
Based on this accumulated evidence, this approach has become the recommended treatment following surgical resection of a brain metastasis. However, optimal management of resected brain metastases remains challenging and several issues remain to be resolved, including the timing of postoperative SRS treatment, optimal radiation dose prescription and fractionation, and target delineation of the surgical bed [bib_ref] Stereotactic radiosurgery for resected brain metastases: new evidence supports a practice shift,..., Minniti [/bib_ref].
We provide a critical overview on current status and recent advances in resection cavity irradiation of brain metastases, with the aim of answering questions relevant to clinical and technical issues, such as the appropriate radiation technique, optimal radiation schedule, risk of leptomeningeal disease and treatment-related toxicity for patients receiving radiotherapy to postoperative resection cavity of brain metastases.
# Methods and materials
A literature search was conducted in MEDLINE PubMed using combinations of the following medical subjects headings (MeSH) and free-text words: "radiotherapy" or "radiosurgery" and "resection", "brain metastasis", "postoperative". We included clinical trials, retrospective studies, and review articles that were published within the past 15 years to reflect modern systemic therapies and neurosurgical and radiosurgical techniques. Articles were selected if they had (1) 1-year local control and/or rates of radiation-induced brain necrosis reported and (2) radiosurgery administered as definitive or postoperative treatment. Articles were excluded from the review if they had a non-English abstract, were not available through Pubmed, were pediatric series or case studies involving less than 30 patients, or were duplicated publications. To identify additional articles, the references of articles identified through the formal searches were scanned for additional sources. Based on the initial searches, a total of 352 articles were identified. Finally, 69 papers containing relevant data on clinical outcomes following postoperative SRS/HSRT in adult patients were chosen for this review.
# Results
## Local control and survival
The clinical success of SRS in patients presenting with a limited number of brain metastases resulted in its application to surgical cavities as an alternative to WBRT. Several retrospective studies reported local control and overall survival rates of 70% to 90% and of 50% to 70% at 12 months, respectively, following either SRS [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] Cavity-directed radiosurgery as adjuvant therapy after resection of a brain metastasis, Jensen [/bib_ref] [bib_ref] Stereotactic radiosurgery to the resection cavity of brain metastases: a retrospective analysis..., Rwigema [/bib_ref] [bib_ref] Current dosing paradigm for stereotactic radiosurgery alone after surgical resection of brain..., Prabhu [/bib_ref] [bib_ref] Radiosurgery to the surgical cavity as adjuvant therapy for resected brain metastasis, Robbins [/bib_ref] [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] A phase 2 trial of stereotactic radiosurgery boost after surgical resection for..., Brennan [/bib_ref] [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Stereotactic radiosurgery to the resection bed for intracranial metastases and risk of..., Ojerholm [/bib_ref] [bib_ref] Stereotactic radiosurgery to the resection cavity for brain metastases: prognostic factors and..., Abel [/bib_ref] [bib_ref] Hypofractionated radiosurgery has a better safety profile than single fraction radiosurgery for..., Eaton [/bib_ref] [bib_ref] Patterns of failure after stereotactic radiosurgery of the resection cavity following surgical..., Strauss [/bib_ref] [bib_ref] Surgical resection of brain metastases and the risk of leptomeningeal recurrence in..., Johnson [/bib_ref] [bib_ref] Feasibility and safety of cavity-directed stereotactic radiosurgery for brain metastases at a..., Rava [/bib_ref] [bib_ref] Impact of regular magnetic resonance imaging followup after stereotactic radiotherapy to the..., Bachmann [/bib_ref] and HSRT [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy to the post-operative cavity for radioresistant and radiosensitive brain..., Ahmed [/bib_ref] [bib_ref] Hypofractionated stereotactic radiation therapy to the resection bed for intracranial metastases, Keller [/bib_ref] [bib_ref] Outcomes of postoperative stereotactic radiosurgery to the resection cavity versus stereotactic radiosurgery..., Minniti [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery for resected brain metastases: a comparison of outcomes for..., Zhong [/bib_ref] [bib_ref] Does size matter? Investigating the optimal planning target volume margin for postoperative..., Jhaveri [/bib_ref] [bib_ref] Comparative effectiveness of multifraction stereotactic radiosurgery for surgically resected or intact large..., Minniti [/bib_ref] [bib_ref] Surgery followed by hypofractionated radiosurgery on the tumor bed in oligometastatic patients..., Navarria [/bib_ref] [bib_ref] Image guided, linac-based, surgical cavity-hypofractionated stereotactic radiotherapy in 5 daily fractions for..., Soliman [/bib_ref] [bib_ref] Stereotactic cavity irradiation or whole-brain radiotherapy following brain metastases resection-outcome, prognostic factors,..., El Shafie [/bib_ref] [bib_ref] Adverse radiation effect after hypofractionated stereotactic radiosurgery in 5 daily fractions for..., Faruqi [/bib_ref] [bib_ref] Evidence of dose-response following hypofractionated stereotactic radiotherapy to the cavity after surgery..., Garimall [/bib_ref] [bib_ref] Multi-institutional analysis of prognostic factors and outcomes after hypofractionated stereotactic radiotherapy to..., Eitz [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastases: single-institutional experience of over 500 cavities, Shi [/bib_ref]. The question on the efficacy and safety of postoperative SRS has been recently addressed in two randomized trials comparing postoperative SRS to observation or to WBRT, respectively [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref].
Mahajan et al.compared adjuvant SRS to observation in 128 patients who underwent gross total resection for 1-3 brain metastases between 2009 and 2016 at The University of Texas M.D. Anderson Cancer Center. The primary endpoint was the local tumor-free recurrence rate. The target volume was defined as the surgical cavity on the volumetric MR imaging with an additional margin of 1 mm. Prescription doses were 16, 14, and 12 Gy for target volumes of ≤ 10 cc, 10.1-15 cc, and > 15 cc, respectively, given in a single session by Gamma Knife. The 12-month tumor-free recurrence rates were 43% in the observation group and 72% in SRS group (p = 0.015), with comparable median overall survival times of 18 and 17 months. Amongst cavities treated with SRS, metastasis size was a significant predictor of local failure; 12-month local control rates were 91% for patients with tumors with a maximal diameter of ≤ 2.5 cm, 40% for patients with tumors > 2.5 to 3.5 cm in diameter, and 46% for patients with tumors > 3.5 cm in diameter. Considering that larger tumors received radiation doses of ≤ 14 Gy, these data indicate that lower SRS doses applied in patients with larger resection cavities, corresponding to a biological effective dose assuming an α/β ratio of 10 Gy for the tumor (BED 10Gy ) < 33.6 Gy , may be not sufficient to control microscopic disease. In addition, the trial confirmed previous evidence that surgical resection alone is insufficient to provide satisfactory local control [bib_ref] Postoperative radiotherapy in the treatment of single metastases to the brain: a..., Patchell [/bib_ref] [bib_ref] Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one..., Kocher [/bib_ref] despite improvements in neurosurgical techniques, such as stereotactic navigation and cortical mapping.
In the NCCTG N107C/CEC.3 prospective randomized trial of 194 patients with one resected brain metastasis and a resection cavity less than 5 cm in maximal size who were randomly assigned to either SRS (12 to 20 Gy) to WBRT Gy in 10-15 daily fractions), Brown et al. [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] reported superior preservation of neurocognitive function and quality of life in patients who received SRS with no negative impact on survival, although adjuvant WBRT was associated with better intracranial control compared to SRS. With similar median survival times of 12.2 months in the SRS arm and 11.6 months in the WBRT arm, median cognitive deterioration-freesurvival was longer in patients randomized to SRS at both 3 and 6 months, reaching statistical significance for immediate memory (p = 0.00062), delayed memory (p = 0.00054), processing speed (p = 0.023), and executive function (p = 0.015). The negative impact of WBRT on cognitive function, quality of life and functional independence remained persistent over time. The prescribed SRS dose was selected based on surgical cavity volume: 20 Gy if the cavity volume was less than 4.2 ml, 18 Gy Selected studies of postoperative stereotactic radiosurgery (SRS) to surgical bed LINAC, linear accelerator; GK, Gamma Knife; CK, CyberKnife; SRS, stereotactic radiosurgery; WBRT, whole brain radiation therapy; HSRT, hypofractionated stereotactic radiation therapy; OBS, observation; p, prospective; BED Gy10, biological equivalent dose with an α/β ratio of 10 Gy; BED Gy2, biological equivalent dose with an α/β ratio of 2 Gy; w, weeks; LC, local control; DP, dostant progression; OS, overall survival; NR, not reported [bib_ref] Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one..., Kocher [/bib_ref] [bib_ref] Effect of radiosurgery alone vs radiosurgery with whole brain radiation therapy on..., Brown [/bib_ref] and suggests that adjuvant SRS should be considered the recommended treatment for surgical bed because of significantly lower risk of cognitive decline and better quality of life compared to WBRT [bib_ref] Effect of radiosurgery alone vs radiosurgery with whole brain radiation therapy on..., Brown [/bib_ref] [bib_ref] A European Organisation for Research and Treatment of Cancer phase III trial..., Soffietti [/bib_ref]. While these randomized studies reported on singlefraction SRS, similar results have been observed following HSRT using different dose and fractionation schedules, typically 24-27 Gy given in three fractions or 25-30 Gy given in 5 fractions . Surveillance imaging following both SRS and HSRT to the resection cavity is important for the increased risk of distant brain failure after focal irradiation as opposed to WBRT. Therefore, frequent magnetic resonance imaging (MRI), typically at regular intervals of 2-3 months after SRS, is strongly recommended.
Several studies have evaluated the impact of different prognostic factors on local tumor control following radiation to the resection cavity. Larger preoperative tumor size and cavity volumes greater than 3 cm [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] Current dosing paradigm for stereotactic radiosurgery alone after surgical resection of brain..., Prabhu [/bib_ref] [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] A phase 2 trial of stereotactic radiosurgery boost after surgical resection for..., Brennan [/bib_ref] [bib_ref] Hypofractionated stereotactic radiation therapy to the resection bed for intracranial metastases, Keller [/bib_ref] [bib_ref] Does size matter? Investigating the optimal planning target volume margin for postoperative..., Jhaveri [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery without whole-brain radiation therapy for brain metastases: potential role..., Hartford [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy for local control of resected brain metastases, Traylor [/bib_ref] , incomplete resection [bib_ref] Stereotactic radiosurgery to the resection cavity for brain metastases: prognostic factors and..., Abel [/bib_ref] [bib_ref] Stereotactic cavity irradiation or whole-brain radiotherapy following brain metastases resection-outcome, prognostic factors,..., El Shafie [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy for local control of resected brain metastases, Traylor [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery for patients with resected brain metastases: a volumetric analysis, Patel [/bib_ref] , lower radiation dose [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Impact of regular magnetic resonance imaging followup after stereotactic radiotherapy to the..., Bachmann [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy for local control of resected brain metastases, Traylor [/bib_ref] [bib_ref] Impact of adjuvant fractionated stereotactic radiotherapy dose on local control of brain..., Musunuru [/bib_ref] , pretreatment tumor volume in contact with dura [bib_ref] A phase 2 trial of stereotactic radiosurgery boost after surgical resection for..., Brennan [/bib_ref] [bib_ref] Hypofractionated stereotactic radiation therapy to the resection bed for intracranial metastases, Keller [/bib_ref] , and longer interval time between surgery and radiation treatment [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Increasing time to postoperative stereotactic radiation therapy for patients with resected brain..., Yusuf [/bib_ref] [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] have been significantly correlated with worse local control. Factors associated with longer survival include Karnofsky Performance Status (KPS) score of 80% or greater, an interval less than 4 weeks between resection and postoperative radiation treatment [bib_ref] Increasing time to postoperative stereotactic radiation therapy for patients with resected brain..., Yusuf [/bib_ref] [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] , and a controlled primary tumor [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Comparative effectiveness of multifraction stereotactic radiosurgery for surgically resected or intact large..., Minniti [/bib_ref] [bib_ref] Multi-institutional analysis of prognostic factors and outcomes after hypofractionated stereotactic radiotherapy to..., Eitz [/bib_ref]. In contrast, combined systemic treatment and histology did not emerge as independent prognostic factors for either local control or survival in most studies [bib_ref] Impact of regular magnetic resonance imaging followup after stereotactic radiotherapy to the..., Bachmann [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy to the post-operative cavity for radioresistant and radiosensitive brain..., Ahmed [/bib_ref] [bib_ref] Evidence of dose-response following hypofractionated stereotactic radiotherapy to the cavity after surgery..., Garimall [/bib_ref] [bib_ref] Multi-institutional analysis of prognostic factors and outcomes after hypofractionated stereotactic radiotherapy to..., Eitz [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy for local control of resected brain metastases, Traylor [/bib_ref].
## Optimal dose and fractionation
Tables 1 and 2 show patient data and clinical outcomes of postoperative radiation to the resection cavity given as SRS or HSRT. Currently, there are several terms that have been used interchangeably for fractionated SRS, including multi-fraction, multi-dose, multi-session SRS, and hypofractionated stereotactic radiotherapy (HSRT) where dose is generally delivered in few, generally 2-5, fractions using frameless, mask-based SRS systems with the same level of accuracy of fixed-frame SRS [bib_ref] American College of Radiology (ACR) and American Society for Radiation Oncology (ASTRO)..., Seung [/bib_ref]. Using single-fraction SRS with doses of about 12 to 20 Gy, 16 studies including 1,556 patients show median local control and overall survival rates of about 60-90% and 50-70% at 12-months, respectively . For 1.749 patients receiving HSRT as postoperative treatment included in 13 studies , 12-month local control and overall survival rates were 88-95% and 58-82% using 24-33 Gy in 3 fractions, respectively, and 84-95% and 62-77% using 25-35 Gy in 5 daily fractions, respectively. Median cavity volumes were 12.7 ml (0.9-83 ml) for patients receiving postoperative SRS and 23.8 ml (2.8-283 ml) for those treated with HSRT. In a recent analysis of 588 resection cavities treated with postoperative irradiation included in nine studies, Lehrer et al. found no significant differences in the estimated 12-month local control between single-fraction SRS and fractionated SRS (68% vs 86.8%; p = 0.1); however, larger cavities were more likely to receive fractionated treatment.
A significant correlation between the radiation dose and local control has been observed for both SRS and HSRT [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Impact of regular magnetic resonance imaging followup after stereotactic radiotherapy to the..., Bachmann [/bib_ref] [bib_ref] Evidence of dose-response following hypofractionated stereotactic radiotherapy to the cavity after surgery..., Garimall [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery for patients with resected brain metastases: a volumetric analysis, Patel [/bib_ref] [bib_ref] Impact of adjuvant fractionated stereotactic radiotherapy dose on local control of brain..., Musunuru [/bib_ref]. In the Mahajan trial, SRS prescription doses were 16, 14, and 12 Gy for target volumes of ≤ 10 ml, 10.1-15 ml, and > 15 ml, respectively, given in a single session by Gamma Knife. Local control rates at 12 months of resection cavities were 91% for 40 patients with tumors with a maximal diameter of ≤ 2.5 cm receiving 16 Gy and 46% for 33 patients with tumors > 3.5 cm in diameter receiving 12 Gy (p = 0.0002). In the Brown trial [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] , local control decreased for postoperative cavity volumes > 20 ml who received radiation doses < 15 Gy, being significantly lower than that observed after WBRT. A significantly better local control with radiation dose ≥ 18 Gy has been observed in other retrospective studies [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Impact of regular magnetic resonance imaging followup after stereotactic radiotherapy to the..., Bachmann [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery for patients with resected brain metastases: a volumetric analysis, Patel [/bib_ref]. It needs to be added that data discussed above do not allow a separate analysis of SRS dose versus volume, as larger volumes were consistently treated with lower SRS doses because of concerns of toxicity.
For HSRT, most common schedules were 24-27 Gy in 3 fractions and 30-35 Gy in 5 fractions with a reported similar 12-month local control of about 85-95%, as shown in ; in contrast, lower doses, such as 5 × 5 Gy or 3 × 7 Gy, were associated with lower local control [bib_ref] Evidence of dose-response following hypofractionated stereotactic radiotherapy to the cavity after surgery..., Garimall [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Impact of adjuvant fractionated stereotactic radiotherapy dose on local control of brain..., Musunuru [/bib_ref]. In a retrospective study of 39 patients with 43 surgical beds treated with postoperative HSRT, Kumar et al. [bib_ref] Postoperative hypofractionated stereotactic brain radiation (HSRT) for resected brain metastases: improved local..., Kumar [/bib_ref] found that 30 Gy in 5 fractions and 27 Gy in 3 fractions provided better local control (93-100%) compared to lower dose 3-and 5-fraction regimens. Using the linear quadratic model to compare radiation doses of different fractionation schedules to predict tumor control probability and normal tissue complication probability [bib_ref] The alfa and beta of tumours: a review of parameters of the..., Van Leeuwen [/bib_ref] , available data indicate that BED 10Gy > 40 Gy should be delivered to the surgical bed to achieve excellent local Selected studies of postoperative hypofractionated stereotactic radiotherapy (HSRT) to surgical bed LINAC, linear accelerator; GK, Gamma Knife; CK, CyberKnife; SRS, stereotactic radiosurgery; WBRT, whole brain radiation therapy; HSRT, hypofractionated stereotactic radiation therapy; OBS, observation; BED Gy10, biological equivalent dose with an α/β ratio of 10 Gy; BED Gy2, biological equivalent dose with an α/β ratio of 2 Gy; w, weeks; LC, local control; DP, dostant progression; OS, overall survival; NR, not reported control. Calculation of BED 10Gy for the tumor and BED 2Gy for brain parenchyma with respective equivalent doses in 2 Gy fractions (EQD2 2 ) using different dose and fractionation is shown in . In the respect of healthy tissue constraints, this means in clinical practice that radiation doses greater than 16 Gy given as single fraction, 24 Gy given in 3 fractions, and > 27.5 Gy given in 5 fractions should be recommended to improve local cavity control, especially in patients with radioresistant tumors.
## Cavity volume dynamics and timing of treatment
Target delineation of a resected brain metastasis is typically represented by the rim of enhancement at the edge of the resection cavity. While accurate contouring can be performed using thin-slice contrast-enhanced T1-weighted MRI sequences, the challenge is that the surgical bed is dynamic after surgery and prone to significant changes in resection cavity dimensions before SRS treatment, subsequently increasing the risk of missing the target and delivering unnecessary high radiation doses to surrounding normal brain parenchyma. Several, but not all, studies reporting on dynamic changes of resection cavity aiming to define the optimal SRS treatment timing indicate a postoperative decrease of the cavity volumes [bib_ref] Fractionated stereotactic radiotherapy to the post-operative cavity for radioresistant and radiosensitive brain..., Ahmed [/bib_ref] [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastasis: cavity dynamics and factors affecting its..., Alghamdi [/bib_ref] [bib_ref] Cavity volume dynamics after resection of brain metastases and timing of postresection..., Atalar [/bib_ref] [bib_ref] Surgical cavity constriction and local progression between resection and adjuvant radiosurgery for..., Shah [/bib_ref] [bib_ref] Cavity volume changes after surgery of a brain metastasisconsequences for stereotactic radiation..., Scharl [/bib_ref]. In a series of 57 patients who received postoperative SRS to the resection cavity, Scharl et al. [bib_ref] Cavity volume changes after surgery of a brain metastasisconsequences for stereotactic radiation..., Scharl [/bib_ref] found significantly average cavity-volume reduction of 23.4% occurring between immediate post-resection MRI and planning MRI (p < 0.01). Regardless of the initial volume, cavity shrinkage occurred in 79.1%, remained stable in 3.5%, and increased in 17.4% of cases at a median time of 4 weeks after surgery. In another series of 59 patients with 61 cavities treated with postoperative SRS to the resection cavity, Alghamdi et al. [bib_ref] Stereotactic radiosurgery for resected brain metastasis: cavity dynamics and factors affecting its..., Alghamdi [/bib_ref] found an average cavity volume reduction of 22.5% at a median time of 4 weeks after surgery, with most changes occurring within 3 weeks. Tumor size > 3 cm, dural involvement and longer time from surgery were significant predictors of cavity volume reduction. Overall, an average cavity volume reduction of 15% to 43% has been reported in several published studies [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] [bib_ref] Cavity volume dynamics after resection of brain metastases and timing of postresection..., Atalar [/bib_ref] [bib_ref] Surgical cavity constriction and local progression between resection and adjuvant radiosurgery for..., Shah [/bib_ref] [bib_ref] Tumor bed dynamics after surgical resection of brain metastases: implications for postoperative..., Jarvis [/bib_ref] , with larger tumor cavities (> 3 cm) that are associated with greater reduction. With regard to the timing, cavity volume reduction occurs within the first 3-4 weeks after surgery in 58-90% of resected brain metastases [bib_ref] Fractionated stereotactic radiotherapy to the post-operative cavity for radioresistant and radiosensitive brain..., Ahmed [/bib_ref] [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastasis: cavity dynamics and factors affecting its..., Alghamdi [/bib_ref] [bib_ref] Surgical cavity constriction and local progression between resection and adjuvant radiosurgery for..., Shah [/bib_ref] [bib_ref] Cavity volume changes after surgery of a brain metastasisconsequences for stereotactic radiation..., Scharl [/bib_ref] ; however, no change or increase in cavity size have been reported in few studies in the first 3-4 weeks [bib_ref] The Relationship between tumor volume and timing of post-resection stereotactic radiosurgery to..., Yuan [/bib_ref] [bib_ref] Cavity volume dynamics after resection of brain metastases and timing of postresection..., Atalar [/bib_ref] [bib_ref] Tumor bed dynamics after surgical resection of brain metastases: implications for postoperative..., Jarvis [/bib_ref]. The reported high incidence of significant changes in the postoperative resection cavity raises the question of the optimal timing for SRS treatment. As the majority of studies indicate that shrinkage occurs consistently over time in a significant proportion of patients, waiting a few weeks to perform SRS may represent an effective strategy to treat a smaller cavity volume, possibly limiting the risk of neurological toxicity while maintaining the efficacy of treatment. However, longer intervals more than 3-4 weeks between surgery and radiation treatment should be avoided because they have been associated with an increased risk of worse local control [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] [bib_ref] Patterns of failure after stereotactic radiosurgery of the resection cavity following surgical..., Strauss [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery without whole-brain radiation therapy for brain metastases: potential role..., Hartford [/bib_ref] [bib_ref] Timing of adjuvant fractionated stereotactic radiosurgery impacts local control of resected brain..., Roth [/bib_ref]. In a retrospective series of 110 patients with 113 cavities treated with postoperative Gamma Knife SRS with a marginal dose of 18 Gy, Iorio-Morin et al. [bib_ref] Early Gamma Knife stereotactic radiosurgery to the tumor bed of resected brain..., Iorio-Morin [/bib_ref] reported local control rates of 73% at 12 months. Lower maximum radiation dose and a surgery-to-SRS delay longer than 3 weeks were risk factors for local recurrence. The estimated 12-month control rates dropped from 87 to 61% if SRS was performed more than 3 weeks after resection. This difference in rates of surgical bed control remained throughout follow-up; at 36 months, the group that received SRS less than three weeks after surgery had a 72% rate of local control compared to 46% for patients who received SRS more than three weeks after surgery. A possible explanation is that a longer delay might lead to an increased spread of microscopic disease that is harder to target because it is not yet radiographically evident. In this regard, other studies have observed a significant correlation between increasing delay between surgery and SRS and local failure [bib_ref] Cavity-directed radiosurgery as adjuvant therapy after resection of a brain metastasis, Jensen [/bib_ref] [bib_ref] Patterns of failure after stereotactic radiosurgery of the resection cavity following surgical..., Strauss [/bib_ref] [bib_ref] The role of radiosurgery to the tumor bed after resection of brain..., Gans [/bib_ref] [bib_ref] Tumor bed radiosurgery after resection of cerebral metastases, Mathieu [/bib_ref]. Overall, the median interval reported in the majority of studies of either SRS or HSRT was 19 days, with few exceptions of patients exceeding 5-6 weeks. Even though protocols are different with regard to technique, dose fractionation, and interval between surgery and radiation treatment, there is a general consensus to perform postoperative SRS/HSRT to the resection cavity within maximum four weeks after Biological equivalent dose (BED) and equivalent dose in 2 Gy per fraction (EQD2) for various SRS/HSRT radiation schedules BED Gy10, biological equivalent dose with an α/β ratio of 10 Gy; BED Gy2, biological equivalent dose with an α/β ratio of 2 Gy; EQD10/2, eqivalent dose in 2 Gy/fractions with a BED Gy10; EQD2/2, eqivalent dose in 2 Gy/fractions with a BED Gy2; SRS, stereotactic radiosurgery; HSRT, hypofractionated stereoactic radiation therapy surgery with planning MRI acquired < 7 days before treatment to limit negative impact of cavity changes on clinical outcomes.
## Target volume delineation and margins
Target delineation of the resection cavity remains challenging and has not yet been defined, especially in the setting of large volumes. This may be the reason that in some studies, long-term local control after SRS has been found to be worse compared to post-resection WBRT [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref]. In a recently published consensus guideline on target delineation of the postoperative cavity, the primary recommendations for CTV delineation using contrastenhancing T1-weighted MRI scan include contouring of the entire surgical cavity with the exclusion of vasogenic edema and include a margin up to 5 mm along the bone flap/meningeal margin [bib_ref] Consensus contouring guidelines for postoperative completely resected cavity stereotactic radiosurgery for brain..., Soliman [/bib_ref]. For tumors in contact with the dura preoperatively, the guidelines recommend a GTV-to-CTV margin up to 10 mm along the bone flap beyond the initial region of preoperative tumor contact. An example of target delineation is shown in (to be chosen). In a study from University of California San Francisco of 58 patients with 60 resection cavity who received postoperative SRS by Sukso et al., preoperative dural contact increased recurrence rate after postoperative SRS and the median distance of marginal recurrences from the target volume was 3 mm, supporting the CTV delineation consensus guidelines. Of note, the addition of a 10-mm dural margin increased the target volume overlap with the recurrence contours for 10 of the 14 recurrences. Further recommendations for CTV delineation comprise the inclusion of the entire surgical tract and a margin of 1 to 5 mm along the sinus for those tumors that were in contact with a venous sinus preoperatively. These consensus guidelines provide suggestions for standardized postoperative cavity contouring indicating that target definition should be guided by both the preoperative volume and location of lesion and also the postoperative changes as seen at MRI scan; however, controversies continue to exist. In this regard, the use of further margins beyond the contouring for the surgical cavity remains to be defined. Use of margins may improve target coverage and compensate contouring inaccuracy, but SRS to large treatment volumes can be associated with an increased risk of radiation necrosis [bib_ref] Post-operative stereotactic radiosurgery following excision of brain metastases: A systematic review and..., Akanda [/bib_ref] [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Does size matter? Investigating the optimal planning target volume margin for postoperative..., Jhaveri [/bib_ref] [bib_ref] Single versus multifraction stereotactic radiosurgery for large brain metastases: an international meta-analysis..., Lehrer [/bib_ref]. In most studies, margins of 0 to 3 mm for GTV/CTV expansion provide equivalent 1-year local tumor control rates with no evidence of a significantly increased risk of radiation necrosis after either SRS or HSRT ; in contrast, a few studies suggested that the use of margins of 2 mm is associated with better local control [bib_ref] Predicting tumor control after resection bed radiosurgery of brain metastases, Luther [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Stereotactic radiosurgery of the postoperative resection cavity for brain metastases, Soltys [/bib_ref]. Another controversial issue is the inclusion of surgical access track. Several studies did not include the surgical tract for deep lesions [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] A phase 2 trial of stereotactic radiosurgery boost after surgical resection for..., Brennan [/bib_ref] [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Outcomes of postoperative stereotactic radiosurgery to the resection cavity versus stereotactic radiosurgery..., Minniti [/bib_ref] [bib_ref] Surgery followed by hypofractionated radiosurgery on the tumor bed in oligometastatic patients..., Navarria [/bib_ref] [bib_ref] Image guided, linac-based, surgical cavity-hypofractionated stereotactic radiotherapy in 5 daily fractions for..., Soliman [/bib_ref] [bib_ref] Multi-institutional analysis of prognostic factors and outcomes after hypofractionated stereotactic radiotherapy to..., Eitz [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastases: single-institutional experience of over 500 cavities, Shi [/bib_ref] [bib_ref] Gamma Knife surgery combined with resection for treatment of a single brain..., Kalani [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastases: does the surgical corridor need to..., Shi [/bib_ref]. In a series of 64 patients with 66 cavities An overview of target volumes for postoperative resection cavity is presented on post-contrast T1-weighted MRI sequences and CT scans. The gross tumor volume (GTV) is presented in red, the clinical target volume (CTV) in blue and the planning tumor volume (PTV) in pink. For this case, CTV was created by 1-mm expansion of the GTV, extended by 5 mm along the bone flap beyond the initial region of preoperative tumor contact. Note that an extension by 10 mm along the meningeal margin for brain metastases with preoperative dural contact [bib_ref] Consensus contouring guidelines for postoperative completely resected cavity stereotactic radiosurgery for brain..., Soliman [/bib_ref] or the inclusion of the entirety of the craniotomy site [bib_ref] Accuracy of F-DOPA PET and perfusion-MRI fo differentiating radionecrotic from progressive brain..., Cicone [/bib_ref] has been suggested by some authors receiving postoperative SRS for a resected brain metastasis with or without inclusion of surgical corridor in the CTV, Shi et al. [bib_ref] Stereotactic radiosurgery for resected brain metastases: single-institutional experience of over 500 cavities, Shi [/bib_ref] showed that omitting the surgical corridor was not associated with statistically significant differences in corridor or cavity recurrence or adverse radiation effects. Overall, current recommendations for accurate target delineation of postoperative resection cavity include the use of thin-sliced contrast-enhancing T1-weighted MRI with the inclusion of generous meningeal margins up to 1 cm in the CTV without any area of the surrounding edema. To accurately identify the preoperative tumor extent and dural involvement, preoperative contrast-enhanced T1-weighted MRI is preferred. The impact of different target volume delineation and margins in terms of local control and increased risk of radiation necrosis remains to be defined.
## Risk of leptomeningeal disease
Leptomeningeal disease (LMD) is defined as the spread of tumor cells within the leptomeninges and the subarachnoid space and occurs approximately in up to 10% of patients with solid cancer during the course of disease, commonly in the context of progressive systemic disease. The diagnosis of leptomeningeal metastases can be challenging. It is based on clinical evaluation, cerebrospinal MRI and cerebrospinal fluid (CSF) analysis [bib_ref] Gamma Knife surgery combined with resection for treatment of a single brain..., Kalani [/bib_ref]. The classification of leptomeningeal metastasis considers also the imaging presentation which guides clinical decisionmaking independently of the identification of tumor cells in the CSF. MRI abnormalities of LMD include enhancement of the leptomeninges of the brain or spinal cord identified as enhancement of the cranial nerves and spinal nerve roots, brain surface, cerebellar folia, and within cerebral sulci.
A risk of LMD development up to 28% has been observed after surgical resection and adjuvant postoperative SRS/HSRT of brain metastases. Recent series observed an incidence of 6-15% at one year , although most studies did not include data on the risk of LMD. Its development is presumably related to iatrogenic dissemination of tumor cells into cerebrospinal fluid and meninges at the time of resection, resulting in nodule forming subsequently. The variable risk reported across all studies may depend on differences in tumor histology, tumor size and location, pial involvement, and type of surgical resection. In addition, differences in imaging follow-up and discordance in physicians' assessment of LMD are potential factors that explain such variable incidence. For example, the reported risk of LMD observed in Mahajanand Brown [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] randomized trials were 28% and 7%, respectively, suggesting diagnostic variability. Factors associated with the development of LMD include breast cancer histology [bib_ref] Stereotactic radiosurgery to the resection bed for intracranial metastases and risk of..., Ojerholm [/bib_ref] [bib_ref] Surgical resection of brain metastases and the risk of leptomeningeal recurrence in..., Johnson [/bib_ref] [bib_ref] Does size matter? Investigating the optimal planning target volume margin for postoperative..., Jhaveri [/bib_ref] [bib_ref] Cavity volume dynamics after resection of brain metastases and timing of postresection..., Atalar [/bib_ref] [bib_ref] Postoperative radiosurgery for the treatment of metastatic brain tumor: Evaluation of local..., Foreman [/bib_ref] , posterior fossa location [bib_ref] Post-operative stereotactic radiosurgery following excision of brain metastases: A systematic review and..., Akanda [/bib_ref] [bib_ref] Stereotactic radiosurgery to the resection bed for intracranial metastases and risk of..., Ojerholm [/bib_ref] , multiple brain metastases [bib_ref] Surgical resection of brain metastases and the risk of leptomeningeal recurrence in..., Johnson [/bib_ref] , type of surgical resection (piecemeal instead of "en bloc") [bib_ref] Comparative risk of leptomeningeal disease after resection or stereotactic radiosurgery for solid..., Suki [/bib_ref] [bib_ref] Comparative risk of leptomeningeal dissemination of cancer after surgery or stereotactic radiosurgery..., Suki [/bib_ref].
An important finding that emerged from some studies is a peculiar pattern of the meningeal spread after postoperative cavity radiation. In a retrospective series of 1,188 patients with newly diagnosed brain metastases managed with neurosurgical resection and stereotactic radiation (n = 318) or radiation alone (n = 870), Cagney et al. [bib_ref] Association of neurosurgical resection with development of pachymeningeal seeding in patients with..., Cagney [/bib_ref] examined two patterns of intracranial recurrence: (1) the "classical" LMD, defined as subarachnoid enhancement involving the sulci of the cerebral hemispheres, cranial nerves, brainstem, cerebellar folia, or ependyma, and (2) pachymeningeal seeding, defined as nodular, enhancing tumors stemming from the pachymeninges (dura and/ or outer arachnoid) extending 1 cm beyond the planning target volume of the stereotactic field. They found that resection was associated with pachymeningeal seeding (36 of 318 patients vs 0 of 870 patients; p < 0.001), but not with leptomeningeal disease (hazard ratio, 1.14; 95% CI, 0.73-1.77; p = 0.56). Prabhu et al. [bib_ref] A multiinstitutional analysis of presentation and outcomes for leptomeningeal disease recurrence after..., Prabhu [/bib_ref] characterized the pattern of intracranial recurrence in 147 patients who developed LMD following surgery and postoperative SRS for at least one brain metastasis. At a median time from postoperative SRS of 5.6 months, 42.9% of patients presented with classical LMD, while 57.1% presented with nodular LMD, defined as new focal extra-axial distinct nodular enhancing lesions located on the meninges or ependyma. Within the nodular LMD, the median number of nodules was two and the median distance between the surgical corridor and the closest nodule was 2 cm, with about 70% of patients having LMD nodules within 5 cm of the surgical corridor. Patients with nodular LMD had significantly longer median overall survival than those with classical LMD (8.2 vs. 3.3 months, p < 0.001). A new classification for intracranial progression which takes into account this peculiar pattern of intracranial nodular LMD following postoperative SRS together with classical LMD, local parenchymal recurrence, and distant intraparenchymal metastases has been suggested by these authors.
The increased shift in the pattern of intracranial recurrence after surgery and postoperative SRS to surgical bed raises the question on the optimal postoperative radiation technique for these patients. Even though adjuvant WBRT is associated with a lower risk of LMD and better local control compared to postoperative SRS, especially in case of large cavities [bib_ref] Stereotactic radiosurgery versus whole-brain radiotherapy after intracranial metastasis resection: a systematic review..., Lamba [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] Cavity volume dynamics after resection of brain metastases and timing of postresection..., Atalar [/bib_ref] [bib_ref] Intracranial control and radiographic changes with adjuvant radiation therapy for resected brain..., Patel [/bib_ref] , it should be noted that randomized studies have not identified a survival benefit with WBRT for either resected or intact metastases [bib_ref] Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one..., Kocher [/bib_ref] [bib_ref] Effect of radiosurgery alone vs radiosurgery with whole brain radiation therapy on..., Brown [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref]. Therefore, the use of postoperative SRS remains a reasonable approach to avoid neurocognitive decline associated with the use of WBRT. Future studies need to evaluate the impact of different focal radiation approaches to postoperative surgical bed in terms of dose delivery and target delineation, with the aim of reducing the high risk of "near target" localized pachymeningeal seeding, while maintaining the superiority of the SRS/ HSRT approach on neurocognitive function and qualityof-life outcomes compared to WBRT. Additionally, data on outcomes of salvage treatment are needed to learn how to optimally treat patients with different patterns of intracranial progression.
## Risk of radiation necrosis
Radiation necrosis is the most significant adverse effect of radiation treatment of brain metastases. Radiation necrosis presents as a focal enhancing lesion at a variable time of 6-15 months following SRS/HSRT; however, the differential diagnosis between tumor progression and radiation necrosis remains challenging. While pathological confirmation remains the gold standard for diagnosis, non-invasive imaging techniques, including perfusion-weighted MRI and PET using amino acid tracers 11C-methionine (ref ), O-(2-18F-fluoroethyl)l-tyrosine (18F-FET) and 6-18-F-fluoro-l-dopa have emerged as highly sensitive diagnostic tools for distinguishing radiation necrosis from tumor recurrence [bib_ref] PET imaging in patients with brain metastasisreport of the RANO/PET group, Galldiks [/bib_ref] [bib_ref] Accuracy of F-DOPA PET and perfusion-MRI fo differentiating radionecrotic from progressive brain..., Cicone [/bib_ref]. Current treatments for symptomatic radiation necrosis include corticosteroids, surgery, bevacizumab, and hyperbaric oxygen. The 12-month estimated risk of radiation necrosis following postoperative radiation of brain metastases ranges from 1.5% to 28% being similar after postoperative SRS and HSRT ; however, HSRT is typically delivered to much larger resection cavities. Although some retrospective series report radiological changes suggestive of radiation necrosis in more than 20% of patients treated with postoperative radiation, the 12-month estimated risk of symptomatic radiation necrosis is about 5-10% in the majority of studies. In a systematic review and meta-analysis on postoperative SRS following excision of brain metastases, Akanda et al. [bib_ref] Multidisciplinary patient-centered management of brain metastases and future directions, Palmer [/bib_ref] observed a similar incidence of less than 10% in 28 out of 36 studies using different imaging modalities. Although there is no head-to-head comparison of postoperative HSRT versus single-fraction SRS to the surgical bed, the relatively low risk of radiation necrosis after HSRT for volumes larger than 20-25 ml suggests that hypofractionation may represent a better approach for large cavities [bib_ref] Hypofractionated radiosurgery has a better safety profile than single fraction radiosurgery for..., Eaton [/bib_ref] [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Surgery followed by hypofractionated radiosurgery on the tumor bed in oligometastatic patients..., Navarria [/bib_ref] [bib_ref] Image guided, linac-based, surgical cavity-hypofractionated stereotactic radiotherapy in 5 daily fractions for..., Soliman [/bib_ref]. A risk of radiation necrosis less than 10% has been generally observed in studies of HSRT using either 24-27 Gy given in three fractions or 30-35 Gy given in 5 fractions, corresponding to an equivalent dose in 2 Gy fractions of 62-78.7 Gy using an alpha/beta of 2 Gy (EQD2 2 ), and to a BED 2Gy of 124-157.5 Gy . In a series of 45 consecutive patients who received fractionated partial brain radiation therapy to the surgical cavity [bib_ref] Comparative effectiveness of multifraction stereotactic radiosurgery for surgically resected or intact large..., Minniti [/bib_ref] [bib_ref] Surgery followed by hypofractionated radiosurgery on the tumor bed in oligometastatic patients..., Navarria [/bib_ref] [bib_ref] Image guided, linac-based, surgical cavity-hypofractionated stereotactic radiotherapy in 5 daily fractions for..., Soliman [/bib_ref] [bib_ref] Stereotactic cavity irradiation or whole-brain radiotherapy following brain metastases resection-outcome, prognostic factors,..., El Shafie [/bib_ref] [bib_ref] Adverse radiation effect after hypofractionated stereotactic radiosurgery in 5 daily fractions for..., Faruqi [/bib_ref] [bib_ref] Evidence of dose-response following hypofractionated stereotactic radiotherapy to the cavity after surgery..., Garimall [/bib_ref] [bib_ref] Multi-institutional analysis of prognostic factors and outcomes after hypofractionated stereotactic radiotherapy to..., Eitz [/bib_ref] [bib_ref] Stereotactic radiosurgery for resected brain metastases: single-institutional experience of over 500 cavities, Shi [/bib_ref] [bib_ref] A European Organisation for Research and Treatment of Cancer phase III trial..., Soffietti [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery without whole-brain radiation therapy for brain metastases: potential role..., Hartford [/bib_ref] [bib_ref] Local recurrence patterns after postoperative stereotactic radiation surgery to resected brain metastases:..., Gui [/bib_ref] [bib_ref] Fractionated stereotactic radiotherapy for local control of resected brain metastases, Traylor [/bib_ref] [bib_ref] Postoperative stereotactic radiosurgery for patients with resected brain metastases: a volumetric analysis, Patel [/bib_ref] Gy in 3-Gy per fraction) at the Massachusetts General Hospital between April 2012 and September 2017, Byrne et al. [bib_ref] Post-operative radiation therapy to the surgical cavity with standard fractionation in patients..., Byrne [/bib_ref] reported 12-month freedom from local failure rates of 88.2% with no events of late radiation necrosis.
Factors correlated with an increased risk of radiation necrosis include higher radiation dose, larger volumes, and combined immunotherapy [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Does size matter? Investigating the optimal planning target volume margin for postoperative..., Jhaveri [/bib_ref] [bib_ref] Comparative effectiveness of multifraction stereotactic radiosurgery for surgically resected or intact large..., Minniti [/bib_ref] [bib_ref] Adverse radiation effect after hypofractionated stereotactic radiosurgery in 5 daily fractions for..., Faruqi [/bib_ref] [bib_ref] Cerebral radiation necrosis: incidence, outcomes, and risk factors with emphasis on radiation..., Ruben [/bib_ref]. Several studies have found a significant correlation between volume of brain receiving high-dose irradiation and the risk of radiation necrosis after either SRS or HSRT for intact and resected brain metastases [bib_ref] Irradiated volume as a predictor of brain radionecrosis after linear accelerator stereotactic..., Blonigen [/bib_ref] [bib_ref] Stereotactic radiosurgery for brain metastases: analysis of outcome and risk of brain..., Minniti [/bib_ref] [bib_ref] Single-fraction versus multifraction (3 × 9 Gy) stereotactic radiosurgery for large (>2..., Minniti [/bib_ref]. For patients undergoing SRS, the volume of normal brain receiving 12 Gy (brain minus GTV; V12 Gy) > 5-10 ml is predictive of a > 10% risk of radiation necrosis [bib_ref] Estimating normal tissue toxicity in radiosurgery of the CNS: application and limitations..., Kirkpatrick [/bib_ref] [bib_ref] Single-and multifraction stereotactic radiosurgery dosevolume tolerances of the brain, Milano [/bib_ref]. In the context of HSRT, volumetric constraints for brain predicting the risk of radiation necrosis include V18 Gy and V24 Gy for 3-fraction regimens and V25 Gy for 5-fraction regimens [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref] [bib_ref] Adverse radiation effect after hypofractionated stereotactic radiosurgery in 5 daily fractions for..., Faruqi [/bib_ref] [bib_ref] Single-and multifraction stereotactic radiosurgery dosevolume tolerances of the brain, Milano [/bib_ref] [bib_ref] Dosimetric predictors of symptomatic radiation necrosis after five-fraction radiosurgery for brain metastases, Andruska [/bib_ref]. In a recent review of single-and multifraction SRS dose/volume tolerances of the brain including 51 studies published from January 1995 through December 2016, Milano et al. [bib_ref] Single-and multifraction stereotactic radiosurgery dosevolume tolerances of the brain, Milano [/bib_ref] reported brain volumes (brain plus target volume) receiving 20 Gy in 3-fractions or V24 Gy in 5-fractions < 20 ml were associated with < 10% risk of any necrosis or edema in patients with brain metastases.
In a series of 101 patients with brain metastases treated with surgery and postoperative SRS (9 Gy × 3) to the resection cavity at University of Rome Sapienza, Sant' Andrea Hospital, the V24 Gy calculated as normal brain less GTV was the most significant factor associated with the development of radiation necrosis. The crude risk of radiation necrosis was 16% for V24 Gy ≥ 16.8 ml and 2% for V24 Gy < 16.8 ml (p = 0.03), with respective 12-month risk of 8% and 3% [bib_ref] Multidose stereotactic radiosurgery (9 Gy × 3) of the postoperative resection cavity..., Minniti [/bib_ref]. No other factors, including histology, site of tumor, PTV, and conformity index were predictive of radiation necrosis. In another series of 289 consecutive patients who received SRS or HSRT (9 Gy × 3) for at least one brain metastasis > 2.0 cm as primary treatment at Sant' Andrea Hospital, University of Rome Sapienza, the 1-year cumulative incidence rate of radionecrosis was 18% for patients undergoing SRS and 9% for those receiving HSRT (p = 0.01), respectively. For patients receiving HSRT, the V18 Gy was the most significant prognostic factor for radiation necrosis; the incidence was 5% for V18 Gy ≤ 30 ml and 14% for V18 Gy > 30 ml (p = 0.04).
In another series of 187 consecutively treated patients with 118 surgical cavities and 132 intact metastases treated with HSRT (30 Gy in 5 fractions), Faruqi et al. [bib_ref] Adverse radiation effect after hypofractionated stereotactic radiosurgery in 5 daily fractions for..., Faruqi [/bib_ref] showed that the total brain minus gross tumor volume (GTV) receiving 30 Gy (V30) was a significant risk factor for symptomatic radiation necrosis with a threshold of 10.5 ml or more (OR 7.2; p = 0.02). The 1-year symptomatic radiation necrosis rate was 13% for V30 < 10.5 ml and 61% for V30 ≥ 10.5 ml. In a multiinstitutional retrospective review of 117 brain metastases from 83 patients treated with 5 fraction HSRT, Andruska et al. [bib_ref] Dosimetric predictors of symptomatic radiation necrosis after five-fraction radiosurgery for brain metastases, Andruska [/bib_ref] found a two-year risk of symptomatic radiation necrosis of 21% for V25 > 16 ml and V30 > 10 ml and 2% for V25 ≤ 16 ml and V30 ≤ 10 ml (p = 0.007). In another series of 55 resected brain metastases that were treated postoperatively with HSRT (25-35 Gy in 5 fractions), Tanenbaum et al. [bib_ref] Dosimetric factors related to radiation necrosis after 5-fraction radiosurgery for patients with..., Tanenbaum [/bib_ref] observed a 1-year incidence of radiation necrosis of 18.2%; hotspots within the PTV expansion margin > 105% and an absolute dose of 33.5 Gy were significantly associated with the development of radiation necrosis, but hotspots within the CTV did not.
## Future directions
The role of SRS in patients with resected brain metastases will continue to evolve. Postoperative SRS to the resection cavity has become the standard of care after surgery, as it provides local control rates comparable to WBRT, better than with surgery alone, and without a negative impact on survival; however, a few studies have suggested worse local control for large brain metastases after SRS compared to WBRT [bib_ref] Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain..., Brown [/bib_ref] [bib_ref] Stereotactic radiotherapy of the tumor bed compared to whole brain radiotherapy after..., Kępka [/bib_ref]. Future research needs to evaluate the impact of different dose and fractionation on the surgical bed in terms of brain control and risk of radiation necrosis, especially for large volumes that are apparently associated with worse local control following SRS. In this regard, it will be important to compare this approach with alternative strategies, such as fractionated partial brain RT with more generous GTV-to-CTV/PTV margins or WBRT with hippocampal avoidance. A phase III trial of post-surgical single fraction SRS compared with HSRT for resected metastatic brain disease evaluating the time to surgical bed failure as primary endpoint is currently recruiting patients in the US (ClinicalTrials.gov, NCT04114981). Other critical areas of research include understanding the pattern of LMD spread and the optimal timing of adjuvant SRS since surgical cavities undergo morphological changes dependent on the time from surgery.
New strategies to enhance local control and minimize the risk of leptomeningeal disease include pre-operative SRS and the use of systemic agents, alone or in combination with radiation therapy. The rationale for pre-operative SRS is to treat tumor cells prior to potential iatrogenic dissemination at the time of surgical resection, potentially decreasing the rate of leptomeningeal disease. In addition, contouring an intact tumor for pre-operative SRS is much less challenging than for a resection cavity and if no added margin is needed, this approach may result in lower risk of radiation necrosis. In this regard, a few studies have demonstrated the safety and efficacy of preoperative SRS, reporting local control rates of 80 to 90% at 1 year with, with respective risk of symptomatic radiation necrosis and development of leptomeningeal disease of 5-10% [bib_ref] Comparing preoperative with postoperative stereotactic radiosurgery for resectable brain metastases: a multi-institutional..., Patel [/bib_ref] [bib_ref] Efficacy of pre-operative stereotactic radiosurgery followed by surgical resection and correlative radiobiological..., Huff [/bib_ref] [bib_ref] Preoperative stereotactic radiosurgery before planned resection of brain metastases: updated analysis of..., Prabhu [/bib_ref] [bib_ref] Preoperative vs postoperative radiosurgery for resected brain metastases: a review, Prabhu [/bib_ref]. Two prospective trials randomizing patients undergoing pre-operative SRS versus post-operative are currently recruiting patients (Clin-icalTrials.gov, NCT03741673 and NCT03398694).
# Conclusions
Overall, just as our paradigm has shifted from WBRT to SRS for patients with a limited number of intact brain metastases, postoperative SRS is replacing WBRT for Summary of imaging modalities for target volumes delineation and dose/fractionations for postoperative resection cavity of brain metastases SRS, stereotactic radiosurgery; HSRT, hypofractionated stereotactic radiation therapy; MRI, magnetic resonance imaging; 3D, 3-dimensional Imaging for target delineation Isotropic post-contrast-enhanced 3D T1-weighted MRI sequences with 1 mm thick slices and T2-weighted images. Additional images include preoperative contrast-enhanced T1-weighted MRI sequences to identify the preoperative tumor extent and dural involvement Gross Tumor Volume (GTV) Surgical cavity on postoperative contrast-enhanced T1-weighted MR images (typically represented by the rim of enhancement at the edge of the resection cavity) with inclusion of any residual nodular enhancement Clinical Tumor Volume (CTV) The CTV is defined as the GTV plus 0-1 mm margins constrained at anatomical barriers such as the skull. GTV-to-CTV margins up to 5-10 mm are applied along the bone flap/meningeal margin, with larger margins used for tumors in contact with the dura preoperatively. Vasogenic edema and surgical corridor (for deep lesions) are not usually included Planning Target Volume (PTV) A margin of up to 3 mm is usually added to the CTV to generate the PTV, depending on the radiation technique. For frame-based SRS, no additional safety margin is necessary; with frameless SRS and SRT, a GTV-to-PTV safety margin of 1-3 mm is usually applied according to Institutional practice Timing of treatment There is a general consensus to perform postoperative SRS/HSRT to the resection cavity within 4 weeks after surgery with planning MRI acquired < 7 days before treatment to limit negative impact of cavity changes on clinical outcomes Dose and fractionation 12-18 Gy using single-fraction SRS; 24-27 Gy in 3 fractions and 30-35 Gy in 5 fractions using HSRT, typically for larger resection cavity; less commonly 30-40 Gy in 10 fractions patients with resected brain metastases as the standard of care. The rationale for delivering focal radiation and not WBRT is to avoid the complications of WBRT while maintaining high local control without negatively impact on survival. Certainly, MRI at regular intervals of 2-3 months is mandatory to offer appropriate salvage therapy in the event of either local or distant brain progression. While both SRS and HSRT have been shown to improve local control in smaller surgical beds, achieving excellent local control rates still remains a challenge in larger ones. Accurate localization and delineation of the surgical cavity after resection of a brain metastasis is a crucial step in the treatment planning process for improving local control. A summary of recommended imaging modalities for target volume delineation and dose fractionation using either HSRT or SRS is reported in . Future research is needed to answer several questions regarding the optimal treatment timing, target delineation, dose/fractionation, and combination with systemic agents. |
The impact of doubling dairy or plant-based foods on consumption of nutrients of concern and proper bone health for adolescent females
Objective: To determine the effects of increasing plant-based foods v. dairy foods on energy and nutrients of concern in adolescent females via diet modelling exercises. Design: Data from the National Health and Nutrition Examination Survey (NHANES) were used to compare nutrient intakes from usual diet with those from three dietary scenarios that increased current intakes by 100 % of the following: (i) plant-based foods; (ii) protein-rich plant-based foods; and (iii) milk, cheese and yoghurt. The first two scenarios had commensurate reductions in animal products. Setting: What We Eat in America, NHANES 2007-2010. Subjects: Female adolescents (n 1594) aged 9-18 years. Results: When currently consumed plant-based foods were increased by 100 %, there were increases in dietary fibre, added sugar, vitamin E, Fe and folate intakes. These increases were accompanied by decreases in total fat, saturated fat, Zn, vitamin D, Ca and protein intakes. Protein-rich plant foods are consumed in very low quantities in this population such that doubling their intake resulted in no real nutritional impact. When dairy products were increased by 100 % there were increases in intakes of vitamin D, Mg, Zn, Ca, K, energy, saturated fat and protein.Conclusions: Non-specific recommendations to increase plant foods can lead to unintended nutritional consequences. For adolescent girls, meeting the dietary recommendation of three daily servings of dairy improved the intake of the identified nutrients of concern while simultaneously providing adequate nutrients essential for proper growth and bone health critical during the adolescent phase.
Adolescence is characterized by accelerated growth, second only to infancy. This growth spurt increases the need for several nutrients, including vitamin D, Fe, Zn, folate, energy and protein [bib_ref] Nutrition and pubertal development, Soliman [/bib_ref]. Additionally, many diet and lifestyle habits acquired during adolescence carry over into adulthood and can last a lifetime. Hence, this period is an important window of opportunity to significantly impact lifelong health.
While both adolescent boys and girls are nutritionally vulnerable, adolescent girls generally fare the worst in not consuming adequate nutrients. Lifestyle choices involving dieting and meal skipping due to body image issues and the desire to be thin are contributing factors. Adolescent girls are less likely to consume three meals per day and less likely to eat breakfast than their male counterparts . Thus, it is crucial that the foods this population consume be nutrient-rich, such as milk, cheese and yoghurt. Milk is a good or excellent source of protein, Ca, vitamin A, vitamin D, vitamin B 12 , riboflavin, niacin, P and pantothenic acid. In addition, some studies have shown that milk and milk product consumption during adolescence is associated with neutral or reduced risk of adiposity [bib_ref] Dairy consumption and pre-school, schoolage and adolescent obesity in developed countries: a..., Dror [/bib_ref] [bib_ref] The role of dairy products in healthy weight and body composition in..., Spence [/bib_ref] [bib_ref] Is dairy consumption associated with low cardiovascular disease risk in European adolescents?..., Bel-Serrat [/bib_ref] [bib_ref] Milk intake is inversely related to body mass index and body fat..., Abreu [/bib_ref] and reduced blood pressure with either higher intake [bib_ref] Dairy food consumption, blood pressure and retinal microcirculation in adolescents, Gopinath [/bib_ref] or as part of a DASH (Dietary Approaches to Stop Hypertension)-type dietary pattern [bib_ref] Dietary Approaches to Stop Hypertension (DASH) eating pattern and risk of elevated..., Moore [/bib_ref] , and some studies link consumption of milk and milk products to dental health [bib_ref] Association between dairy intake and caries among children and adolescents. Results from..., Lempert [/bib_ref]. However, current estimates of total dairy intakes by adolescent girls are well below the 2015 Dietary Guidelines for Americans' recommendation of 3 servings/d as well as the American Academy of Pediatrics' recommendation of 4 dairy servings/d [bib_ref] Optimizing bone health and calcium intakes of infants, children, and adolescents, Greer [/bib_ref].
The 2015 Dietary Guidelines for Americans report that vitamin D, Ca, P, fibre and Fe are nutrients of concern for adolescent females . Based on data from the third National Health and Nutrition Examination Survey (NHANES III), the prevalence of Fe deficiency in US girls aged 12-19 years was between 8 and 10 % of the population. During adolescence there is an increase in dietary Fe requirement due to the expansion of total blood volume associated with growth, and in adolescent girls this requirement is greater to account for the amount of Fe lost in menses [bib_ref] Symposium: improving adolescent iron status before childbearing iron requirements in adolescent females, Beard [/bib_ref]. The National Growth and Health Study evaluated the longitudinal food intake of 2379 girls during the periods of 9-13, 14-18 and 19-20 years of age and found at each age group that more than 90 % did not consume the recommended amounts of fruit, vegetables or dairy [bib_ref] Food group intake and micronutrient adequacy in adolescent girls, Moore [/bib_ref]. Inadequate intakes of Ca, Mg, K and vitamins D and E were found in the majority of girls in all age groups. Generally, as the girls aged micronutrient inadequacy worsened.
Recently, there has been increased discussion around the role of animal v. plant foods and their associated environmental impacts (e.g. impact on greenhouse gas emissions or land use) to ensure that current and future generations are able to meet their dietary needs for a healthy lifestyle. A plant-centred diet has been shown to lower the risk of diet-related chronic disease [bib_ref] Associations between food and beverage groups and major diet-related chronic diseases: an..., Fardet [/bib_ref] [bib_ref] Towards an even healthier Mediterranean diet, Estruch [/bib_ref] and has been suggested to be more environmentally sustainable [bib_ref] Sustainability of meatbased and plant-based diets and the environment, Pimentel [/bib_ref] [bib_ref] The importance of reduced meat and dairy consumption for meeting stringent climate..., Hedenus [/bib_ref] [bib_ref] How can the EU climate targets be met? A combined analysis of..., Bryngelsson [/bib_ref]. Furthermore, a key message from the Institute of Medicine workshop on healthy sustainable diets stated that plant-based protein sources are worth considering as an alternative to animal-based protein sources. However, it is unclear what the nutritional impact will be on adolescent females if they attempt to reduce the consumption of animal products and increase plant-based foods in their diet. Hence, the objective of the present study was to determine the effects of increasing plantbased foods v. increasing dairy foods on energy and nutrient adequacy in adolescent girls. Diet modelling exercises were conducted based on the following scenarios: (i) increasing the amount of plant-based foods by 100 % based on current consumption patterns; (ii) increasing the amount of protein-rich plant foods (i.e. legumes, nuts, seeds, soya) that are currently consumed by 100 %; and (iii) increasing the amount of dairy foods currently consumed by 100 %. Our hypothesis was that doubling foods that are already consumed, in specific food groups, would be easier to achieve than imposing a behaviour change that required increasing the intakes of foods not currently consumed by adolescents.
# Methods
Data from What We Eat in America, the dietary component of the NHANES 2007-2010, were used in all of our analyses. NHANES is a nationally representative cross-sectional study that utilizes a stratified, multistage probability sample of the non-institutionalized US population. Food and nutrient intakes were determined from two non-consecutive 24 h dietary recalls. The first was conducted by an in-person interview at the mobile examination centre and the second was obtained via a telephone interview a few days later. Only data coded as reliable by the US Department of Agriculture for adolescent females aged 9-18 years were used (n 1594). The National Cancer Institute method was used for estimating usual intakes and the covariates used were day sequence, Dietary Reference Intake age groups and recall weekday/ weekend indicators [bib_ref] A mixed-effects model approach for estimating the distribution of usual intake of..., Tooze [/bib_ref].
Three dietary scenarios were examined in the present study and compared with the usual diet. To model what could happen if adolescent females followed a blanket recommendation of increasing plant-based foods at the expense of animal protein, scenario 1 increased the currently consumed plant-based foods (fruits, vegetables, total grains, legumes, nuts, seeds and soya) by 100 % while animal protein intake (eggs, meat, poultry, fish and dairy) was proportionately decreased based on usual consumption on a gram-per-gram basis. This approach was undertaken because dietary habits are typically difficult to change and it may be easier for individuals to double the plant-based foods they currently consume instead of introducing new plant-based foods. Since it was anticipated that overall protein intake would decrease when currently consumed plant-based foods were increased and animal products were reduced (scenario 1), scenario 2 modelled the impact of increasing currently consumed protein-rich plant-based foods (i.e. beans, peas, legumes, nuts, seeds and soya products) by 100 % with commensurate decreases in animal products on a gram-pergram basis.
For scenario 3, currently consumed dairy foods (milk, cheese and yoghurt) were increased by 100 %. Dietary guidelines in the USA and elsewhere recommend adolescents consume 3-4 servings dairy products/d for optimal nutrition; therefore, scenario 3 modelled the impact of meeting these dietary recommendations by doubling the dairy foods currently consumed without any further adjustments to the diet. This scenario used the US Department of Agriculture's dairy composite nutrient profile for the increased dairy products.
Mixed dishes containing dairy products (e.g. macaroni and cheese), plant foods (e.g. bean soup) or eggs, meat, poultry and fish were not increased or decreased in any of the scenarios due to the difficulty associated with accurately disaggregating NHANES data to identify which nutrients are coming from which ingredients in these mixed dishes. Only those foods eaten as consumed were increased or decreased in the analysis presented herein, which allowed for a direct comparison between changes in food intake and nutrient adequacy.
The Food Pattern Equivalent Database for each NHANES release was used to define various types of foods in the analyses [bib_ref] USDA Food and Nutrient Database for Dietary Studies, 5.0. Beltsville, MD: US..., Ahuja [/bib_ref]. Plant-based foods were defined as food codes with zero servings of dairy, meat/poultry/fish and eggs with non-zero servings for the sum of servings of fruits, vegetables, legumes, grains, soya products and nuts/seeds. Animal-based foods were defined as food codes with zero servings for the sum of servings of fruits, vegetables, legumes, grains, soya products and nuts/seeds with non-zero servings for the sum of dairy, meat/poultry/ fish and eggs. Higher protein-based foods were defined using the US Department of Agriculture's food category 'Plant-based Protein Foods'. This category includes 'Beans, peas and legumes', 'Nuts and seeds' and 'Processed soy products'. Nutrient composition of dairy was based on the dairy food composite used by the US Department of Agriculture for dietary pattern analyses [bib_ref] Milk proteins, peptides, and oligosaccharides: effects against the 21st century disorders, Hsieh [/bib_ref]. Adequate protein intake was based on grams of protein per kilogram of body weight and grams of protein per kilogram of ideal body weight as recommend by the Dietary Reference Intakes.
The statistical software packages SAS 9.2 and SUDAAN 11 were used for all analyses. NHANES survey weights strata and primary sampling units were also used in all calculations. The National Cancer Institute method was used to estimate usual intakes of the food groups and selected nutrients to assess the percentage of the population meeting nutrient adequacy [bib_ref] A mixed-effects model approach for estimating the distribution of usual intake of..., Tooze [/bib_ref]. Given that we know there are differences in modelling scenarios by design, typical statistical testing was not appropriate. Thus, to assess meaningful differences in changes of the means for the modelling scenarios we examined means and their 95th percentile confidence limits. Non-overlapping confidence intervals we deemed meaningful. This approach has been utilized previously in dietary pattern studies [bib_ref] Consuming the daily recommended amounts of dairy products would reduce the prevalence..., Quann [/bib_ref]. [fig_ref] Table 1: Usual daily food group intakes prior to dietary modelling in adolescent females... [/fig_ref] lists the food groups usually consumed by adolescent females prior to dietary modelling. Grain-based foods comprised the largest food group, followed by the meat, poultry and fish group and the dairy group. The least consumed food groups were legumes, soya, and nuts and seeds. Due to the low intakes of foods in these categories, the corresponding decrease in animal-based foods was lower in scenario 2 (where currently consumed protein-rich plant-based foods, such as legumes, nuts and seeds, and soya products, were increased by 100 %) compared with scenario 1 (where all plant-based foods were increased by 100 %). Based on their usual intake, adolescent girls consumed 1·8 cup-equivalents dairy/d. In scenario [bib_ref] Nutrition and pubertal development, Soliman [/bib_ref] , currently consumed milk, cheese and yoghurt was increased by 100 %, increasing the total dairy to approximately 3 cup-equivalents/d, meeting the recommendation of the 2015-2020 Dietary Guidelines for Americans. Changes in individual food group intakes were based on each food as consumed and did not account for those foods in mixed dishes. As such, there was a range of increases and decreases, which reflects differences in how each food is consumed. [fig_ref] Table 2: Daily macronutrient intakes from the usual diet compared with a 100 %... [/fig_ref] show the usual consumption of macronutrients, the nutrients identified to be of concern for adolescent females and the shifts observed with each of the three dietary modelling exercises. A 100 % increase in all plant foods currently consumed (scenario 1) resulted in decreased intakes of protein, total fat and saturated fat and increased intakes of carbohydrate, dietary fibre and added sugars compared with the usual diet. Protein intake was lower by almost 10 g/d in scenario 1 when compared with the usual intake [fig_ref] Table 2: Daily macronutrient intakes from the usual diet compared with a 100 %... [/fig_ref] and the proportion of adolescent females not meeting the Estimated Average Requirement (EAR) for protein almost tripled in scenario 1. Total fat and saturated fat each decreased by about 4 g/d. Dietary fibre increased by about 4 g/d, increasing the proportion of adolescent females who had intakes above the Adequate Intake (AI) by 8 %. Added sugar intake increased by about 1 g/d. Increased consumption of plant foods also impacted vitamin D, vitamin E, folate, Zn and Ca intakes. There was an improvement, or a decrease, between the usual diet and scenario 1 for the percentage of adolescent females not meeting the EAR for vitamin E and folate. However, the percentage of adolescent females not meeting the EAR increased for Ca, Zn and vitamin D between the usual diet and scenario 1. Fe intake increased but the percentage not meeting the EAR did not significantly differ from usual intakes.
# Results
In scenario 2, which increased the consumption of protein-rich plant foods by 100 %, there were no meaningful changes in macronutrients or the specific micronutrients examined compared with usual intakes. One exception was vitamin E intake, which showed an improvement for the percentage of girls not meeting the EAR via a decrease of 3·7 %.
A 100 % increase in currently consumed dairy (scenario 3) led to an increase in energy, protein, total fat, saturated fat and added sugars, compared with usual intake levels. Protein consumption rose by 10·8 g/d and the percentage of girls not meeting the EAR fell by 48 %. Intakes of vitamin D, Mg, Zn, Ca and K were also increased in this scenario. Those not meeting the EAR decreased for vitamin D, Zn and Mg in this model. The largest improvement in the percentage not meeting the EAR was observed for Ca in the dairy model. K intake was marginally improved, while Na intake increased but did not meaningfully impact the proportion of the population that had intakes above the AI v. baseline.
# Discussion
Our modelling exercises compared the effect of doubling plant-based foods v. the effect of doubling dairy foods on nutritional adequacy in adolescent females. It was hypothesized that doubling foods normally consumed in these food categories would be easier to achieve than imposing a behaviour change by doubling foods that are not frequently consumed. Additionally, if nutrient adequacy improved using this approach, it would provide a relatively simple message to advocate for improved health in this vulnerable population. Scenario 1 doubled the intake of plant-based foods (fruits, vegetables, total grains, legumes, nuts, seeds and soya), while eggs, meat, poultry, fish and dairy intakes were decreased. Grain-based foods comprised the largest
[formula] 2 ·8 0 ·4 5 1 ·7 0 ·3 Protein (g) 63·7 0·9 5 4 ·0 † 1·0 6 4 ·2 1 ·0 7 4 ·5 † 1·1 % Energy from protein 14·1 0·2 1 1 ·3 † 0·1 1 4 ·1 0 ·2 1 4 ·7 † 0·2 Total fat (g) 68·0 1·0 6 3 ·5 † 1·4 6 9 ·6 1 ·1 7 7 ·8 † 1·1 % Energy from fat 32·6 0·3 2 8 ·7 † 0·4 3 2 ·8 0 ·3 3 3 ·3 0 ·3 SFA (g) 23·4 0·4 1 9 ·0 † 0·5 2 3 ·4 0 ·4 2 9 ·3 † 0·4 % Energy from SFA 11·2 0·1 8 ·7 † 0·1 1 1 ·1 0 ·1 1 2 ·4 † 0·1 Dietary fibre (g) 13·0 0·3 1 6 ·9 † 0·4 1 3 ·5 0 ·3 1 3 ·5 0 ·3 Added sugars (teaspoon-equivalents) 19·2 0·4 2 0 ·2 † 0·4 1 9 ·2 0 ·4 2 1 ·2 † 0·4 [/formula]
*Each of the modelling exercises was based on plant-based foods, protein-rich plant-based foods or dairy foods as consumed. The models increased/ decreased foods as consumed and did not change the individual foods found in mixed dishes. †Meaningfully different from baseline due to non-overlapping 95th percentile confidence interval. UI, usual intake; BW, body weight. *Each of the modelling exercises was based on a 100 % increase in plant-based foods, protein-rich plant-based foods or dairy foods as consumed. The models increased/decreased foods as consumed and did not change the individual foods found in mixed dishes. †Meaningfully different from baseline due to non-overlapping 95th percentile confidence intervals.
food group consumed within the plant-based food category. Total fat and saturated fat intake levels dropped as expected, but there was a small increase in added sugars. Dietary fibre consumption increased by 30 %, but over 90 % of adolescent girls still had intakes below the AI. There was a significant improvement in folate intake as the percentage not meeting the EAR dropped by almost 60 %. Vitamin E intakes were also improved. However, in this scenario, the proportion of girls not meeting the EAR for protein, Ca, Zn and vitamin D increased, two of which have been identified as nutrients of public health concern. Our estimates of usual intakes indicated that 76 % of the adolescent girls have Ca intakes below the EAR, but in this scenario Ca intakes worsened, resulting in 92 % of adolescent girls not meeting the EAR. Consuming adequate amounts of dietary Ca during childhood and adolescence is critical for reaching peak bone mass, which may be important in reducing the risk of fractures and osteoporosis later in life [bib_ref] Optimizing bone health and calcium intakes of infants, children, and adolescents, Greer [/bib_ref]. Another striking observation of the plant-based food model (scenario 1) was that protein intakes were inadequate for 14·5 % of the population. This could be a cause for concern as protein intake is required to support numerous physiological processes and growth [bib_ref] USDA Food and Nutrient Database for Dietary Studies, 5.0. Beltsville, MD: US..., Ahuja [/bib_ref].
Since we anticipated that protein intake would decrease in scenario 1, the second model increased the intake of protein-rich foods. Surprisingly, only vitamin E levels were different from baseline in this model. This finding is likely due to the fact that adolescent girls currently have very low intakes of legumes, nuts, seeds and soya products such that even doubling their intake had no real effect on nutrient intakes. The present study shows that adolescent girls consume little protein-rich plant foods; therefore there may be difficulty in getting them to increase their consumption to an amount that will lead to meaningful changes in nutrient intakes. This highlights the importance of ensuring that dietary recommendations are realistic and consider the behaviour changes required to meet recommendations.
Doubling the currently consumed dairy intake (scenario 3) resulted in a meaningful increase in total protein intake, reducing the proportion of girls not meeting the EAR by 48 %. Dairy is a source of high-quality protein, containing many peptides and bioactive factors that may have specific effects on growth [bib_ref] Dairy product intake in children and adolescents in developed countries: trends, nutritional..., Dror [/bib_ref] , as well as antihypertensive and antiinflammatory activity [bib_ref] Milk proteins, peptides, and oligosaccharides: effects against the 21st century disorders, Hsieh [/bib_ref]. There has been some concern that increased protein intake increases urinary Ca excretion by increasing the total acid load from protein metabolism; however, more recent research indicates this perception is not accurate in healthy individuals [bib_ref] Dairy in adulthood: from foods to nutrient interactions on bone and skeletal..., Bonjour [/bib_ref] [bib_ref] Dietary protein and skeletal health: a review of recent human research, Kerstetter [/bib_ref]. Furthermore, recommended intakes of dietary Ca take into account the typical protein intake of the population and adjusting Ca intake on the basis of protein intake generally is not recommended for children and adolescents [bib_ref] Optimizing bone health and calcium intakes of infants, children, and adolescents, Greer [/bib_ref].
Increasing dairy intake reduced the proportion of adolescent girls not meeting the EAR for Ca from 76 to 41 %.
Studies in children and adolescents have demonstrated that supplemental dietary Ca intake from dairy foods contributes to increases in bone mineral density, particularly in children with low baseline Ca intakes [bib_ref] Effects of diary food supplements on bone mineral density in teenage girls, Merrilees [/bib_ref] [bib_ref] Effects of dairy products on bone and body composition in pubertal girls, Chan [/bib_ref] [bib_ref] Impact of dairy products and dietary calcium on bone-mineral content in children:..., Huncharek [/bib_ref]. Furthermore, it has been observed in a nationally representative sample of women that low milk intake during childhood and adolescence was associated with low bone mineral content or bone mineral density of the hip in adulthood [bib_ref] Milk intake during childhood and adolescence, adult bone density, and osteoporotic fractures..., Kalkwarf [/bib_ref]. While adequate Ca intake during adolescence is essential, and dairy products provide the greatest amount of Ca in the diet [bib_ref] Dietary sources of nutrients among US children, Subar [/bib_ref] , there has been concern about dietary Ca negatively impacting the absorption of Fe [bib_ref] Calcium: effect of different amounts on nonheme-and heme-iron absorption in humans, Hallberg [/bib_ref]. A multi-country cross-sectional study showed a significant inverse correlation for dietary Ca intake (predominantly from dairy foods) and Fe stores; however, the quantitative effect was relatively small [bib_ref] Calcium intake is weakly but consistently negatively associated with iron status in..., Van De Vijver [/bib_ref]. Additionally, results from a young adult female cohort who consumed a glass of milk with breakfast, lunch and dinner for four consecutive days showed no difference in the amount of non-haem Fe absorbed [bib_ref] Calcium from milk or calcium-fortified foods does not inhibit nonheme-iron absorption from..., Grinder-Pedersen [/bib_ref].
Nutrients other than Ca are also involved in skeletal growth, including Mg, Zn and vitamin D [bib_ref] Meeting and exceeding dairy recommendations: effects of dairy consumption on nutrient intakes..., Rice [/bib_ref]. Results from our dietary modelling exercises indicate that increasing dairy intake increased the intake of these nutrients, resulting in fewer adolescent girls falling below the EAR. Very few foods have naturally occurring vitamin D. Fortified foods are the primary source of vitamin D in the American diet, with 60 % of vitamin D provided by milk for children aged 2-18 years [bib_ref] Food sources of energy and nutrients among children in the United States:..., Keast [/bib_ref]. However, results from our dairy model indicate that 80 % of adolescent girls were still not meeting their EAR for vitamin D. Similarly, while K intakes increased with increased dairy consumption, 99 % of the adolescents were still consuming amounts below the AI. It has been noted that less than 2 % of Americans have K intakes at or above the AI for K [bib_ref] Consuming the daily recommended amounts of dairy products would reduce the prevalence..., Quann [/bib_ref]. Yet the importance of dietary K in the adolescent female population was recently highlighted in a study by Moore et al., who reported that higher K intakes were inversely associated with blood pressure change [bib_ref] Dietary Approaches to Stop Hypertension (DASH) eating pattern and risk of elevated..., Moore [/bib_ref]. The percentage of girls falling below the EAR for Fe, a nutrient singled out as being deficient in the diets of teenage girls, was not meaningfully improved by any of the modelling scenarios with the exception of a small decrease in scenario 1.
Although doubling dairy foods increased the intake of many key nutrients, it also increased the intakes of energy and saturated fat. Increased energy intake is a concern due to the rise in childhood and adolescent obesity. However, systematic reviews of the available data from observational and randomized controlled studies on children and adolescents indicate either an inverse or neutral relationship between dairy product and/or Ca intake and body weight and/or body fat [bib_ref] The role of dairy products in healthy weight and body composition in..., Spence [/bib_ref] [bib_ref] The role of dairy foods and dietary calcium in weight management, Van Loan [/bib_ref]. The increased saturated fat intake associated with increased dairy intake may be less of a concern than previously believed. Several recent systematic reviews [bib_ref] Meeting and exceeding dairy recommendations: effects of dairy consumption on nutrient intakes..., Rice [/bib_ref] [bib_ref] Yogurt and dairy product consumption to prevent cardiometabolic diseases: epidemiologic and experimental..., Astrup [/bib_ref] [bib_ref] Influence of dairy product and milk fat consumption on cardiovascular disease risk:..., Huth [/bib_ref] [bib_ref] A reappraisal of the impact of dairy foods and milk fat on..., German [/bib_ref] and meta-analyses [bib_ref] Dairy consumption and incidence of hypertension: a doseresponse meta-analysis of prospective cohort..., Soedamah-Muthu [/bib_ref] [bib_ref] The consumption of milk and dairy foods and the incidence of vascular..., Elwood [/bib_ref] have found either no association or an inverse association between the intake of dairy foods and cardiovascular risk, regardless of milk fat levels. This is consistent with a recent review that found no clear association between higher intake of saturated fat and all-cause mortality, CHD, CHD mortality, ischaemic stroke or type 2 diabetes among apparently healthy adults [bib_ref] Intake of saturated and trans unsaturated fatty acids and risk of all..., Souza [/bib_ref].
Strengths of the present study include the use of a large, nationally representative database for determining the nutritional impact of changing the consumption patterns of plant and dairy foods; the use of the Food Pattern Equivalent Database to develop the diet modelling scenarios; use of usual intake methodology to assess nutrient adequacy; and modelling dietary changes that would be relatively simple to advocate. Limitations of the study include the use of self-reported intake data, which in the case of adolescent girls has been indicated to be underreported [bib_ref] Self-reported dietary energy intake of normal weight, overweight and obese adolescents, Vance [/bib_ref]. None of our dietary models looked at the effect of doubling the intake of only fruits and vegetables v. all plant-based foods. Additionally, because it is difficult to disaggregate the individual foods from mixed dishes in NHANES, our models only reflect changes made to individual foods as consumed; foods in mixed dishes were not doubled. Accordingly, some food groups were not exactly doubled because a portion of their consumption occurs in mixed dishes. Data from NHANES 2009-2012 show that the total energy contribution of mixed dishes for adolescent females aged 9-18 years (n 1163) is approximately 1770 (SE 75) kJ/d (423 (SE 18) kcal/d), which translates to 23 (SE 0·87) % of all energy in their diet (data not shown). Accordingly, future research should assess the impact of mixed dishes on diet quality as well as how changes in the foods associated with each mixed dish differ after the scenarios presented herein.
# Conclusions
Overall, our diet modelling exercises in adolescent girls demonstrate that general non-specific recommendations to increase plant-based foods at the expense of all animal foods lead to some nutritional benefits but can also result in unintended consequences. Doubling plant-based foods lowered intakes of total fat and saturated fat while increasing the intakes of dietary fibre, folate and vitamin E, but also resulted in lowered intakes for protein, Ca and vitamin D. If adolescent females were to follow this dietary pattern, only two of the five nutrients of public health concern identified by the 2015 Dietary Guidelines for Americans for adolescent females improved (fibre and Fe), while the intakes of the other three (Ca, vitamin D and K) worsened. Increasing protein-rich plant foods had no real impact because legumes, nuts, seeds and soya products are consumed at very low levels in this population. Doubling dairy foods led to dairy intakes that met recommended levels and helped to increase the intakes of protein, Ca, vitamin D, Mg and Zn; nutrients that have been singled out as important for adequate growth and bone health during the adolescent years. Additionally, the dairy model also meaningfully improved intakes for three of the nutrients of public health concern (Ca, vitamin D and K). Our data reinforce that dietary intake needs to be assessed as a whole and that dietary recommendations should consider the feasibility of realistically implementing the suggested behaviour changes required to meet nutrient needs.
[table] Table 1: Usual daily food group intakes prior to dietary modelling in adolescent females aged 9-18 years, National Health and Nutrition [/table]
[table] Table 2: Daily macronutrient intakes from the usual diet compared with a 100 % increase in plant foods, a 100 % increase in protein-rich plant foods and a 100 % increase in dairy foods in adolescent females aged 9-18 years, National Health and Nutrition Examination Survey (NHANES) 2007-2010* [/table]
|
Drop the Needle; A Temperature Stable Oral Tablet Vaccine Is Protective against Respiratory Viral Pathogens
# Introduction
The recent pandemic caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) demonstrated that public health systems throughout the world were inadequately prepared to respond to an outbreak of a novel pathogen. Although innovative vaccines were quickly developed and manufactured in record time, the storage requirements and slow distribution continue to be a significant obstacle to achieving rapid vaccine rollout. Currently, SARS-CoV-2 vaccines in use around the world require cold chain storage and robust infrastructure that mobilizes healthcare providers to administer vaccines. Many developing countries lack the public health resources and infrastructure to quickly implement new strategies in response to emergent infections, posing challenges to equitable healthcare delivery [bib_ref] Building the evidence base for global health policy: The need to strengthen..., Jakab [/bib_ref] [bib_ref] Ensuring health equity during the COVID-19 pandemic: The role of public health..., Benjamin [/bib_ref]. To rapidly respond to future pandemics in a manner that ensures global uptake and equitable health outcomes, new vaccine technologies must be simple to administer and ideally maintain potency for long periods of time at room temperature [bib_ref] Grand Challenges in Pharmaceutical Research Series: Ridding the Cold Chain for Biologics, Yu [/bib_ref].
In addition to the challenges posed by vaccine storage and distribution, vaccine hesitancy within industrialized nations, especially regarding injections, impedes adequate immunization rates. A recent poll conducted by Quadrant Strategies reveals that a phobia of injections is a deterrent to those deciding whether to become immunized. The poll showed that 32% of individuals who didn't plan on obtaining a SARS-CoV-2 vaccine indicated they would be more likely to opt for pill option rather than a parenteral immunization. Furthermore, 70% of those polled would prefer taking a pill on their own schedule rather than making an appointment with a provider, which can be time consuming and inconvenient and may require time off work. This highlights the fact that a self-administered tablet vaccine could lead to better vaccination coverage. necessity of administration by healthcare professionals. rAd-based vaccine approaches are also modular and can be easily modified to carry different vaccine antigens. The flexibility of this technology allows for the rapid development of new vaccines following the emergence of unique influenza virus strains or novel pathogens. Another benefit of rAd vector vaccines is their quick production timeline by cell culture manufacturing techniques, which allows for the generation of clinical material in less than 9 weeks [fig_ref] Figure 1: Next generation vaccine production, manufacturing, and distribution timelines during a pandemic [/fig_ref]. While mRNA formulations can be manufactured rapidly and currently have the shortest time to market, their inherent instability at temperatures above −70 - C results in numerous discarded doses [bib_ref] Addressing the Cold Reality of mRNA Vaccine Stability, Crommelin [/bib_ref]. It is estimated that about a million SARS-CoV-2 vaccine doses were wasted in the United States between December 2020 and July 2021 due to temperature deviations. This limitation does not apply to an oral vaccine tablet, which is designed to be room temperature-stable for multiple years. Furthermore, injected vaccines generate excessive biohazardous waste, such as syringes and sanitation materials during vaccine campaigns [bib_ref] COVID-19 pandemics Stage II-Energy and environmental impacts of vaccination, Klemes [/bib_ref] , which can be greatly decreased by employing a tablet-based system. Most importantly, an oral tablet approach reduces the need for skilled practitioners to administer the vaccine, as it can be easily distributed and self-administered [fig_ref] Figure 1: Next generation vaccine production, manufacturing, and distribution timelines during a pandemic [/fig_ref]. This fundamental change in delivery can accelerate vaccine rollout during pandemics by eliminating cold storage and infrastructure requirements while also reducing biowaste and unused vaccine doses. Lastly, another advantage of rAd oral tablet administration is the adverse events experienced after vaccination appear to be quite mild, without any signs of injection site pain or fever when compared to a placebo group [bib_ref] High titre neutralising antibodies to influenza after oral tablet immunisation: A phase..., Liebowitz [/bib_ref]. That is not the case with the current mRNA vaccines or injected adenovirus vaccines, where fever, headache and chills are quite common outcomes post immunization [bib_ref] Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine, Baden [/bib_ref] [bib_ref] Phase 3 Safety and Efficacy of AZD1222 (ChAdOx1 nCoV-19) COVID-19 Vaccine, Falsey [/bib_ref] [bib_ref] Safety and Efficacy of the BNT162b2 mRNA COVID-19 Vaccine, Polack [/bib_ref] [bib_ref] Safety and Efficacy of Single-Dose Ad26.COV2.S Vaccine against COVID-19, Sadoff [/bib_ref].
Vaccines 2022, x FOR PEER REVIEW 3 of 13 plemented in developed nations, and low-income countries have been left behind, amplifying healthcare disparities and emphasizing the urgent need for improvements in vaccine storage and global distribution [bib_ref] COVID-19 Healthcare Inequity: Lessons Learned from Annual Influenza Vaccination Rates to Mitigate..., Bazan [/bib_ref] [bib_ref] Disparities in global COVID-19 vaccination rates & allocation of resources to countries..., Sen-Crowe [/bib_ref]. An oral rAd tablet vaccine that can be rapidly produced and deployed, is suited to transforming global vaccination strategies by removing cold-chain requirements and the necessity of administration by healthcare professionals. rAd-based vaccine approaches are also modular and can be easily modified to carry different vaccine antigens. The flexibility of this technology allows for the rapid development of new vaccines following the emergence of unique influenza virus strains or novel pathogens. Another benefit of rAd vector vaccines is their quick production timeline by cell culture manufacturing techniques, which allows for the generation of clinical material in less than 9 weeks [fig_ref] Figure 1: Next generation vaccine production, manufacturing, and distribution timelines during a pandemic [/fig_ref]. While mRNA formulations can be manufactured rapidly and currently have the shortest time to market, their inherent instability at temperatures above −70 °C results in numerous discarded doses [bib_ref] Addressing the Cold Reality of mRNA Vaccine Stability, Crommelin [/bib_ref]. It is estimated that about a million SARS-CoV-2 vaccine doses were wasted in the United States between December 2020 and July 2021 due to temperature deviations. This limitation does not apply to an oral vaccine tablet, which is designed to be room temperature-stable for multiple years. Furthermore, injected vaccines generate excessive biohazardous waste, such as syringes and sanitation materials during vaccine campaigns [bib_ref] COVID-19 pandemics Stage II-Energy and environmental impacts of vaccination, Klemes [/bib_ref] , which can be greatly decreased by employing a tablet-based system. Most importantly, an oral tablet approach reduces the need for skilled practitioners to administer the vaccine, as it can be easily distributed and self-administered [fig_ref] Figure 1: Next generation vaccine production, manufacturing, and distribution timelines during a pandemic [/fig_ref]. This fundamental change in delivery can accelerate vaccine rollout during pandemics by eliminating cold storage and infrastructure requirements while also reducing biowaste and unused vaccine doses. Lastly, another advantage of rAd oral tablet administration is the adverse events experienced after vaccination appear to be quite mild, without any signs of injection site pain or fever when compared to a placebo group [bib_ref] High titre neutralising antibodies to influenza after oral tablet immunisation: A phase..., Liebowitz [/bib_ref]. That is not the case with the current mRNA vaccines or injected adenovirus vaccines, where fever, headache and chills are quite common outcomes post immunization [bib_ref] Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine, Baden [/bib_ref] [bib_ref] Phase 3 Safety and Efficacy of AZD1222 (ChAdOx1 nCoV-19) COVID-19 Vaccine, Falsey [/bib_ref] [bib_ref] Safety and Efficacy of the BNT162b2 mRNA COVID-19 Vaccine, Polack [/bib_ref] [bib_ref] Safety and Efficacy of Single-Dose Ad26.COV2.S Vaccine against COVID-19, Sadoff [/bib_ref]. mRNA-based vaccines have the shortest development and manufacturing timelines; however, these formulations are the most difficult to transport and store [bib_ref] mRNA vaccines-A new era in vaccinology, Pardi [/bib_ref] [bib_ref] Addressing the Cold Reality of mRNA Vaccine Stability, Crommelin [/bib_ref]. (B) An oral tablet rAd5 vaccine that is room temperature-stable can be shipped without cold chain and does not need to be administered by trained health care professionals. Distribution of an oral vaccine can exponentially speed up vaccine rollout compared to injected vaccines, especially in areas with considerably fewer resources, as tablets can be shipped directly to individuals and self-administered. Traditional needle-based immunization approaches have an inherent distribution bottleneck due to temperature constraints for shipping and storage and the need for health care professionals to administer the vaccines. Administering needle-based vaccines seems to be the greatest bottleneck and can take more than 6 months, even with substantial investment in infrastructure.
Room temperature-stable vaccines could also be beneficial for stockpiling vaccine supply, allowing for rapid distribution during a pandemic. The potency of the oral rAd5 tableted vaccine drug product has been evaluated at multiple temperatures over extended storage times. Vaccine drug products for human use must be manufactured under regulations known as good manufacturing practices (GMP). In preparation for a phase II clinical influenza challenge study, a GMP-manufactured monovalent hemagglutinin vaccine drug product (VXA-A1.1) was placed at +25 - C at 60% relative humidity for a period exceeding two years. Tablets used in the influenza challenge study retained potency, as measured by viral infectivity assay, which showed that the drug product remained within all tested specifications for up to 426 days [fig_ref] Table 1: Oral rAd5 vaccine tablet stability [/fig_ref]. GMP tablets used in norovirus clinical trials, have also been extensively tested and remained within specification for 246 days at 30 - C and can tolerate higher temperatures [fig_ref] Table 1: Oral rAd5 vaccine tablet stability [/fig_ref]. These results demonstrate that an oral rAd5 vaccine tablet is notably stable at room temperature and could simplify vaccine distribution to lower-resource areas where cold chain transport is unavailable.
## Advantages of orally delivered rad vaccines
Non-replicating rAd vectors are attractive vaccine delivery platforms due to their effective antigen expression and presentation, which elicit innate and adaptive cellular and humoral immune responses [bib_ref] Replication-incompetent adenoviral vaccine vector elicits effective anti-immunodeficiency-virus immunity, Shiver [/bib_ref]. When rAd vaccines are delivered to the host, target antigens are expressed intracellularly and can be displayed on the cell surface or secreted, depending on the targeting sequence. Intracellular vaccine antigen expression also leads to constitutive processing and the presentation of peptides through MHC-class I and II, like natural infection, provoking robust T cell responses [bib_ref] Cross-Presentation of Cell-Associated Antigens by MHC Class I in Dendritic Cell Subsets, Gutierrez-Martinez [/bib_ref] [bib_ref] Epithelial MHC Class II Expression and Its Role in Antigen Presentation in..., Wosen [/bib_ref]. Mimicking the natural protein conformation of vaccine antigens, rAd vectors elicit broadly protective antibodies that are not easily produced from injected protein or subunit vaccines, where the protein can refold during manufacturing to exist in the lowest energy state [bib_ref] Influence of protein fold stability on immunogenicity and its implications for vaccine..., Scheiblhofer [/bib_ref]. Several influenza-based studies demonstrate that native stable HA homotrimer conformation enhances immunogenicity and elicits broadly protective antibody responses [bib_ref] A stable trimeric influenza hemagglutinin stem as a broadly protective immunogen, Impagliazzo [/bib_ref] [bib_ref] Enhanced immunogenicity of stabilized trimeric soluble influenza hemagglutinin, Weldon [/bib_ref]. However, intramuscular parenteral delivery of Ad vectors does not imitate the natural infection of respiratory pathogens and therefore does not elicit mucosal responses. The high number of breakthrough cases of SARS-CoV-2 following parenteral vaccine administration demonstrates the benefits of enhancing mucosal immune responses to reduce viral transmission [bib_ref] Mucosal Immunity in COVID-19: A Neglected but Critical Aspect of SARS-CoV-2 Infection, Russell [/bib_ref].
The rAd vaccines that can be delivered to the mucosa are ideal for displaying vaccine antigens in a manner similar to what the immune system encounters during infection. However, oral delivery of Ad vectors in humans has been challenging to develop due to oral tolerance mechanisms that hinder immunogenicity [bib_ref] First-in-human randomized controlled trial of an oral, replicating adenovirus 26 vector vaccine..., Stephenson [/bib_ref] , a major obstacle in the development of orally based vaccines [bib_ref] Mucosal vaccines-Fortifying the frontiers, Lavelle [/bib_ref] [bib_ref] The biology of oral tolerance and issues related to oral vaccine design, Poonam [/bib_ref]. The oral rAd5 vaccine platform uses a molecular TLR3 adjuvant to overcome oral tolerance in the small intestine (reviewed by [bib_ref] The biology of oral tolerance and issues related to oral vaccine design, Poonam [/bib_ref]. TLR3 is a pattern recognition receptor that senses dsRNA, which can be made by many pathogenic viruses during intracellular replication [bib_ref] Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3, Alexopoulou [/bib_ref]. In the human small intestine, TLR3 protein is highly expressed by enterocytes, primarily in the villi [bib_ref] The Expression of Toll-like Receptors in Normal Human and Murine Gastrointestinal Organs..., Huhta [/bib_ref] [bib_ref] Differential alteration in intestinal epithelial cell expression of toll-like receptor 3 (TLR3)..., Cario [/bib_ref]. When TLR3 is activated, inflammatory cascades and anti-viral responses are elicited, producing the release of both IFN types I and III. These potent, soluble, anti-viral mediators have been shown to enhance plasmablast differentiation and T cell activation [bib_ref] Interferon-lambda Enhances the Differentiation of Naive B Cells into Plasmablasts via the..., Syedbasha [/bib_ref] [bib_ref] Interferon-lambda modulates dendritic cells to facilitate T cell immunity during infection with..., Hemann [/bib_ref]. Preclinical murine models have demonstrated synthetic double-stranded RNA as an effective adjuvant when administered intranasally in conjunction with an inactivated split HA influenza vaccine [bib_ref] Synthetic double-stranded RNA poly(I:C) combined with mucosal vaccine protects against influenza virus..., Ichinohe [/bib_ref].
This study also demonstrated that IgA secretion in nasal wash and systemic IgG responses were both enhanced by mucosal delivery of TLR3 and vaccine antigens. Tissue-resident dendritic cells also respond to TLR3 stimulation and have been shown to be key in inducing influenza mucosal vaccine-specific responses including CD8 T cells [bib_ref] Toll-like receptor 3 in nasal CD103( + ) dendritic cells is involved..., Takaki [/bib_ref]. This approach, tested in a preclinical model, demonstrated the administration of intranasal polyinosine-polycytidylic (polyI:C) in conjunction with an influenza subunit vaccine, enhanced antigen-specific IgA responses [bib_ref] Toll-like receptor 3 in nasal CD103( + ) dendritic cells is involved..., Takaki [/bib_ref]. By including a molecular TLR3 adjuvant that is expressed by the viral vector, this vaccine technology has overcome a major barrier for effective oral rAd delivery.
## Mucosal vaccination elicits cellular and humoral responses
The generation of pathogen-specific antibodies in the serum that are capable of neutralizing pathogens is one of the main hallmarks of measuring vaccine-induced immunogenicity. Circulating B cells that produce antibodies are known as antibody-secreting cells (ASCs) and can differentiate into plasmablasts that can traffic to immune sites, such as bone marrow, lymph nodes, and the spleen. In mucosal tissues, plasmablasts predominantly secrete dimeric IgA into the lumen, and due to increased valency, this immunoglobulin has been shown to have higher cross-reactive neutralizing activity against invading pathogens and viral variants than IgG [bib_ref] Comparison of antiviral activity between IgA and IgG specific to influenza virus..., Muramatsu [/bib_ref] [bib_ref] IgA tetramerization improves target breadth but not peak potency of functionality of..., Saito [/bib_ref]. Lymphocytes express various integrins that assist circulating immune cells in homing to various tissues in the body. The differential upregulation of surface integrins on circulating ASCs has been demonstrated to be important for lymphocyte homing [bib_ref] The immune geography of IgA induction and function, Macpherson [/bib_ref]. Mucosal homing B and T cells express the integrin α4β7+, which recognizes the mucosal addressin cellular adhesion molecule-1 (MAdCAM-1), highly expressed on endothelial cells lining the mucosa [bib_ref] Distribution of alpha 4 beta 7 and alpha E beta 7 integrins..., Andrew [/bib_ref] [bib_ref] Differential expression of tissue-specific adhesion molecules on human circulating antibody-forming cells after..., Quiding-Jarbrink [/bib_ref]. A clinical study comparing vaccination administration routes found that mucosal delivery elicited α4β7+ ASCs, whereas injected preparations upregulated the lymph node homing marker L-selectin (CD62L+) [bib_ref] Differential expression of tissue-specific adhesion molecules on human circulating antibody-forming cells after..., Quiding-Jarbrink [/bib_ref]. Following exposure to vaccine antigens in the small intestine, lamina propria resident dendritic cells and epithelial cells can cross present antigens to naïve T cells through MHC I and II [bib_ref] Cross-Presentation of Cell-Associated Antigens by MHC Class I in Dendritic Cell Subsets, Gutierrez-Martinez [/bib_ref] [bib_ref] Epithelial MHC Class II Expression and Its Role in Antigen Presentation in..., Wosen [/bib_ref]. Activated T cells migrate to intestinal lymphoid tissue where T follicular helper cells (Tfh) support the B cell maturation and ASC development. Lastly, studies have shown that immune signals from the gut mucosa can support immune responses in the respiratory tract, providing evidence that there is cross talk between distal mucosal sites [bib_ref] Mucosal vaccines-Fortifying the frontiers, Lavelle [/bib_ref] [bib_ref] Microbiota regulates immune defense against respiratory tract influenza A virus infection, Ichinohe [/bib_ref]. Therefore, mucosal immunization can offer a significant protective benefit over injected vaccines by enhancing IgA production and cellular immune responses at the principal sites of infection.
Oral administration of influenza antigens using the rAd5 tablet vaccine elicits both humoral and cellular mucosal pathogen-specific responses, leading to a working model base in experimental and clinical studies [fig_ref] Figure 2: Working model of oral rAd5 vaccine tablet, illustrating how antigen delivery to... [/fig_ref]. A single dose of oral rAd5 vaccine tablet containing the influenza HA antigen, when delivered specifically to the ileum by radiocontrolled capsule, elicited strong antigen-specific systemic and mucosal responses [bib_ref] Systemic and mucosal immune responses following oral adenoviral delivery of influenza vaccine..., Kim [/bib_ref]. Immunized subjects showed an increase in α4β7+, IgA+ B cells and increased HA-specific IgGand IgA-specific ASC's [bib_ref] Systemic and mucosal immune responses following oral adenoviral delivery of influenza vaccine..., Kim [/bib_ref]. This demonstrated that oral vaccination by a non-replicating vector can shift ASC immunophenotyping profiles and enhance plasmablast homing to mucosal sites. Furthermore, HA-specific IgA in fecal and nasal samples was elevated following vaccination, an observation that was mirrored by preclinical studies in ferret models of enteric vaccine delivery [bib_ref] An adenovirus-based vaccine with a double-stranded RNA adjuvant protects mice and ferrets..., Scallan [/bib_ref]. In a phase I study, subjects immunized with an oral rAd5 HA-expressing vaccine tablet elicited IgG and IgA ASC's, which correlated with increases in hemagglutination inhibition and microneutralization assays [bib_ref] High titre neutralising antibodies to influenza after oral tablet immunisation: A phase..., Liebowitz [/bib_ref]. Together, these data established that the rAd5-based oral vaccine platform elicits immunogenic mucosal humoral responses.
In addition to antibody-mediated immunity, antiviral responses by T cells have been shown to be important for protection against influenza [bib_ref] Preexisting influenza-specific CD4 + T cells correlate with disease protection against influenza..., Wilkinson [/bib_ref] , particularly in the elderly where immunosenescence may impact immune responses [bib_ref] What does it mean to health outcomes in older adults?, Mcelhaney [/bib_ref]. Oral vaccination using rAd5 increased the CD8+ and CD4+ T cell responses (memory and effector populations) post vaccination, as measured using a 40-antibody panel and mass-cytometry [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. Multiple T cell subsets meaningfully contributed to protection by multivariate analysis [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. Additionally, in a recent phase I clinical safety trial, individuals who received first generation oral SARS-CoV-2 rAd5 vaccine had elevated anti-viral circulating CD8+ T cellsand manuscript in preparation). S-specific increases in IFNγ, CD107α, and TNFα were detected in CD8+ T cells by intracellular flow cytometry analysis following prime and boost immunizations 30 days apart. Furthermore, CD8+ T cell responses in individuals that received the oral rAd5 vaccine series were of higher magnitude than currently licensed vaccines when run in a comparator study. Additional studies are currently underway to further understand T cell responses elicited by oral rAd5 vaccine tablets and how these potent cells contribute to vaccine-mediated protection.
Vaccines 2022, x FOR PEER REVIEW 6 of 13
Together, these data established that the rAd5-based oral vaccine platform elicits immunogenic mucosal humoral responses. In addition to antibody-mediated immunity, antiviral responses by T cells have been shown to be important for protection against influenza [bib_ref] Preexisting influenza-specific CD4 + T cells correlate with disease protection against influenza..., Wilkinson [/bib_ref] , particularly in the elderly where immunosenescence may impact immune responses [bib_ref] What does it mean to health outcomes in older adults?, Mcelhaney [/bib_ref]. Oral vaccination using rAd5 increased the CD8+ and CD4+ T cell responses (memory and effector populations) post vaccination, as measured using a 40-antibody panel and mass-cytometry [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. Multiple T cell subsets meaningfully contributed to protection by multivariate analysis [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. Additionally, in a recent phase I clinical safety trial, individuals who received first generation oral SARS-CoV-2 rAd5 vaccine had elevated anti-viral circulating CD8+ T cellsand manuscript in preparation). S-specific increases in IFNγ, CD107α, and TNFα were detected in CD8+ T cells by intracellular flow cytometry analysis following prime and boost immunizations 30 days apart. Furthermore, CD8+ T cell responses in [fig_ref] Figure 2: Working model of oral rAd5 vaccine tablet, illustrating how antigen delivery to... [/fig_ref]. Working model of oral rAd5 vaccine tablet, illustrating how antigen delivery to the small intestine generates protective mucosal immune responses. The non-replicating rAd5 vector contains both the vaccine antigen transgene and the molecular TLR3 adjuvant, which are delivered together to epithelial and resident immune cells. Once released in the gut ileum, the tablet's enteric coating dissolves and releases rAd5 into the lumen. Following translation of the transgene, the vaccine protein antigen can be displayed on target cells in native confirmation or cross-presented to T cells as peptides on MHC I and II. Constantly sampling the intestinal lumen are other antigen-presenting immune cells, such as dendritic cells (DCs), that can display non-self antigens to naïve T cells and drive T follicular (TfH) expansion. TfH cells provide co-stimulation that enhances B cell differentiation, maturation, and class-switch recombination to IgA-expressing activated B cells. Activated B cells mature into IgA-secreting plasmablasts and enter the lymph and blood, where they traffic to lymphoid tissues and mucosal sites. Following oral vaccination, circulating IgA antibody-secreting cells upregulate the mucosal homing marker α4β7+, traffic to the respiratory tract, and secrete IgA. IgA transcytosis through the epithelial layer is mediated by the polymeric Ig receptor (pIgR) and released into the airway. Resident T cells may also recognize vaccine antigens displayed on the surface of gut epithelial cells or dendritic cells and elicit effector T cell maturation. These effector T cells can provide additional mucosal protection by identifying naturally infected cells and releasing effector molecules, such as IFNγ and granzyme (GzmB), in response to infection.
## Mucosal iga and influenza vaccine efficacy
Eliciting pathogen-specific mucosal immune responses is ideal for effective prevention of influenza and other microorganisms that enter the host through mucosal surfaces [bib_ref] Intestinal epithelial cells: Regulators of barrier function and immune homeostasis, Peterson [/bib_ref]. One of the hallmarks of mucosal responses is dimeric secretory IgA (SIgA), which plays an important role in prohibiting pathogen adherence and excluding infectious particles from gaining access to the host. Due to its valency, SIgA has more potent neutralizing and cross-reactive activity than IgG and has been shown to provide enhanced protection against influenza antigenic drift [bib_ref] Mechanisms of broad cross-protection provided by influenza virus infection and their application..., Tamura [/bib_ref] [bib_ref] Mechanism of neutralization of influenza virus by secretory IgA is different from..., Taylor [/bib_ref] [bib_ref] Relationship of the quaternary structure of human secretory IgA to neutralization of..., Suzuki [/bib_ref]. This immunoglobulin is secreted at mucosal tissues including the nasopharynx, which is the first line of defense the host has for protection against influenza or other airborne pathogens. SIgA contains a J chain that binds the polymeric immunoglobulin receptor pIgR, allowing for transcytosis from the basolateral to the apical side of epithelial cells [bib_ref] Role of J chain in secretory immunoglobulin formation, Johansen [/bib_ref]. IgA also has intracellular neutralizing activity, another important host defense function at the mucosal surface [bib_ref] Colocalization of intracellular specific IgA (icIgA) with influenza virus in patients' nasopharyngeal..., Kok [/bib_ref] [bib_ref] Translocalized IgA mediates neutralization and stimulates innate immunity inside infected cells, Bidgood [/bib_ref] [bib_ref] Intracellular neutralization of virus by immunoglobulin A antibodies, Mazanec [/bib_ref]. There have been several vaccine candidates that elicit antigen-specific nasal IgA responses, which correlates with protection against influenza virus infection [bib_ref] An Intranasal Proteosome-Adjuvanted Trivalent Influenza Vaccine Is Safe, Immunogenic & Efficacious in..., Lambkin-Williams [/bib_ref]. Additionally, methods have not yet been established for standardizing measurements for accurately quantifying mucosal IgA responses, making it challenging to implement in large clinical studies.
Most of the influenza vaccines currently on the market received licensure based on hemagglutination inhibition assays (HAI), which measure the functional activity of serum antibodies. While a serum HAI titer of greater than 40 is conventionally accepted as sufficient for protection against influenza [bib_ref] The role of serum haemagglutination-inhibiting antibody in protection against challenge infection with..., Hobson [/bib_ref] [bib_ref] Relationship between haemagglutination-inhibiting antibody titres and clinical protection against influenza: Development and..., Coudeville [/bib_ref] , immunized individuals with high serum HAI are not always protected from disease [bib_ref] Influenza hemagglutination-inhibition antibody titer as a correlate of vaccine-induced protection, Ohmit [/bib_ref]. Influenza virus diversity and antigenic drift among circulating strains impedes HAI as a predictive measurement for vaccine efficacy, as vaccine antigens do not always match circulating strains. Furthermore, HAI is not always an accurate predictor of vaccine efficacy, especially when other routes of administration are used [bib_ref] Influenza hemagglutination-inhibition antibody titer as a correlate of vaccine-induced protection, Ohmit [/bib_ref]. For example, the intranasal live attenuated influenza vaccines elicit considerable nasal IgA responses but less serum antibody [bib_ref] Immune responses after live attenuated influenza vaccination, Mohn [/bib_ref]. Other humoral responses to influenza include non-neutralizing Fc-effector functions that induce antibody-mediated cellular cytotoxicity, and these also have been shown to be important for protection [bib_ref] Effector mechanisms of influenza-specific antibodies: Neutralization and beyond, Sicca [/bib_ref]. Therefore, solely utilizing HAI as the key metric for evaluating new mucosal vaccine technologies will continue to be a significant hurdle, as serum and blood collection remains the preferred method for tracking efficacy. Therefore, gauging influenza vaccine efficacy by only measuring systemic humoral responses does not accurately reflect all correlates of protection, especially for mucosal vaccine approaches.
## Humoral reponses in the mucosa are protective against challenge and reduce transmission
Parenteral vaccines are designed to induce systemic pathogen-specific IgG and memory responses and reduce severe disease. However, they do not induce antigen-specific mucosal responses, which may be necessary to reduce community-wide transmission during a pandemic. The ongoing SARS-CoV-2 pandemic has created a situation where parenterally immunized individuals, while protected from severe disease and death, are still able to become infected and transmit the virus to other people. A mucosal antibody response in the upper respiratory tract can block viral transmission more effectively than a vaccine that only elicits serum antibodies [bib_ref] Immune responses after live attenuated influenza vaccination, Mohn [/bib_ref]. This observation has been demonstrated in preclinical models where dimeric IgA responses in the mucosa are more effective for reducing transmission than serum IgG responses. An influenza transmission study using a guinea pig model demonstrated that a recombinant neutralizing IgA antibody was highly effective at blocking transmission, but recombinant IgG with the same variable region administered to animals by injection was not effective [bib_ref] Recombinant IgA is sufficient to prevent influenza virus transmission in guinea pigs, Seibert [/bib_ref]. Additionally, the recombinant IgG administered intranasally was able to block infection, indicating that the location of the neutralizing antibody is an important factor, not just the isotype. Therefore, strategically designing vaccines to elicit protective mucosal IgA responses may increase vaccine efficacy.
Mucosal delivery of vaccine antigens also has been demonstrated to be effective for reducing viral transmission to naïve animals. In a hamster study, animals mucosally immunized with rAd5-expressing SARS-CoV-2 spike protein and subsequently infected with SARS-CoV-2 virus had a significant reduction in their ability to transmit the virus to naïve animals via unidirectional aerosol airflow. This study also demonstrated that vaccination routes of the same antigen resulted in significant differences in viral transmission. Viral copies measured in naïve animals' nasal passages following exposure to mucosally immunized index animals was significantly reduced compared to naïve animals exposed to index animals vaccinated with the same vaccine by intramuscular injection. These studies illustrate that oral immunization with an oral rAd5 vaccine can elicit antigen-specific responses in the mucosa and reduce transmission to naïve animals.
## Evaluating novel immune correlates for mucosal vaccines
Correlates of protection for mucosal vaccines may likely be different for injected immunizations (reviewed by [bib_ref] Mucosal vaccines-Fortifying the frontiers, Lavelle [/bib_ref]. Most influenza vaccines are evaluated based on serum HAI titers; however, this presents an obstacle for oral formulations that are designed to generate immune responses at the mucosa. To determine the appropriate immune correlate to properly assess efficacy of the oral rAd5 influenza vaccine tablet, a human challenge study was conducted [bib_ref] Efficacy, immunogenicity, and safety of an oral influenza vaccine: A placebo-controlled and..., Liebowitz [/bib_ref]. Prior to intranasal infection with the H1N1 virus, subjects were randomized to receive either a placebo, commercially available quadrivalent FluZone (IIV), or an oral rAd5 tablet H1N1 influenza vaccine (VXA-A1.1) [bib_ref] Efficacy, immunogenicity, and safety of an oral influenza vaccine: A placebo-controlled and..., Liebowitz [/bib_ref]. Although both vaccines protected the majority of subjects from illness [fig_ref] Table 2: Vaccine efficacy following H1N1 challenge [/fig_ref] , the group that received VXA-A1.1 had lower symptomatic disease compared to subjects immunized intramuscularly with FluZone (29%, 35% respectively). Furthermore, in a post-hoc analysis, viral shedding was decreased in subjects immunized with VXA-A1.1 compared to IIV (80% probability of superiority based on Bayesian analysis), demonstrating that administration of an oral rAd5 vaccine may further reduce viral shedding of the influenza virus compared to current licensed vaccines. Serum HAI responses were higher in the IIV-immunized cohort compared to the oral VXA-A1.1 group, indicating that serum antibody titers may not be the most important correlate for assessing the efficacy of the oral rAd5 vaccine. By utilizing a random forest analysis, a predictive machine learning algorithm that examines complex and diverse datasets as an ensemble, cellular responses measured by flow cytometry were found to be the most important correlate for VXA.A1-1 efficacy. More specifically, in subjects immunized with VXA.A1-1, circulating antigen-specific ASCs and α4β7+ mucosa homing lymphocytes were significantly elevated and correlated with a reduction in symptomatic disease. This study illustrates that only measuring serum-based responses is not sufficiently predictive of protective efficacy for mucosal vaccines. Developing novel cellular immunogenicity assays that can be used in large scale clinical trials to measure efficacy is critical for the development of next-generation vaccines. Measuring antigen-specific cellular responses is challenging in clinical trials; however, more recent technological advances have been facilitating the development of new correlates in clinical trials. The use of fixed whole blood and other advanced screening techniques has further enhanced the identification and quantification of important cellular phenotypic markers that can be used to evaluate protective efficacy [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. In a follow-up to the human influenza challenge phase II study, fixed whole blood and mass cytometry analysis were used to measure B and T cell activation 8 days after vaccination [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. In the VXA.A1.1 vaccine cohort, elevated α4β7+, phospho-Stat5+, and HA+ specific B cells highly correlated with a reduction in illness. Furthermore, central and effector CD8+ T cells in VXA-A1.1 recipients expressed higher β7 integrin and displayed an activated phenotype compared to placebo and FluZone recipients on day 8. Lastly, individuals in the VXA-A1.1 cohort who shed lower levels of the virus had elevated β7 integrin and pSTAT5 and low CD62L, indicating a correlation between mucosal homing and reduced viral load. This study illustrates that oral rAd5 vaccine administration induces protective levels of mucosal homing lymphocytes, which correlates with decreased viral shedding.
In addition to β7, α 4 can also dimerize with integrin β 1, and α 4 β 1+ B cells have been shown to migrate to various mucosal sites, including the bronchus-associated lymphoid tissue [bib_ref] The immune geography of IgA induction and function, Macpherson [/bib_ref]. Both homing markers β 1 and β7 have been found in human peripheral blood ASCs following rAd5 oral vaccination, demonstrating that this vaccine technology induces a combination of intestinal and non-intestinal homing B cells [bib_ref] Systemic and mucosal immune responses following oral adenoviral delivery of influenza vaccine..., Kim [/bib_ref] [bib_ref] Human influenza virus challenge identifies cellular correlates of protection for oral vaccination, Mcilwain [/bib_ref]. In summary, to accurately measure mucosal vaccine efficacy, new cellular immunogenicity measurements need to be further developed for implementation in large clinical studies.
## Anti-vector responses
One concern regarding the use of Ad vector vaccine platforms is the high rate of natural adenovirus seroprevalence and pre-existing immunity, which may impede efficacy. Several preclinical and clinical experiments were conducted to test if oral delivery of rAd5 to the small intestine generates anti-vector responses that interfere with vaccine antigen immunogenicity. In a ferret influenza challenge model utilizing H5N1, neutralizing antibody titers to rAd5 were quantified following a prime/boost regimen administered by injection or orally four weeks apart. The animals that received the vaccine via intramuscular administration developed substantial anti-Ad neutralizing antibodies. However, those animals vaccinated by oral administration with the same vector did not induce detectable anti-Ad5 responses [bib_ref] An adenovirus-based vaccine with a double-stranded RNA adjuvant protects mice and ferrets..., Scallan [/bib_ref]. Furthermore, preexisting immunity in humans to oral rAd5 had no effect on the ability to elicit neutralizing antibody responses or T cell responses following oral rAd5 immunization expressing influenza HA as an antigen nor did it have an effect on specific serum antibody responses in norovirus clinical trials [bib_ref] High titre neutralising antibodies to influenza after oral tablet immunisation: A phase..., Liebowitz [/bib_ref]. Therefore, anti-vector responses that impede immunogenicity in injected rAd-based vaccines have less of a role when immunizing with oral rAd5.
# Conclusions
The human mucosa interacts with numerous environmental and microbial agents every day, and as such, it is the first line of defense against respiratory pathogens, such as influenza and SARS-CoV-2. An oral rAd5 tablet vaccine described in this review has several advantages compared to traditional needle-based immunization approaches, including eliciting immune protective responses at primary sites of infection. Multiple clinical and pre-clinical studies have demonstrated that immunization with oral rAd5 elicits ASCs and T cells that express mucosal homing integrins and enhances pathogen-specific IgA responses. The combination of mucosal-directed humoral and cellular responses provides enhanced protection to respiratory pathogens, as shown by an influenza human challenge study. Recent studies have established that correlates of protection for an oral rAd5 mucosal vaccine are unique compared to currently licensed parenteral vaccines. Therefore, multiple serum, cellular, and mucosal endpoints will be evaluated in upcoming clinical trials to demonstrate immunogenicity and their correlation to efficacy.
While the systemic administration of vaccines via needle-based delivery has been successful for developing long-lasting immunity to many human pathogens, an oral rAd5 tablet-based vaccine platform has distinct advantages over the current parenteral vaccines, as it can be easily administered, stored at room temperature, and rapidly produced. An oral rAd5 vaccine has proven to be well-tolerated, safe, and immunogenic while also having a distinct distribution advantage as a room temperature-stable tablet that can be shipped globally and self-administered. Currently, countries with underdeveloped healthcare infrastructure are struggling to manage the cold-chain requirements, dosing schedules, and trained personnel required for the successful distribution and administration of COVID-19 vaccinations. The implementation of an oral rAd5 room temperature-stable, tablet-based vaccine would be a revolutionary tool that has the potential to make a significant impact on reducing the global burden of pathogenic viral diseases.
[fig] Figure 1: Next generation vaccine production, manufacturing, and distribution timelines during a pandemic. (A) Non-replicating recombinant adenovirus serotype 5 (rAd5) vaccine constructs can be developed and used in preclinical testing in as little as three weeks. Large scale manufacturing and release timelines for rAd-based vaccines are comparable and can be completed within 9 weeks, as demonstrated during the 2020 pandemic (S. Gilbert, personal communication, 28 October 2021). mRNA-based vaccines have the shortest development and manufacturing timelines; however, these [/fig]
[fig] Figure 2: Working model of oral rAd5 vaccine tablet, illustrating how antigen delivery to the small intestine generates protective mucosal immune responses. The non-replicating rAd5 vector contains both the vaccine antigen transgene and the molecular TLR3 adjuvant, which are delivered together to epithelial and resident immune cells. Once released in the gut ileum, the tablet's enteric coating dissolves and releases rAd5 into the lumen. Following translation of the transgene, the vaccine protein antigen can be displayed on target cells in native confirmation or cross-presented to T cells as peptides on MHC I and II. Constantly sampling the intestinal lumen are other antigen-presenting immune cells, such as dendritic cells (DCs), that can display non-self antigens to naïve T cells and drive T follicular (TfH) expansion. TfH cells provide co-stimulation that enhances B cell differentiation, maturation, and class-switch recombination to IgA-expressing activated B cells. Activated B cells mature into IgA-secreting plasmablasts and enter the lymph and blood, where they traffic to lymphoid tissues and mucosal sites. Following oral vaccination, circulating IgA antibody-secreting cells upregulate the mucosal homing marker α4β7+, traffic to the respiratory tract, and secrete IgA. IgA transcytosis through the epithelial layer is mediated by the polymeric Ig receptor (pIgR) and released into the airway. Resident T cells may also recognize vaccine antigens displayed on the surface of gut epithelial cells or dendritic cells and elicit effector T cell maturation. These effector T cells can provide additional mucosal protection by identifying naturally infected cells and releasing effector molecules, such as IFNγ and granzyme (GzmB), in response to infection. [/fig]
[fig] Author: Contributions: Original draft preparation, B.A.F., M.R.B. and S.N.T.; Review and editing, B.A.F., M.R.B. and S.N.T. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. [/fig]
[table] Table 1: Oral rAd5 vaccine tablet stability. [/table]
[table] Table 2: Vaccine efficacy following H1N1 challenge. Detectable viral shedding by qPCR in nasopharyngeal swab on any day after the first 36 h after challenge. 2 Illness was defined based on self-reported symptoms, along with laboratory-confirmed infection. [/table]
|
Inhibitors of the Influenza A Virus M2 Proton Channel Discovered Using a High-Throughput Yeast Growth Restoration Assay
The M2 proton channel of the influenza A virus is the target of the anti-influenza drugs amantadine and rimantadine. The effectiveness of these drugs has been dramatically limited by the rapid spread of drug resistant mutations, mainly at sites S31N, V27A and L26F in the pore of the channel. Despite progress in designing inhibitors of V27A and L26F M2, there are currently no drugs targeting these mutated channels in clinical trials. Progress in developing new drugs has been hampered by the lack of a robust assay with sufficient throughput for discovery of new active chemotypes among chemical libraries and sufficient sensitivity to provide the SAR data essential for their improvement and development as drugs. In this study we adapted a yeast growth restoration assay, in which expression of the M2 channel inhibits yeast growth and exposure to an M2 channel inhibitor restores growth, into a robust and sensitive high-throughput screen for M2 channel inhibitors. A screen of over 250,000 pure chemicals and semi-purified fractions from natural extracts identified 21 active compounds comprising amantadine, rimantadine, 13 related adamantanes and 6 non-adamantanes. Of the non-adamantanes, hexamethylene amiloride and a triazine derivative represented new M2 inhibitory chemotypes that also showed antiviral activity in a plaque reduction assay. Of particular interest is the fact that the triazine derivative was not sufficiently potent for detection as an inhibitor in the traditional two electrode voltage clamp assay for M2 channel activity, but its discovery in the yeast assay led to testing of analogues of which one was as potent as amantadine.
# Introduction
Influenza A viruses are highly infectious pathogens responsible for seasonal epidemics and for pandemics. Worldwide, seasonal epidemics result in 3-5 million cases of severe illness, and 250,000-500,000 deaths yearly , while pandemics such as the 1918 Spanish Flu, 1957 Asian Flu, 1968 Hong Kong Flu, and 2009 Swine Flu have resulted in millions of deaths [bib_ref] Pandemic Influenza: The Inside Story, Nicholls [/bib_ref].
Vaccination is the primary strategy for prevention, but antiviral agents are needed to manage seasonal influenza in vulnerable patients and are essential if generation of an appropriate vaccine is not rapid enough during a new pandemic. Only four drugs are currently approved in the USA for influenza A treatment: the viral neuraminidase inhibitors oseltamivir and zanamivir and the viral M2 proton channel inhibitors amantadine and its methyl derivative rimantadine [bib_ref] Antivirals Targeting Influenza A Virus, Das [/bib_ref]. Of these agents, only amantadine, rimantadine and oseltamivir are orally administered. Strains resistant to the M2 inhibitors are now predominant [bib_ref] ADamantane resistance among influenza a viruses isolated early during the 2005-2006 influenza..., Bright Ra [/bib_ref] and resistance to oseltamivir is increasingly encountered [bib_ref] Emerging Influenza Antiviral Resistance Threats, Hayden [/bib_ref] [bib_ref] Evidence of Person-to-Person Transmission of Oseltamivir-Resistant Pandemic Influenza A(H1N1) 2009 Virus in..., Moore [/bib_ref] [bib_ref] Dual Resistance to Adamantanes and Oseltamivir Among Seasonal Influenza A(H1N1) Viruses: 2008-2010, Sheu [/bib_ref]. Emergence of strains with resistance to all approved drugs is a distinct possibility and could have particularly serious repercussions in the event of a new pandemic.
Progress is being made in developing new neuraminidase inhibitorsbut there has been less progress with M2 proton channel inhibitors [bib_ref] Inhibitors of the M2 channel of influenza A virus, Duque [/bib_ref]. The M2 proton channel is required for virus replication and maturation. After the virus is taken up into the host cell by endocytosis, the low pH of the endosome activates the M2 channel to allow proton flux from the endosome into the viral interior. This acidification dissociates the viral RNA from its bound matrix proteins and permits release of the viral genetic material to the cytoplasm for replication [bib_ref] Unpacking the incoming influenza virus, Helenius [/bib_ref]. The M2 protein also equilibrates the pH gradient between the Golgi lumen and the cytoplasm to prevent premature conformational changes of hemagglutinin during viral maturation [bib_ref] Maturation of influenza A virus hemagglutininestimates of the pH encountered during transport..., Grambas [/bib_ref] [bib_ref] The ion channel activity of the influenza virus M(2) protein affects transport..., Sakaguchi [/bib_ref]. M2 is a homotetramer with each chain consisting of a short unstructured extracellular N-terminal domain (residues 1-24) that is important for incorporation into the virion; a single transmembrane domain that is necessary and sufficient for tetramerization, proton conductance and drug binding; an amphiphilic membraneassociated a-helix (residues 47-61) that is important for viral budding and scission; and a unstructured C-terminal cytoplasmic tail (residues 62-97) that interacts with matrix protein M1 [bib_ref] Nuclear transport of influenza virus ribonucleoproteins: The viral matrix protein (M1) promotes..., Martin [/bib_ref]. Amantadine binds the transmembrane region with its charged amino group mimicking hydronium [bib_ref] Structural and dynamic mechanisms for the function and inhibition of the M2..., Wang [/bib_ref]. Because the proton conductance rate of the channel has to match the pH sensitivity of hemagglutinin [bib_ref] Maturation of influenza A virus hemagglutininestimates of the pH encountered during transport..., Grambas [/bib_ref] [bib_ref] Influence of amantadine resistance mutations on the pH regulatory function of the..., Grambas [/bib_ref] , of the large number of amantadineresistant mutations that have been identified in vitro, only three have been identified in transmissible viruses [bib_ref] Evolution of the M gene of the influenza A virus in different..., Furuse [/bib_ref] [bib_ref] Large-Scale Sequence Analysis of M Gene of Influenza A Viruses from Different..., Furuse [/bib_ref]. These are single amino acid substitutions in the channel -V27A, L26F or S31N.
The development of inhibitors that are active against amantadine-resistant channels is highly desirable but it is very challenging to produce and assay the M2 membrane protein under native-like conditions. Methods of assaying ion channel activity are labor intense and technically demanding [bib_ref] Novel assay for the influenza virus M2 channel activity, Giffin [/bib_ref] [bib_ref] Solidsupported membrane technology for the investigation of the influenza A virus M2..., Balannik [/bib_ref] [bib_ref] Proton and cation transport activity of the M2 proton channel from influenza..., Leiding [/bib_ref] and therefore unsuitable for high-throughput screens of chemical libraries. Krystal et al. showed that when the M2 channel gene was expressed in the yeast Saccharomyces cerevisiae, a proton-selective channel was formed that disrupted the electrochemical potential across the yeast membrane and inhibited yeast growth, and that compounds that block M2 channel activity were able to restore growth [bib_ref] Growth impairment resulting from expression of influenza virus M2 protein in Saccharomyces..., Kurtz [/bib_ref]. The same principle was also applied using E. Coli but was not implemented for high throughput screening [bib_ref] Quantitative analysis of influenza M2 channel blockers, Astrahan [/bib_ref]. In this study, we developed the yeast growth restoration assay into a highthroughput screen for inhibitors of the M2 channel. We used it to discover not only additional analogs of known M2 channel inhibitors, but also amiloride derivatives and substituted triazines that represent chemotypes previously unrecognized as M2 channel inhibitors and that provide new starting points for influenza drug development.
# Results
## Development and validation of a yeast growth
Restoration Assay to Detect Inhibitors of the A/M2 Channel S. cerevisiae strains were generated containing a multicopy plasmid for expression of the wild-type, S31N-mutated, or V27A-mutated M2 gene from the Udorn strain of influenza A controlled by the inducible GAL1 promoter (designated WT, S31N and V27A respectively), or an empty plasmid. The growth of the four strains was monitored over time by turbidimetry following induction of the GAL1 promoter by galactose. Expression of WT M2 considerably slowed yeast growth to 27% of the growth of the control strain at 48 h [fig_ref] Figure 1: Effect of M2 expression and amantadine on yeast growth [/fig_ref]. Expression of amantadineresistant S31N M2 reduced yeast growth to 60% of the control strain while expression of V27A M2 reduced growth to 55% of control [fig_ref] Figure 1: Effect of M2 expression and amantadine on yeast growth [/fig_ref]. Amantadine was used to test whether the observed growth inhibition was caused by M2 proton channel activity. The growth of the yeast strain containing the empty plasmid was not affected by amantadine at 0.3, 1 or 3 mM [fig_ref] Figure 1: Effect of M2 expression and amantadine on yeast growth [/fig_ref]. By contrast, amantadine considerably increased the growth of the strain expressing WT M2 from 27% of control without amantadine to 65% of control at 0.3 mM amantadine, and 95% of control at 1 mM and 3 mM [fig_ref] Figure 1: Effect of M2 expression and amantadine on yeast growth [/fig_ref]. Amantadine did not increase the growth of the S31N and V27A strains (Figs 1C, 1D), as expected since these mutated channels are amantadineresistant.
To test whether growth inhibition by amantadine-resistant M2 mutants was also dependent on proton channel activity, we used a spiroadamantane amine [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref]. This compound inhibits the V27A mutant channel as well as the WT channel in the two electrode voltage clamp (TVEC) assay of M2 channel conductance with IC 50 of 0.3 and 18.7 mM respectively, and also inhibits replication of recombinant viruses bearing the V27A mutation in a plaque reduction assay [bib_ref] Molecular Dynamics Simulation Directed Rational Design of Inhibitors Targeting Drug-Resistant Mutants of..., Wang [/bib_ref]. As discussed previously, these assays are conducted at a short time scale of 2 min acidification (due to the limited stability of oocytes at low pH), prior to the establishment of equilibrium. The IC 50 values obtained with the TEVC assay are therefore one or two orders of magnitude larger than in cellular assays that are conducted at longer time scales [bib_ref] Design and Pharmacological Characterization of Inhibitors of Amantadine-Resistant Mutants of the M2..., Balannik [/bib_ref]. In accordance with these expectations, spiroadamantane amine clearly increased growth of the V27A strain at 0.01 mM and above, with an EC 50 of 0.3 mM [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref]. It also restored growth of the WT M2 strain in the same concentration range [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref]. Interestingly, the spiroadamantane amine was 3-fold more potent against the WT M2 strain (EC 50 = 0.03 mM) [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref] than amantadine (EC 50 = 0.1 mM).
These results with drug-sensitive and drug-resistant mutants and a mutant-selective inhibitor showed that yeast growth restoration results from inhibition of M2 proton channel activity. Moreover, the ability of low concentrations of the inhibitors to significantly restore growth indicated the potential of this assay for development into a sensitive screen for M2 inhibitors.
## High-throughput screen for m2 inhibitors
To determine whether this assay could be used to discover new inhibitors, a simple benchtop 96-well plate assay was developed (see Materials and Methods). This assay showed good discrimina- tion between the growth of strains expressing WT M2 or empty plasmid [fig_ref] Figure 3: 96-well pilot screen results [/fig_ref] , with an acceptable Z' factor of 0.50 [bib_ref] Highthroughput screening assays for the identification of chemical probes, Inglese [/bib_ref].
We used this assay in a pilot screen of ,30,000 commercially available pure chemicals (Prestwick, BioMol, Sigma LOPAC, Microsource Spectrum, Maybridge Hitfinder and ChemBridge DIVERset). A representative example of a screen of a batch of 1,440 compounds and their corresponding control wells tested on the same day is shown in [fig_ref] Figure 3: 96-well pilot screen results [/fig_ref]. Many of the compounds inhibited yeast growth but only one showed substantial growth restoration. This compound, rimantadine derivative 6, had no effect on the growth of the yeast strain carrying an empty plasmid but it increased the growth of the WT M2 strain in a concentration-dependent manner with an EC 50 of 0.2 mM [fig_ref] Figure 3: 96-well pilot screen results [/fig_ref].
Overall, this pilot screen identified 14 active compounds (1-8, 14-16, 18) as well as amantadine and rimantadine .
Encouraged by these findings, we sought to increase the capacity and automation of this assay to enable high-throughput screening. A similar assay using 384-well plates and automated liquid handling and plate reading was established. In this assay, yeast expressing WT M2 treated or not with amantadine was used to establish the assay Z' factor, which averaged 0.54. This assay was used to screen 50,000 pure compounds from Chembridge, yielding 7 additional active compounds . To enable discovery of structurally novel compounds, we also carried out a screen of 176,000 semi-purified fractions from the Nature Bank collection of natural products extracts [bib_ref] Developing a Drug-like Natural Product Library, Quinn [/bib_ref]. These fractions, which typically contain several compounds, were tested at a dilution expected to result in individual compounds being present at 0.5-5 mM, assuming a MW of 300. Few fractions showed any level of growth restoration. Nevertheless, ,500 fractions showing .1.8-fold increase in growth (1 mM amantadine caused a 3-fold growth increase) were retested but none showed activity.
## Characterization of active compounds
The yeast growth restoration screen identified 21 active compounds, including amantadine, rimantadine, 13 adamantanes and 6 non-adamantanes. The activity of these compounds was confirmed in concentration curves and their names, structures and EC 50 values are shown in . The EC 50 of these active compounds ranged from 0.1 to 30 mM. Four adamantanes (6, 9, 10 and 12) and two non-adamantanes (15 and 17) had EC 50 values below 1 mM.
All of the compounds with the exception of the weak inhibitor 18 had the bipartite structure of previous M2 inhibitors which includes a hydrophobic scaffold (e.g., adamantane, bicycloheptane, or cyclooctane) and a polar head group [bib_ref] Exploring the Requirements for the Hydrophobic Scaffold and Polar Amine in Inhibitors..., Wang [/bib_ref] [bib_ref] Discovery of highly potent agents against influenza A virus, Zhao [/bib_ref] [bib_ref] Identification of Hits as Matrix-2 Protein Inhibitors through the Focused Screening of..., Hu [/bib_ref]. Examination of the active adamantanes shows that a wide variety of substitutions are tolerated at either the 1 or 2 position of adamantane, including substituted benzyl (2, 3, 4 and 5), fivemembered heterocycles (6, 9, 10, 11, 12 and 13) and a sixmembered heterocycle substitution (1). Compounds 3, 4, 5, 8, 15, 16, 17 and 19 were primary or secondary amines that are likely protonated at the pH of the assay medium (pH = 6.5). This finding is consistent with earlier studies that suggest that the ammonium serves to mimic the hydronium ion in the M2 channel [bib_ref] Structure and mechanism of proton transport through the transmembrane tetrameric M2 protein..., Acharya [/bib_ref]. Additionally, compounds 10, 12, 13, 14 and 18 bear basic heterocycles, acylguanidine or tertiary amines that might also be protonated at near neutral or slightly acidic pH. The remaining compounds 1, 2, 6, 7, 9 and 11 contain neutral polar groups, particularly cyclic or acyclic thiourea (1 and 6) and sulfonamide [bib_ref] Pandemic Influenza: The Inside Story, Nicholls [/bib_ref] , some of them seen in previously reported M2 inhibitors [bib_ref] Exploring the Requirements for the Hydrophobic Scaffold and Polar Amine in Inhibitors..., Wang [/bib_ref] [bib_ref] Discovery of Spiro-Piperidine Inhibitors and Their Modulation of the Dynamics of the..., Wang [/bib_ref]. Compound 4 was selected to be tested in the TEVC assay and found to inhibit 43.5% of the WT channel activity at 100 mM , which correlates with the yeast screening result with EC 50 of 5 mM .
Among the active compounds not containing an adamantane moiety , bicyclic compounds such as derivatives of pinene 15, camphor 16 and bicycloheptane 17 had been shown previously to be M2 inhibitors [bib_ref] Discovery of highly potent agents against influenza A virus, Zhao [/bib_ref] [bib_ref] Identification of Hits as Matrix-2 Protein Inhibitors through the Focused Screening of..., Hu [/bib_ref] [bib_ref] Synthesis of substituted 1-norbornylamines with antiviral activity, Martinez [/bib_ref] supporting the statement that the M2 WT channel is able to accommodate hydrophobic molecules with cage shapes other than that of adamantane [bib_ref] Inhibitors of the M2 channel of influenza A virus, Duque [/bib_ref]. Pyrrole-substituted cyclooctylamine 19 was active, but with a poor EC 50 of 30 mM. Cyclooctylamine had been found by Shroeder's group to block M2 channel proton flux in a liposome assay [bib_ref] Different modes of inhibition by adamantane amine derivatives and natural polyamines of..., Lin [/bib_ref] and its binding affinity was later determined to be 6-fold weaker than amantadine [bib_ref] Interaction of aminoadamantane derivatives with the influenza A virus M2 channel-Docking using..., Eleftheratos [/bib_ref]. Together, these results show that WT M2 channel is able to accommodate compounds with diverse structures.
Significantly, two active compounds, hexamethylene amiloride 14 and triazine 18 (EC 50 of 1.1 and 12 mM respectively), have not been reported previously as M2 inhibitors. To determine whether restoration of yeast growth by these compounds was truly a result of M2 channel inhibition, we purchased or synthesized several analogs of 14 and 18 and tested them in the TEVC assay, which directly measures proton channel conductance.
Hexamethylene amiloride 14 was highly active against WT M2 with 89.9% channel conductance inhibition at 100 mM, as active as amantadine . Three close structural analogs 20, 21 and 22 were also tested and these showed only moderate inhibition . However, a clear structure-activity relationship was observed in this series of amiloride analogs -the more hydrophobic the molecule is, the more potent it is in inhibiting M2. This was anticipated based on earlier SAR results that all active compounds have cLogP.1.5 [bib_ref] Exploring the Requirements for the Hydrophobic Scaffold and Polar Amine in Inhibitors..., Wang [/bib_ref].
Triazine 18 was not active (4.5%) in the TEVC assay at 100 mM . Twelve closely related structural analogs of 18 (compounds 23-34) were tested in the TVEC assay. Interestingly, the two compounds with hydrophobic scaffolds of more than seven carbon atoms were active while all others were essentially inactive in this assay (,20% inhibition at 100 mM) . Active compounds 33 and 34 were tested in the yeast growth restoration assay and showed EC 50 values of 0.7 mM and 0.3 mM respectively, compared with an EC 50 of 12 mM for triazine 18. Given the good correspondence between inhibition at 100 mM in the TEVC assay and potency in the yeast assay for compounds identified here, and the fact that the TEVC assay requires much higher inhibitor concentrations than the yeast assay to show activity, we surmise that triazine 18 is not a false positive from the yeast assay but rather a true M2 inhibitor that is not sufficiently potent to be detected by the TVEC assay. Taken together, these results point to the importance of the contribution of the hydrophobic scaffolds to the high activity of the inhibitors (compound 14, 33 and 34) of WT M2 rather than the polar headgroup, which explains why a wide variety of polar headgroups can be tolerated.
Given the high prevalence of amantadine-resistant mutations in seasonal flu strains, it was of interest to determine whether any of the compounds found to be active against WT M2 also showed activity towards the S31N or V27A mutant channels. Compounds 1-19, 33 and 34 were tested against these two mutants in the yeast assay. Compound 16 showed significant activity against the V27A mutant (EC 50 = 7 mM) and was inactive against S31N while all other compounds were inactive against V27A or S31N. Compounds 14, 18, and 20-34 were also tested at 100 mM against the S31N and V27A mutant channels in the TEVC assay and were found to be inactive and . Two potent adamantanes (6 and 10) and two non-adamantanes [bib_ref] Maturation of influenza A virus hemagglutininestimates of the pH encountered during transport..., Grambas [/bib_ref] [bib_ref] Discovery of Spiro-Piperidine Inhibitors and Their Modulation of the Dynamics of the..., Wang [/bib_ref] were further tested for inhibition of influenza A replication in a MDCK plaque reduction assay [bib_ref] Design and Pharmacological Characterization of Inhibitors of Amantadine-Resistant Mutants of the M2..., Balannik [/bib_ref] , with amantadine as a positive control. All four compounds completely inhibited plaque formation at 10 mM [fig_ref] Figure 6: Antiviral activity and cytotoxicity of selected compounds [/fig_ref]. At 1 mM, compounds 6 and 10 almost completely inhibited plaque formation, comparable to 1 mM amantadine, compound 34 showed moderate inhibition, and compound 14 had nearly no effect [fig_ref] Figure 6: Antiviral activity and cytotoxicity of selected compounds [/fig_ref]. The potency of the antiviral effect of 6, 10 and 34 was consistent with their low EC 50 in the yeast growth restoration assays [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref] , compared with an EC 50 of 0.1 mM for amantadine. The good potency of compound 34 agreed with its TEVC assay result showing 89.7% inhibition at 100 mM (92.0% for amantadine at 100 mM). Hexamethyleneamiloride 14, which was less potent in the yeast assay (EC 50 = 1.1 mM) was also less potent in the plaque reduction assay. The four compounds were also tested for cytotoxicity towards uninfected MDCK cells. Compounds 6, 10 and 34 showed very low cytotoxicity, indicative of a good activity window, while hexamethyleneamiloride was more cytotoxic, with a CC 50 of 16 mM [fig_ref] Figure 6: Antiviral activity and cytotoxicity of selected compounds [/fig_ref] , indicative of a less favorable activity window. Overall, the results demonstrate that the yeast growth restoration assay can identify M2 inhibitors with antiviral activity.
# Discussion
Progress in finding new inhibitors targeting drug resistant M2 channels has been slow but recent advances in understanding the structure and dynamic properties of the M2 channel in a lipid bilayer environment [bib_ref] The Flu's Proton Escort, Fiorin [/bib_ref] [bib_ref] Mechanisms of Proton Conduction and Gating in Influenza M2 Proton Channels from..., Hu [/bib_ref] [bib_ref] Insight into the Mechanism of the Influenza A Proton Channel from a..., Sharma [/bib_ref] , and the interaction of amantadine with the channel [bib_ref] Structure of the amantadine binding site of influenza M2 proton channels in..., Cady [/bib_ref] [bib_ref] Structure and mechanism of the M2 proton channel of influenza A virus, Schnell [/bib_ref] [bib_ref] Structural basis for the function and inhibition of an influenza virus proton..., Stouffer [/bib_ref] [bib_ref] Specific Binding of Adamantane Drugs and Direction of Their Polar Amines in..., Cady [/bib_ref] have spurred structure-based drug design and virtual screening efforts. Several new inhibitors were described in the last few years, many closely resembling amantadine [bib_ref] Inhibitors of the M2 channel of influenza A virus, Duque [/bib_ref]. However, inhibitors targeting drug resistant M2 mutants have not been reported. A notable exception is the spiroadamantane amine shown in [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref] that is active against WT, L26F and V27A M2 [bib_ref] Molecular Dynamics Simulation Directed Rational Design of Inhibitors Targeting Drug-Resistant Mutants of..., Wang [/bib_ref]. A robust and sensitive high-throughput screening assay would enable sampling of an expanded chemical diversity repertoire and guide SAR studies.
In this study, we modified an assay described by Krystal et al [bib_ref] Growth impairment resulting from expression of influenza virus M2 protein in Saccharomyces..., Kurtz [/bib_ref] and optimized it for high-throughput screening, notably by expressing M2 in a yeast strain with defective drug efflux systems [bib_ref] Screening for chemical inhibitors of heterologous proteins expressed in yeast using a..., Balgi [/bib_ref] , and by simplifying and automating assay conditions. The resulting growth restoration assay is robust, highly sensitive, quantitative, cost-effective and technically simple. We found that the yeast growth inhibition was greatest for WT, giving rise to a good Z' factor for screening. Because a successful drug should inhibit WT M2 as well as amantadine-resistant mutated forms, we use WT in the primary screen followed by testing of active compounds against mutants in secondary assays. Our screen of over 250,000 pure chemicals and semi-purified fractions from natural extracts identified only 21 active compounds. This 0.008% hit rate is 10-to 100-fold lower than typically encountered in screens for other targets. The assay shows high selectivity for M2 channel inhibitors as further testing of the 21 active compounds revealed no false positives. In positive readout cell-based assays, high selectivity is imparted largely by the requirement for active compounds to both inhibit their target and not hamper cell growth, effectively eliminating many non-selective active chemicals. However, since we have observed higher hit rates in yeast growth restoration assays with other targets [bib_ref] Inhibitors of human indoleamine 2,3-dioxygenase identified with a target-based screen in yeast, Vottero [/bib_ref] , the very low hit rate and the paucity of chemical classes identified indicates that M2 is not easily inhibited by the drug-like synthetic chemicals assembled in commercially available libraries. Natural products are known to encompass a much wider range of biologically relevant chemical space, so we were surprised that our large-scale screen of semi-purified fractions obtained from natural extracts did not yield a single hit. We do not believe this is due to technical factors such as testing of the fractions at too low a concentration to reveal activity [bib_ref] Drug-like Properties: Guiding Principles for the Design of Natural Product Libraries, Camp [/bib_ref]. Rather, we think this indicates that the M2 channel can be selectively inhibited by very few pharmacophores and that it is a challenging target for drug development. Nonetheless, the screen identified two compounds with novel structures for M2 channel inhibitors -hexamethylene amiloride 14 and triazine 18. Further testing in the TEVC assay confirmed that hexamethylene amiloride 14 inhibited the M2 channel, and subsequent SAR evaluation of amiloride analogs showed that hydrophobicity played a critical role in M2 channel inhibition. Hexamethylene amiloride 14 is a known inhibitor of the mammalian Na + /H + exchanger [bib_ref] Amiloride and its analogs as tools in the study of ion transport, Kleyman [/bib_ref] , and is also active against the HCV proton channel p7, the HIV-1 Vpu ion channel and the SARS Coronavirus envelope protein ion channel activity [bib_ref] Amiloride derivatives block ion channel activity and enhancement of virus-like particle budding..., Ewart [/bib_ref] [bib_ref] Cation-selective ion channels formed by p7 of hepatitis C virus are blocked..., Premkumar [/bib_ref] [bib_ref] Structure and Inhibition of the SARS Coronavirus Envelope Protein Ion Channel, Pervushin [/bib_ref]. Hexamethylene amiloride 14 inhibits the mammalian Na + /H + exchanger with an apparent K i of 0.2 mM [bib_ref] Inhibition of chemotactic factor-activated Na+/ H+ exchange in human neutrophils by analogues..., Simchowitz [/bib_ref] , more potently than it inhibits M2 in the yeast assay (EC 50 = 1 mM) and therefore does not appear to be a good drug candidate. An azolo-1,2,4-triazine derivative has been described as an inhibitor of influenza A and B virus replication [bib_ref] Antiviral properties, metabolism, and pharmacokinetics of a novel lead azolo-1,2,4-triazine-derived inhibitor of..., Karpenko [/bib_ref] , with an uncharacterized mechanism of action. Triazines 18, 33 and 34 described here are considerably different and have not previously been linked to the inhibition of ion channels. Triazine 34 completely inhibited influenza A replication at 10 mM and was not very toxic to MDCK cells (CC 50 = 75 mM). These compounds may be good starting points for drug discovery. The high sensitivity of this yeastbased assay also revealed that the spiroadamantane amine shown in [fig_ref] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT... [/fig_ref] is not only highly potent against the amantadine-resistant V27A mutant as reported [bib_ref] Molecular Dynamics Simulation Directed Rational Design of Inhibitors Targeting Drug-Resistant Mutants of..., Wang [/bib_ref] , but that it is a 3-fold more potent inhibitor of the WT M2 channel than amantadine and rimantadine, and may therefore be a candidate for drug development. Together, these results demonstrate the value of the yeast growth restoration assay for high throughput screening of M2 inhibitors and discovery of anti-influenza agents.
# Materials and methods
## Yeast strains and growth measurement
The multicopy expression plasmids, the yeast strains and the assay to measure yeast growth used in this study are described in detail in a previous publication [bib_ref] Screening for chemical inhibitors of heterologous proteins expressed in yeast using a..., Balgi [/bib_ref]. The coding sequences of the Udorn WT, S31N and V27A M2 genes were introduced downstream of the GAL1 promoter by PCR amplification followed by recombination cloning in yeast. Plasmids were rescued from the yeast strains and coding sequences were sequenceverified. The yeast strains were propagated in synthetic complete (SC) medium pH 6.5 lacking leucine and containing glucose to repress M2 gene expression. To measure growth under M2 expression conditions, yeast were harvested by centrifugation, the pellets were washed with water to remove traces of glucose and they were suspended to A 600 = 0.008 in liquid SC medium pH 6.5 lacking leucine and containing 2% galactose. 100 ml samples were distributed into 96-well plates and incubated at 30uC in a humidified chamber, without agitation. A 600 was measured in a microplate reader at different times after suspending the cells by gentle but thorough vortexing.
## 96-well medium-throughput screening assay
WT M2 yeast were grown as above, suspended at A 600 = 0.01 in liquid SC medium without leucine and with 2% galactose, and distributed at 100 ml per well into sterile transparent 96-well plates using 8-channel dispensing pipetors. Screening chemicals were transferred from 5 mM stock solutions in DMSO to yeast plates at a final concentration of ,15 mM using a Biorobotics TAS1 robot equipped with a 0.7 mm diameter 96-pin tool. The plates were incubated at 30uC in a humidified incubator. After 40-42 h cells were suspended by vortexing and A 600 was measured. The percentage of growth restoration was calculated as described [bib_ref] Screening for chemical inhibitors of heterologous proteins expressed in yeast using a..., Balgi [/bib_ref].
Chemicals showing growth restoration were retested in triplicate at different concentrations against WT M2 and empty plasmid strains to confirm activity and determine EC 50 .
## 384-well plate high throughput screening assay
The 96-well plate assay was adapted to 384-well format as follows: 30 ml of SC medium lacking leucine and containing 2% galactose was distributed into wells of Corning #3680 clear polystyrene 384-well plates using a Matrix Wellmate microplate dispenser. Screening chemicals and extracts were added using a Matrix Platemate Plus automated liquid handling system equipped with a 0.7 mm diameter FP3 384-pin tool transferring ,80 nl. 20 ml yeast at A 600 = 0.00625 in the same medium was then added using the matrix Wellmate dispenser. The plates were incubated at 30uC in a humidified Fisher isotemp incubator for 44 h and A 600 was measured without shaking using a Biotek Powerwave plate reader integrated with a Biotek Biostack plate stacking system. Plates were assayed in batches of 40-50 plates. Each 384-well plate contained 16 control wells not exposed to drug and 16 control wells into which 1.8 mM amantadine had been dispensed using the Matrix Wellmate.
## Tvec assay
Compounds were tested via a two-electrode patch clamp assay using Xenopus laevis oocytes microinjected with RNA for expression of the WT, S31N or V27A M2 gene [bib_ref] Design and Pharmacological Characterization of Inhibitors of Amantadine-Resistant Mutants of the M2..., Balannik [/bib_ref].
## Chemical synthesis
Compounds listed in and were obtained either from commercial sources and used without further purification or synthesized according to a literature procedure [bib_ref] Chemical genomics in Escherichia coli identifies an inhibitor of bacterial lipoprotein targeting, Pathania [/bib_ref]. The following compounds were ordered from Sigma-Aldrich: 14 (cat. [fig_ref] Figure 1: Effect of M2 expression and amantadine on yeast growth [/fig_ref]. Amine (1 mmol) was mixed with formaldehyde (2 mmol) in dioxane (5 ml). Isothiourea (1 mmol) was subsequently added. The mixture was heated until a clear solution was formed and allowed to stir at room temperature overnight. The mixture was extracted with dichloromethane and saturated NaHCO 3 solution. The organic phase was combined and dried over MgSO 4 , filtered and the solvents were removed under reduced pressure. The crude product was purified by silica gel flash column chromatography with a gradient of 10-20% CH 3 OH/dichloromethane. The synthesized compounds were characterized by 1 HNMR and MS. Selected compounds were also characterized with 13 CNMR.
6-(methylthio)-3-(thiophen-2-ylmethyl)-1,2,3,4-tetrahydro-1,3,5triazine [bib_ref] Proton and cation transport activity of the M2 proton channel from influenza..., Leiding [/bib_ref]. 3-(cyclohexylmethyl)-6-(ethylthio)-1,2,3,4-tetrahydro-1,3,5-triazine [bib_ref] Structure and mechanism of proton transport through the transmembrane tetrameric M2 protein..., Acharya [/bib_ref]. Table S1 Activity of hexamethylene amiloride and triazine analogs in the TEVC assay.
## Supporting information
## (pdf)
Author Contributions
[fig] Figure 1: Effect of M2 expression and amantadine on yeast growth. Yeast strains containing an empty plasmid (A) or plasmid bearing WT M2 (B), S31N M2 (C) or V27A M2 (D) were distributed into 96-well plates and their growth was measured over time following transfer at 0h to medium containing galactose and addition of the indicated concentrations of amantadine. doi:10.1371/journal.pone.0055271.g001 [/fig]
[fig] Figure 2: Effect of a spiroadamantane amine on the growth of yeast expressing WT and mutated M2. Yeast strains containing the indicated plasmids were distributed in 96-well plates in medium containing galactose and exposed to the indicated concentrations of the spiroadamantane amine (structure shown) for 40 h. doi:10.1371/journal.pone.0055271.g002 [/fig]
[fig] Figure 3: 96-well pilot screen results. (A) Yeast expressing WT M2 were exposed to 1,440 screening chemicals (M2+ Drugs) with 8 untreated wells per plate as negative controls (M2, n = 144). Each plate also contained 8 positive controls not treated with drugs (Empty plasmid, n = 144). Growth was measured after 40 h and the % growth restoration was calculated as described in Materials and Methods. The arrow points to the single active compound found in this batch. (B) the active compound found in A was retested at different concentrations against yeast expressing WT M2 or bearing the empty plasmid and growth was measured after 40 h. doi:10.1371/journal.pone.0055271.g003 [/fig]
[fig] Figure 4, Figure 5: StructuresActivity of hexamethylene amiloride and selected triazine analogs in the TEVC assay. Each point is the mean and standard deviation of five to eight oocytes. doi:10.1371/journal.pone.0055271.g005 [/fig]
[fig] Figure 6: Antiviral activity and cytotoxicity of selected compounds. The effect of the compounds on influenza A/Udorn/72 WT virus replication was evaluated by plaque formation in MDCK cells. Cytotoxicity towards MDCK cells was assayed after 48 h drug exposure with the MTT tetrazolium dye assay as described[55]. doi:10.1371/journal.pone.0055271.g006 [/fig]
[fig] Figure S1: Procedure used to synthesize compounds 23, 24, 28, 30, 31, 32, 33 and 34. (PDF) [/fig]
|
Talaromyces marneffei infection in a non-HIV non-endemic population
# Introduction
Talaromyces (Penicillium) marneffei is a dimorphic fungus and the only member in its genus known to be pathogenic to humans. It can cause both localized as well as overwhelming disseminated infection. T. marneffei is endemic to Southeast Asia [bib_ref] Penicillium marneffei infection and recent advances in the epidemiology and molecular biology..., Vanittanakom [/bib_ref] [bib_ref] Talaromyces (Penicillium) marneffei infection in non-HIV-infected patients, Chan [/bib_ref]. It is an important opportunistic infection in HIV-infected patients who live in or have traveled to endemic regions and is considered an acquired immunodeficiency syndrome (AIDS) defining illness. Recently, there have been several case reports of T. marneffei infection in HIV-negative patients who are otherwise not immunosuppresssed [bib_ref] Talaromyces (Penicillium) marneffei infection in non-HIV-infected patients, Chan [/bib_ref]. We report a case of a HIV negative Vietnam war veteran with advanced chronic obstructive pulmonary diseases (COPD) who presented with pulmonary T. marneffei lung infection which reactivated 40 years after a one year tour in Vietnam.
## Case
Our patient was a 75-year-old Caucasian man, a Navy veteran, who T bullous emphysema and presented to his primary care provider with uncontrollable nonproductive cough, for which he was prescribed oral levofloxacin and steroids with no improvement. He was then referred to pulmonary medicine because of an abnormal chest computerized tomography (CT), which showed a cavitary left lower lobe (LLL) infiltrate and several lung nodules [fig_ref] Figure 1: a & b [/fig_ref].
Bronchoscopy with bronchoalveolar lavage (BAL) yielded no specific etiology and he was treated symptomatically. After transient improvement, there was recurrence of productive cough, night sweats, malaise, and new hemoptysis. A second chest CT showed increase in the size of the LLL cavitary infiltrate, a new small necrotizing lesion in the left lung, and a new small left pleural effusion. Repeat bronchoscopy with transbronchial biopsy and BAL as well as thoracentesis was performed, revealing a sterile exudate and nondiagnostic findings on transbronchial biopsy and BAL. The patient was empirically started on itraconazole suspension on suspicion of an indolent fungal infection even though serologies for Histoplasma, Cryptococcus, and Coccidiomycosis were negative. Therapy had to be stopped due to gastrointestinal (GI) side effects. The patient was then changed to voriconazole but his symptoms continued unabated.
At this point, the patient was referred to Infectious Disease. Initial evaluation revealed a man with a severe cough, hemoptysis, and purulent sputum. Sputum samples were obtained for culture. Voriconazole was continued but only for short time because of side effects. The initial BAL grew a mold. Because he remained ill, the patient was started on liposomal amphotericin B (LAMB) and isavuconazole. The other bronchoscopy samples grew the same mold, which was identified as Talaromyces (Penicillium) marneffei. He was given 28 days of LAMB and isavuconazole with an excellent response and was discharged home on daily isavuconazole. On follow-up, he did very well and the antifungal was stopped at day 90. However, soon thereafter productive cough and malaise recurred. He was then readmitted and both isavuconazole and LAMB were resumed. His symptoms slowly worsened and isavuconazole was changed to voriconazole, but he continued to worsen. New lesions developed in his left lung as did a right lung cavitary opacity. Micafungin was added empirically and despite triple therapy, the patient progressed clinically and radiographically, and eventually succumbed to his infection. An extensive workup for an immunosuppressive state was nondiagnostic; HIV testing was repeatedly negative, including HIV viral load.
# Discussion
T. marneffei infection is an important systemic mycosis in HIV infected individuals who reside in or have traveled to Southeast Asia. It was described in many regions of Thailand, Vietnam, northeastern India, Southern China, Hong Kong, Taiwan, Laos, Malaysia, Myanmar, and Cambodia [bib_ref] Talaromyces (Penicillium) marneffei infection in non-HIV-infected patients, Chan [/bib_ref]. The prevalence of T. marneffei infection in non-HIV, non-endemic population is very low, only 4 cases were identified in the literature [bib_ref] Penicillium marneffei infection in a lung transplant recipient, Stathakis [/bib_ref] [bib_ref] Chronic pulmonary penicilliosis due to Penicilliom marneffei: late presentation in a French..., De Monte [/bib_ref] [bib_ref] Travel-related disseminated Penicillium marneffei infection in a renal transplant patient, Hart [/bib_ref].
The first reported case of T. marneffei infection was laboratory acquired in France in 1959 and the first case due to natural exposure was reported in 1973, in a patient who had Hodgkin's disease and lived in North America and who had traveled to Southeast Asia [bib_ref] Penicillium marneffei infection and recent advances in the epidemiology and molecular biology..., Vanittanakom [/bib_ref]. However, it was not until the late 1980s, with the arrival of the HIV epidemic in Asia, that the incidence of T. marneffei infection increased in the endemic population and among visitors traveling to Southeast Asia [bib_ref] Penicilliosis marneffei-west meets east, Wong [/bib_ref]. Hence, T. marneffei infection is considered an AIDS defining opportunistic infection.
With the use of potent antiretroviral therapy, the incidence of HIVassociated T. marneffei infection decreased considerably [bib_ref] Talaromyces (Penicillium) marneffei infection in non-HIV-infected patients, Chan [/bib_ref]. At the same time, the incidence increased in HIV-negative patients with other immunocompromising conditions who traveled to Southeast Asia. This change in epidemiology poses a diagnostic challenge. As a result, there may be delays initiating therapy which may result in disease progression and poor outcomes [bib_ref] Penicilliosis marneffei-west meets east, Wong [/bib_ref] [bib_ref] Clinical and laboratory characteristics of penicilliosis marneffei among patients with and without..., Kawila [/bib_ref]. Often, the only clue a physician may have to make the appropriate diagnosis is a history of travel to an endemic area.
The pathogenesis of T. marneffei infection is unclear. The main mode of acquisition is inhalation of conidia from the environment. Phagocytic cells are the primary host defense against the fungus, resulting in granulomatous and suppurative reactions in immunocompetent patients and necrotizing reactions in those who are immunocompromised [bib_ref] Penicillium marneffei infection and recent advances in the epidemiology and molecular biology..., Vanittanakom [/bib_ref]. Disseminated disease via the reticuloendothelial system is common in immunosuppressed individuals, in whom the severity of disease depends on the degree of immunosuppression. Five cases have been reported to date among non-HIV non-endemic patients, including the case described here [bib_ref] Penicillium marneffei infection in a lung transplant recipient, Stathakis [/bib_ref] [bib_ref] Chronic pulmonary penicilliosis due to Penicilliom marneffei: late presentation in a French..., De Monte [/bib_ref] [bib_ref] Travel-related disseminated Penicillium marneffei infection in a renal transplant patient, Hart [/bib_ref] [bib_ref] Penicillium marneffei presenting as an obstructing endobronchial lesion in an immunocompetent host, Joosten [/bib_ref]. [fig_ref] Table 1: Talaromyces marneffei Infection in non-HIV, non-Endemic Infected Adult Patients [/fig_ref] - Biopsy: negative condition should be encouraged to improve early diagnosis and treatment. It is very important to remember that T. marneffei is a primary pulmonary pathogen which may well use macrophages to proliferate and later cause acute pulmonary, disseminated and/or reactivation disease. The laboratory diagnosis is performed by identifying the fungus by microscopy and culture from a variety of specimens including blood, skin, bone marrow, lymph nodes, respiratory sources, liver biopsies, cerebrospinal fluid, urine, stool, kidney, pericardium, intestine or gastric specimens. A definitive diagnosis of T. marneffei infection by fungal culture has been reported with 100%, 90% and 76% sensitivity from bone marrow, skin biopsy and blood, respectively [bib_ref] Penicillium marneffei infection and recent advances in the epidemiology and molecular biology..., Vanittanakom [/bib_ref]. T. marneffei can be found on direct examination of a peripheral blood smear in patients with disseminated disease. Histopathological sections stained with hematoxylin and eosin, Grocott methenamine silver or periodic acid Schiff stain will demonstrate T. marneffei in biopsy specimens. In tissue, the yeast forms can be found with clearly defined central septae, characteristic of T. marneffei.
T. marneffei is generally susceptible to miconazole, itraconazole, ketoconazole and flucytosine, whereas amphotericin B has only intermediate fungicidal activity for this pathogen [bib_ref] Penicillium marneffei infection and recent advances in the epidemiology and molecular biology..., Vanittanakom [/bib_ref]. Current infectious disease guidelines recommend liposomal amphotericin B for 2 weeks, followed by oral itraconazole for ten weeks, followed by secondary prophylaxis. This regimen has resulted in a 97.3% cure rate in a non-randomized study [bib_ref] Epidemiology and management of penicilliosss in human immunodeficiency virus-infected patients, Sirisanthana [/bib_ref]. In our review, 3 patients who were treated initially with IV amphotericin B and later discharged on itraconazole had complete resolution. Our patient succumbed to this invasive fungal infection despite 6 months of intensive antifungal therapy. Initial therapy consisting of 30 days of amphotericin B plus isavuconazole achieved an excellent response. Unfortunately, the infection recurred 3 months later. Voriconazole plus amphotericin B trial was started then, and ultimately, triple therapy with micafungin, voriconazole and amphotericin B was administered with no positive outcome. Although we believe isavuconazole in combination with LAMB had clinical success, time proved that we did not accomplish microbiological eradication since in his recurrence 4 months later this combination was not able to contain the invasive and progressive fungal respiratory infection, nor did any combination that we implemented thereafter. We believe the patient may have had an underlying immunosuppressive disorder that was neither HIV, hypogammaglobinemia or any other obvious hematological disorder which remained undefined even after a reasonable extensive investigation.
Although there are only a handful of T. marneffei cases described in the literature among non-HIV, non-endemic patients, this should not keep health care providers from considering the diagnosis in the appropriate clinical context. Early diagnosis and appropriate treatment should result in clinical success and microbiological eradication, but delays in a timely diagnosis may result in a fatal outcome. Penicillium spp often grow from bronchoscopic samples and are typically disregarded as contaminants, but sub-speciation should be pursued in the appropriate clinical setting so as not to miss this potentially lethal pathogen.
We speculate that our patient is not the first Vietnam veteran who has developed late reactivation of this fungus, which we suspect remains all too frequently undiagnosed. We hope this case report will raise awareness of the potential for late reactivation of this dimorphic fungus in our Vietnam veteran population and promote its inclusion in the differential diagnosis of atypical lung infections in these patients.
[fig] Figure 1: a & b: Computerized tomography of the patient's lung showing the pulmonary lesion. [/fig]
[table] Table 1: Talaromyces marneffei Infection in non-HIV, non-Endemic Infected Adult Patients. [/table]
|
Reduced Prevalence of Oral Human Papillomavirus (HPV) 4 Years after Bivalent HPV Vaccination in a Randomized Clinical Trial in Costa Rica
Title:A double blind, controlled, randomized, phase III study of the efficacy of an HPV16/18 VLP vaccine in the prevention of advanced cervical intraepithelial neoplasia (CIN2, CIN3, adenocarcinoma in situ [AIS] and invasive cervical cancer) associated with HPV16 or HPV18 cervical infection in healthy young adult women in Costa Rica NCI Sponsor signatory: NCI Protocol number: 04-C-N191 GSK Protocol number: 580299/009 (HPV-009) IND number: BB-IND 7920 (held by GSK Biologicals) Title: A double blind, controlled, randomized, phase III study of the efficacy of an HPV16/18 VLP vaccine in the prevention of advanced cervical intraepithelial neoplasia (CIN2, CIN3, adenocarcinoma in situ [AIS] and invasive cervical cancer) associated with HPV16 or HPV18 cervical infection in healthy young adult women in Costa Rica I agree:- To assume responsibility for the proper conduct of the study at this site.- To conduct the study in compliance with this protocol and with any mutually agreed future protocol amendments.
- Not to implement any changes to the protocol without agreement from the sponsors (NCI and GSK Biologicals) (Amended: 29 Mar 2005) and prior review and written approval from the Institutional Review Board (IRB) or Independent Ethics Committee (IEC), except where necessary to eliminate an immediate hazard to the subjects, or for administrative aspects of the study (where permitted by all applicable regulatory requirements).
- That I am thoroughly familiar with the appropriate use of the vaccine(s), as described in this protocol, and any other information provided by the sponsor, including, but not limited to, the following: the current Investigator's Brochure (IB) or equivalent document, IB supplement (if applicable), prescribing information (in the case of a marketed vaccine) and/or Master Data Sheet (if the Master Data Sheet exists and serves as reference document for the vaccine in the case of a marketed vaccine).
- That I am aware of, and will comply with, "Good Clinical Practices" (GCP) and all applicable regulatory requirements.
- That I have been informed that certain regulatory authorities require the sponsor (GSK Biologicals) (Amended: 29 Mar 2005) to obtain and supply, as necessary, details about the investigator's ownership interest in the sponsor or the investigational product, and more generally about his/her financial ties with the sponsor (GSK Biologicals) (Amended: 29 Mar 2005). GSK Biologicals will use and disclose the information solely for the purpose of complying with regulatory requirements.
Hence I:
- Agree to supply GSK Biologicals with any necessary information regarding ownership interest and financial ties (including those of my spouse and dependent children).
- Agree to promptly update this information if any relevant changes occur during the course of the study and for 1 year following completion of the study.
- Agree that GSK Biologicals may disclose any information it has about such ownership interests and financial ties to regulatory authorities.
- Agree to provide GSK Biologicals with an updated Curriculum Vitae and other FDA required documents.
## Glossary of terms
Adverse event Any untoward medical occurrence in a patient or clinical investigation subject, temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product.
An AE can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease (new or exacerbated) temporally associated with the use of a medicinal product.
## Blinding
A procedure in which one or more parties to the trial are kept unaware of the treatment assignment in order to reduce the risk of biased study outcomes. When the subjects, the investigator, and the sponsor staff who are involved in the treatment or clinical evaluation of the subjects and review/analysis of data are unaware of the treatment assignments, as is the case for the present trial, the study is double blind.
## Eligible
Qualified for enrolment into the study based upon strict adherence to inclusion/exclusion criteria.
## Evaluable
Meeting all eligibility criteria, complying with the procedures defined in the protocol, and, therefore, included in the according-to-protocol (ATP) analysis.
Investigational product A pharmaceutical form of an active ingredient or placebo being tested or used as a reference in a clinical trial, including a product with a marketing authorization when used in a way different from the approved form, or when used for an unapproved indication, or when used to gain further information about an approved use.
## Medical monitor
An individual medically qualified to assume the responsibilities of the sponsor (GSK Biologicals) especially in regards to the ethics, clinical safety of a study and the assessment of adverse events.
Protocol amendment ICH defines a protocol amendment as "A written description of a change(s) to or formal clarification of a protocol." This is further detailed to include a change to an approved protocol that affects the safety of subjects, scope of the investigation, study design, or scientific integrity of the study.
## Protocol administrative change
A protocol administrative change addresses changes to only logistical or administrative aspects of the study. N.B. Any change that falls under the definition of a protocol amendment (see definition above) must be prepared as an amendment to the protocol.
## Randomization
Process of random attribution of treatment to subjects in order to reduce bias of selection.
Solicited adverse event Adverse events (AEs) to be recorded as endpoints in the clinical study. The presence/occurrence/intensity of these events is actively solicited from the subject or an observer during a specified post-vaccination follow-up period.
## Subject
Term used throughout the protocol to denote an individual that has been contacted in order to participate in the clinical study, either as a recipient of the investigational product(s) or as a control.
Unsolicited adverse event Any adverse event (AE) reported in addition to those solicited during the clinical study. Also any "solicited" symptom with onset outside the specified period of follow-up for solicited symptoms will be reported as an unsolicited adverse event. [fig_ref] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations... [/fig_ref] Power
## Synopsis title
A double blind, controlled, randomized, phase III study of the efficacy of an HPV16/18 VLP vaccine in the prevention of advanced cervical intraepithelial neoplasia (CIN2, CIN3, AIS, and invasive cervical cancer) associated with HPV16 or HPV18 cervical infection in healthy young adult women in Costa Rica Study population Healthy young adult women between and including 18-25 years of age at the time of first vaccination
## Rationale
Cervical cancer is one of the most common cancers among women worldwide. Approximately 400,000 new cases are diagnosed each year, resulting in about 200,000 deaths. Cervical infection with one of a dozen or so oncogenic human papillomavirus (HPV) types is now known to be the cause of nearly all cervical tumors, with a single HPV type, HPV16, accounting for approximately half of the cases worldwide and an additional type, HPV18 accounting for an additional 10-20% of cases.
Although cytological screening programs have been in place for several decades, these programs have not been effective at controlling cervical cancer in developing countries, where the vast majority of cervical cancer cases are diagnosed. In developed countries such as the United States, where screening programs have been successful, the costs associated with screening and subsequent colposcopic evaluation and treatment are very high (>$5 billion annually). An effective vaccine against HPV, the virus that causes cervical cancer, is needed.
Pre-clinical studies have convincingly demonstrated the potential prophylactic efficacy of virus-like particle (VLP) based HPV vaccines discovered at the National Cancer Institute (NCI). The efficacy of the VLP-based vaccines in animal trials has been shown to be due to their correct three-dimensional conformation and to their ability to induce strong neutralizing antibody responses. Human safety and immunogenicity trials conducted in the United States and elsewhere have suggested that VLP-based vaccines are safe and that they induce a strong immune response. Recent data further indicate that vaccination with VLP-based HPV vaccines protects against persistent HPV infections [bib_ref] A controlled trial of a human papillomavirus type 16 vaccine, Koutsky [/bib_ref].
Based on these promising findings, the NCI and the Foundation for the Costa Rican Institute for Research and Training in Nutrition and Health (FUNIN) plan to conduct a double-blinded, controlled phase III pivotal efficacy trial in Guanacaste, Costa Rica, to evaluate the prophylactic efficacy of the HPV16/18 L1 papillomavirus structural protein VLP-based vaccine manufactured and developed by GlaxoSmithKline (GSK) Biologicals/MedImmune Inc. The vaccine is formulated with investigational adjuvant AS04 (previously SBAS4) comprised of aluminum hydroxide (Al(OH) 3 ) and monophosphoryl lipid A (MPL®).
## Rationale (continued)
The trial will enroll approximately 15,000 healthy young women ages 18-25 that will be randomly assigned to one of two arms. Participants will be vaccinated three times intramuscularly over a six month period (i.e., at 0, 1, and 6 months) with a 40 µg dose of the HPV16/18 L1 VLP vaccine or with 720 enzyme-linked immunoabsorbent assay (ELISA) units of GSK Biologicals' (Amended: 29 Mar 2005) hepatitis A vaccine, inactivated (Havrix®) prepared as an investigational formulation. Each woman will be followed with yearly clinic visits and pelvic exams for four years to obtain information on long-term adverse events (AEs) and to determine whether vaccination protects against the development of incident, histopathologically confirmed cervical intraepithelial neoplasia (CIN) 2+ (defined as advanced cervical intraepithelial neoplasia that includes CIN2, CIN3, adenocarcinoma in situ , and invasive cervical cancer) associated with HPV16 or HPV18 cervical infections. The six secondary objectives of this trial are listed below. In keeping with the mission of NCI to produce hypothesisgenerating data, numerous exploratory tertiary objectives and ancillary studies are also planned.
## Objectives primary
To demonstrate efficacy of the candidate vaccine compared with control in the prevention of histopathologically confirmed CIN2+ associated with an HPV16 or HPV18 cervical infection post dose 3 (from Month 6 to Month 48), in young adult women negative for HPV deoxyribonucleic acid (DNA) by polymerase chain reaction (PCR) at Months 0 and 6 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS, or invasive cervical cancer.
## Secondary
1. To evaluate the safety of the candidate vaccine in young adult women in the entire randomized cohort, regardless of their initial HPV16/18 DNA (by PCR) status, and also stratified according to initial status. This will be done by strict monitoring and documentation of adverse events.
2. To evaluate the duration of protection conferred by the candidate vaccine against cervical infection with HPV16 or HPV18 post dose 3 (from Month 6 to Month 48), in young adult women negative for HPV DNA (by PCR) at Months 0 and 6 for the corresponding HPV type.
3. To evaluate the efficacy of the candidate vaccine compared with control in the prevention of CIN2+ associated with cervical infection by any oncogenic HPV type, post dose 3 (from Month 6 to Month 48), in young adult women negative for HPV DNA (by PCR) at Months 0 and 6 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS, or invasive cervical cancer.
4. To evaluate the efficacy of the candidate vaccine compared with control in the prevention of histopathologically confirmed CIN2+ associated with an HPV16 or HPV18 cervical infection post dose 3 (from Month 6 to Month 48) detected within the lesional component of the cervical tissue specimen by PCR, in young adult women negative for HPV DNA (by PCR) at Months 0 and 6 for the corresponding HPV type and seronegative by ELISA at Month 0 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS, or invasive cervical cancer.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -18 -
## Objectives (continued)
5. To evaluate the efficacy of the candidate vaccine compared with control in the prevention of persistent HPV16 or HPV18 cervical infection post dose 3 (from Month 6 to Month 48), in uninfected young adult women negative for HPV DNA (by PCR) at Months 0 and 6 for the corresponding HPV type.
Persistent HPV16 or HPV18 cervical infection is defined as detection of the same HPV type (by PCR) in cervical samples from all consecutive evaluations over approximately 12 months.
6. To evaluate the vaccine immunogenicity (by ELISA and V5/J4 monoclonal antibody inhibition enzyme immunoassay in a subcohort of (Amended: 19 May 2006) 600 subjects enrolled.
## Study design
Experimental design: Phase III, single center study, active control, with two parallel groups Study site: Single center with up to five satellite sites. Sites located in one region of Costa Rica (Guanacaste Province and adjacent areas, population ~300,000) Enrollment: Population-based, ages 18-25, N= ~15,000 (of ~20,000 who will be invited, based on 2000 census data). Randomization of 7,500 women will provide a sufficient study size to evaluate the main trial objectives. HPV assessments: HPV DNA assessments on cervical cell suspensions collected for cytology at Months 0, 12, 24, 36, and 48. In addition, there will be HPV DNA assessments on cervical cell suspensions collected for cytology at 6 monthly intervals in those women who are being followed for abnormal cytology (see next bullet point). Moreover, all subjects will submit a self-collected cervicovaginal sample at Month 6.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
## Study design (continued)
Management of women with abnormal cytology: Women detected as having evidence of a low-grade squamous intraepithelial lesion (LSIL) (regardless of HPV testing results) or atypical squamous cells of undetermined significance (ASC-US) concomitant with detection of an oncogenic HPV type by Hybrid Capture II (HCII) will have cytology performed at 6 monthly intervals (rather than yearly intervals). Those whose ASC-US/oncogenic HPV positive lesion or LSIL persists for 2 evaluations (whether consecutive or intermittent) will be referred for colposcopy for evaluation and excisional treatment, if warranted. Those with 3 consecutive normal cytology evaluations will return to an annual cytology evaluation schedule. For purposes of this management strategy, women with ASC-US whose concomitant specimen tests negative for HPV will be considered normal and followed at annual intervals.
Women detected as having evidence of a high-grade squamous intraepithelial lesion (HSIL), atypical glandular cells (endocervical, endometrial, or glandular; favor neoplastic or not otherwise specified) (AGC), atypical squamous cells, cannot exclude HSIL (ASC-H), or cancer will be immediately referred for colposcopic evaluation and excisional treatment, if warranted.
At the Month 48 visit we will apply the same criteria applied throughout the trial to define women who require colposcopic evaluation. Specifically, all women with evidence of cancer, HSIL, glandular lesions, ASC-H, or persistent ASC-US+/LSIL will be referred to colposcopy. Close-out of individual study participants will only occur after colposcopy referral, for women who require such referral.
It should be noted, that after participant close-out, participants will be offered additional clinical care if needed. Case management after study close-out will be defined under a separate protocol. (Amended: 13 Nov 2008)
## Safety assessments:
- In all subjects during the 60-minute (Amended: 31 Oct 2005) postvaccination period, solicited local AEs (pain, redness, swelling and temperature at the injection site) and solicited general AEs with new onset or aggravation since vaccination (fever, fatigue, myalgia, arthralgia, gastrointestinal symptoms, headache, rash [other than urticaria] and urticaria) will be assessed.
- In 10% of subjects randomly selected, on one day from Day 3 to Day 6 post-vaccination (day of vaccination is Day 0), local and solicited general AEs will be collected during a home visit.
- In all subjects, non-serious AEs that are reported spontaneously or through solicitation at every study visit will be collected throughout the trial.
- In all subjects, serious adverse events (SAEs) and pregnancy outcomes will be collected throughout the trial.
## Study design (continued)
Immunogenicity subcohort: A subcohort (Amended: 19 May 2006) of 600 subjects enrolled, regardless of the study site at which they are enrolled, will have blood samples taken at Months 0, 6, 7, 12, 24, 36 and 48 for immunogenicity assessments. The immunogenicity subcohort will be comprised of these 600 subjects.
Note: All other (Amended: 19 May 2006) subjects enrolled will have blood drawn at Months 0, if seen at those time intervals) to allow exploratory analyses and ancillary studies by NCI.
Type of study: Self-contained, Investigational New Drug (IND).
Data collection: Conventional (paper) case report form (CRF). In most instances CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 Amendment 5 10 September 2009 -20 -the CRF will act as a source document.
Duration of the study: A total of 48 months of follow-up per subject is planned for all subjects. Although participants will each be followed for 48 months, final analysis of CIN2+ will be triggered when one of the following set of conditions is met:
- 48 months of follow-up per participant is achieved, or
- At least 30 confirmed cases of CIN2+ associated with HPV16 or HPV18 cervical infection that are eligible for the ATP analysis are observed and the last woman recruited into the study has completed a minimum follow-up period of two years, pending review and approval of the final statistical analysis plan by the DSMB. Total enrollment was 7,466, which is lower than initial estimates. The final accrual of 7,466 provides 91% power to detect a vaccine efficacy of 80% for the outcome of CIN2+. (Amended: 13 Nov 2008)
- Should analysis occur prior to completion of 48 months of follow-up (under the second scenario above), subjects will continue to be followed for the full 48 months and an annex report will be issued to describe any additional data collected.
## Number of subjects
Target 15,000 subjects enrolled and randomly assigned to one of two arms, 7,500 per group (vaccine/control). 13,000 subjects are expected to be evaluable for the primary According to Protocol (ATP) analysis (6,500 evaluable subjects per group).
## Endpoints primary
Histopathologically confirmed CIN2+ associated with HPV16 or HPV18 infection detected by PCR in the preceding cervical cytology specimen in young adult women negative for HPV DNA by PCR at Months 0 and 6 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS or invasive cervical cancer.
Preceding cervical cytology means the last cervical cytology specimen collected before the histopathology specimen was obtained.
## Secondary
## Safety
- Occurrence and intensity of solicited local AEs and occurrence (either onset or aggravation) of solicited general AEs including urticaria within 60 (Amended: 31 Oct 2005) minutes after vaccination.
- Occurrence and intensity of solicited local AEs and solicited general AEs on a 10% random subset of participants on one day from Day 3 to Day 6.
- after each vaccination and over all vaccinations combined.
- Occurrence of unsolicited AEs and SAEs throughout the entire study (Month 0 up to Month 48).
- Outcome of all pregnancies.
2. Time to occurrence of incident cervical infection with HPV16 or HPV18 (by PCR).
3. Histopathologically confirmed CIN2+ associated with infection by any oncogenic HPV type (Another way of describing this endpoint is when a subject has 2 or more cervical cytology specimens positive for the same HPV type, either HPV16 or HPV18, over an approximately 12-month period without any intervening negative specimen. The term approximately is used because ATP visit intervals may vary as shown in [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref] in Section 3.11).
6. HPV16 and HPV18 ELISA and V5/J4 monoclonal antibody inhibition EIA titers in the 600 subjects enrolled into the immunogenicity subcohort.
# Background
DCEG has a longstanding interest in studies of HPV and cervical neoplasia. In the 1980's two case-control studies were conducted, one in the United States and one in Latin America, to examine risk factors associated with cervical cancer. The Latin American study was the first large-scale epidemiological study to demonstrate an association between HPV infection and cervical cancer [bib_ref] Human papillomavirus infection and cervical cancer in Latin America, Reeves [/bib_ref]. Subsequently, three cohort studies (two in the United States and one in Costa Rica) were established to examine the natural history of HPV and cervical neoplasia [bib_ref] Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia, Schiffman [/bib_ref] [bib_ref] Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa..., Herrero [/bib_ref] , The Atypical Squamous Cells of Undetermined Significance/Low-Grade Squamous Intraepithelial Lesions Triage Study (ALTS) ]. These studies have jointly provided critical data linking HPV infection to the development of cervical cancer and its precursor lesions (known cytologically as LSIL for low-grade squamous intraepithelial lesions and HSIL for high-grade SIL; and known histopathologically as condyloma and CIN1-3 for cervical intraepithelial neoplasia grades 1-3 -see . In fact, it is now understood that LSIL is the cytological manifestation of HPV infection [bib_ref] Epidemiologic Support for a Simplified View of Cervical Carcinogenesis, Schiffman [/bib_ref]. It is also known that while the majority of HPV infections are self-limited, persistent infection with one of about 15 oncogenic HPV types is required for the development of nearly all cases of cervical cancer and its immediate precursor, HSIL [bib_ref] Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa..., Herrero [/bib_ref] [bib_ref] International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic..., Munoz [/bib_ref] [bib_ref] and the International biological study on cervical cancer (IBSCC) study group. Prevalence..., Bosch [/bib_ref]. HPV16 is the most important type, accounting for about half of all cancer cases, while HPV18 accounts for another 10-20% of all cases [bib_ref] and the International biological study on cervical cancer (IBSCC) study group. Prevalence..., Bosch [/bib_ref]. Available evidence suggests that infections with different HPV types are independent and that infection with one type does not modulate infections with other HPV types [bib_ref] Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa..., Herrero [/bib_ref] [bib_ref] A prospective study of human papillomavirus (HPV) type 16 DNA detection by..., Liaw [/bib_ref]. Prevention of CIN and cervical cancer (one of the most common cancers among women worldwide) is therefore theoretically possible via prevention of oncogenic HPV infection and/or its associated disease.
One of the cohort studies sponsored by DCEG is being conducted in Guanacaste, Costa Rica, among 10,000 adult women [bib_ref] Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa..., Herrero [/bib_ref] [bib_ref] Design and methods of a population-based natural history study of cervical neoplasia..., Herrero [/bib_ref] [bib_ref] HPV DNA testing in cervical cancer screening: results from women in a..., Schiffman [/bib_ref]. Started in 1993, this population-based cohort was designed to elucidate the origins of HSIL and cervical cancer. Response rates during the enrollment phase of this study were 93% and over the past seven years close to 95% of women selected for active follow-up have continued to participate. This makes it an attractive region in which to evaluate the efficacy of a vaccine developed to prevent HPV infection and SIL, not only because of the high compliance of this rural population, but also because of the high rates of cervical neoplasia in Guanacaste (despite rates of HPV infection comparable to those in the USa phenomenon likely due to differences between the two countries in screening frequency and quality) and our understanding of the expected distribution and rates of HPV infections and cervical neoplasia.
## Figure 1 the cervical cancer continuum
In parallel with the epidemiological work conducted by DCEG in the 1980's and 90's, investigators in the Center for Cancer Research (CCR) at NCI developed an in vitro method to produce papillomavirus structural proteins (L1 proteins) capable of selfassembling into viral particles (called L1 virus-like particles or L1 VLPs) [bib_ref] A virus-like particle enzyme-linked immunosorbent assay detects serum antibodies in a majority..., Kirnbauer [/bib_ref]. These particles consist of the outer coat of the virus in its correct threedimensional configuration but are free of nucleic acids and are therefore noninfectious. When used as vaccines in animal studies, L1 VLPs were shown to induce high levels of neutralizing antibodies against the papillomavirus L1 structural protein and to protect against direct experimental challenge with the virus [bib_ref] Immunization with viruslike particles from cottontail rabbit papillomavirus (CRPV) can protect against..., Breitburd [/bib_ref] [bib_ref] Viruslike particles of bovine papillomavirus type 4 in prophylactic and therapeutic immunization, Kirnbauer [/bib_ref]. It has further been demonstrated that protective antibody levels are seen for at least one year following vaccination and that systemic inoculation induces not only circulating antibodies but also antibodies in the genital tract [bib_ref] Immunization with viruslike particles induces long-term protection of rabbits against challenge with..., Christensen [/bib_ref] [bib_ref] Human papillomavirus type 11 (HPV-11) neutralizing antibodies in the serum and genital..., Lowe [/bib_ref] ]. There are currently three animal models (rabbits, cattle, and dogs) supporting a prophylactic effect of vaccination with papillomavirus L1 VLP. VLP antibody-mediated protection from infection was type-specific both in the animal vaccine studies and in in vitro assays using HPV VLPs.
Based on these initial pre-clinical studies, safety and immunogenicity studies were initiated in humans. the primary target of these vaccines since it is known to be associated with half of all cervical cancer cases detected worldwide. HPV18, the other oncogenic HPV type included in some of the vaccines tested to date accounts for an additional 10-20% of cervical cancer cases worldwide. HPV types 6 and 11 are non-oncogenic HPV types that are responsible for genital warts. The presence of these two types should not impact cervical cancer rates.
Without exception, results from early phase human trials have indicated that VLP-based vaccines are well tolerated and that they induce high levels of neutralizing antibodies [bib_ref] A controlled trial of a human papillomavirus type 16 vaccine, Koutsky [/bib_ref] [bib_ref] Safety and immunogenicity trial in adult volunteers of a human papillomavirus type..., Harro [/bib_ref] [bib_ref] A Phase 1 study of a recombinant viruslike particle vaccine against human..., Evans [/bib_ref]. Antibodies have been detected both systemically and in the genital tract of vaccinated women [bib_ref] Specific antibody levels at the cervix during the menstrual cycle of women..., Nardelli-Haefliger [/bib_ref] ]. Furthermore, initial results from a phase IIb trial of an HPV16 VLP vaccine conducted by Merck Pharmaceuticals has recently demonstrated that vaccination effectively protects against 6 month persistent infection with HPV16 with a high level of efficacy [bib_ref] A controlled trial of a human papillomavirus type 16 vaccine, Koutsky [/bib_ref]. The antigen concentration and adjuvant for the vaccine formulation advanced to phase III development was selected mainly based on these clinical studies. In summary, the vaccine at all antigen concentrations and formulations (with or without various adjuvants), when administered as 3 doses over 6 months, was well-tolerated and induced potent immune responses, measured as neutralizing antibodies against HPV16 and HPV18. A formulation containing 20 µg of each of the two HPV types (40 µg total) plus the adjuvants aluminum hydroxide (Al(OH) 3 ) and 3-deacylated monophosphoryl lipid A (MPL®) (in combination designated AS04) elicited the highest geometric mean titers of virus-specific antibody while retaining a favorable reactogenicity profile. This vaccine candidate was used in a recently completed phase IIb trial and will be used in the Costa Rican phase III efficacy trial.
As indicated above, the candidate HPV16/18 VLP vaccine that will be used in our trial is formulated with an investigational adjuvant (AS04). AS04 is an adjuvant that is not used in any licensed vaccines. Nevertheless, GSK Biologicals has compiled considerable information regarding its safety during clinical trials of other vaccines. The information in this section reflects knowledge that was current at the time of study start. Please refer to the most recent edition of the Investigator Brochure for a review of clinical and preclinical studies (Amended: 13 Nov 2008).
## The need for an hpv vaccine
Cervical cancer is the second or third most common female tumor worldwide. In many developing countries, where rates are highest, cervical cancer is the most common malignancy diagnosed among women. Over 400,000 new cases of cervical cancer are diagnosed each year, resulting in approximately 200,000 deaths. While Pap smear screening programs in developed nations have been shown to be effective at reducing the incidence of invasive cervical cancer, difficulties remain in providing underserved populations within these developed countries with access to routine cytological screening programs. Also, in developing countries, it has proven difficult to implement effective screening programs despite numerous attempts over the past 30+ years. In Costa Rica, for example, rates of disease have remained high despite screening of approximately 80% of adult women in the country [bib_ref] Determinants of the geographic variation of invasive cervical cancer in Costa Rica, Herrero [/bib_ref]. Furthermore, even in developed nations, where cervical cancer rates have dropped significantly since Pap smear screening programs were implemented, the cost associated with these efforts is large. As an example, in the United States, the cost associated with Pap smear screening and subsequent colposcopic follow-up and treatment is greater than five billion dollars annually . An effective HPV vaccine could, in the long run, prove to be a more effective way to reduce cervical cancer incidence and mortality in developing countries, and also to reduce the costs associated with cervical cancer prevention efforts in the United States and other developed and industrialized nations.
## Defining cervical abnormalities
Before describing the proposed efficacy trial itself, it is important that we carefully define the terminology used to classify women with respect to degree of cervical abnormality. Based on our understanding of the natural history of cervical cancer, we know that cervical cancer begins with infection with one of about 15 oncogenic HPV types [bib_ref] International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic..., Munoz [/bib_ref] , IARC Working Group on the Evaluation of Carcinogenic Risks to Humans, 1995]. As mentioned above, most of these HPV infections are transient but a small proportion of infections persist and are at risk of progressing to more severe disease [bib_ref] Persistence of type-specific human papillomavirus infection among cytologically normal women, Hildesheim [/bib_ref] [bib_ref] Persistent genital human papillomavirus infection as a risk factor for persistent cervical..., Ho [/bib_ref] [bib_ref] Relation of human papillomavirus status to cervical lesions and consequences for cervical-cancer..., Nobbenhuis [/bib_ref].
Using the latest cytological Bethesda 2001 classification system [bib_ref] for the Forum Group Members and the Bethesda 2001 Workshop. The 2001..., Solomon [/bib_ref] In addition to these LSIL lesions, equivocal lesions (called ASC-US for atypical squamous cells of undetermined significance) are commonly diagnosed. It is now possible to classify ASC-US lesions into the benign reactive atypias or true SILs by using oncogenic HPV testing [bib_ref] ALTS Study Group. Comparison of three management strategies for patients with atypical..., Solomon [/bib_ref].
Persistent oncogenic HPV infections can progress to more worrisome high-grade disease known as high-grade squamous intraepithelial lesions (HSIL), usually within a period of many years (5-10 years, on average) [IARC Working Group on the Evaluation of Carcinogenic Risks to . HSIL, a cytological diagnosis, comprises the histopathological diagnoses of CIN2 and CIN3. These lesions are often, but not always, preceded by abnormal cytology [IARC Working Group on the Evaluation of Carcinogenic Risks to [bib_ref] A cohort study of the risk of cervical intraepithelial neoplasia grade 2..., Koutsky [/bib_ref]. These high-grade lesions, which include carcinoma in situ (CIS), are the immediate precursors to cervical cancer.
Based on this understanding of the natural history of cervical cancer, we propose to use the following cytological classification to guide clinical management of participants in our trial:
1. Normal: Normal cytology (including benign reactive changes) or atypical squamous cells of undetermined significance (ASC-US) that tests negative for oncogenic HPV types.
2. ASC-US+/LSIL: LSIL by cytology or ASC-US that tests positive for oncogenic HPV types.
3. HSIL: HSIL by cytology. Atypical squamous cells, cannot exclude HSIL (ASC-H) will be included in this group due to their high risk of underlying HSIL. Atypical glandular cells (AGC) and AIS, rare forms of cervical glandular abnormalities, will also be included in this group due to their high risk of progression.
4. Cancer: Cancer by cytology (very few, if any, cancers are expected in our cohort of young women).
Additional details regarding our clinical management plan are provided in Section 3.15
For purposes of defining endpoints for our primary and secondary analyses, histopathological criteria will be used. Only histopathologically confirmed CIN2+ will be considered a true outcome for analysis purposes. Additional details regarding our definition of trial endpoints are provided in Section 3.13.
## Summary
We now know that HPV is involved in the pathogenesis of nearly all cervical tumors and that HPV16 and HPV18 combined account for 60% or more of all cervical cancer cases worldwide. Pre-clinical and safety/immunogenicity human trials (in >3,000 participants) have shown that vaccination against papillomaviruses using VLP-based vaccines protects against infection and tumor development in animals, and that the vaccines are both safe and immunogenic in humans. Results from proof-of-principle trials also indicate that VLP-based vaccines are effective at protecting against persistent infection in humans. While alternative methods exist to control cervical cancer (i.e., Pap smear screening), these methods have been unsuccessful in developing countries. In the few developed countries in which screening has been successful, the costs associated with these efforts are high. The development of an HPV vaccine that protects against the development of cervical cancer and its precursors is therefore desirable. We propose to conduct a pivotal efficacy trial to evaluate the effectiveness of the HPV16/18 VLP-based vaccine manufactured by GSK Biologicals to protect against the development of histopathologically confirmed, incident CIN2+ associated with either HPV16 or HPV18 infections. The trial will be conducted in Guanacaste, Costa Rica, an area with high rates of cervical cancer and the site of an ongoing NCI-sponsored 10,000-woman, populationbased natural history study.
## Study objectives
## Primary objective
The main objective of the present randomized clinical trial is to:
Demonstrate efficacy of the candidate vaccine compared with control in the prevention of histopathologically confirmed CIN2+ associated with an HPV16 or HPV18 cervical infection post dose 3 (from month 6 to month 48), in young adult women negative for HPV DNA (by PCR) at months 0 and 6 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS, or invasive cervical cancer. Refer to Section 3.13.1 for the definition of the primary endpoint.
## Secondary objectives
In addition to the primary objective listed above, this trial will address a few selective secondary objectives of importance to determine potential clinical uses of the vaccine. These secondary objectives are to:
1. Evaluate the safety of the candidate vaccine in young adult women in the entire randomized cohort, regardless of their initial HPV16/18 DNA (by PCR) status, and also stratified according to initial status. This will be done by strict monitoring and documentation of adverse events.
2. Evaluate the duration of protection conferred by the candidate vaccine against cervical infection with HPV16 or HPV18 post dose 3 (from month 6 to month 48), in young adult women negative for HPV DNA (by PCR) at months 0 and 6 for the corresponding HPV type.
3. Evaluate the efficacy of the candidate vaccine compared with control in the prevention of CIN2+ associated with cervical infection by any oncogenic HPV type, post dose 3 (from month 6 to month 48), in young adult women negative for HPV DNA (by PCR) at months 0 and 6 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS or invasive cervical cancer.
4. Demonstrate efficacy of the candidate vaccine compared with control in the prevention of histopathologically confirmed CIN2+ associated with an HPV16 or HPV18 cervical infection post dose 3 (from month 6 to month 48) detected within the lesional component of the cervical tissue specimen by PCR, in young adult women negative for HPV DNA (by PCR) at months 0 and 6 for the corresponding HPV type and seronegative by ELISA at month 0 for the corresponding HPV type.
CIN2+ is defined as CIN2, CIN3, AIS, or invasive cervical cancer.
## Tertiary objectives
The NCI and Costa Rican researchers are committed to addressing several other objectives of interest to the scientific community. While a comprehensive listing of all tertiary objectives and analyses is beyond the scope of this protocol, to follow is a summary of some of the topics that are being considered for exploratory evaluation by NCI and Costa Rica. [Note: It is understood that the power to address some of these objectives may be limited].
## Ancillary analyses & studies
The present clinical trial provides a unique opportunity for NCI, Costa Rican, and other investigators to answer various other immunological and natural history questions using information and materials collected as part of this trial. Opportunities might also arise to conduct ancillary studies that require the collection of additional information and/or biological specimens not planned for as part of the current investigation. Although the exact nature of these ancillary analyses and studies is currently unknown, we plan to set up a mechanism by which investigators can propose to perform ancillary studies within the context of our trial. These ancillary studies will be judged on their scientific merit and on their potential impact on the primary and secondary objectives of our trial. Efforts will be made to ensure that no ancillary study interferes with our ability to successfully evaluate the primary and secondary objectives listed above.
Any ancillary study or protocol requiring contacts with the participants not described in the current protocol or requiring collection of additional specimens will be submitted for review by the IRBs involved.
Examples of ancillary analyses and studies being considered at present include:
1. A family-based study to examine host genetic factors involved in the immunological response to HPV vaccination.
## 2.
A nested case-control study to evaluate the risk factors associated with the development of "rapid onset" HSIL among young women.
3. A longitudinal study of women who are sexually inexperienced (virginal) at study entry and who subsequently become sexually active, to examine the early natural history of HPV infection in unvaccinated women and the host immune response to these early infections.
# Methods
A detailed operating procedures manual called the "Field Procedures Manual" has been written for this study. The Field Procedures Manual has been reviewed to assure that it reflects the procedures described in this protocol. The investigators, study staff, and site monitors will be trained to use the Field Procedures Manual.
## Overview of the trial
Approximately 20,000 women will be invited to participate in a double blinded, randomized clinical trial to evaluate whether vaccination with the bivalent HPV16/18 VLP-based vaccine manufactured by GSK Biologicals protects against the development of histopathologically confirmed, incident CIN2+. Women between 18-25 years of age at entry who are residents of Guanacaste Province or surrounding areas will be eligible. Participants will be required to not be pregnant at the time of vaccination, to agree to use an effective birth control method 30 days before vaccination until 60 days after the last vaccination (approximately 9 months total), and to be in good health as determined by an overall physical examination administered by a study physician at entry into the trial.
Women will be identified through a census of the region performed in February/July 2000, and will be invited to participate in the trial by visiting one of the study clinics set up specifically for this trial. While randomization (into HPV or HAV vaccination) of 15,000 women was initially anticipated, randomization of 7,500 women will provide a sufficient study size to evaluate the main trial objectives based on revised statistical assumptions, see Section 4.7. (Amended: 13 Nov 2008)
Selected women who are eligible and agree to participate will be given a pelvic examination (if they report previous sexual activity) to collect baseline cervical specimens prior to vaccination, and will be administered 3 doses of a 0.5 mL vaccine containing either HAV (control vaccine) or 40 µg HPV16/18 VLP (20 µg each of HPV16 VLP and HPV18 VLP; VLP vaccine). The initial vaccination will be performed at entry, and additional vaccine doses will be given at one and six months following the initial vaccination. Women will be monitored at the clinic for a minimum of 60 (Amended: 31 Oct 2005) minutes following each vaccination, during which study personnel will collect information on reactogenicity of the vaccine. Appropriate medical treatment will be readily available in case of a rare anaphylactic reaction. A home visit in the week after each vaccination will also be performed by a specially-trained outreach worker on a 10% random sample of participants, to collect additional information on vaccine side effects. A toll-free number will be staffed throughout the duration of the trial so that volunteers may contact the vaccine trial clinical team at any time in case of an emergency.
Long-term follow-up will include clinic visits at yearly intervals starting on the first anniversary of vaccination and continuing through the fourth year anniversary. Follow-up visits will include a pelvic examination (among sexually experienced women) to permit a careful evaluation of the cervix and determination of HPV and cytological status.
Questionnaires and biological specimens will be obtained at the enrollment and follow-up visits. The questionnaire will elicit information on risk factors thought to be associated with HPV infection, SIL, or immune response to vaccination. Biological specimens collected during the trial will include blood (plasma, serum, red blood cells, and buffy coat on all participants, and cryopreserved PBMCs on a subset of participants), cervical cells used for cytology and HPV DNA testing and other research assays, and cervical secretions used to measure levels of mucosal antibodies and other immunological/inflammatory parameters.
Each volunteer is expected to have, on average, seven clinic visits (three vaccination visits and four annual follow-up visits). A 10% random sample of women will also have three home visits performed to monitor vaccine side effects. A subset of 600 women will have a visit at month 7 to assess immunogenicity a minimum of 30 days following the last vaccination. Women diagnosed with LSIL or HPV positive ASC-US will be scheduled for clinic visits with pelvic examinations at six-month intervals to assure adequate clinical management. Women with evidence of HSIL, ASC-H, AGC (Amended: 29 Mar 2005) or persistent ASC-US+/LSIL at any time during the study will be referred for colposcopic evaluation and subsequent excisional treatment, if needed. Treatment standards will be consistent with those in the United States and Costa Rica. To assist the reader, three summary tables/figures are provided. The first summarizes the expected number of women identified, enrolled, vaccinated and followed as part of our trial. The second is an overview diagram of the trial . The third is a table depicting the timeline and procedures planned for the trial. Number of women (from census) invited to participate 19,000 -21,000 Number of eligible women who agree to be randomly assigned to one of two study arms 7,500
Number of women who will be randomly assigned to one of two study arms, will receive 3 doses of vaccine and will be evaluable in ATP analysis 5,300 Analysis (Section 4.0)
Note: There is only one set of study procedures, however, the analyses will vary depending on the initial HPV DNA status of the subject.
## Confidential
NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -33 - - is used to indicate a study procedure that requires documentation in the individual CRF. Note: The double-line borders following Months 0, 1 and 6 indicate that a home visit by a specially-trained outreach worker to monitor AEs will occur during the week following vaccination (any day from Day 3 to Day 6) in a random subset of 10% of the women. ***This blood draw will be performed on a subset of women, and at selected study visits, as follows: A) A random sample of women at 0, 12, and 36 months; B) All women referred to accelerated screening visits per management algorithm, at yearly intervals before March 31, 2009 (Amended: 10 September 2009); C) All women referred for colposcopic evaluation who require a biopsy or LEEP, at the time of the colposcopy visit; D) women enrolled in the immunogenicity subcohort after . †Procedure will not be performed on virginal women. † †Pelvic examination will be performed if indicated by the case management algorithm. The protocol-specified clinical management algorithm is summarized in Appendix A. ¶Reflex HCII HPV testing will be performed automatically for all subjects with results of ASC. HCII HPV testing will also be performed on all women who have a pelvic examination at the 0 month visit. §HCII testing for Chlamydia trachomatis and Neisseria gonorrhoeae will be performed on samples obtained at the 0 month visit (or the first visit with a pelvic examination for women who are virginal at entry) only. ^ HPV testing by HCII will be performed on all specimens obtained at the 48 month visit. #Every effort will be made to schedule study participants within the preferred Month 48 study visit window (1381 to 1740 days post Month 0). However, the window for the last visit might be extended until October 3, 2010, which is the scheduled date of last subject last visit (i.e, 1740 days after the 4 year anniversary date for the last subject first visit).
## Study population
The trial will be conducted in the Guanacaste Province of Costa Rica and surrounding areas. Potential participants were initially identified through a door-to-door census of women in the region. The census was performed by a professional survey team in February/July 2000 and obtained the following information on all female residents between the ages of 12 and 22:
- Complete name
- Date of birth
- National ID number (if available)
- Nationality
## - address
- Telephone number (if available)
- Contact person outside the household.
While women recruited into our trial will be 18-25 years of age at enrollment (see eligibility criteria below), the census enumerated women aged 12-22. This was done to account for the lag time between the census and the initiation of the trial, and to allow for the possibility of extending enrollment if necessary to achieve the desired sample size. If the number of women to be invited is modified, approval will be sought by the IRBs reviewing the trial.
## Eligibility criteria
## Inclusion criteria
To be eligible for participation in our trial, participants must fulfill all of the following inclusion criteria:
- Gender: Female
- Age: 18 to 25 years (inclusive) at the time of first vaccination (i.e., enrollment)
- Planned residence: Guanacaste Province and surroundings (including parts of Puntarenas and Alajuela Provinces) for the 6 months following first vaccination We have chosen to restrict the trial to women since the prevalence of detectable HPV infection and clinical outcomes is higher and more easily and reliably detectable in women than in men. It is understood, however, that should the vaccine prove effective, vaccination programs might include both males and females in order to achieve "herd immunity". Separate trials will be required to evaluate the vaccine in men.
Our trial is further restricted to young, adult women (18-25 years) to maximize the number of participants unexposed to HPV at entry who are likely to be exposed to the virus during the follow-up period. We know from natural history studies in Guanacaste and elsewhere that prevalence of HPV peaks in the late teens and early 20s, following initiation of sexual activity, and that LSIL rates peak approximately 1-2 years later.
Restriction of our trial to the Guanacaste Province and surroundings was dictated by our understanding of this population, their cooperativeness in previous studies, and the known high rates of cervical cancer in this area. The high rates of disease in this population maximize the potential benefit to individual participants, should the vaccine prove effective. Women with definite plans to out-migrate will not be eligible to avoid loss to follow-up, since this could introduce bias into our trial.
Women who cannot speak or understand Spanish and those who are mentally incompetent will not be included in our trial because of concerns that they may not be capable of providing adequate informed consent.
Finally, to assure safety of women enrolled in our trial, we plan to collect medical history information and to perform a physical and pelvic examination (if sexually experienced) on all women at the time of their enrollment into the study. Women with a history of chronic conditions requiring treatment (such as cancer, chronic hepatic or kidney diseases, diabetes, or autoimmune disorders) will not be eligible for our trial.
## Exclusion criteria
The following criteria will be checked at the time of enrollment. If any apply, the participant will not be included in the study.
1. History of chronic condition(s) requiring treatment such as cancer, chronic hepatic or kidney diseases, diabetes, or autoimmune disorders.
2. Acute or chronic, clinically significant pulmonary, cardiovascular, neurologic, hepatic or renal functional abnormality, as determined by physical examination or previous laboratory tests carried out by the participant's doctor.
3. History of allergic reaction (e.g., difficulty breathing) after receipt of any vaccine.
4. History of vaccination against hepatitis A (the control vaccine is a hepatitis A vaccine) or a known history of hepatitis A infection.
5. History of chronic administration (>14 days) of immunosuppressants or immunemodulating drugs within 6 months prior to the first vaccine dose. (For corticosteroids, this will mean prednisone, or equivalent, ≥0.5 mg/kg/day. Inhaled and topical steroids are allowed.)
6. Are sexually experienced and of childbearing potential (i.e., not surgically sterilized), and are unwilling to use an effective method of birth control for 30 days before vaccination until 60 days after the last vaccination (approximately 9 months) 7. Uterus has been removed.
8. Use of any investigational or non-registered product (drug or vaccine) other than the study vaccines within 30 days preceding the first dose of study vaccine, or planned use during the study period. 12. Hypersensitivity to latex (found in the tip cap and syringe plunger).
13. Any medically diagnosed or suspected immunodeficient condition based on medical history and physical examination (no lab testing required).
Those women who report having a history of chronic conditions requiring treatment, an acute or chronic functional abnormality noted upon physical examination, an allergic response to any previous vaccination or to components of the vaccine/syringe, medically diagnosed or suspected immunodeficient conditions (note that this assessment is based on medical history and physical examination; no lab testing is required), or a recent history of chronic use of immunosuppressant or immune-modulating drugs will also be excluded, to assure safety of the participants. Women with a history of hepatitis A vaccination or of acute hepatitis A disease will also be excluded, since HAV will be used as the control vaccine in our trial and revaccination is not currently indicated in these groups.
Although there is no indication to date that the HPV16/18 VLP vaccine might adversely affect pregnant women or their unborn children, the possible risks to pregnant women and unborn babies are not known. Therefore, we intend to exclude from the trial sexually experienced women who do not agree to use an effective form of birth control from 30 days before vaccination until 60 days after the last vaccination (approximately 9 months total). Acceptable forms of birth control include abstinence; surgical sterilization; oral contraceptives; the intrauterine contraceptive device; a diaphragm or condom; or injectable contraceptives including implants, monthly injectables, and three-monthly injectables.
Women without an intact uterus will be excluded from the trial since they are not at risk of developing cervical HPV infection or CIN2+.
Finally, women who report recent (within 30 days) or planned (during the study period) use of an investigational or non-registered product (drug or vaccine, including MPL, AS04, and HPV vaccine) will be excluded. This will be done to avoid having an adverse event occurring in response to one of these other agents being confused with an adverse event resultant from study vaccination. - Are pregnant, less than three months post-partum, or breastfeeding at the time of enrollment/vaccination visit. If the participant reports a menstrual delay and her pregnancy test is negative, she will be deferred for a minimum of seven days. Women with a second negative pregnancy test after the minimal deferral period would be eligible for enrollment/vaccination. A menstrual delay is defined as a delay of more than five days for women with regular menses and more than 35 days for women with irregular menses. (Amended: 31 Oct 2005)
- Are sexually experienced and using an effective method of birth control (including abstinence) for less than 30 days at the time of enrollment/vaccination visit.
(Amended: 31 Oct 2005)
- Are sexually experienced but cannot have a pelvic exam performed due to heavy bleeding (menstruation or other) or heavy vaginal discharge. This criteria applies only to sexually experienced women at enrollment or those who require a pelvic exam at the Month 6 visit. (Amended: 31 Oct 2005)
Women who have received immunoglobulins (within 90 days of enrollment) and/or other vaccinations (within 30 days of enrollment) or who have current acute illnesses will be deferred to avoid confounding of our findings by these other vaccinations/conditions (for example, an adverse event to another vaccination).
Although there is no reason to believe that HPV16/18 VLP vaccination would adversely affect pregnant women or their unborn child, the possible risks to pregnant women and unborn babies are not known. Therefore, we have chosen to defer enrollment of pregnant women. A sensitive pregnancy test will be performed at the study clinic prior to each vaccination, and women found to be pregnant will have their enrollment deferred until they are at least three months post-partum and no longer breastfeeding. It should be noted that women who are found to be pregnant at the time of their one or six month vaccinations, despite use of an acceptable form of birth control, will not receive their subsequent vaccinations and will have their follow-up pelvic examinations and other follow-up activities deferred until they are at least three months post-partum.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -41 -Sexually experienced women who are not pregnant but report not using an acceptable form of birth control (including abstinence; surgical sterilization; oral contraceptives; the intrauterine contraceptive device; a diaphragm or condom; or injectable contraceptives including implants, monthly injectables, and three-monthly injectables) for at least 30 days will be deferred until they begin practicing birth control.
Women will also be deferred if, for any reason, a pelvic examination cannot be performed at the time of the initial visit. A pelvic examination will allow us to determine whether a woman has prevalent HPV infection and/or SIL at the time of initial vaccination. Although initial cytological and virological status will not be used to define a woman's eligibility for participation in the trial, it is important that cytological and HPV status at enrollment be obtained for sexually experienced women since the vaccine may not protect against an already established infection/lesion. It can be assumed that virginal women are unexposed to HPV and not at risk of SIL. The low rates of exposure to HPV among virginal women in Guanacaste have been confirmed by serological testing of self-reported virginal women in our natural history study.
## Elimination criteria from the atp analyses during the study
For the primary according to protocol (ATP) analysis planned for our study, the following elimination criteria may apply. These are also discussed in the statistical analysis section of this protocol (Section 4.2).
- Women who do not receive all three doses of the vaccine within the protocolspecified intervals (see [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref].
- Women who are found to be HPV16 and HPV18 DNA positive (for the corresponding type considered in the analysis) during the vaccination phase, by clinician and/or self-administered collection (months 0 or 6) (Amended: 10 September 2009).
- Women who indicate chronic administration (defined as more than 14 days) of immunosuppressants or other immune-modifying drugs occurring in the 90 days prior to blood sampling. For corticosteroids, this will mean prednisone, or equivalent, ≥0.5 mg/kg/day. Inhaled and topical steroids are allowed. It shall be noted that concomitant medications administered during this study will be collected only in the context of assessing SAEs. In specific, we will collect information on medications that are taken in the month preceding an SAE and those that are taken to treat an SAE if, in the investigator's judgement, the medication used to treat the SAE can help clarify the diagnosis of the SAE or if the medication is an immunosuppressant.
## Contraindications to subsequent vaccination
The following events constitute a contraindication to further vaccination. If any of the events listed below occur, women will not receive further doses of the vaccine but will continue to be followed for the full four years or until resolution of the AE, whichever occurs last. The decision to discontinue vaccination will be carefully documented on the appropriate CRF, along with the date and reason for vaccine discontinuation. Women CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -42 -who fail to receive one or more doses of vaccine will not be included in the ATP analysis but will be included in the ITT analysis (see Section 5).
Hypersensitivity reaction following vaccine administration
## - pregnancy
- Any medically diagnosed or suspected immunodeficient condition based on medical history or physical examination (no lab testing required)
- Any other acute or newly acquired chronic condition at the time of scheduled vaccination as determined by the medical history, which in the opinion of the study doctor precludes further administration of the study vaccine
- Any SAE judged to be possibly, probably, or definitely related to study vaccine
- Other significant reactions including severe pain, severe swelling, severe limitation of motion, persistent high fever, severe headache or other systemic or local reactions, which in the opinion of the study doctor preclude further administration of the study vaccine
## Trial arms & randomization
The present trial will be a double blinded, randomized trial consisting of two arms. Women randomly assigned to the experimental arm will receive the HPV16/18 VLP vaccine while those randomly assigned to the control arm will receive the HAV vaccine, as described more fully in Section 3.6, Section 3.7, and Section 3.10. Women who agree to visit one of the study clinics and who fulfill the eligibility criteria described earlier will be randomly assigned to one of the two trial arms at an overall ratio of 1:1. Because the HPV16/18 VLP and HAV vaccines are indistinguishable in their appearance, we are able to and will mask investigators and participants to trial arm assignment.
A blocked randomization procedure will be used, with random block sizes used to avoid the possibility of unmasking at the clinic site. Details of the vaccine labeling and subject randomization process are provided in Appendix B.
Since knowledge of vaccine arm assignment would not affect treatment of participants who have an SAE, no provision is made to allow for unblinding on site in Costa Rica. Unblinding of individual participants, at the request of the DSMB, the IRBs, the GSK Global Clinical Safety Department or the NCI Medical Monitor as part of reporting requirements to the FDA (e.g., for rapid reporting of an unexpected SAE associated with vaccination) will be allowed. This individual unblinding will be performed in a manner that assures that the overall study blinding is maintained, that staff involved in the conduct, analysis, or reporting of the study remain blinded except where absolutely impractical, and that any unblinding be appropriately documented (see also Section 3.16.3, Exiting Women from the Trial) (Amended: 10 September 2009). Treatment codes will be kept at NCI and GSK under controlled/secured access.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -43 -
## Study vaccines
The candidate HPV16/18 VLP-based vaccine to be used in our trial has been manufactured by GSK Biologicals. The control vaccine (investigational formulation of Havrix hepatitis A vaccine -HAV) has been developed and manufactured by GSK Biologicals. The Quality Control Standards and Requirements for the candidate vaccine and the control vaccine are described in separate release protocols and the required review and agreement of the FDA will be obtained before initiation of the trial.
The candidate HPV16/18 virus-like particle (VLP)/AS04 vaccine will be supplied as a liquid in mono-dose syringes. HPV16/18 VLPs at a concentration of 40 µg/0.5 mL are formulated with AS04 adjuvant. AS04 is an investigational adjuvant comprised of 500 µg of aluminum hydroxide (Al(OH) 3 ) and 50 µg of monophosphoryl lipid A (MPL). The control HAV vaccine will be supplied as a liquid in mono-dose syringes and will have an indistinguishable appearance to the HPV16/18 VLP vaccine. Inactivated hepatitis A viral antigen at a concentration of 720 ELISA units/0.5 mL is formulated with 500 µg aluminum salts (as hydroxide: AL(OH) 3 ), 0.5% 2-phenoxyethanol (w/v), and contains trace amounts of neomycin sulfate (no more than 20 ng per 0.5 mL dose).
Three lots of HPV16/18 VLP vaccine have been prepared to support the Phase III efficacy trial. Each HPV16/18 VLP vaccine lot is manufactured using one single batch of HPV16 L1 VLP antigen and one single batch of HPV18 L1 VLP antigen. Four lots of an investigational formulation of Havrix control vaccine have been prepared to support the Phase III efficacy trial. Formulation of the investigational Havrix control vaccine lots was performed using one batch of purified inactivated Hepatitis A antigen.
It should be noted that after completion of this trial and under a separate protocol, participants will be offered cross-over vaccination. This will include offering the licensed formulation of Havrix (1440 ELISA units/0.5 mL) hepatitis A vaccine to women who received the HPV16/18 VLP vaccine and the licensed HPV16/18 VLP vaccine to women who received the investigational hepatitis A vaccine (Amended: 10 September 2009). In addition, all women will be offered vaccination with Engerix-B® (Hepatitis B Vaccine ; Manufactured by GlaxoSmithKline Biologicals, Rixensart, Belgium).
## In order to reduce clinic visits, women who are eligible to recieve both the hepatitis a vaccine and hepatitis b vaccine will be offered vaccination with twinrix® [hepatitis a inactivated & hepatitis b (recombinant) vaccine] (amended: 10 september 2009).
The Material Safety Data Sheets (MSDSs) for HPV16/18 VLP vaccine and the investigational formulation Havrix that will be administered during this study are provided as Appendix E of this protocol.
At the time of study start, five Phase I-IIb clinical trials were conducted that evaluated vaccine safety.(Amended: 13 Nov 2008). During these trials, the 40 µg HPV16/18 VLP AS04 vaccine (i.e., the formulation planned for use in the current trial) was administered to approximately 690 young women living in Brazil, Canada or the United States who ranged in age from 15 to 30 years. No Serious Adverse Events associated with this formulation were considered by the investigators to be related to vaccination.
The single largest study evaluating HPV16/18 VLP vaccine safety was GSK study HPV-001. In this double-blind randomized trial, 1113 subjects (560 vaccine, 553 placebo) were evaluated. Subjects who received vaccine reported slightly higher rates of solicited local adverse events (injection site pain, redness, and swelling) compared with the placebo group; however, the rates of solicited general adverse events were similar between the vaccine and placebo groups. Overall, most solicited local and solicited general adverse events were low-grade in intensity and transient. Rates of unsolicited symptoms with onset within 30 days after each vaccination were similar in both groups.
The results of these clinical trials suggest that the HPV16/18 VLP vaccine is safe and well tolerated. However, with the administration of any vaccine there may be unanticipated allergic reactions (which may include rash and/or other allergic manifestations) and, as with the administration of any foreign protein, there may be rare occurrences of anaphylactic reaction.
## Selection of an investigational formulation of havrix as the control vaccine
The purpose of an active control is to confer benefit to recipients while maintaining the study blind. Therefore, the control vaccine must match the HPV16/18 VLP candidate vaccine in appearance, reactogenicity, injection volume, and number of doses. Moreover, the control vaccine must be manufactured by GSK Biologicals to enable labeling (for blinding and randomization). Another important consideration is limited prior use of the active control vaccine in the population targeted for enrollment, since individuals previously vaccinated with the active control vaccine would be ineligible for our study.
In addition, Costa Rica, and particularly Guanacaste, is considered an intermediate to high incidence area for hepatitis A and a low risk area for hepatitis B .
Based on the above criteria, and in consultation with the trial DSMB, we have selected an investigational formulation of Hepatitis A vaccine (HAV) called Havrix as the control vaccine for our trial. A 3-dose, 720 ELISA unit (EL.U.) formulation of Havrix is registered for use in Costa Rica, although it has been supplanted by a more convenient 2dose 1440 EL.U. formulation in recent years. All marketed Havrix products have a 1.0 mL injection volume. To maintain the blind, lots of Havrix 720 EL.U. investigational formulation will be prepared with a 0.5 mL injection volume; they will be labelled "For Investigational Use Only". The safety, immunogenicity, and efficacy of the Havrix investigational formulation is expected to match that of the Havrix 720 EL.U. formulation presently registered in Costa Rica.
In the United States, Havrix is licensed for use in children and adults. In adults, the vaccine is administered as a single dose of 1440 ELISA units (EL.U.) adsorbed on 0.5 mg Alum in 1.0 mL, followed by a booster dose 6 to 12 months after the primary dose.
There is a sufficient body of data to support that a 720 EL.U. dose of Havrix, administered to adults on a 0, 1, 6 month schedule, will be safe and immunogenic. This data is summarized below. surveillance data indicate that this dose and schedule of hepatitis A virus antigen is both safe and immunogenic in adults. It is reasonable to expect that the same dose and schedule of Havrix (720 EL.U. given on a 0, 1, 6 month schedule) administered as a monovalent vaccine will be safe and immunogenic in adults.
In summary, there is a preponderance of evidence to suggest that investigational Havrix 720 EL.U. in 0.5 mL will be safe and immunogenic when administered to adults. This makes it a suitable active control for this HPV vaccine efficacy study. The use of investigational Havrix as a control will allow this study to be conducted in a fully blinded fashion and will offer benefit for the women allocated to the control group.
## Central storage of vaccine
Vaccine syringes will be shipped from GSK Biologicals in Rixensart to the study site in Costa Rica at 4°C (range 2°C to 8°C). Once in Costa Rica, the vaccine shipment QC procedures detailed in the Field Procedures Manual will be followed to ensure that the vaccine has been maintained at the proper temperature during shipment. At the Costa Rican repository, boxes containing vaccine syringes will be transferred and stored in a walk-in refrigerator located at the central study repository and kept at 4°C (range 2°C to 8°C) and will never be frozen. The walk-in refrigerator will be equipped with temperature monitors, with a back-up refrigeration unit and generator, and will be monitored on a daily basis by trained staff. Access to the vaccine storage facility will be restricted to authorized personnel only.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -46 -Deviations in temperature beyond those allowable by protocol that occur either during vaccine storage, vaccine transport or vaccine use will be documented and the NCI principal investigator or his designee will be notified. NCI will also notify GSK Biologicals regarding such violations. Vaccines affected by such deviations will not be used for the trial without joint approval by NCI and GSK Biologicals.
## Concomitant medications/vaccination
We plan to provide women in our rural population with two 500 mg acetaminophen pills after each vaccination with an indication to take the medication in case of a mild fever or pain. The use of this acetaminophen during the trial will be followed in the 10% random subset of women for whom a home visit is performed on Day 3 to Day 6 following each vaccination. Also, concomitant medications administered during this study will be collected in the context of assessing AEs. In specific, we will collect information on medications/vaccinations that are taken in the month preceding an AE and those that are taken to treat an AE if it is felt that the medication used to treat the AE can help clarify the diagnosis of the AE or if the medication is an immunosuppressant. (Amended: 29 Mar 2005)
## Vaccination route and schedule
Participants will be vaccinated intramuscularly in the deltoid muscle (upper arm) and will receive three doses at 0, 1, and 6 months, as summarized in [fig_ref] Table 3: Vaccine Dosage and Administration [/fig_ref] below. Three clinic visits are planned for vaccination at 0, 1, and 6 months. At the first clinic visit, we will collect information for determining eligibility, ensure that written informed consent has been obtained prior to any study procedure, perform urine pregnancy testing, assign treatment arm by randomization, give the initial vaccination, and monitor women for at least 60 (Amended: 31 Oct 2005) minutes following vaccination with appropriate medical treatment readily available in case of a rare anaphylactic reaction. At the two subsequent visits at 1 and 6 months, information on AEs will be collected, subsequent vaccinations will be administered, and women will be monitored for a minimum of 60 (Amended: 31 Oct 2005) minutes following vaccination with appropriate medical treatment readily available in case of a rare anaphylactic reaction. After each vaccination visit, women will be provided with two 500 mg acetaminophen pills with an indication to take the medication in case of a mild fever or pain. The use of acetaminophen during the trial will be followed in the 10% home visit subset. A summary of the procedures performed at each of these visits is presented in. Intervals between study visits during the vaccination phase of the trial are provided in [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref].
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
Blood specimens will be collected at each of these three timepoints to permit evaluation of antibody titers at the time of each vaccination and to allow for research-based testing. Blood collected at entry will also be used to measure hematocrit levels, as a service to participants. Cervical secretions and cervicovaginal cells will be collected from nonvirginal women at entry and at the six-month visit, to assess HPV status at the beginning and end of the vaccination schedule and to allow for research tests. Cervicovaginal cells collected at the first pelvic examination will also be used for Chlamydia trachomatis and Neisseria gonorrhoeae testing using the Hybrid Capture test (Digene Corp., Gaithersburg, MD, USA). A ThinPrep® (a standardized cervical specimen preparation method; Manufactured by Cytyc Corporation, Boxborough, MA, USA) Pap smear will also be prepared (non-virginal women only) before the first vaccination, to assess cytological status at entry. An interview will be conducted at entry and at the six-month visit to assess exposure to known and suspected risk factors for HPV infection and cervical cancer.
To permit characterization of maximal antibody titers achieved after all three doses of vaccine are administered, a blood draw is also planned a minimum of 30 days after the third vaccination (7 month visit) on a subset of (Amended: 19 May 2006) 600 women randomly assigned to one of the two arms of our trial. vaccination serves as the reference date for intervals between study visits. Also, for Intervals 1 and 2, the length of interval listed represents the desired interval. Broader, allowable intervals are defined in. (Amended: 31 Oct 2005) #Every effort will be made to schedule study participants within the preferred Month 48 study visit window (1381 to 1740 days post Month 0). However, the window for the last visit might be extended until October 3, 2010, which is the scheduled date of last subject last visit (i.e, 1740 days after the 4 year anniversary date for the last subject first visit).
## Active follow-up of participants (months 12-48)
Long-term follow-up is planned for a period of four years from the time of entry into the trial. Follow-up will be performed to monitor trial outcomes, to collect additional information on long-term AEs, and as part of NCI-sponsored ancillary studies to evaluate issues of immunological and etiological interest. A summary of procedures performed during follow-up is presented in. Intervals between study visits during the followup phase of the trial are provided in [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref].
Follow-up will consist of yearly clinic visits, at which time the following activities will be completed:
- An interview to collect information on risk factors known or suspected to be associated with HPV infection, SIL, or the immune response to vaccination.
- An interview to obtain information on AEs (Amended: 29 Mar 2005), including hospitalizations, incapacitation lasting more than one day, pregnancies, and other events reported by participants. AEs and SAEs are further defined in Appendix D.
- A pelvic examination (if the participant reports previous sexual activity) to collect cervical specimens for cytological and virological evaluation. Cervical secretions will also be collected at this time for ancillary evaluation of local immune responses.
- A blood specimen will be collected for HPV antibody and other testing. On an approximately (Amended: 19 May 2006) 10% subset of women at enrollment and at the 7 (Amended: 19 May 2006), 12 and 36 month visits, additional blood for ancillary studies will be collected to permit cryopreservation of lymphocytes for immunological assessment.
Women who at any of the yearly screening visits are found to have evidence of ASC-US+/LSIL will be seen at six-month intervals (rather than yearly intervals), and will continue to be seen every six months until there is evidence of disease persistence/progression (in which case they will be referred for colposcopic evaluation) or three consecutive normal cytologies are observed (in which case they will return to a yearly screening interval). Women who at any of the clinic visits are found to have evidence or suspicion of AGC/ASC-H/HSIL/cancer will be immediately referred for colposcopic evaluation and excisional treatment, if needed (Amended: 13 Nov 2008). Participants referred for colposcopic evaluation will be seen by an expert study colposcopist. Detailed colposcopy referral and colposcopy management algorithms are provided in Appendix A. Women who are referred for colposcopic evaluation, whether treated or not, will continue to be actively followed at six month intervals using the criteria defined above for women diagnosed with ASC-US+/LSIL. CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
Women who are determined to be pregnant at the time of their scheduled follow-up visit will be deferred until they are three months post-partum. Please note that pregnancy outcomes will be documented for all pregnancies among trial participants, via participant report and review of medical charts.
A small proportion of cytology and HPV specimens is expected to be inadequate and/or unsatisfactory for evaluation (<5%). In such cases, we do not intend to resample the women until her next scheduled clinic visit. However, to ensure optimal clinical management of women with inadequate/unsatisfactory results, we will presume an LSIL cytology and HPV DNA positivity. This will effectively mean that women with inadequate/unsatisfactory results are either rescreened at 6-month intervals (if no previous history of ASC-US+/LSIL), or referred for colposcopic evaluation (if previous history of ASC-US+/LSIL).
## Trial endpoints
## Primary endpoint
The primary endpoint for our trial will be histopathologically confirmed, incident CIN2+ (defined as CIN2, CIN3, AIS, or invasive cervical cancer) associated with an HPV16 or HPV18 infection. To define histopathological endpoints, a panel of expert pathologists will review all histopathological material collected during the course of our trial. This expert panel review will supercede the initial local histopathological diagnosis for purposes of defining the primary endpoint for US FDA submission. Panel members will review histopathological material blinded to trial arm assignment and to HPV DNA status or other participant characteristics. They will also review specimens blinded to the interpretation given by other panel members. Should the initial independent assessment differ between panel members, adjudication will be performed by consensus. Additional details of the plan for histology review to define study outcomes are provided in Section 3.19 (Laboratory Tests). Women diagnosed with cytological evidence of HSIL but without histopathological confirmation of CIN2+ will not be considered for analysis of the primary endpoint.
HPV16 and HPV18 status will be determined by testing cervical cells collected into PreservCyt® (Cell preservative solution; Manufactured by Cytyc Corporation, Boxborough, MA, USA) medium using PCR-based general primers (SPF 10 ) followed by LiPA and HPV16 and HPV18 type-specific PCR-based primer systems. The algorithm for PCR-based testing is provided as Appendix C. Additional details of the plan for HPV testing are provided in Section 3.19 (Laboratory Tests). For women who test positive for multiple HPV types, including HPV16 or HPV18, we will assume that HPV16 or HPV18 are associated with the CIN2+ lesions (i.e., will count as an outcome).
## Secondary endpoints
In addition to histopathologically confirmed CIN2+ and HPV16 and HPV18 infection, which are described above, other endpoints required to achieve the six secondary objectives of our trial are discussed below. 2. Occurrence and intensity of solicited local adverse events and solicited general adverse events in a 10% random subset of participants on one day from Day 3 to Day 6 after each vaccination and over all vaccinations combined.
3. Occurrence of unsolicited AEs and SAEs throughout the entire study (Month 0 up to Month 48).
## Outcome of all pregnancies.
To evaluate the duration of protection conferred by the candidate vaccine (Secondary Objective #2), we will examine the time to occurrence of incident cervical infection with HPV16 or HPV18 (by PCR). HPV16 and HPV18 infection will be defined as described in Section 3.19 of this protocol (Laboratory Tests).
To evaluate the efficacy of the candidate vaccine to protect against infection with any oncogenic HPV type (Secondary Objective #3), we will use as an endpoint histopathologically confirmed CIN2+ associated with infection by any oncogenic HPV type detected by PCR in the preceding cervical cytology specimen. Whenever oncogenic and non-oncogenic HPV types are detected in a woman diagnosed with CIN2+, we will assume that the lesion was caused by one of the oncogenic types detected. Additional details of the plans for histological assessment and HPV testing are provided in Section 3.19 (Laboratory Tests).
To more strictly evaluate whether the candidate vaccine protects against incident HPV16 or HPV18 associated CIN2+ among unexposed individuals (Secondary Objective #4), we will evaluate vaccine efficacy against the primary study endpoint while further restricting the evaluation to women who are HPV DNA (by PCR) negative at months 0 and 6 for the corresponding HPV type, seronegative by ELISA at month 0 for the corresponding HPV type, and who have evidence of HPV16 or HPV18 infection detected within the lesional component of the cervical tissue specimen by PCR. Details of the plans for HPV DNA and serologic testing are provided in Section 3.19 (Laboratory Tests).
To evaluate the efficacy of the candidate vaccine to protect against persistent infection with HPV16 or HPV18 (Secondary Objective #5), we will define as persistent infection any HPV16 or HPV18 infection that persists (same type persistence) for approximately 12 months or longer. The word "approximately" is used here because our annual visits can occur slightly less (or more) than 12 months apart based on our protocol-defined visit intervals (see [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref]. For women who are being followed at 6-month intervals, the intervening HPV result collected at the 6-monthly visit will also be required to be positive for the same HPV type for a woman to be defined as having a persistent HPV infection. Details of the plans for HPV testing are provided in Section 3.19 (Laboratory Tests).
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -51 -Finally, to confirm that immunogenicity in our population is comparable to that seen in previous Phase I/II trials and to define long-term immunogenicity of the candidate vaccine (Secondary Objective #6), we plan to measure HPV16 and HPV18 antibody levels by ELISA and V5/J4 monoclonal antibody inhibition EIA on blood specimens collected from the 600 women recruited into the immunogenicity subcohort of (Amended: 19 May 2006) our trial. Details of the serology testing plan are provided in Section 3.19 (Laboratory Tests)
## Tertiary and exploratory endpoints (amended: 13 nov 2008)
In addition to the primary and secondary endpoints listed above, NCI and Costa Rican investigators plan to evaluate numerous additional exploratory virological and immunological endpoints as surrogate endpoints in our trial. A full description of these endpoints is beyond the scope of this protocol, but includes:
- Persistent HPV16 or HPV18 infection at intervals other than the 12-month interval defined for secondary objective #5 and defined based on intratypic variant testing.
- High viral load HPV16 or HPV18 infection (whether transient or persistent).
- HPV16 or HPV18 VLP antibodies detected in serum by ELISA, pseudovirion neutralization, or other methods that might be developed during the course of the trial.
- HPV16 or HPV18 VLP antibodies detected in cervical secretions by ELISA or other methods that might be developed during the course of the trial.
- T-cell proliferative assays for HPV16 or HPV18 VLP.
- T-cell cytokine production assays for HPV16 or HPV18 VLP.
- T-cell cytotoxicity assays for HPV16 or HPV18 VLP.
## Adverse events monitoring
One of the major objectives of the present trial is to obtain additional information on the AEs associated with HPV16/18 VLP vaccination. While safety and immunogenicity trials in the United States and elsewhere have suggested the overall safety of the HPV16/18 VLP-based vaccine, rare AEs cannot be ruled out in these early phase studies given the relatively Our plan to monitor all trial participants for solicited and unsolicited AEs (whether expected or unexpected) is summarized below.
- AEs occurring immediately following vaccination. We will observe all vaccinees for at least 60 (Amended: 31 Oct 2005) minutes at the clinic following each CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -52 -vaccination with appropriate medical treatment readily available in case of a rare anaphylactic reaction. This will be performed under the supervision of trained medical staff. During this 60 (Amended: 31 Oct 2005) minute observation period, we will collect information on vaccine reactogenicity, including both solicited and unsolicited events, as summarized on the Vaccine Administration and Reactogenicity Monitoring (VAR) and AER (Amended: 31 Oct 2005) forms that are provided separately from this protocol.
- AEs occurring in the week following vaccination. We plan to visit a random sample of 10% of vaccinees once during the week following each vaccination to collect information on solicited AEs, unsolicited AEs, and use of study-provided acetaminophen. This visit will be conducted by a specially trained outreach health worker between Days 3 and 6 after each vaccination. The information collected during this visit is summarized on the Home Visit Reactogenicity Monitoring and AER (Amended: 31 Oct 2005) forms that are provided separately from this protocol.
- Long-term active monitoring. We will schedule all participants for annual clinic visits for a period of 4 years from enrollment. At the time of each clinic visit, we will collect information on long-term, unsolicited AEs, and SAEs. The clinic visit will be performed by a trained study doctor or nurse and will include a pelvic examination for sexually experienced women. The Post-Enrollment Eligibility Screener and Adverse Event Report forms will be used to document all reported AEs. These forms are provided separately from this protocol.
- Long-term passive monitoring. We will provide trial participants with a toll-free number that can be used at any time during our trial visit to report AEs to our trained emergency response team. Participants will be instructed to call this toll-free number should they have any medical problems, whether they think the problem is related or not to vaccination. The phone will be staffed 24 hours a day by a physician who will triage all calls and refer as needed. The AER (Amended: 29 Mar 2005) form will be used to document all AEs reported through this mechanism. This form is (Amended: 29 Mar 2005) provided separately from this protocol.
- Monitoring pregnancy outcomes. Information regarding pregnancies and their outcomes will be obtained at every follow-up visit performed during the trial. For women who are lost to follow-up, information will be captured via linkage to hospital records. We plan to document outcomes for all pregnancies occurring among vaccinees during the course of our study. For deliveries, we will review and abstract relevant information from standardized forms that are completed for each delivery at all hospitals in Costa Rica. Review will be conducted by trained abstractors from our staff at regional hospitals in Guanacaste and hospitals outside of Guanacaste to which women from Guanacaste are sometimes referred. Alternatively, the same information can be obtained from the "baby card" provided by the hospital to the mothers of newborns. Miscarriages (whether spontaneous or voluntary) resulting in hospitalizations or emergency room visits will be monitored as described below. Miscarriages occurring outside of a hospital setting will be documented by self-report during follow-up visits.
- Monitoring specific AEs of interest in offspring. At the request of the Costa Rican IRB specific adverse events occurring in subject's offspring who were conceived CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -53 -within one month prior to first vaccination to one year after last vaccination will be reported at the final visit (Month 48). This includes the following medical event categories in the offspring:
- Endocrinologic and metabolic conditions
[formula] - Autoimmune diseases - Sensory impairment - Intellectual impairment - Death [/formula]
The occurrence of these specific events in offspring will be included in the clinical database. A child case will be created and linked to the mother case. Medical judgment will be exercised in deciding whether any disorders/diseases reported in offspring are included in these specific categories and should be reported as appropriate. Of note congenital malformations in offspring, as per original protocol, are to be documented and reported as SAEs and included in the pregnancy outcome analysis. (Amended: 13 Nov 2008)
- Monitoring of hospitalizations and emergency room visits. Hospitalization listings from all hospitals and emergency rooms in our study area will be reviewed by trained study staff. Listings from selected hospitals outside of Guanacaste to which women from Guanacaste are referred will also be reviewed by our staff. This will allow us to identify women in our trial who are hospitalized during the study period and who might not have contacted us via the toll-free number listed above. Medical records from these women will be reviewed and abstracted to obtain detailed information about the condition that led to hospitalization and/or the emergency room visit. This source of information will be important for the small fraction of women (<10% based on 10 years of experience from our natural history study in 10,000 women in Guanacaste) that we expect to be lost to follow-up during our study.
All AEs reported on the AER form (Amended: 31 Oct 2005) will be followed through resolution by a staff physician. Non-serious AEs will be reviewed by a staff clinician before participant close-out; participants with ongoing conditions will be referred to the appropriate referral center if needed. (Amended: 13 Nov 2008) AEs will be evaluated for their intensity, seriousness, and relationship to vaccination. In addition, the nature of each event (including a diagnosis using the ICD-10 coding structure), date and time of onset (where appropriate), and outcome (including resolution date) will also be documented. As outlined in Appendix D, all SAEs, regardless of expectedness will be reported by one of the Costa Rican investigators to the NCI Medical Monitor (who will in turn report to the DSMB) and the GSK safety group, in a timely manner, in compliance with FDA regulations (21 CFR 312.32). Definitions of seriousness, relationship to vaccination, intensity, and the procedures for and timing of reporting AEs are also provided in Appendix D.
It should be noted that all medical staff will receive special training to respond to clinical emergencies. Emergency resuscitation equipment will also be available at all study clinics. All participants are expected to have at least seven visits over the course of our trial (see also. This will include:
- An initial enrollment visit, at which time potential participants will be screened for eligibility, will be randomly assigned to one of two arms, and will receive their first vaccination
- Two clinic visits for the administration of additional vaccine doses (one and six months following the enrollment visit)
- A clinic visit will be conducted at seven months on a subset of (Amended: 19 May 2006) 600 women enrolled to allow for immunogenicity assessment a minimum of 30 days following the last vaccination
- Four long-term follow-up clinic visits at yearly intervals starting 12 months after enrollment, and continuing through their four-year anniversary
In addition to the clinic visits listed above, women diagnosed with ASC-US+/LSIL at entry or during follow-up will be followed at six-month intervals and referred for colposcopic evaluation if the lesion persists for six months or longer. Also, women found to have prevalent or incident HSIL or cancer at any time during the study will be referred for colposcopic evaluation and excisional treatment, if necessary.
## Enrollment physical examination
Women selected for the trial will be invited to visit one of the study clinics to participate in the trial. At the time of the initial clinic visit, medical history information will be obtained and a complete physical examination will be performed to determine the overall health status and eligibility for the study. The Enrollment General Physical Exam and Medical History form that will be used to record results from this examination is provided separately from this protocol.
Immediately preceding the physical examination, and before each vaccination (at 0, 1, and 6 months), all participants will be asked to collect urine specimens for a urine pregnancy test. Only women with a negative pregnancy test and who report use of an acceptable form of birth control for at least 30 days and a willingness to continue use of an acceptable form of birth control until 60 days after the last dose of vaccine is administered (i.e., for the first eight months of the trial) will be enrolled. The enrollment of women who are pregnant will be deferred until they are at least three months postpartum and no longer breastfeeding.
## Pelvic examinations
Pelvic examinations will be performed at enrollment and during follow-up visits, starting at 12 months (i.e., at 12, 24, 36, and 48 months). These examinations will be performed CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
to permit a visual examination of the external genitalia and cervix, to collect cells for cytology and HPV DNA testing (and for C. trachomatis and N. gonorrhoeae testing by Hybrid Capture at the first pelvic exam), and to collect cervical secretions for HPV antibody and other testing. Conditions identified during these procedures will be documented on the appropriate CRF and will result in treatment by study staff or referral to the Costa Rican National Social Security Administration Health System. No pelvic examinations will be performed on virginal women and these women will be assumed to be HPV and SIL negative, since virginal women are known to be at extremely low risk of HPV infection and SIL. The low rates of exposure to HPV among virginal women in Guanacaste have been confirmed by serological testing of self-reported virginal women in our natural history study. No pelvic examinations will be performed on pregnant women, as discussed previously. Pregnant women will have their enrollment and/or follow-up visits deferred until they are at least three months post-partum. During enrollment, women who are more than three months post-partum but still breastfeeding will also be deferred until they are no longer breastfeeding.
The enrollment pelvic examination will be performed to assess whether a woman has prevalent HPV infection and/or ASC-US/SIL at baseline. As discussed earlier, cytological findings at entry will not be used to determine eligibility for vaccination. However, women with advanced disease (defined as women with suspected or confirmed HSIL and/or cancer and those with lesions of glandular origin) will be referred for colposcopic examination and, if confirmed, the lesion will be treated. Women with evidence of advanced disease during the first six months of the trial will receive no additional vaccinations until after colposcopic assessment and treatment (if necessary), but will continue to be followed for the duration of the trial, as described in the next section. Also, the presence of HPV16 or HPV18 DNA, antibodies to HPV16 or HPV18, and/or ASC-US/LSIL at entry will not result in exclusion from the trial, but assessment of baseline status is important to permit planned primary and secondary analyses restricted to individuals uninfected with HPV16 or HPV18 at entry (see Section 4).
While a pelvic examination will not be performed on women at their six month visit (unless there is evidence of ASC-US+/LSIL at the enrollment visit, as discussed below), a self-administered cervicovaginal cell collection will be performed at this visit to collect cells for HPV DNA testing. This will allow for a separate evaluation of women who develop incident HPV infection before the full vaccination schedule is completed. This will be important to determine whether vaccination failures are more likely before all vaccine doses are administered and "maximal" immunological protection is achieved. The self-administered collection proposed for our trial has previously been compared against clinician-administered collections and the two types of collection were found to correlate very well (88.1% agreement; kappa = 0.73) . To ensure that selfadministered specimens collected during our trial are comparable to clinicianadministered sampling for HPV DNA testing, we plan a re-evaluation of these two collection methods within the context of our trial. Both specimen types will be collected and tested for HPV DNA (by PCR) from the subset of women in our trial who require a pelvic examination at their six-month visit due to evidence of ASC-US+/LSIL at the enrollment visit.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
Follow-up pelvic examinations conducted at 12, 24, 36, and 48 months will permit active surveillance of trial participants for the primary and secondary outcomes of our trial. As mentioned previously, additional pelvic examinations at six-month intervals will be scheduled for women with cytological evidence of ASC-US+/LSIL at any of the routine annual screening visits (Appendix A). Women whose ASC-US+/LSIL lesion persists for six months or longer will be referred for colposcopic evaluation and excisional treatment, if needed.
Cytological specimens obtained during pelvic examinations will be reviewed as described in the Laboratory Tests section of this protocol (Section 3.19).
## Referral for colposcopic evaluation
To assure prompt treatment of precancerous lesions of trial participants, women with cytological evidence of ASC-H by cytology, HSIL/cancer by cytology, with evidence of a glandular lesion by cytology, or with visual abnormalities suggestive of cancer will be rapidly referred for colposcopic evaluation and excisional treatment, if needed (Appendix A). If all doses of the vaccine have not been completed at the time the diagnosis of one of these advanced conditions is made, the woman will not receive any additional vaccinations until after colposcopic evaluation and treatment (if needed) is completed. Women referred for colposcopic evaluation (regardless of treatment) will continue to be followed actively throughout the four-year trial period to assure adequate clinical management and to collect information on potential vaccine AEs. These women will be followed at six-month intervals until there is evidence of persistent or progressive disease, or they have three consecutive normal screening visits.
Women with cytological evidence of ASC-US+/LSIL at any point during the trial will be seen for repeat screening visits every six months (Appendix A). Should the cytological abnormality persist for two visits (whether consecutive or not), these women will be referred for colposcopic evaluation and excisional treatment, if needed. As was the case for women diagnosed with advanced disease (ASC-H/HSIL/glandular lesions/cancer), those who have persistent ASC-US+/LSIL and are referred for colposcopic evaluation will continue to be followed actively at six-month intervals until there is evidence of persistent or progressive disease, or they have three consecutive normal screening visits.
Biopsy and LEEP specimens obtained during the trial will be retrieved for histopathological review by the study pathologists in Costa Rica and the NCI, as described in the Laboratory Tests section (Section 3.19). These histopathological specimens will also be used for research purposes.
## Exiting women from the trial (amended: 13 nov 2008)
At the Month 48 visit we will apply the same criteria applied throughout the trial to define women who require colposcopic evaluation. Specifically, all women with evidence of cancer, HSIL, glandular lesions, AGC, ASC-H, or persistent ASC-US+/LSIL will be referred to colposcopy. Close-out of individual study participants will only occur after colposcopy referral, for women who require such referral. See Appendix A for more details.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
It should be noted, that after participant close-out, participants will be offered additional clinical care if needed. Case management after study close-out will be defined under a separate protocol.
## Activities at study completion (amended: 10 september 2009) the data and safety monitoring board (dsmb) for our trial has recommended crossover immunization of both treatment and control recipients. the participants who received the hpv vaccine will be offered vaccination against hepatitis a and hepatitis b and participants who received the control vaccine will be offered vaccination against hpv-16/18 and hepatitis b, as appropriate. vaccines to be used during cross-over immunization are the formulations licensed for use in costa rica. cross-over immunization will be implemented by nci and costa rican investigators under a separate protocol they have developed.
## Implementation of cross-over immunization will require unblinding. unblinding information and cross-over will not be offered to participants until they have been exited (closed out) from the present study and their data for this study have been frozen in the study databases. exiting of participants in the trial will occur in an ongoing/batched manner as participants complete their 4-year participation in our study and after ongoing saes or pregnancies are resolved, or on october 3, 2010 (10 months after the 4-year anniversary for the last woman enrolled), whichever comes first. this will ensure that study blind is maintained for individual participants until their data for the final analysis is frozen.
If cross-over immunization cannot be offered to a subject (e.g., because of medical reasons) or the subject refuses cross-over immunization, unblinding information will be provided after the participant has exited the current study and her data has been frozen. This will occur at the time of the first cross-over implementation visit, at which time participation/eligibility for cross-over immunization will be determined.
## Other procedures and data collection
## Overview
In addition to the clinical examinations described above, the following data and specimen collection is planned:
- Interviews at enrollment and during follow-up to ascertain information regarding HPV exposure risk (i.e., sexual behavior) and other socio-demographic, behavioral and exogenous factors that are either believed to be associated with the development of SIL or thought to potentially modulate the immune response to vaccination.
- Blood specimens will be collected at every clinic visit for HPV antibody testing and testing for other SIL risk factors (e.g., other sexually transmitted diseases) and/or factors postulated to modulate immune response to vaccination (e.g., HLA typing and T-cell responses to HPV). Blood specimens collected at the enrollment visit will also be used to perform a hematocrit as a benefit to participants. It should be noted that, for sexually experienced women, both the interview and blood collection will be deferred whenever a woman's pelvic examination is deferred (e.g., due to pregnancy).
## Interview questionnaires
A brief, 10-15 minute questionnaire will be administered to all participants at enrollment. This questionnaire will assess information on factors believed to be associated with either SIL or immune response to vaccination. More specifically, the questionnaire will obtain information on:
- A copy of the study enrollment questionnaire is provided separately from this protocol.
During follow-up (starting at the six month visit), additional questionnaires will be administered to participants at each clinic visit to ascertain changes in the factors listed above. Should some of the specific ancillary studies listed above be approved, additional questions relating to the specific hypotheses being addressed in the ancillary studies may also be added. A copy of the study follow-up questionnaire is provided separately from this protocol.
## Blood collection
Blood specimens will also be collected from all participants at each of the clinic visits (see. The major purpose of these blood collections will be to monitor immune response to vaccination, both humoral (i.e., antibody) and cell-mediated (i.e., T-cells). Aliquots of plasma, serum, buffy coat, peripheral blood mononuclear cells (PBMCs) and red blood cells (RBCs) will also be saved for other research-based assays, including antibody tests against sexually transmitted diseases other than HPV, hormone assays, nutritional assays (particularly those nutrients believed to impact the immune response), host DNA testing (to evaluate host susceptibility factors associated with SIL and the immune response to vaccination, such as HLA), and others. The blood specimen collected at study entry will also be used to perform a hematocrit as a benefit to participants. Women with evidence of anemia will be provided appropriate initial treatment and referral to the Social Security system. CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -59 -Three types of blood draws are planned. These include the collection of 3 mL of blood for hematocrit testing (done as a benefit to participants), 10-20 mL of blood that will be kept cool after collection and processed locally within a few hours of collection, and the collection of an additional 40 mL of blood on a subset of participants that will be kept at room temperature and shipped overnight to a specialized laboratory set up to cryopreserve viable lymphocytes. The 3 mL blood draw will be obtained at enrollment only. The 10-20 mL blood draw will be collected from all participants at all clinic visits (20 mL at the enrollment and exit visits and 10 mL at all other intervening visits) and will be used for HPV antibody, host DNA, hormonal, nutritional, and other research-based assays. The 40 mL blood draw will be collected to allow for cryopreservation of PBMCs required to perform functional immunological testing planned as part of efforts by NCI and Costa Rican investigators to monitor the immune response of trial participants. This third blood draw will be collected for ancillary research purposes on a subset of participants, including a random sample of 10% of participants at enrollment and at their 12 and 36 month visits ( While the primary focus of this larger blood draw will be on the cryopreservation of PBMCs for immunological monitoring, residual plasma will also be saved for use in possible future assays.
While the primary focus of this larger blood draw will be on the cryopreservation of PBMCs for immunological monitoring, residual plasma will also be saved for use in possible future assays.
To assure that participation and retention rates in our trial are not negatively affected by the blood draws, refusal to have blood collected will not result in exclusion from the trial.
## Extra-cervical collection
Oral, anal and vulvar specimen collection at the 48-month visit is being added to allow for the evaluation of the effect of vaccination on oral, anal, and vulvar HPV infection rates. Prior to the pelvic exam, an oral specimen will be collected. During the pelvic examination, a vulvar and an anal specimen will be collected prior to cervical sampling. Each specimen will be collected according to a standard procedure and will be stored in separate tubes. (Amended: 13 Nov 2008)
## Biospecimen processing, storage, & shipment
Collection of numerous types of biological specimens is planned, as mentioned in previous Sections. The types of specimens we propose to collect and plans for processing, storage and shipment are summarized in this Section.
The individual types of specimens we propose to collect and the expected number of aliquots resultant from this collection and requiring storage/shipment are summarized in CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 [fig_ref] Table 5 Types: participants [/fig_ref] to allow NCI to plan for adequate repository space at its main repository in Maryland, USA. In brief, we propose to collect cervical cells for liquid-based cytology, additional cervical cells for HPV DNA, RNA and other testing, cervical secretions for immunological assessment, cytological material (slides) for evaluation, histopathological material (blocks and slides) for review and testing, and blood for various immunological and other research-based tests, and oral, anal and vulvar samples (Amended: 13 Nov 2008). We have had experience collecting all of these biological specimens in Costa Rica on several thousand women over more than seven years as part of our natural history study.
Several biological specimens we propose to collect require little or no processing and can be stored at air-conditioned room temperature until shipment to the USA. These include the cells for liquid-based cytology and cytopathological materials. Other specimens require little or no processing but need to be frozen on the day of collection. These include the cervical cells for HPV DNA and RNA testing, and cervical secretions, and oral, anal and vulvar samples (Amended: 13 Nov 2008). Blood specimens will require more extensive processing. For all participants at enrollment and during follow-up, we propose to collect 10-20 mL of blood and to separate and store its various components (plasma, serum, buffy coat, and red blood cells). Processing of these specimens will be performed by trained personnel on site who are devoted to this task. They will use specific Standard Operating Procedures (SOPs) developed for our trial.
Finally, we propose to collect additional blood specimens on a random sample of participants to permit cryopreservation of lymphocytes for use in functional immunological and other testing. Cryopreservation of lymphocytes is technically more challenging than the separation of whole blood into its components, and so collaborative arrangements have been established with investigators at the University of Costa Rica in San Jose to have specimens from our trial cryopreserved. We will have specimens requiring cryopreservation transported daily by a study driver to a university lab devoted to the processing of our trial specimens. Trained personnel will utilize the separation and cryopreservation procedures established by Biotech Research Laboratories, and used CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -61 -widely on other DCEG studies including our Costa Rican natural history study. Standardized SOPs have been developed for use by the processing laboratory.
Biological specimens collected during our trial will be temporarily stored in a repository located adjacent to the trial administrative offices. This repository is outfitted with air conditioning, a walk-in refrigerator, and various liquid nitrogen tanks, and is staffed by trained personnel devoted to maintaining the repository.
Shipment of biological specimens will be performed periodically to one of the NCI repositories in Maryland, USA and further distributed, as needed, to collaborating laboratories for testing. Materials that can be stored at room temperature will be shipped on a monthly basis by air. Shipments will be made by cargo plane, since some of the materials being shipped contain methanol and are considered flammable. Materials that need to be stored frozen will be shipped in liquid nitrogen tanks. Given the large volume of specimens to be shipped as part of our study, we plan to have frozen specimens shipped in large liquid nitrogen tanks that are specifically manufactured to allow for air transport. A specialized shipping and customs agent will be responsible for these shipments. The agent we propose to use has been responsible for the successful shipment of biological specimens collected during our natural history study. Hundreds of shipments have been made from Costa Rica since 1993, and to date only one shipment (in the early phase of our study) was lost in transit.
## Laboratory tests
## Cytology
Cervical cytology will be performed using the ThinPrep PapTest ™ (Cytyc Corporation, Boxborough, MA, USA). Cervical cells for ThinPrep cytology will be collected at each clinic visit where a pelvic exam is performed using a Cervex® brush (Rovers Medical Devices BV, Oss, The Netherlands) rinsed into a collection vial containing PreservCyt medium. Two 0.5 mL aliquots of the 20 mL in the PreservCyt vial will be withdrawn for PCR testing (see below), prior to preparation of ThinPrep slides using the ThinPrep 2000 Processor. ThinPrep slides will be prepared by qualified trial staff at the local cytology staining laboratory, will be screened sequentially by two cytotechnologists specifically trained to screen ThinPrep slides, and will be interpreted by the Costa Rican study cytopathologist, Mario Alfaro, using the Bethesda 2001 classification system.
The clinical management of the study participants will rely on the Costa Rican cytopathology interpretation, with the exception of HSIL noted below. As part of NCI's ongoing quality control efforts (led by cytotechnologist Claire Eklund and pathologist Martha Hutchinson who are affiliated with the Women's and Infants' Hospital in Providence, RI), all slides that read as abnormal in Costa Rica and a 10% sample of the slides read as negative in Costa Rica will be re-screened and re-interpreted in the United States. At the study exit (Month 48), in addition to 100% of slides read as abnormal in Costa Rica and 10% of the slides read as negative in Costa Rica mentioned above, 100% of slides that are read in Costa Rica as negative with evidence of reactive changes and that are concurrently HPV positive by the HCII will be re-interpreted in the United States.(Amended: 13 Nov 2008) The 10% set of negatives selected for evaluation in the United States will be randomly selected, without replacement. This scheme will result in approximately 50% of subjects with a negative cytology having an expert review over the course of the 4-year study follow-up period. The purpose of this review is to maximize the accuracy of the Costa Rican interpretations and the comparability of the Costa Rican readings to a US standard, realizing nonetheless that cytopathology is inherently and unavoidably subjective to some extent and that there is no absolute reference standard of cytopathology, particularly for equivocal interpretations such as ASC-US.
The cytological interpretation given by the Costa Rican cytopathologist will be used in clinical management and in the analysis, with one exception. In cases where the Costa Rican interpretation is <HSIL and the US interpretation is HSIL or more severe disease, the interpretation from the US will be communicated to Costa Rica, entered into the CRF, and will lead to colposcopic referral. Also, for clinical management purposes, should ThinPrep slides be inadequate for interpretation (estimate <2%), the subject will be managed as if she were LSIL positive, rather than delay management by recall of the patient for repeat cytology.
## Histology
(to define trial endpoints)
Biopsy and excisional treatment (loop electrosurgical excisional procedure, or LEEP) specimens obtained during the trial will be fixed in buffered formalin, and sent to the Costa Rican pathologist, Diego Guillen, for processing and diagnosis. LEEP specimens will be sectioned extensively, and up to three levels will be cut from each block. The diagnosis made by the Costa Rican pathologist will be used to determine clinical management of participants.
In addition to the Costa Rican diagnosis, an NCI designated pathologist (Amended: 13 Nov 2008, Amended: 10 September 2009) with expertise in cervical pathology will periodically diagnose all histological specimens, blinded to the initial Costa Rican diagnosis. If the diagnoses are concordant (concordance will be required at the level of negative or atypical metaplasia/CIN1/CIN2/CIN3 or in situ/cancer), that will establish the NCI provisional study diagnosis. Discrepant findings between the Costa Rican and NCI pathologist will lead to review by the second NCI pathologist. Agreement between two of the three pathologists at the <CIN2/CIN2+ level will establish the NCI study diagnosis. These diagnoses will not be available within a timeframe that would allow their use for clinical management purposes. Therefore, NCI pathology review will be performed to define histological diagnoses for analysis purposes. This review process and its results will be performed and documented in the US. A copy of the results of the US pathology review will be sent to the clinical staff in Costa Rica. Periodic review of proficiency slide sets will be performed throughout the trial to assure comparability of the Costa Rican/NCI panel consensus readings against other expert pathologists involved in HPV vaccine studies.
The primary trial endpoint is CIN2+ (associated with HPV16 and/or HPV18). CIN2 and CIN3 will be separately diagnosed for scientific interest, but combined as study endpoints. No cancers are anticipated and, in our experience in Guanacaste, CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -63 -adenocarcinoma in situ will be extremely rare. If these rare outcomes do occur, they will be included as cases.
## Pcr typing of hpv dna from preservcyt specimens
(to define trial endpoints) Two 0.5 mL aliquots of PreservCyt specimens will be withdrawn prior to ThinPrep preparation to avoid possible specimen-to-specimen contamination. The removal of the two small aliquots is not expected to impact on the cellular adequacy of the PreservCyt specimen, which typically contains many more cells than required for a ThinPrep and subsequent Hybrid Capture® 2 (a signal amplification assay; manufactured by Digene Corporation, Gaithersburg, MD, USA) (HCII HPV DNA test) testing.
The PCR testing used to establish the presence of HPV16 and/or HPV18 for the efficacy endpoints will be performed by GSK Biologicals or DDL Diagnostic Laboratories (Amended: 13 Nov 2008) on one PreservCyt aliquot, as follows (see also Appendix C).
To test for HPV DNA positivity, GSK Biologicals or DDL Diagnostic Laboratories (Amended: 13 Nov 2008) will use SPF 10 primers that amplify a 65 nucleotide region of the HPV L1 gene for most of the known HPV isolates. The read-out for this test is a DNA enzyme linked immunoassay. Positive specimens will be typed by reverse hybridization line probe assay (LiPA) using 25 type-specific hybridization probes (14 oncogenic HPV types and 11 non-oncogenic HPV types . If LiPA detects no HPV16, the specimen will be tested by HPV16 specific PCR using primers that amplify a 92 nucleotide segment of the E6/E7 gene. If LiPA detects no HPV18, the specimen will be tested by HPV18 specific PCR using primers that amplify a 126 nucleotide segment of the L1 gene. Redundant testing using general PCR with LiPA, followed by type-specific PCR to confirm absent HPV16 and HPV18 affords maximum test sensitivity while preserving high specificity. To monitor laboratory reproducibility, NCI will insert masked QC replicates in each batch sent for testing. The results of this testing algorithm will be considered definitive for all HPV16 and HPV18 related study endpoints.
In order to establish the presence of an HPV type other than HPV16 or HPV18, the second 0.5 mL aliquot of PreservCyt will be sent by NCI to a collaborating contract laboratory to test for a full range of genital HPV types using PGMY primers and reverse line blot typing assay (LBA). To monitor laboratory reproducibility, NCI will insert masked QC replicates in each batch sent for testing. Samples positive for HPV types other than HPV16 and HPV18 in either the PGMY/LBA or SPF/LiPA assay will be considered as HPV positive for the appropriate other HPV type.
## Pcr typing of hpv dna from histopathological specimens
(to define outcome for secondary objective #4 defined in Section 2.3.) Tissue-based HPV PCR testing will be performed on sections obtained from formalinfixed & paraffin-embedded tissue specimens collected from participants referred to colposcopy who are found to have histopathological evidence of CIN2+.
Clean techniques will be implemented when sectioning tissue blocks to avoid contamination. Furthermore, lesional material will be selected for analysis using microdissection techniques, as appropriate. Testing will be performed at a laboratory designated by NCI and GSK Biologicals after assay validation.
To test for HPV DNA, SPF 10 primers which amplify a 65 nucleotide region of the HPV L1 gene for most of the known HPV isolates will be used. The read-out for this test is a DNA enzyme linked immunoassay. Positive specimens will be typed by reverse hybridization line probe assay (LiPA), using 25 type-specific hybridization probes. This typing process enables detection of 14 oncogenic HPV types and 11 non-oncogenic HPV types . If LiPA detects no HPV16, the specimen will be tested by HPV16 specific PCR using primers that amplify a 92 nucleotide segment of the E6/E7 gene. If LiPA detects no HPV18, the specimen will be tested by HPV18 specific PCR using primers that amplify a 126 nucleotide segment of the L1 gene. Redundant testing using generic PCR with LiPA, followed by type-specific PCR to confirm absence of HPV-16 and HPV-18 affords maximum test sensitivity while preserving high specificity.
To monitor laboratory reproducibility, NCI will insert QC control specimens as appropriate. The results of this testing algorithm will be considered definitive for the study endpoint used to assess secondary objective #4 defined in Section 2.2
## Hpv dna testing by hybrid capture 2 (hcii)
(for baseline/exit assessment and triage of ASC cytology)
After ThinPrep cytology slides have been prepared, residual PreservCyt specimens on slides read as ASC will be sent to the HPV DNA testing laboratory (Enrique Freer) at the University of Costa Rica for HPV testing using the Hybrid Capture 2 test (HCII; Digene Corp., Gaithersburg, Maryland, USA). The assay uses a cocktail of labeled RNA probes (Probe B) designed to detect 13 oncogenic HPV types .
An ASC-US (Amended: 29 Mar 2005) cytologic interpretation will lead to "reflex" HCII testing and for the purposes of clinical management subjects who have HCII positive (or HCII-inadequate) ASC-US will be treated the same as an LSIL. Should residual PreservCyt material be insufficient for HCII testing, or should HCII testing fail for any other reason, specimens will be assumed to be HCII positive for clinical management purposes and treated accordingly (<5% expected).
HCII testing will also be performed on all subjects at baseline for NCI research purposes and at exit (Month 48 visit) to assist in defining the exit management strategy. Testing will be performed according the manufacturer's specifications at the time of each woman's first pelvic examination for NCI research purposes and as a benefit to study participants. The same cervical sample used for the HCII-HPV assay will be used for the HCII CT/GC assay.
## Serology testing by elisa
Serological assays for HPV16/18 by ELISA will be performed at GSK Biologicals laboratories, Rixensart, Belgium or in a laboratory sponsored by the NCI in Costa Rica, pending validation. Baseline ELISA anti-HPV16 and anti-HPV18 testing will be performed on all women enrolled. In addition, anti-HPV16 and anti-HPV18 ELISA will be performed on all blood samples (Month 0 through Month 48) collected from the 600 women enrolled into the immunogenicity subcohort within the (Amended: 19 May 2006) trial. Women in this immunogenicity subcohort will have an extra clinic visit performed a minimum of 30 days following the last vaccination (Month 7 visit) to permit immunological assessment after all three vaccine doses are administered.
A 1 mL aliquot of serum obtained from 10 mL whole blood will be used for HPV16/18 serology testing. Quantitation of antibodies to HPV16 and HPV18 VLPs (anti-VLP-16 and anti-VLP-18) will be performed by ELISA using either HPV16 or HPV18 VLPs as coating antigens. ELISA titers of VLP-specific antibodies will be calculated relative to the optical density readings obtained with serial dilutions of a positive control pooled reference serum sample using a four-parameter equation. Midpoint titers (expressed in EL.U./mL) will be obtained by averaging the values from all dilutions that fall within the working range of the standard positive control titration curve.
In case of insufficient blood sample volume to perform assays for all antibodies, the samples will be analyzed according to the following priority ranking (ranked in order of decreasing priority):
1. HPV16 antibody (by ELISA)
2. HPV18 antibody (by ELISA).
Serology testing of follow-up specimens collected from women not included in the 600woman immunogenicity subset or testing for additional antigens contained in the study vaccines will be performed if deemed necessary by the NCI, or if any findings in the CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -66 -present study or in other studies necessitate additional investigation of the immunogenicity of the vaccine. In this case, the ranking above may also be changed.
## V5/j4 monoclonal inhibition enzyme immunoassay (eia) antibodies on a subset of subjects
V5/J4 monoclonal inhibition EIA will be performed at GSK Biologicals laboratories, Rixensart, Belgium or in a laboratory designated by NCI and GSK Biologicals after assay validation on all blood samples (Month 0 through Month 48) collected from the 600 women enrolled into the immunogenicity subcohort of the trial.
These EIA inhibition assays are based on the indirect measure of serum antibodies specifically directed either against V5 or J4 epitopes that are described in the literature as the major neutralizing epitopes of HPV16 and HPV18, respectively.
These assays are based on the competitive test principle. Briefly, HPV16 or 18 VLPs are coated onto 96-well microplates, serial dilutions of human sera are added to the VLPcoated plates and incubated. After washing, biotinylated V5 or J4 monoclonal antibodies are added into the wells. The plate is washed and peroxidase (POD)-conjugated streptavidin is added. The binding of streptavidin is detected by the addition of tetra methyl benzidine (TMB). Colorimetric change induced will be inversely proportional to the concentration of anti-V5 or -J4 antibodies in the serum. ELISA titers are calculated from a reference curve using a four parameters equation and expressed in ELISA Units per milliliter (EU/mL).
## Other assays
Several other laboratory assays will be performed for research purposes to enable exploratory analyses of tertiary objectives and ancillary studies.
## Efforts to maximize participation/retention
Although a randomized clinical trial is the ideal method to assess the effectiveness of any intervention, it is well recognized that subject withdrawal or loss-to-follow-up subsequent to randomization can introduce bias if losses are differential in the different trial arms. Thus, issues of retention subsequent to enrollment are important and have been given emphasis in planning for the present trial.
Efforts aimed at minimizing withdrawal and losses during follow-up are listed below:
- Selection of Guanacaste as the trial site because of high participation and retention rates in previous studies.
- Community outreach and education efforts are planned to assure that adequate information regarding the trial is disseminated throughout the region. These efforts will target both the population at large and the public health community in the region.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
- We will invite 20,000 women to participate, but our power estimates (see [fig_ref] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations... [/fig_ref] Section 4.7) indicate the need for 7,500 women (Amended: 13 Nov 2008). Only women who, in principle, agree to comply with the full protocol will be enrolled.
- We plan to send reminder letters before each appointment. For appointments preceding vaccination visits we will also send an outreach worker to the home of participants to remind them of their scheduled appointments.
- Reimbursement for transportation costs will be provided, if necessary and, when needed, transportation to the clinic in a study vehicle will be provided.
- Women who, at enrollment, have definite plans to move out of the study area within six months after enrollment will be excluded.
- Women who move within Guanacaste will be given the option to have their clinic assignment changed to a clinic closer to their new home. Women who work will also have the option to be seen at a clinic that is close to their work location. Several clinics throughout Guanacaste and surroundings will be used for our trial, allowing for such flexibility.
- Despite our attempt to exclude women with plans to migrate outside of our study region, some women are likely to move during the four-year study follow-up period.
Women who move will be offered transportation or reimbursement of expenses to visit the nearest study clinic. During our Natural History Cohort study women who had migrated outside of Guanacaste often agreed to return to our clinics for their appointments.
Efforts described above should minimize withdrawals from the study prior to the completion of 48 months of follow-up. In the event that a woman withdraws from the study due to participant refusal, relocation, death, or other reason, the specific reason for withdrawal will be carefully specified on a CRF designed for this purpose. Discontinuation of vaccination (i.e., withdrawal from vaccination) will not constitute grounds for withdrawal from study follow-up. The occurrence of an AE will also not constitute grounds for withdrawal. However, should a woman choose to refuse further participation as a result of an AE, this will be carefully documented and all attempts will be made to track the AE to its resolution.
## Statistical analysis plan & power
## Overview
The main analyses planned for our trial will be conducted by the NCI and Costa Rican investigators and will focus on the primary outcome, viz., HPV16 or HPV18 (combined) associated, histopathologically confirmed CIN2+ among uninfected individuals (i.e., HPV DNA negative for HPV16 or HPV18 at entry and at 6 months), as defined in Section 2. The primary analyses will be based on the difference in the cumulative proportions of women who develop an outcome between the treatment group and the control group (i.e., difference between attack rates). The test statistic will be the Fisher's exact test for difference between two proportions. Analysis will be triggered by the total number of events observed, and will be conducted after at least 30 confirmed cases are CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 observed and the last woman recruited into the study has completed a follow-up time of two years, pending review and approval of the final statistical analysis plan by the DSMB.
The primary analysis for this trial will be "According-to-Protocol" (ATP). The ATP analysis will exclude women otherwise eligible for the trial who do not receive all three doses of vaccine or who test positive for HPV16 and HPV18 DNA before or at the time of the last vaccination (scheduled for 6 months after randomization). An Intent-to-Treat (ITT) analysis is also planned to confirm ATP findings. For the ITT analysis, all women randomly assigned to one of the two study arms and confirmed by HPV DNA testing to be HPV16 or HPV18 negative at randomization will be included.
Secondary analyses for efficacy will be performed to examine duration of the effect of the vaccine in the prevention of incident infection with HPV16 or HPV18. These analyses will focus on events that occur at one, two, three, and four years after randomization in women who are under follow-up and had not previously had the event of interest. Additional secondary analyses will evaluate the efficacy of the candidate vaccine in the prevention of CIN2+ associated with all incident oncogenic HPV types including HPV16 and HPV18, in the prevention of persistent HPV16 or HPV18 infection (persistence defined as 12+ months of infection with same type infection). Analyses of the 600 women in the immunogenicity subset will also be performed and will focus on defining the serological antibody profile observed in vaccinated individuals over time.
Additional exploratory tertiary and ancillary analyses will be conducted by the NCI and Costa Rican investigators and will evaluate numerous issues, as described in the Tertiary Objectives and Ancillary Analyses & Studies sections of this protocol (Section 2.3 and Section 2.4).
No interim statistical analyses for the primary efficacy endpoint are planned at this time (Amended: 10 September 2009).
Adverse events (AEs) observed in the course of the trial will be carefully monitored. Both expected and unexpected AEs will be examined. Rates of specific solicited, unsolicited, and serious AEs (SAEs) will be compared by treatment arm and evaluated. These analyses will be performed throughout the trial by the DSMB (study personnel will remain blinded) and will not affect the overall alpha level of the study.
Power calculations are provided at the end of this section (Section 4.7). Assumptions inherent in our calculations are specified and are based largely on results of our ongoing population-based natural history study in Guanacaste. Assuming conservatively that 12,000 (6,000 per arm) of the 20,000 women invited to participate are randomly assigned to one of the two arms of the trial, we estimate an overall power of 91% to detect an 80% reduction in incidence of HPV16 or HPV18 (combined) associated CIN2+ over the four years of follow-up planned for this trial.
## Primary analysis for efficacy
The primary analysis will assess the candidate vaccine's ability to protect against incident, histopathologically confirmed HPV16 or HPV18 associated CIN2+ among uninfected individuals (i.e., HPV DNA negative for HPV16 or HPV18 at entry and at 6 months). An expert panel of pathologists will review all histopathological specimens blinded to vaccination status to determine CIN2+ diagnoses. Presence of HPV16 or HPV18 DNA will be determined by testing cervical cells collected into liquid-based PreservCyt medium by PCR using the SPF 10 primers typed using the LiPA method and followed by HPV16 and HPV18 type-specific primers (Section 3.19). For women with multiple type infections including HPV16 or HPV18, we will assume that HPV16 or HPV18 is associated with the CIN2+ lesion.
Final analysis for Biologic License Application (BLA) submission will be triggered when one of the following sets of conditions is met: Assuming 15,000 women enroll, it is expected that a total of approximately 50 HPV16/18+ CIN2+ events will occur across both trial arms over the course of the trial under the null hypothesis. A minimum threshold for analysis of 30 events was chosen since this number of events would provide adequate power (96%) to observe a statistically significant difference between the two trial arms under the assumption of 80% efficacy [fig_ref] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations... [/fig_ref]. The requirement of a minimum (Amended: 29 Mar 2005) followup time of two years was chosen to guard against analyses that are biased by early events only, which in the case of CIN2+ among young women might not be representative of true cancer precursors.
Should analysis occur prior to completion of the full 48 months of follow-up (under the second scenario above), women will continue to be followed through 48 months and an annex report will be issued to describe any additional data collected.
The main analysis of this primary endpoint will compare the cumulative proportions (i.e., attack rates) of women who develop an event among those who were randomly assigned to one of the two arms of the trial and found to be HPV16 or HPV18 DNA negative and without CIN2+ at baseline.
## Baseline characteristics
An initial evaluation will examine baseline characteristics (age, residence, education, virginity) of each trial group to assure comparability between groups.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
## Exclusion from analysis based on pre-enrollment testing
Women will be tested at entry for the presence of HPV16 or HPV18 DNA; only those who test negative for HPV16 or HPV18 will be included in the main ATP and ITT analyses described below. Women infected with HPV16 and HPV18 DNA prior to vaccination will be excluded from the main analysis to focus the analysis on the prophylactic (not therapeutic) efficacy of our vaccine. All decisions about presence of HPV16/18 at baseline will be made by protocols that ensure blindness to assigned treatment arm and any information or materials collected at or subsequent to randomization.
## According to protocol (atp) analysis
The primary analysis will be an ATP analysis. Criteria for elimination from the ATP 3.3.4 analyses are listed in Section of this protocol. ATP analysis will be performed using a Fisher's exact test for equality of proportions. Kaplan Meier analysis will also be performed (see plans for secondary analyses below). The log-rank test is not being proposed for this primary analysis since a vaccine that significantly delays onset of the primary outcome without significantly affecting the cumulative difference in attack rates should not be considered an efficacious vaccine.
## Intent to treat (itt) analysis
An ITT analysis will also be performed to confirm the ATP findings. For the ITT analysis, all eligible women who consent and are randomly assigned to one of the two arms of the trial (and all events occurring after randomization) will be included, regardless of whether or not all three doses of the HPV16/18 VLP or control vaccine are administered.
## Secondary analyses for efficacy
## Duration of protection
An important objective of this trial is to evaluate the duration of protection provided by the candidate vaccine. Short-term protection from the vaccine is sufficient to provide evidence for efficacy and will constitute the primary justification for licensure. However, evidence of longer-term protection is necessary to justify vaccination as a viable alternative to currently available screening practices for the prophylaxis of cervical cancer, particularly in developing countries. Therefore, we will consider analyses aimed at evaluating long-term efficacy of the vaccine in the prevention of incident infection with HPV16 or HPV18. To this end, we plan to examine Kaplan Meier curves and stratified log-rank test among those free of the outcome (left-truncation) at one-, two-and three-years after completion of vaccination (for ATP analyses) and randomization (for ITT analyses).
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
## Protection against any oncogenic hpv types
An additional important secondary objective of our trial is to evaluate whether the candidate vaccine protects against incident CIN2+ associated with any oncogenic HPV type. Both ATP and ITT analyses will be conducted, using the same methods proposed for the primary and secondary analyses described above (i.e., comparison of attack rates and of Kaplan Meier curves).
## Protection against hpv16 or hpv18 associated cin2+ among women with no evidence of exposure to hpv16 or hpv18 at entry
To more stringently evaluate whether the candidate vaccine protects against incident, HPV16 or HPV18 associated CIN2+ among unexposed individuals, secondary analyses will be performed restricted to women who are not only HPV DNA (by PCR) negative at months 0 and 6, but also seronegative by ELISA at month 0 for the corresponding HPV type. In addition, HPV16 or HPV18 positivity will be defined by PCR on the lesional component of the cervical tissue specimen. Statistical analysis methods identical to those proposed for the primary analysis will be used for this purpose.
## Protection against persistent hpv16 or hpv18 infection
To confirm published findings that immunization with VLP-based HPV vaccines protects against persistent infection, secondary analyses will also be performed to evaluate whether the candidate vaccine protects against long-term type-specific persistence (defined as 12+ months of infection) with HPV16 or HPV18. Both ATP and ITT analyses will be conducted, using the same methods proposed for the primary and secondary analyses described above (i.e., comparison of attack rates and of Kaplan Meier curves).
## Vaccine immunogenicity
To directly assess HPV16/18 VLP vaccine immunogenicity (as measured by anti-HPV16 and anti-HPV18 ELISA and V5/J4 monoclonal antibody tests) in our population, secondary analyses will be performed with the specific intent of confirming that HPV16 immunogenicity in our population mirrors that seen in other Phase I/II trials conducted to date, evaluating whether HPV18 immunogenicity seen after vaccination is comparable to that seen for HPV16, and evaluating patterns of antibody levels over time since vaccination.
## Evaluation of adverse events (aes)
In addition to vaccine efficacy, we plan to carefully monitor and evaluate AEs observed in the course of the trial. Our plan for detecting, documenting, and classifying AEs is discussed in Section 3.14. In this section we describe our plans for evaluation of these AEs. Analysis of AE data will be based on the total vaccinated group without exclusions. An additional analysis of safety will be performed based on an ATP safety cohort. The ATP safety cohort will be defined as all subjects who have received at least one dose of CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -72 -study vaccine/control according to their randomization assignment (i.e., excluding individuals who received no vaccination or who received a vaccine other than the one specified by randomization), who did not receive a vaccine forbidden in the protocol, for whom the administration site is known, and for whom sufficient data is available to perform an analysis of safety. It should be noted that for these analyses to evaluate AEs, study personnel will not be unblinded, analyses will initially be presented to the DSMB by group without specifying which group is the vaccinated or control group, and unblinding of the groups by the DSMB will only occur at the request of the DSMB should safety issues arise.
Both expected AEs, as defined in the GSK Investigator Brochure, and unexpected AEs will be evaluated. Solicited, unsolicited, and serious AEs (SAEs) will be evaluated separately. AEs and SAEs will be evaluated throughout the study. SAEs will be reported to the DSMB immediately, whereas non-serious AEs will be summarized and reported to the DSMB on a periodic basis. Expedited reporting of SAEs to the US FDA, other national regulatory authorities, and Investigators will be performed by GSK, in accordance with relevant regulations. Details of the plan for US FDA reporting are provided in Appendix D. Our plan for the analysis of AEs by the DSMB is summarized below.
1. Descriptive analysis will be performed to quantify the proportion (along with 95% confidence intervals) of vaccinated women who experience specific AEs. Solicited and unsolicited AEs will be tabulated separately by intensity. SAEs will also be evaluated separately. Relationship to vaccination will be examined. The proportion of individuals ever experiencing AEs and the proportion experiencing AEs at specific times (e.g., proportion of vaccinated women who experience pain at the injection site on the day of vaccination and in the week following vaccination) will be computed. Analysis stratified by dose number will also be performed.
2. The proportion of women who experience specific solicited AEs, unsolicited AEs, and SAEs will also be compared between the vaccinated group and the control group to determine whether there is evidence for significantly increased reactogenicity to the vaccine compared to control. Again, both cumulative and time-specific proportions will be computed, as will analysis stratified by dose number.
3. Pregnancies and pregnancy outcomes will be monitored throughout the study, and comparisons between the vaccine and control groups will be carried out in cumulative and time specific analyses.
4. Finally, since the incidence of new or exacerbated chronic illnesses with a possible immune pathogenesis is elevated among women 18-25 years of age, separate analyses will focus on an evaluation of new or exacerbated chronic illnesses such as autoimmune diseases (e.g., systemic lupus erythematosus, thyroiditis), insulindependent diabetes mellitus, and asthma. A similar approach to that described above will be used for this assessment.
## Other analyses
In keeping with NCI's research mission, we plan multiple analyses to evaluate issues of scientific relevance to the research community in addition to the analyses summarized CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -73 -above. These include analyses of vaccine effects using Cox regression that are stratified by specific subpopulations (e.g., seronegative and seropositive individuals at entry; users and non-users of oral contraceptives), analyses restricted to women who are virgins at baseline and therefore have never been previously exposed to HPV16 or HPV18. NCI will also perform analyses that adjust for any effect of these factors if they are not distributed in a balanced way between vaccination groups despite randomization. Additional analyses will look at endpoints for evaluation of HPV16 and HPV18 effects separately, of the ability of the candidate vaccine to protect against transient and persistent infections with HPV16 and HPV18, and of the therapeutic effect of the candidate vaccine among HPV16 or HPV18 positive individuals at entry. The large size of the present trial (including the large number of women who will receive control vaccinations) also provides us the unique opportunity to evaluate various hypotheses related to the natural history of HPV infection and SIL and to immunological response to HPV and to HPV vaccination. These ancillary analyses will be performed by NCI and Costa Rican investigators to increase our knowledge of HPV, cervical neoplasia, and immune response rather than to evaluate the efficacy of the vaccines to protect against the primary trial outcomes. Therefore, although results from these analyses will be shared with the DSMB and with the research community as a whole, they will not necessarily be used to determine efficacy of the vaccine. A detailed description of these natural history and etiological analyses is beyond the scope of this protocol.
# Interim analysis
There will be no restrictions on the number of analyses that can be conducted by the DSMB to examine safety of the candidate vaccine. Unscheduled analyses of safety may be conducted by the DSMB and will not affect the alpha for testing vaccine efficacy. Safety concerns with the vaccine shall constitute grounds for stopping enrollment or vaccination at any time during the trial.
No interim statistical analyses for efficacy were planned at the beginning of the trial (Amended: 10 September 2009). A need for an interim analysis for efficacy may arise during the trial. Should this occur, the study protocol will be amended, and the amendment submitted to the US FDA for review. To assist the DSMB in making decisions regarding interim analyses in the future, a set of guidelines are provided separately from this protocol in the DSMB Charter.
## A statistical analysis plan (sap) for initial analyses that focuses on virological endpoints has been submitted to the fda. the analyses proposed within are largely exploratory in nature, but there is one overlap between an objective in the sap for initial analysis and secondary objective #5 listed in this protocol (i.e., for secondary objective 5 listed in this protocol, we are proposing to conduct an interim analysis). given this overlap, for secondary objective #5 listed in this protocol, the following alpha adjustment plan will be implemented: the overall alpha of 0.025 (one-sided) will be split into 0.001 for the interim analysis proposed in the sap for initial analyses and 0.024 for the final analysis. as specified in more detail in the sap for initial analysis (table 3 of that document), power (assuming 80% vaccine efficacy and 20 or more total events) should remain above 85-90% for the final analysis of secondary objective #5 (amended: 10 september 2009).
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -74 -
## No stopping rule will be applied at this stage; blinding will be maintained and followup will continue for all subjects until month 48 (amended: 10 september 2009).
## Power to determine vaccine efficacy
The powers of our trial to detect vaccine efficacy rates ranging from 30-90% and for sample sizes ranging from 10,000-15,000 are summarized in [fig_ref] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations... [/fig_ref]. These calculations were performed using the PASS program (Hintze J. 2001. NCSS and PASS Number Cruncher Statistical Systems, Kaysville, UT, USA). The assumptions used in estimating power were based largely on results from our population-based natural history study in Guanacaste, Costa Rica [bib_ref] Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa..., Herrero [/bib_ref] and from the British natural history study published by Woodman and colleagues . We estimated the expected person-years and rates as follows for the ATP analysis assuming 15,000 women randomly assigned to one of the two arms of the trial:
1. Person-years: A total of 3,087 person-years/1,000 evaluable women are expected over the course of the four-years of follow-up in the control group. The expected person-years/1,000 evaluable women in the HPV16/18 VLP vaccine group is slightly higher (3,090/1,000 evaluable women) since we expect fewer CIN2+ events that terminate a woman's follow-up in the HPV16/18 VLP vaccine arm. These estimates account for the exclusion of person-years accrued:
- by women in the first 6 months of the study (since events in that time are excluded from ATP analysis),
- by women exposed to both HPV16 and HPV18 at entry or at the time of the third vaccination (month 6),
- by women diagnosed with CIN2+ in the first 6 months of the study,
- by women who do not receive all three doses of vaccine,
- by women who are virginal since virginal women are not at risk of CIN2+,
- by women who are lost to follow-up during the study 2. Outcome Rates: The expected rate of histopathologically confirmed CIN2+ associated with HPV16 or HPV18 occurring over the 42 months post third vaccination follow-up period among evaluable women is estimated to be 0.5% (.00238 annual rate of CIN2+ * 0.6 proportion of CIN2+ associated with HPV16/18 * 3.5 for 42 months of follow-up in ATP).
To estimate the person-years and rates described above, the following specific assumptions were made, based largely on data from our natural history study in Costa Rica. All of the assumptions used to calculate power are conservative to avoid underpowering our study.
1. The annual incidence of histopathologically confirmed CIN2+ is estimated to be 0.238%.
2. The percent of non-virginal women with histopathologically confirmed CIN2+ at enrollment is estimated to be 1.5%.
3. The percent of CIN2+ with detectable HPV16 or HPV18 DNA is estimated to be 60%.
4. The proportion of women who are sexually inexperienced at entry is estimated to be 23.0%. The proportion of women who are sexually experienced at entry is estimated to be 77.0%.
5. The proportion of virginal women who initiate sexual activity during follow-up is estimated to be 43.2%. We assumed that initiation of sexual activity among virginal women at entry will occur evenly throughout the four years of follow-up.
6. The proportion of these sexually experienced women who are both HPV16 DNA positive and HPV18 DNA positive at entry is estimated to be 1%.
7. The proportion of sexually experienced women who remain HPV16 or HPV18 negative at the end of the vaccination phase (month 6) among those who are HPV16 and HPV18 DNA negative at entry is 98.5%.
8. The proportion of women diagnosed with CIN2+ in the first 6 months of the study is estimated to be 0.2%.
9. We expect that 10% of women will receive less than three vaccine doses.
10. Women will be followed for a total of four years from first vaccination.
11. We expect 10% of participants to be lost to follow-up during our study. The dropout rate is expected to be evenly distributed throughout the four years of follow-up.
12. Two-sided alpha of 0.05. Total accrual into the Costa Rica Trial (HPV-009) was lower than initial estimates. For the outcome of histological CIN2+ (expected # events = 24), the final accrual of 7,466 provides 91% power to detect a vaccine efficacy of 80% under the following assumptions:
## Oversight by the data and safety monitoring board & advisory board
A Data and Safety Monitoring Board (DSMB) has been established by NCI to review and monitor the trial. The Charter Document for this DSMB is provided separately from this protocol. The DSMB membership consists of individuals with expertise in statistics, gynecology, pathology, clinical trials, and health advocacy. A list of the DSMB members is provided separately from this protocol in the DSMB Charter. Members of the DSMB are from outside the NIH to assure an independent review of the trial. DSMB members have no conflicts of interest with pharmaceutical companies currently involved in the development of prophylactic HPV vaccines. The DSMB is composed of both United States and Costa Rican nationals to assure joint review of the trial by representatives of both countries involved.
The primary role of the DSMB will be to:
- Monitor accrual and event rates to confirm that assumptions underlying the trial design are adequate.
- Monitor adverse event data at yearly intervals, or more frequently if needed, to assure the safety of participants.
- Monitor efficacy, should the need for an interim analysis arise, considering the guidelines which are provided separately from this protocol as part of the Charter Document for the DSMB. It should be noted that no interim analysis is planned at present.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -77 -After each meeting the DSMB will submit its recommendations to the NCI Principal Investigator, the DCEG Scientific Director, the Costa Rican Principal Investigator, and the NCI and Costa Rican IRBs. The NCI Principal Investigators will share these recommendations with GSK Biologicals.
In addition to the DSMB, the NCI has formed an External Advisory Group comprised of experts in virology, vaccinology, gynecology, oncology, pathology, epidemiology, statistics, medical ethics, and health advocacy. This group will advise the NCI Principal Investigator and the DCEG Scientific Director on other issues that will undoubtedly arise during the course of the trial. For example, the External Advisory Group may be asked to:
- Review and approve the trial protocol and procedures prior to initiation of the trial.
- Review concepts for ancillary studies to assure that their implementation will not negatively impact the primary or secondary objectives of our trial.
- Review major publications relating to the primary or secondary objectives of our trial.
Members of the External Advisory Group are from outside the NIH to assure independence. Members have no conflicts of interest with pharmaceutical companies currently involved in the development of prophylactic HPV vaccines. The External Advisory Group is composed of both United States and Costa Rican nationals to assure joint review of the trial by representatives of both countries involved. ]. The principles of respect for persons, beneficence, and justice will be paramount in planning our trial. Participants will be recruited into our trial only after they have been informed about the trial procedures, have understood what is being asked of them, and signed an informed consent form. A draft of the Informed Consent form is provided separately from this protocol. The subject information sheet and written informed consent form as well as the informed consent process employed during our trial will comply with Section 4.8 (Informed Consent of Trial Subjects) of the ICH guideline for Good Clinical Practices. Our trial will be limited to women, since the goal of the trial is to investigate vaccine efficacy against a disease that occurs among women only. Women 18-25 years of age at enrollment will be eligible to participate. The lower age limit of 18 was chosen since this is the age at which individuals in Costa Rica are considered adults and the age at which they can legally self-consent to participation in research projects. Since the US National Institutes of Health (NIH) defines women younger than 21 as children, our trial will include "children" (women ages 18-20 years) based on the US NIH definition.
## Human subjects protection
Trial plans and procedures will be reviewed by the NCI Institutional Review Board (IRB). A local IRB in Costa Rica will also review this project. The Costa Rican IRB The trial will be conducted in a manner that maximizes patient confidentiality. A trialspecific patient identification number will be assigned to each participant to avoid use of name or other patient identifiers (e.g., national ID number). Separate specimen identification numbers will also be used (distinct from the patient identification numbers) to provide an additional level of privacy protection. Interviews and clinical examinations will be conducted in a private setting by study staff that have been carefully trained and are aware of the need to maintain patient confidentiality. Data collected as part of this trial will be coded and keyed into a secure data management system developed specifically for this trial. Information that is sent in an expedited manner to NCI and GSK (e.g., SAE reports) will contain only the study-specific identification number and will not contain names or other patient identifiers (e.g., national ID number). Analysis datasets generated for this trial will not contain patient names, and will instead rely on the trialspecific patient and specimen identification numbers.
Since our trial will involve a medical intervention on a large number of women, an independent DSMB has been established to monitor adverse events and trial progress (see Section 5). The DSMB will perform periodic review of adverse events reported during the trial. After each review, members of the DSMB will summarize their findings and will report them to the NCI Principal Investigator and to the NCI and local Costa Rican IRBs. Should the NCI or local IRB have any concerns regarding trial adverse events, these will be addressed directly to the DSMB with a copy to the Costa Rica and NCI Principal Investigator. Full communication between the human subjects review committees and the DSMB is essential and will be the responsibility of the NCI Principal Investigator.
## Vaccine manufacturing & quality control
The vaccines to be used in our trial will be manufactured by GSK Biologicals and provided to the NCI under a Clinical Trials Agreement (CTA). The HPV16/18 L1 VLP and control (HAV) vaccines will be manufactured under strict GMP conditions. Three lots of HPV16/18 VLP vaccine have been prepared to support the Phase III efficacy trial. Each HPV16/18 VLP vaccine lot is manufactured using one single batch of HPV16 L1 VLP antigen and one single batch of HPV18 L1 VLP antigen. Four lots of the investigational formulation of Havrix control vaccine have been prepared to support the Phase III efficacy trial. Formulation of the Havrix control vaccine lots was performed using one batch of purified inactivated Hepatitis A antigen. Individual use syringes will CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -79 -be prepared and pre-labeled with individual specimen ID numbers. Exhaustive quality control procedures are implemented during the manufacturing, syringe filling, and labeling process, and include an assessment of product purity, identity, strength/potency, and safety. Details regarding vaccine manufacturing and quality control procedures can be found in the GSK Investigator Brochure.
Each lot of vaccine prepared for the Costa Rica trial will undergo extensive stability studies in which both the bulk and final product packaged in syringes will be tested to assure continued purity, stability, and strength/potency of the vaccine. Such testing will be performed at predefined intervals and at temperatures ranging from 37°C to 4°C.
Studies to date indicate that the vaccine is stable for two years when stored at 2°C to 8°C. Moreover, experience during Phase II demonstrated that vaccine handled in the field at more than 50 different study centers in the USA, Canada, and Brazil was highly immunogenic; overall, more than 98% of women receiving 3 doses of HPV16/18 VLP vaccine developed antibodies to HPV16 and HPV18, compared to less than 5% of women receiving 3 doses of control vaccine. Nonetheless, results from stability studies performed at GSK Biologicals on the lots of vaccine manufactured for the Costa Rica trial will be reviewed by the NCI and Costa Rica Principal Investigators on an ongoing basis. Should evidence of >50% VLP disassociation or loss in potency be observed, a new lot of vaccine will be manufactured by GSK Biologicals to replace the original lot in Costa Rica.
## Collaborators and consultants
A list of collaborators and consultants, the medical and scientific contributions that they will provide to the NCI HPV16/18 VLP Vaccine Trial in Costa Rica, and their institutional affiliations is provided separately from this protocol.
The main collaborating institutions involved in this trial and their role in the trial are summarized below.
## U.s. national cancer institute (nci):
The NCI, in close collaboration with its Costa Rican collaborators, is responsible for the overall design and implementation of the trial. NCI will also be responsible, in collaboration with its Costa Rican collaborators, for the analysis of data obtained from this trial. The trial is funded by the NCI.
## Proyecto epidemiologico guanacaste (peg): peg investigators, in close collaboration
with its NCI collaborators, are responsible for the overall design of the trial and for conducting the field activities required for its implementation. PEG investigators are also responsible, along with its NCI collaborators, for the analysis of data obtained from this trial.
## Glaxosmithkline biologicals (gsk):
Under a Clinical Trials Agreement with the NCI, GSK has agreed to provide the vaccines to be used for the trial. The trial will be conducted under a GSK IND with the US FDA and as such GSK will be responsible for reporting to the US FDA. Data generated by this trial will be used by GSK for submission to regulatory authorities in the United States and elsewhere.
## Information management services (ims):
Under contract with the NCI, IMS is responsible for the development and maintenance of the data management system used for our trial. IMS will also assist NCI in the preparation of reports for the DSMB and other review entities, in running of edit reports on data received from Costa Rica, and in data analysis.
## Boston biomedica inc.-biotech research laboratories (bbi):
Under contract with the NCI, BBI will serve as the primary repository for specimens collected from this trial. BBI will also provide assistance to the NCI and Costa Rican investigators in the set-up and maintenance of the local repository in Costa Rica. Biopsy Biopsy Colpo 6 months Colpo within 6 months *Unsatisfactory means no lesion was seen in ectocervix but a portion of the endocervical canal was not visible. If unsatisfactory for other reasons, reschedule colposcopy visit. **If the reason for the unsatisfactory colposcopy can be modified (e.g., heavy discharge, etc) the woman will be appointed for a new colposcopy as soon as possible. If the reason can not be modified (TZ inside the canal with narrow cervical os) the participant will be given a new appointment to colposcopy (including cytology with a Cytobrush) within 2 months. If the high-grade lesion is confirmed LEEP will be performed. If the second test is negative, a third colposcopy will be done six months later.
## Projected timetable
## Review of protocol by external advisory group
## Approach to labeling and distribution of randomized vaccine and control syringes
Women identified in the enumeration survey ages 18 to 25 at the time of enrollment will be invited to participate in the trial. Enrollment will occur at study clinics in the Guanacaste province. Subjects will not be randomly assigned to a particular arm (i.e., material type) prior to study start. Instead, randomization to one of the two trial arms will occur when the subject receives her first vaccination. The procedure for vaccine labeling and distribution is described below.
Vaccine and control syringes will be appropriately labeled using a randomized number list associated with the vaccine and the control syringes. Vaccine identification numbers will be randomized to HPV16/18 VLP and control vaccine syringes by NCI using a standard SAS (Statistical Analysis System) program. The vaccine identification numbers will uniquely identify the vaccine doses to be administered to the same subject. The randomization listing will be used by the NCI to generate vaccine labels. Labels containing the randomized numbers will be provided to GSK Biologicals in Rixensart, where they will be applied to the respective syringes. Labeled HPV16/18 VLP vaccine and control syringes will be combined, sorted in numerical order, and delivered in sequentially numbered boxes to the study site in Costa Rica. Syringes will be pulled in numeric order for vaccination in the study clinics from these boxes.
In preparation for vaccine distribution to the clinics at the enrollment visit, first dose vaccine syringes will be pulled in numeric order and placed in boxes for transport to each clinic according to the number of subjects appointed to each clinic. The vaccine syringes to be used in administration at enrollment are already labeled with a label containing the sample vaccine Identification number (vaccine ID) and the sequence number representing the dose (e.g., VX 21234-001) of the vaccine being administered. As an example, for an enrollment visit at Clinic 1 where 25 subjects have been invited, vaccines starting with a vaccine ID of VX12345-001 and continuing through VX12369-001 will be placed in a transport box in a serpentine fashion by ascending number. Boxes filled with syringes for Clinic 2 will be filled with syringes VX12370-001 to VX12394-001 for 25 potential subjects, and so on. As a subject is determined to be eligible, the next available syringe with the lowest value vaccine ID number will be pulled for application. In this manner, as subjects randomly enter the study, the integrity of randomization carried out during the numbering and labeling process is maintained and study site personnel and subjects are blinded to material type.
However, it should be noted that while this pattern of placing syringes by ascending numeric order will be followed for distribution of vaccine and control syringes, the syringes may not be sequentially ordered after the first day of the trial. This is due to the fact that not all invited subjects will become eligible to participate in the trial, and some first dose syringes in each batch distributed to a clinic may be left unused. Unused syringes will be returned to inventory on a daily basis. Upon assignment of subjects to the next day's clinics, the next available first dose syringe with the lowest vaccine ID will be used until a sufficient number of syringes are packed for distribution at each clinic. Due to this process, the vaccine IDs in each batch of syringes sent to a clinic could have some numeric gaps between syringe identification numbers. However, the pattern of placing the lowest vaccine ID followed by ascending vaccine IDs will continue, thereby ensuring the assignment of a syringe in an ordered fashion, thus preserving randomization.
As the trial progresses, the second and third doses of vaccine and control syringes will be needed. The syringes for the second and third doses will be selected based on linkage of the vaccine ID to the first dose so as to maintain the same material type for each participant. A subject who received a vaccination from a syringe with a vaccine ID label of VX12345-001, would receive her second and third doses from syringes with a vaccine ID of VX12345-002 and VX12345-003, respectively. The second and third doses of the vaccine and control syringes will also be distributed in each clinic in ascending numeric order. This method of placement in transport boxes and use in the field allows easy accessibility for study personnel. Moreover, this procedure provides for the initial randomization of material type to a subject and the assurance that the second and third dose for each subject is the same material type as the first dose received.
In the event that an assigned syringe (i.e., a dose 2 or dose 3 syringe) is determined to be unusable, additional doses will be available on site to be used as replacements. In case an assigned syringe dose is broken or unusable, the investigator will replace it with a replacement syringe, using a web-based system developed by trial Data Management Center (DMC), under contract with the NCI. This system will ensure a fast, secure, and masked process for assigning replacement syringes. Only authorized personnel with valid logon and password information will be allowed to access the web-based replacement system; no treatment information will be accessible via this system. The replacement process will be documented on the vaccine administration page of the CRF and on the web-based replacement application. c. Requires hospitalization or prolongation of existing hospitalization NOTE: In general, hospitalization signifies that the subject has been detained (usually involving at least an overnight stay) at the hospital or emergency ward for observation and/or treatment that would not have been appropriate in the physician's office or outpatient setting. Complications that occur during hospitalization are AEs. If a complication prolongs hospitalization or fulfils any other serious criteria, the event is serious. When in doubt as to whether "hospitalization" occurred or was necessary, the AE should be considered serious.
Hospitalization for elective surgery related to a pre-existing condition that did not worsen from baseline is not considered an AE.
d. Results in persistent or significant disability/incapacity, e. Is a congenital anomaly/birth defect in the offspring of a study subject, NOTE: Medical or scientific judgement should be exercised in deciding whether reporting is appropriate in other situations, such as important medical events that may not be immediately life-threatening or result in death or hospitalization but may jeopardize the subject or may require medical or surgical intervention to prevent one of the other outcomes listed in the above definition. These should be considered serious. Examples of such events are invasive or malignant cancers, intensive treatments in an emergency room or at home for allergic bronchospasm, blood dyscrasia, or convulsion but do not result in hospitalization.
## Lack of efficacy
"Lack of efficacy" per se will not be reported as an AE. The signs and symptoms or clinical sequelae resulting from lack of efficacy will be reported if they fulfil the AE or SAE definition.
4. Clinical laboratory parameters and other abnormal assessments qualifying as adverse events and serious adverse events CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -96 -Abnormal laboratory findings (e.g., clinical chemistry, cytology) or other abnormal assessments (e.g., vital signs) that are judged by the investigator to be clinically significant will be reported as AEs or SAEs if they meet the definition of an AE, as defined in Section 1, or SAE, as defined in Section 2. Clinically significant abnormal laboratory findings or other abnormal assessments that are detected during the study or are present at baseline and significantly worsen following the start of the study will be reported as AEs or SAEs. The investigator will exercise his or her medical and scientific judgement in deciding whether or not an abnormal laboratory finding or other abnormal assessment is clinically significant.
## Time period, frequency, and method of detecting adverse events and serious adverse events
The standard time period for collecting and recording SAEs will begin after receipt of the participant's informed consent signature and will end at the concluding visit, which will occur at approximately 48 months following administration of the first dose of study vaccine for each subject. (Amended: 29 Mar 2005) See Section 8 for instructions for reporting and recording SAEs.
Additionally, in order to fulfill international reporting obligations, SAEs that are related to study participation (e.g., procedures, invasive tests, a change from existing therapy) or are related to a concurrent medication will be collected and recorded from the time of consent until the subject completes her concluding visit or withdraws from the study (e.g., early withdrawal from the study, loss to follow-up). (Amended: 29 Mar 2005)
The investigator will record all AEs during the entire active phase (including the approximately 48 month follow-up after consent (Amended: 29 Mar 2005), SAEs during the entire study period, and the onset of chronic illness(es) (diabetes, autoimmune disease, asthma, and allergies), emergency room (ER) visits and physician office visits that are not related to routine physical examination, vaccination, or common acute illnesses such as upper respiratory infection, sinusitis, pharyngitis, gastroenteritis and injury.
The AER and any other relevant forms (e.g., VAR, HV2) (Amended: 29 Mar 2005) will be completed for every report that meets the definition of an adverse event. The investigator will attempt detection of AEs and SAEs in the following ways: For all subjects the investigator will monitor and record the occurrence of reactions in the first 60 (Amended: 31 Oct 2005) minutes following administration of the vaccine at every vaccination visit (baseline [0 month], 1 month and 6 months) using the Vaccine Administration and Reactogenicity Monitoring (VAR) form. In addition, 10% of women will receive home visits from outreach workers 3 to 6 days post vaccination to monitor AEs (vaccination day is defined as day zero for this purpose). The Home Visit Reactogenicity Monitoring (HV2) form will be used for these visits. Spontaneous reports of solicited adverse events that are reported by phone to the physician on-call or inperson during a vaccination visit at the clinic will be documented on the AER (Amended: 29 Mar 2005) form. In addition, the AER form will be completed by fieldworkers during home visit to participants that had unsolicited AEs or SAEs during a vaccination clinic visit or other adverse event for which the physician orders follow-up to be conducted. Adverse events will also be monitored on the Post-Enrollment Eligibility Screener (PES) form which is administered at the one month and six month vaccination visits and at follow-up visits occurring at , and 48 months after the first vaccination visit and on the Colposcopy Intake Form (CIF) for women who are referred for colposcopic evaluation. Another surveillance effort will be periodic query to identify study subjects on summary lists of women in the same age range who were admitted to hospitals or ER clinics in Guanacaste and selected referral hospitals outside of Guanacaste. The investigator will evaluate all AEs observed by the investigator or one of his/her clinical collaborators or reported by the subject, whether spontaneously, or in response to a direct question. AEs not previously documented in the study will be recorded on the AER form. The nature of each event, date and time (where appropriate) of onset, outcome, resolution date, intensity and relationship to vaccination will be established.
When an AE/SAE occurs, it is the responsibility of the investigator to review all documentation (e.g., hospital progress notes, laboratory, and diagnostics reports) relative to the event. The investigator will then record all relevant information regarding an AE/SAE on the AER form. There may be instances when the NCI Medical Monitor will request additional abstraction of information from the participant's medical records. GlaxoSmithKline (GSK) Biologicals, hereafter referred to as the IND holder, may also request additional information, but will do so via the NCI Medical Monitor or NCI Project Officer (Amended: 31 Oct 2005). In these instances, all subject identifiers will be removed by the investigator prior to document transmission.
The investigator will attempt to establish a diagnosis of the event based on signs, symptoms, and/or other clinical information. In such cases, the diagnosis should be documented as the AE/SAE and not the individual signs/symptoms.
## Solicited adverse events
For all subjects the investigator will monitor and record the occurrence of reactions in the first 60 (Amended: 31 Oct 2005) minutes following administration of the vaccine at every vaccination visit (baseline [0 month], 1 month and 6 months) using the VAR form.
In addition, a random sampling of 10% of the women enrolled in the trial will receive home visits from outreach workers 3 to 6 days after each vaccination to monitor for AEs (vaccination day is defined as day zero). The same randomly selected 10% sample of women will receive home visits after each vaccination. The HV2 form will be used during these 3-6 day post vaccination home visits. With the exception of this sample of 10% of participants who are scheduled to receive a home visit, the AER (Amended: 29 Mar 2005) form will be used to follow up on all other AEs that occur in days 0-6 after vaccination. For example, the AER (Amended: 29 Mar 2005) form will be used to follow up on spontaneously reported AEs as well as AEs that are identified during physician directed home visits.
## Solicited local (injection site) aes
Standardized data will be collected for local injection site AEs on the VAR and HV2 (Amended: 29 Mar 2005) for pain, redness, swelling, and temperature at the injection site. Subjects will be asked, "Do you feel pain at the injection site"? Those who answer "Yes", will be asked to rate the intensity of the pain at the injection site. Possible The following observations will also be made of the injection site:
Area of redness __ __ mm Area of swelling __ __ mm Temperature at injection site ("Normal" or "Warmer than surrounding area")
## Other solicited general aes
## Sixty-minute (amended: 31 oct 2005) post-vaccination monitoring
Pre-vaccination oral temperature and 60-minute (Amended: 31 Oct 2005) postvaccination oral temperature are recorded on the VAR form. Standardized data are collected on the VAR form to document the following general solicited AEs: fatigue, myalgia, arthralgia, GI symptoms (nausea, vomiting, diarrhea, and/or abdominal pain), headache, rash (other than urticaria), and urticaria. Subjects reporting any of these solicited symptoms will be asked to rate the intensity of each symptom (mild, moderate, severe). If a solicited general AE meets the criteria of an SAE this is indicated on the form and an AER form is completed.
## Home visit reactogenicity monitoring
Oral temperature is recorded on the HV2 form and any temperature above 39°C is documented on the AER form and followed to resolution. Fever can be classified as an SAE at the discretion of the physician. Subjects providing responses to home visit interviews will be asked whether they experienced the following solicited symptoms since their most recent vaccination: fatigue, pain in their muscles, aches or pain in their joints, GI symptoms (nausea, vomiting, diarrhea, and/or abdominal pain), headache, rash (other than urticaria), and urticaria. Subjects reporting any of these solicited symptoms will be asked to rate the maximum intensity of each symptom (mild, moderate, severe). If a solicited general AE meets the criteria of an SAE, this is indicated on the HV2 form and an AER form is completed.
Possible responses to each solicited symptom: 1 (Yes), 2 (No)
Next they will be asked, "Do you still have this adverse event?"
Possible responses: 1 (Yes), 2 (No)
They will be asked about the maximum intensity of the event: 1=Mild, 2=Moderate, 3=Severe.
Participants will also be asked, "How long have you had/did you have this condition?" 0 (<1 day), 1 (1-2 days), 2 (≥3 days) CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -99 -
## Spontaneous reports of solicited aes
Spontaneous reports of AEs will be captured directly on the AER form. (Amended: 29 Mar 2005)
## Unsolicited aes
The VAR and HV2 forms have fields for the collection of information on unsolicited conditions that are observed or reported after vaccination (VAR form) or at the 3-6 day home visit following vaccination for 10% of the women (HV2 form). When completing these forms, if any unsolicited AEs are reported, the person completing the form is instructed to complete an AER form. Also, unsolicited AEs reported outside the context of a clinic visit or a home visit on Days 3-6 after vaccination for 10% of the women will be documented directly on the AER form. (Amended: 29 Mar 2005)
The Post Enrollment Eligibility Screener form (PES) is used at all scheduled postenrollment clinic appointments. The Colposcopy Intake Form (CIF) is used at the intake visit to the colposcopy clinic. The PES and CIF forms contain essentially identical questions about any hospitalizations, doctor visits, or illnesses lasting more than one day that have occurred since the last study visit. An AER form is to be completed for any such occurrences.
6. Evaluating adverse events and serious adverse events
## Assessment of intensity
The site investigator or designee(s) will be responsible for direct measurements related to several solicited AEs: oral temperature (measured in Celsius), the area of redness and area of swelling at the injection site (measured in millimeters), and temperature at the injection site.
## Temperature at injection site
(0) Normal: The injection site is no warmer than the surrounding area.
(1) Warmer than the surrounding area
## Other signs and symptoms
The area of local injection site redness/swelling will each be measured in millimeters and recorded on the relevant forms (i.e., the VAR or HV2 form). (Amended: 29 Mar 2005)
The maximum intensity of local injection site redness/swelling will be scored for purposes of analysis as follows:
(0) None
(1) >0 mm to ≤20 mm
(2) >20 mm to ≤50 mm
(3) >50 mm
In addition, the intensity of symptoms will be solicited for the following AEs: pain at the injection site, fatigue, myalgia, arthralgia, GI symptoms (nausea, vomiting, diarrhea and/or abdominal pain), headache, rash (other than urticaria), and urticaria. The following severity/intensity scale will be used for this purpose:
(0) None
(1) (Mild) = An AE which is easily tolerated by the subject, causing minimal discomfort and not interfering with everyday activities.
(2) (Moderate) = An AE which is sufficiently discomforting to interfere with normal everyday activities.
(3) (Severe)=An AE which prevents normal, everyday activities. (In adults/ adolescents, such an AE would, for example, prevent attendance at work/school and would necessitate the administration of corrective therapy.)
For urticaria, maximum intensity of the event will be classified as:
(1) Urticaria occurring on 1 body area only.
(2) Urticaria occurring on 2 or 3 body areas.
(3) Urticaria occurring on at least 4 body areas.
NOTE: An AE that is assessed as severe should not be confused with a Serious Adverse Event. "Severe" is a category utilized for rating the intensity of an event, and both AEs and SAEs can be assessed as "severe". An event is defined as "serious" when it meets one of the pre-defined outcomes as described in Section 2 above (e.g., adverse events that are life-threatening, fatal, or result in prolonged hospitalization, persistent or significant disability/incapacity, or congenital anomaly).
## Fever
Fever will be recorded on the VAR or HV2 (Amended: 29 Mar 2005) form, as appropriate, and will be categorized into severity/intensity categories for analysis purposes according to the following oral temperature cut-offs: CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
(0) <37.5 °C
(1) ≥37.5 °C to <38.0 °C
(2) ≥38.0 °C to 39.0 °C
(3) >39.0 °C Fever will be classified as a serious adverse event based on the medical and scientific judgement of the site investigator or designee(s).
## Assessment of causality
The investigator is obligated to assess the relationship between investigational product and the occurrence of each AE/SAE. The investigator will use clinical judgement to determine the relationship. In doing so, alternative causes, such as the natural history of the underlying diseases, concomitant therapy, other risk factors and the temporal relationship of the event to the investigational product will be considered. The investigator will provide this information by completing Part B of the AER form (section to be completed by the physician regarding diagnosis, relationship, additional information, etc.).
The relationship of the adverse event to the vaccination will be coded as follows:
(0) None
(1) Possibly related
(2) Probably related
(3) Definitely related
The code and signature of the physician, as well as the date that Part B was completed by the physician will be entered on the AER form.
There may be situations when an SAE has occurred and the investigator has minimal information to include in the initial report to the Westat Regulatory Associate; however, it is important that the investigator make an initial assessment of causality for an event prior to transmission of the AER form.
## Follow-up of adverse events and serious adverse events and assessment of outcome
After the initial AER form is completed, the investigator will follow up and provide information on the subject's condition. The AER form (question 5) identifies the report as either a new event or a follow-up to a previous adverse event.
All AEs and SAEs documented at a previous visit/contact and designated as unresolved will be reviewed at subsequent visits/contacts. CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -102 -Investigators will follow up subjects:
- with SAEs, until the event has resolved, subsided, stabilized, or disappeared; the event is otherwise explained; or the subject is lost to follow-up;
- or, in the case of other non-serious AEs, until the event has resolved, subsided, stabilized, or disappeared; the subject completes the study; or the subject is lost to follow-up; whichever occurs first.
Clinically significant laboratory abnormalities will be followed up until they have returned to baseline, or a satisfactory explanation has been provided. Reports relative to the subsequent course of an SAE noted for any subject must be submitted to the Westat Regulatory Associate.
The NCI Medical Monitor or IND holder, via the NCI Medical Monitor or NCI Project Officer (Amended: 31 Oct 2005), may request that the investigator obtain additional information from the participant's clinical records to elucidate as fully as possible the nature and/or causality of the AE or SAE. The investigator is obliged to assist. If a subject dies during participation in the study or during a recognized follow-up period, the investigator will obtain copies of any available post-mortem findings, including histopathology, and will forward them to Westat's Regulatory Associate for appropriate distribution.
Updated information will be recorded on new AER forms linked to the original form with an AE number provided by the study data management system. The updated SAE report form should be resent to the Westat Regulatory Associate within 24 hours of receipt of the follow-up information as outlined in Section 8. (Amended: 13 Nov 2008)
Outcome of any non-serious AE (i.e., unsolicited AE) or any SAE reported during the entire study will be assessed as indicated by AER form question 8 (Resolution Code). SAEs will be reported promptly to the Westat Regulatory Associate once the investigator determines that the event meets the protocol definition of an SAE. In the event of a life-threatening or fatal SAE, the investigator or designate will telephone the Westat Regulatory Associate to inform him/her about the SAE and also fax the completed AER form and supporting documentation to the Westat HPV Regulatory Associate WITHIN 24 HOURS OF HIS/HER BECOMING AWARE OF THESE EVENTS. All other SAEs, regardless of relationship, will be reported to the Westat Regulatory Associate within three calendar days of the initial report via email (notification only) and fax (AER form and supporting documentation). The Westat Regulatory Associate will forward these expedited SAE reports to the NCI Medical Monitor and IND holder within one business day of receipt from Costa Rica, for their review and assessment. Additional or follow-up information relating to the initial SAE report is also reported to Westat within 24 hours of investigator receipt and forwarded from Westat to the NCI Medical Monitor and IND holder within one business day of receipt of such additional information from Costa Rica.
## = recovered
## Completion and transmission of serious adverse event reports to westat
When an investigator becomes aware of an SAE for a study subject, she/he will report the information to Westat's Regulatory Associate who is responsible for coordinating receipt and distribution of SAE reports to the NCI Medical Monitor and IND holder within the time-frames outlined in Section 8.1 and. The AER form will always be completed with all available details of the event, signed by the investigator (or designee), and forwarded to Westat within the designated time-frame. If the investigator does not have all information regarding an SAE, he/she will not wait to receive additional information before notifying the Westat Regulatory Associate of the event by faxing the completed form. The form will be updated when additional information is received and forwarded to Westat as outlined in Section 8.1. The investigator will provide an assessment of causality at the time of the initial report as described in Section 6.2.
Facsimile (fax) transmission of the AER Report Form is the preferred method to transmit SAE information to the Westat Regulatory Associate. In rare circumstances (e.g., the absence of facsimile equipment), notification by telephone is acceptable, with a copy of the AER form sent by overnight mail. Initial notification via the telephone does not replace the need for the investigator to complete, sign, and transmit the AER form as outlined in Section 8.1.
## Regulatory reporting requirements for serious adverse events
The investigator promptly reports SAEs to Westat's Regulatory Associate, who forwards them to NCI's Medical Monitor and the IND holder as outlined in Section 8. The holder of the IND is responsible for notifying the Food and Drug Administration (FDA) and other regulatory bodies about the safety of a product under clinical investigation. Prompt reporting of an SAE to the IND holder is essential to meet legal and ethical responsibilities for the safety of all subjects. The IND holder reports SAEs to FDA following these timelines:
## Event category reporting timelines
Non life-threatening Within 15 calendar days from the day the sponsor first received knowledge of the event Life-threatening or Fatal Notify FDA by phone or fax within 7 calendar days from the day the sponsor first received knowledge of the event, followed by as complete a report as possible within 8 additional calendar days
The investigator, or responsible person according to local requirements, will comply with the applicable local regulatory requirements related to the reporting of SAEs to local regulatory authorities and the IRB/IEC.
## Post study adverse events and serious adverse events
A post-study AE/SAE is defined as any event that occurs outside of the AE/SAE detection period defined in Section 5. Investigators are not obligated to actively seek AEs or SAEs in former study participants. If the investigator learns of any SAE, including a death, at any time after a subject has been discharged from the study, and considers the event reasonably related to the investigational product, the investigator will notify the sponsor. (Amended: 13 Nov 2008)
## Pregnancy
Urine pregnancy tests are performed at the 0, 1, and 6-month vaccine clinic visits. The investigator, or his/her designee, will record the pregnancy test result for the 0 month vaccination clinic visit on the Initial Visit Eligibility Screener (ES) form and the pregnancy test result for the 1 and 6-month visits on the Post-Enrollment Eligibility Screener (PES) form. Subjects whose pregnancy test yields a positive result at enrollment cannot receive the study vaccine. The participant will be examined for eligibility for the study once again when she is at least three months postpartum and no longer breastfeeding. If the participant is pregnant at the 1-month or 6-month visit she will have her participation in the study deferred until she is at least three months postpartum and no longer breastfeeding. At that time she will be eligible to continue in the study.
The PES, PO (Amended: 13 Nov 2008), PED (Amended: 29 Mar 2005) and CIF forms contain questions about pregnancies and pregnancy outcomes that have occurred since the last study visit. The principal investigator can also receive information about new pregnancies and pregnancy outcomes through other direct or indirect contacts with participants in the study (e.g., at home reminder visits). All new pregnancy and pregnancy outcome information for study participants, whether or not an SAE is involved, will be transmitted to the Westat Regulatory Associate via a read-only electronic mail report within 5 business days (i.e., no longer than 40 business hours) (Amended: 13 Nov 2008, Amended: 10 September 2009) of receipt by the principal investigator or designee. The Westat Regulatory Associate will, in turn, transmit the electronic mail to the NCI Medical Monitor and GSK Global Clinical Safety and Pharmacovigilance within one business day of receipt of the electronic mail from Costa Rica, documenting the receipt date, file name and transmittal date. While pregnancy itself is not considered an AE or SAE, all pregnancies will be followed to term.
A complication of pregnancy or delivery will be recorded as an AE or an SAE, as described in Section 1 and Section 2, and will be followed as described in Section 8. (Amended: 13 Nov 2008) A miscarriage is always considered an SAE and will be reported as outlined in Section 8. (Amended: 13 Nov 2008) Furthermore, SAEs occurring as a result of a post-study pregnancy AND which the investigator considers reasonably related in time to receipt of the investigational product, will be reported to the Westat Regulatory Associate as in Section 8. (Amended: 13 Nov 2008) While the investigator is not obligated to actively seek this information from former study participants, he/she may learn of a pregnancy through spontaneous reporting. Information on pregnancies identified during the screening phase/prior to first vaccine administration does not need to be communicated as a safety issue.
## Treatment of adverse events
Treatment of adverse events is at the discretion of the investigator and according to current good medical practice. Medications taken by a subject for treatment of an AE that are either immunosuppressive or that the physician deems may help clarify some aspect of the AE, whether or not administered by a study physician, will be recorded in Part C, section II of the AER form.
## Section 3 hazards identification
## Fire and explosion
This product is classified as non-flammable.
## Health
Handling this product in its final form presents minimal risk from occupational exposure. Health effects information is based on hazards of components.
## Environment
No information is available about the potential of this product to produce adverse environmental effects.
## Section 4 first-aid measures
Ingestion Never attempt to induce vomiting. Do not attempt to give any solid or liquid by mouth if the exposed subject is unconscious or semi-conscious. Wash out the mouth with water. If the exposed subject is fully conscious, give plenty of water to drink. Obtain medical attention.
## Inhalation
Physical form suggests that risk of inhalation exposure is negligible.
## Skin contact
Using appropriate personal protective equipment, remove contaminated clothing and flush exposed area with large amounts of water. Obtain medical attention if skin reaction occurs, which may be immediate or delayed. The disposal method for rejected products/returned goods must ensure that they cannot be re-sold or reused.
## Regulatory requirements
Observe all local and national regulations when disposing of this product.
## Section 14 transport information
The SDS should accompany all shipments for reference in the event of spillage or accidental release. Only authorised persons trained and competent in accordance with appropriate national and international regulatory requirements may prepare dangerous goods for transport.
## Un classification and labelling
Transport Information Transportation and shipping of this product is not restricted. It has no known, significant hazards requiring special packaging or labelling for air, maritime, US or European ground transport purposes.
## Section 15 regulatory information
The information included below is an overview of the major regulatory requirements. It should not be considered to be an exhaustive summary. Local regulations should be consulted for additional requirements.
## Eu classification and labelling exempt from requirements of eu dangerous
Preparations directive -product regulated as a medicinal product, cosmetic product or medical device. The information and recommendations in this safety data sheet are, to the best of our knowledge, accurate as of the date of issue. Nothing herein shall be deemed to create any warranty, express or implied. It is the responsibility of the user to determine the applicability of this information and the suitability of the material or product for any particular purpose.
## Health
Handling this product in its final form presents minimal risk from occupational exposure. May produce allergic skin reactions. Possible effects of overexposure in the workplace include: symptoms of hypersensitivity (such as skin rash, hives, itching, and difficulty breathing). Health effects information is based on hazards of components.
Environment No information is available about the potential of this product to produce adverse environmental effects.
## Section 4 first-aid measures
Ingestion Never attempt to induce vomiting. Do not attempt to give any solid or liquid by mouth if the exposed subject is unconscious or semi-conscious. Wash out the mouth with water. If the exposed subject is fully conscious, give plenty of water to drink. Obtain medical attention.
## Inhalation
Physical form suggests that risk of inhalation exposure is negligible.
Skin Contact Using appropriate personal protective equipment, remove contaminated clothing and flush exposed area with large amounts of water. Obtain medical attention if skin reaction occurs, which may be immediate or delayed.
Eye Contact Wash immediately with clean and gently flowing water. Continue for at least 15 minutes. Obtain medical attention.
## Notes to health professionals
Medical Treatment None. Conditions to Avoid None for normal handling of this product.
## Section 11 toxicological information
Oral Toxicity Not expected to be toxic following ingestion.
Inhalation Toxicity No studies have been conducted.
Skin Effects Irritation is not expected following direct contact.
Eye Effects Irritation is not expected following direct contact with eyes.
Sensitisation Allergic skin reactions might occur following dermal exposure. Assessment based upon effects of structurally similar substances. Respiratory sensitisation (allergic) reactions might occur following exposure.
Genetic Toxicity Not expected to be genotoxic under occupational exposure conditions. CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
Carcinogenicity Not expected to produce cancer in humans under occupational exposure conditions. No components are listed as carcinogens by GSK, IARC, NTP or US OSHA.
Reproductive Effects Not expected to produce adverse effects on fertility or development under occupational exposure conditions.
## Section 12 ecological information
## Summary
No information is available about the potential of this product to produce adverse environmental effects. Local regulations and procedures should be consulted prior to environmental release.
## Section 13 disposal considerations
Disposal Recommendations The disposal method for rejected products/returned goods must ensure that they cannot be re-sold or re-used.
## Regulatory requirements
Observe all local and national regulations when disposing of this product.
## Section 14 transport information
The SDS should accompany all shipments for reference in the event of spillage or accidental release. Only authorised persons trained and competent in accordance with appropriate national and international regulatory requirements may prepare dangerous goods for transport.
## Un classification and labelling
Transport Information Transportation and shipping of this product is not restricted. It has no known, significant hazards requiring special packaging or labelling for air, maritime, US or European ground transport purposes.
## Section 15 regulatory information
The information included below is an overview of the major regulatory requirements. It should not be considered to be an exhaustive summary. Local regulations should be consulted for additional requirements.
## Eu classification and labelling
Exempt from requirements of EU Dangerous Preparations directive -product regulated as a medicinal product, cosmetic product or medical device. The information and recommendations in this safety data sheet are, to the best of our knowledge, accurate as of the date of issue. Nothing herein shall be deemed to create any warranty, express or implied. It is the responsibility of the user to determine the applicability of this information and the suitability of the material or product for any particular purpose.
## List of abbreviations
## Indsr: investigational new drug safety report (amended: 29 mar 2005)
## Msds: material safety data sheet (amended: 29 mar 2005)
## Sr: spontaneous report form (deleted: 29 mar 2005)
## Synopsis rationale
The trial will enroll approximately 15,000 healthy young women… or with 720 enzyme-linked immunoabsorbent assay (ELISA) units of GSK Biologicals' (Amended: 29 Mar 2005) hepatitis A vaccine, inactivated (Havrix®) prepared as an investigational formulation.
## Study design (continued)
- Duration of the study:
- At least 30 confirmed cases of CIN2+ associated with HPV16 or HPV18 cervical infection are observed and a minimum (Amended: 29 Mar 2005) follow-up period of two years is achieved, pending review and approval of the final statistical analysis plan by the DSMB.
# Background
## Results from a recently completed phase iib trial conducted by gsk
## Active follow-up of participants (months 12-48)
- An interview to obtain information on potential vaccine (Amended: 29 Mar 2005) AEs, including hospitalizations, incapacitation lasting more than one day, pregnancies, and other events reported by participants. AEs and SAEs are further defined in APPENDIX IV.
## Adverse events monitoring
- Long-term passive monitoring. The Spontaneous Report and Adverse Event Report (Amended: 29 Mar 2005) form will be used to document all AEs reported through this mechanism. This These forms is are provided separately from this protocol.
## (amended: 29 mar 2005)
## Hpv dna testing by hybrid capture® 2 (hcii)
(for baseline/exit assessment and triage of ASC cytology)
An ASC-US cytologic interpretation will lead to "reflex" HCII testing and for the purposes of clinical management subjects who have HCII positive (or HCII-inadequate) ASC-US will be treated the same as an LSIL and subjects who have a HCII positive (or HCII-inadequate) ASC-H will be treated the same as an HSIL (Amended: 29 Mar 2005).
## Primary analysis for efficacy
- At least 30 confirmed cases of CIN2+ associated with HPV16 or HPV18 cervical infection that are eligible for the ATP analysis are observed and the last woman recruited into the study has completed a minimum median follow-up period of two years is achieved, pending review and approval of the final statistical analysis plan by the DSMB.
## Solicited local (injection site) aes
Standardized data will be collected for local injection site AEs on the VAR and, HV2 and SR (Amended: 29 Mar 2005) for pain, redness, swelling, and temperature at the injection site.
## Spontaneous reports of solicited aes
Spontaneous reports of AEs will be captured directly on the AER SR form. For each solicited general AE that is reported, the same format that is used on the HV2 form will be used to record AE information on the SR form (i.e., does the woman still have the condition, what was the maximum intensity of the condition, and how long did she have the condition). If a solicited general AE meets the criteria of a SAE, this is indicated on the SR form and an AER form is completed. (Amended: 29 Mar 2005)
## Unsolicited aes
The VAR, SR and HV2 forms have fields for the collection of information on unsolicited conditions that are observed or reported after vaccination (VAR form) or since the last study visit at the 3-6 day home visit following vaccination for 10% of the women (HV2, SR forms). When completing these forms, if any unsolicited AEs are reported, the person completing the form is instructed to complete an AER.
## Also, unsolicited aes reported outside the context of a clinic visit or a home visit on days 3-6 after vaccination for 10% of women will be documented directly on the aer form. (amended: 29 mar 2005)
## 6.
Evaluating adverse events and serious adverse events
## Other signs and symptoms
The area of local injection site redness/swelling will each be measured in millimeters and recorded on the relevant forms (i.e., the VAR, SR or HV2 form).
# Methods
## Overview of the trial
…The initial vaccination will be performed at entry, and additional vaccine doses will be given at one and six months following the initial vaccination. Women will be monitored at the clinic for a minimum of 30 60 (Amended: 31 Oct 2005) minutes following each vaccination, during which study personnel will collect information on reactogenicity of the vaccine. … - Are sexually experienced and using an effective method of birth control (including abstinence) for less than 30 days at the time of enrollment/vaccination visit (Amended: 31 Oct 2005).
## Deferral criteria
## Selection of an investigational formulation of havrix as the control vaccine
Second, Havrix 720 EL.U./0.25 mg Alum in 0.5 mL (pediatric formulation) is currently licensed in the US for use in children aged 12 months (Amended: 31 Oct 2005) 2 to 18 years on a 0, 6-12 month schedule. The 720 EL.U. dose has been demonstrated to be safe for use in children in numerous clinical trials and through post-marketing surveillance. This suggests that the formulation can be safely administered to adults.
## Clinic visits during vaccination phase (0, 1, 6, & 7 months)
Three clinic visits are planned for vaccination at 0, 1, and 6 months. At the first clinic visit, we will collect information for determining eligibility, ensure that written informed consent has been obtained prior to any study procedure, perform urine pregnancy testing, assign treatment arm by randomization, give the initial vaccination, and monitor women for at least 30 60 (Amended: 31 Oct 2005) minutes following vaccination with appropriate medical treatment readily available in case of a rare anaphylactic reaction. At the two subsequent visits at 1 and 6 months, information on AEs will be collected, subsequent vaccinations will be administered, and women will be monitored for a minimum of 30 60 (Amended: 31 Oct 2005) minutes following vaccination with appropriate medical treatment readily available in case of a rare anaphylactic reaction. After each vaccination visit, women will be provided with two 500 mg acetaminophen pills with an indication to take the medication in case of a mild fever or pain. The use of acetaminophen during the trial will be followed in the 10% home visit subset. A summary of the procedures performed at each of these visits is presented in. Intervals between study visits during the vaccination phase of the trial are provided in [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref].
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
## Solicited adverse events
For all subjects the investigator will monitor and record the occurrence of reactions in the first 30 60 (Amended: 31 Oct 2005) minutes following administration of the vaccine at every vaccination visit (baseline [0 month], 1 month and 6 months) using the VAR form.
## Thirty sixty-minute (amended: 31 oct 2005) post-vaccination monitoring
Pre-vaccination oral temperature and 30 60-minute (Amended: 31 Oct 2005) postvaccination oral temperature are recorded on the VAR form. …
## Follow-up of adverse events and serious adverse events and assessment of outcome
Clinically significant laboratory abnormalities will be followed up until they have returned to baseline, or a satisfactory explanation has been provided. Reports relative to the subsequent course of an SAE noted for any subject must be submitted to the Westat Regulatory Associate.
## The nci medical monitor or ind holder, via the nci medical monitor or nci project officer (amended: 31 oct 2005)
Westat Regulatory Associate, may request that the investigator obtain additional information from the participant's clinical records to elucidate as fully as possible the nature and/or causality of the AE or SAE. The investigator is obliged to assist. If a subject dies during participation in the study or during a recognized follow-up period, the investigator will obtain copies of any available post-mortem findings, including histopathology, and will forward them to Westat's Regulatory Associate for appropriate distribution.
Updated information will be recorded on new AER forms linked to the original form with an AE number provided by the study data management system. The updated SAE report form should be resent to the Westat Regulatory Associate within 24 hours of receipt of the follow-up information as outlined in Section 8. - is used to indicate a study procedure that requires documentation in the individual CRF. Note: The double-line borders following Months 0, 1 and 6 indicate that a home visit by a specially-trained outreach worker to monitor AEs will occur during the week following vaccination (any day from Day 3 to Day 6) in a random subset of 10% of the women. ***This blood draw will be performed on a subset of women, and at selected study visits, as follows: A) A random sample of women at 0, 12, and 36 months; B) All women referred to accelerated screening visits per management algorithm, at yearly intervals; C) All women referred for colposcopic evaluation who require a biopsy or LEEP, at the time of the colposcopy visit; D) women enrolled in the immunogenicity subcohort after March 1, 2006. (Amended: 19 May 2006) †Procedure will not be performed on virginal women. † †Pelvic examination will be performed if indicated by the case management algorithm. The protocol-specified clinical management algorithm is summarized in Appendix A. ¶Reflex HCII HPV testing will be performed automatically for all subjects with results of ASC. HCII HPV testing will also be performed on all women who have a pelvic examination at the 0 month visit. §HCII testing for Chlamydia trachomatis and Neisseria gonorrhoeae will be performed on samples obtained at the 0 month visit (or the first visit with a pelvic examination for women who are virginal at entry) only. ^ HPV testing by HCII will be performed on all specimens obtained at the 48 month visit. To permit characterization of maximal antibody titers achieved after all three doses of vaccine are administered, a blood draw is also planned a minimum of 30 days after the third vaccination (7 month visit) on a subset of the initial (Amended: 19 May 2006) 600 women randomly assigned to one of the two arms of our trial.
## Overview
All participants are expected to have at least seven visits over the course of our trial (see also. This will include:
- A clinic visit will be conducted at seven months on a subset of the initial (Amended: 19 May 2006) 600 women enrolled to allow for immunogenicity assessment a minimum of 30 days following the last vaccination.
## Blood collection
Three types of blood draws are planned… This third blood draw will be collected for ancillary research purposes on a subset of participants, including a random sample of 10% of participants at enrollment and at their 12 and 36 month visits (to document the distribution of the immune responses in our population before and after vaccination), women referred to accelerated screening visits at six-month intervals (this group would have the additional blood sample collected annually), women who are referred for colposcopic evaluation and require a biopsy or LEEP, and
## Women enrolled in the immunogenicity subcohort of our trial after march 1, 2006. (amended: 19 may 2006)
While the primary focus of this larger blood draw will be on the cryopreservation of PBMCs for immunological monitoring, residual plasma will also be saved for use in possible future assays.
## Serology testing by elisa
Serological assays for HPV16/18 by ELISA will be performed at GSK Biologicals laboratories, Rixensart, Belgium or in a laboratory sponsored by the NCI in Costa Rica, pending validation. Baseline ELISA anti-HPV16 and anti-HPV18 testing will be performed on all women enrolled. In addition, anti-HPV16 and anti-HPV18 ELISA will be performed on all blood samples (Month 0 through Month 48) collected from the 600 women enrolled into the immunogenicity subcohort within the To update the Protocol to reflect the actual enrollment figure of 7,466 subjects as well as revision of statistical assumptions.
To clarify procedures at the final visit (Month 48 visit). Given the rapid changes occurring in clinical management of cervical abnormalities, we stated in the original protocol for our trial that the colposcopy referral algorithm used for the final screening visit under this protocol (Month 48) would be reviewed and updated, as needed, prior to implementation of the 48-month visit. We have updated the protocol to reflect that the same criteria used to define women who require colposcopy evaluation used throughout the trial will be applied at the Month 48 visit. Any additional care deemed necessary by the local clinicians will be conducted after participant close-out under a separate protocol.
To include the collection of extra-cervical specimens for HPV testing (oral, anal, vulvar) at the Month 48 visit, and to include its evaluation as a tertiary objective.
To revise the cytology QC process at the Month 48 visit. We propose to add slides interpreted as negative in Costa Rica with evidence of reactive changes and that are concurrently HPV positive by HCII to the set of slides selected for 100% review in the United States. This increase in the proportion selected for review is being proposed because in an evaluation of the enrollment (pre-vaccination) cytology QC effort we observed that up to 3% (3 out of 114) of slides with reactive changes that are HPV positive might have histologically evident CIN2+. Conversely, we propose to continue to review at the previous 10% level (i.e., not to increase to 100% review) those slides that are classified in Costa Rica as either negative or negative with evidence of reactive changes without concurrent evidence of HPV positivity by HCII because our evaluation of the enrollment (pre-vaccination) cytology QC effort revealed that 0% (0 of 397) of slides in these two categories were found to have histologically evident CIN2+.
To modify the date range for the Month 48 visit window. The Month 48 window has been modified to be consistent with the length of the other routine visit windows, 360 days, from day 1381-1740. Since the 48-month visit is the last routine study visit, allowances are also made to allow women who cannot otherwise be seen within the pre-defined window.
To clarify the procedures for reporting adverse events in offspring of participants (Section 3.14-Adverse Event Monitoring).
To implement administrative changes including (1) update of sponsor signatory and laboratory name (2) clarification stating that background information in the protocol is reflective of knowledge at study start (3) removing mention of specific names of NCI pathologists responsible for histology review (4) updates to Appendix D including correction of typographical errors in the pregnancy and adverse event reporting sections.
Other minor revisions and corrections have been made and are noted throughout the protocol in bold, italic text.
Details of the revision and identification of the affected protocol sections are provided below. Amended text has been included in bold, italic text and deleted text is shown with strikethrough. Capture II (HCII) will have cytology performed at 6 monthly intervals (rather than yearly intervals). Those whose ASC-US/oncogenic HPV positive lesion or LSIL persists for 2 evaluations (whether consecutive or intermittent) will be referred for colposcopy for evaluation and excisional treatment, if warranted. Those with 3 consecutive normal cytology evaluations will return to an annual cytology evaluation schedule. For purposes of this management strategy, women with ASC-US whose concomitant specimen tests negative for HPV will be considered normal and followed at annual intervals.
Women detected as having evidence of a high-grade squamous intraepithelial lesion (HSIL), atypical glandular cells (endocervical, endometrial, or glandular; favor neoplastic or not otherwise specified) (AGC), atypical squamous cells, cannot exclude HSIL (ASC-H), or cancer will be immediately referred for colposcopic evaluation and excisional treatment, if warranted.
At the Month 48 visit we will apply the same criteria applied throughout the trial to define women who require colposcopic evaluation. Specifically, all women with evidence of cancer, HSIL, glandular lesions, ASC-H, or persistent ASC-US+/LSIL will be referred to colposcopy. Close-out of individual study participants will only occur after colposcopy referral, for women who require such referral.
## It should be noted, that after participant close-out, participants will be offered additional clinical care if needed. case management after study close-out will be defined under a separate protocol. (amended: 13 nov 2008)
In addition to the management described above, exit colposcopy will be performed following the Month 48 visit in all women who, at that time are found to have atypical squamous cells (includes ASC-H and ASC-US) (ASC), LSIL, or more severe disease; women with normal cytology at the Month 48 visit who have cytologically evident abnormalities (ASC or LSIL) detected in the 12 months preceding the Month 48 visit; and a random subset of 400 women selected from among the remaining women with normal cytology at Month 48.
Note: Given that clinical management options are rapidly evolving, this preliminary exit strategy will be reviewed with the Data and Safety Monitoring Board (DSMB) prior to its implementation. Should the final exit strategy differ from that described here, a protocol amendment will be implemented detailing the exit visit strategy and its rationale.
Duration of the study: Total enrollment was 7,466, which is lower than initial estimates. The final accrual of 7,466 provides 91% power to detect a vaccine efficacy of 80% for the outcome of CIN2+.
## (amended: 13 nov 2008)
Assuming 15,000 women enroll, it is expected that a total of approximately 50 HPV16/18+ CIN2+ events will occur across both trial arms over the course of the trial under the null hypothesis and that 30 events will occur across both trial arms over the course of the trial in the event that the vaccine is 80% effective.
# Background
This experience is summarized in the GSK Investigator Brochure. The information in this section reflects knowledge that was current at the time of study start. Please refer to the most recent edition of the Investigator Brochure for a review of clinical and preclinical studies (Amended: 13 Nov 2008).
## Tertiary objectives
## 5.
## An evaluation of vaccine efficacy of the candidate vaccine against hpv infections that occur at extra-cervical sites. (amended: 13 nov 2008)
## Overview of the trial
Approximately 20,000 women will be invited to participate in a double blinded, randomized clinical trial to evaluate whether vaccination with the bivalent HPV16/18 VLP-based vaccine manufactured by GSK Biologicals protects against the development CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 of histopathologically confirmed, incident CIN2+. Women between 18-25 years of age at entry who are residents of Guanacaste Province or surrounding areas will be eligible. Participants will be required to not be pregnant at the time of vaccination, to agree to use an effective birth control method 30 days before vaccination until 60 days after the last vaccination (approximately 9 months total), and to be in good health as determined by an overall physical examination administered by a study physician at entry into the trial. Women will be identified through a census of the region performed in February/July 2000, and will be invited to participate in the trial by visiting one of the study clinics set up specifically for this trial. While randomization (into HPV or HAV vaccination) of 15,000 women was initially anticipated, randomization of 7,500 women will provide a sufficient study size to evaluate the main trial objectives based on revised statistical assumptions, see Section 4.7. (Amended: 13 Nov 2008) 12,000-15,000 women are expected to be eligible and agree to be randomly assigned to one of two study arms: VLP vaccination or vaccination with a control vaccine (HAV). Number of women (from census) invited to participate 19,000 -21,000 Number of eligible women who agree to be randomly assigned to one of two study arms 7,500 12,000 -15,000 Number of women who will be randomly assigned to one of two study arms, will receive 3 doses of vaccine and will be evaluable in ATP analysis 5,300 10,400 -13,000 Estimated Nnumber of women who complete four years of follow-up 6,375 9,400 -11,800 CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -144 - persistence/progression (in which case they will be referred for colposcopic evaluation) or three consecutive normal cytologies are observed (in which case they will return to a yearly screening interval). Women who at any of the clinic visits are found to have evidence or suspicion of AGC/ASC-H/HSIL/cancer will be immediately referred for colposcopic evaluation and excisional treatment, if needed. - Monitoring specific AEs of interest in offspring:
## At the request of the costa rican irb specific adverse events occurring in subject's offspring who were conceived within one month prior to first vaccination to one year after last vaccination will be reported at the final visit (month 48). this includes the following medical event categories in the offspring:
-
## Endocrinologic and metabolic conditions
## Death
The occurrence of these specific events in offspring will be included in the clinical database. A child case will be created and linked to the mother case. Medical judgment will be exercised in deciding whether any disorders/diseases reported in offspring are included in these specific categories and should be reported as appropriate. Of note congenital malformations in offspring, as per original protocol, are to be documented and reported as SAEs and included in the pregnancy outcome analysis. (Amended: 13 Nov 2008)
All AEs reported on the AER form (Amended: 31 Oct 2005) will be followed through resolution by a staff physician. Non-serious AEs will be reviewed by a staff clinician before participant close-out; participants with ongoing conditions will be referred to the appropriate referral center if needed. (Amended: 13 Nov 2008)
## Exiting women from the trial (amended 13 nov 2008)
At the Month 48 visit we will apply the same criteria applied throughout the trial to define women who require colposcopic evaluation. Specifically, all women with CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 evidence of cancer, HSIL, glandular lesions, AGC, ASC-H, or persistent ASC-US+/LSIL will be referred to colposcopy. Close-out of individual study participants will only occur after colposcopy referral, for women who require such referral. See Appendix A for more details.
It should be noted, that after participant close-out, participants will be offered additional clinical care if needed. Case management after study close-out will be defined under a separate protocol.
Before the trial is terminated, it is important that we assure all women in the trial that they are free of disease. Therefore, at the last routine screening visit proposed as part of our trial (at 48 months), the criteria used to determine whether a woman should be referred for colposcopic evaluation will differ from that implemented during the initial years of the trial.
More specifically, in addition to referring all women with evidence of cancer, HSIL, glandular lesions, ASC-H, or persistent ASC-US+/LSIL to colposcopy, we propose to refer to colposcopy women with evidence of newly diagnosed ASC/LSIL (irrespective of HPV status) and women with normal cytology at the Month 48 visit who have cytologically evident abnormalities (ASC or LSIL) detected in the 12 months preceding the Month 48 visit. A random subset of 400 women will also be referred for colposcopic evaluation to confirm the sensitivity of our clinical management algorithm for the detection of CIN2+. Given that clinical management options are rapidly changing, we further propose that this preliminary exit strategy be reviewed and discussed by the IRBs and DSMB prior to its implementation. The revised strategy will be submitted for review before the first woman recruited reaches her 48 month visit. Should the final exit strategy differ from that summarized here, an amendment will also be filed with the US FDA detailing the exit visit strategy and its rationale. This will assure that state-of-the-art knowledge is incorporated into our exit visit strategy.
## Overview
## - oral, anal and vulvar specimen collection at the 48-month visit is being added to allow for the evaluation of the effect of vaccination on oral, anal, and vulvar hpv infection rates. (amended: 13 nov 2008).
## Extra-cervical collection
## Oral, anal and vulvar specimen collection at the 48-month visit is being added to allow for the evaluation of the effect of vaccination on oral,anal, and vulvar hpv infection rates.
Prior to the pelvic exam, an oral specimen will be collected. During the pelvic examination, a vulvar and an anal specimen will be collected prior to cervical sampling. Each specimen will be collected according to a standard procedure and will be stored in separate tubes. (Amended: 13 Nov 2008)
## Biospecimen processing, storage, & shipment
The individual types of specimens we propose to collect and the expected number of aliquots resultant from this collection and requiring storage/shipment are summarized in CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 [fig_ref] Table 5 Types: participants [/fig_ref] to allow NCI to plan for adequate repository space at its main repository in Maryland, USA. In brief, we propose to collect cervical cells for liquid-based cytology, additional cervical cells for HPV DNA, RNA and other testing, cervical secretions for immunological assessment, cytological material (slides) for evaluation, histopathological material (blocks and slides) for review and testing, blood for various immunological and other research-based tests, and oral, anal and vulvar samples (Amended: 13 Nov 2008).
We have had experience collecting all of these biological specimens in Costa Rica on several thousand women over more than seven years as part of our natural history study.
Several biological specimens we propose to collect require little or no processing and can be stored at air-conditioned room temperature until shipment to the USA. These include the cells for liquid-based cytology and cytopathological materials. Other specimens require little or no processing but need to be frozen on the day of collection. These include the cervical cells for HPV DNA and RNA testing, cervical secretions, and the oral, anal and vulvar samples (Amended: 13 Nov 2008) .
## Cytology
The clinical management of the study participants will rely on the Costa Rican cytopathology interpretation, with the exception of HSIL noted below. As part of NCI's ongoing quality control efforts (led by cytotechnologist Claire Eklund and pathologist Martha Hutchinson who are affiliated with the Women's and Infants' Hospital in Providence, RI), all slides that read as abnormal in Costa Rica and a 10% sample of the slides read as negative in Costa Rica will be re-screened and re-interpreted in the United States. In addition, all slides from the study exit visit (Month 48) will be re-screened and re-interpreted in the United States. At the study exit (Month 48), in addition to 100% of slides read as abnormal in Costa Rica and 10% of the slides read as negative in Costa Rica mentioned above, 100% of slides that are read in Costa Rica as negative with evidence of reactive changes and that are concurrently HPV positive by the HCII will be re-interpreted in the United States (Amended: 13 Nov 2008). The 10% set of negatives selected for evaluation in the United States will be randomly selected, without replacement. This scheme will result in approximately 50% of subjects with a negative cytology having an expert review over the course of the 4-year study follow-up period. The purpose of this review is to maximize the accuracy of the Costa Rican interpretations CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5 and the comparability of the Costa Rican readings to a US standard, realizing nonetheless that cytopathology is inherently and unavoidably subjective to some extent and that there is no absolute reference standard of cytopathology, particularly for equivocal interpretations such as ASC-US.
## Histology
In addition to the Costa Rican diagnosis, an NCI pathologist (Mark Sherman and/or Diane Solomon) (Amended: 13 Nov 2008) with expertise in cervical pathology will periodically diagnose all histological specimens, blinded to the initial Costa Rican diagnosis. If the diagnoses are concordant (concordance will be required at the level of negative or atypical metaplasia/CIN1/CIN2/CIN3 or in situ/cancer), that will establish the NCI provisional study diagnosis. Discrepant findings between the Costa Rican and NCI pathologist will lead to review by the second NCI pathologist. Agreement between two of the three pathologists at the <CIN2/CIN2+ level will establish the NCI study diagnosis. These diagnoses will not be available within a timeframe that would allow their use for clinical management purposes. Therefore, NCI pathology review will be performed to define histological diagnoses for analysis purposes. This review process and its results will be performed and documented in the US. A copy of the results of the US pathology review will be sent to the clinical staff in Costa Rica. Periodic review of proficiency slide sets will be performed throughout the trial to assure comparability of the Costa Rican/NCI panel consensus readings against other expert pathologists involved in HPV vaccine studies.
## Pcr typing of hpv dna from preservcyt specimens
(to define trial endpoints) Two 0.5 mL aliquots of PreservCyt specimens will be withdrawn prior to ThinPrep preparation to avoid possible specimen-to-specimen contamination. The removal of the two small aliquots is not expected to impact on the cellular adequacy of the PreservCyt specimen, which typically contains many more cells than required for a ThinPrep and subsequent Hybrid Capture® 2 (a signal amplification assay; manufactured by Digene Corporation, Gaithersburg, MD, USA) (HCII HPV DNA test) testing.
The PCR testing used to establish the presence of HPV16 and/or HPV18 for the efficacy endpoints will be performed by GSK Biologicals or Delft DDL (Amended: 13 Nov 2008) Diagnostic Laboratories on one PreservCyt aliquot, as follows (see also Appendix C).
To test for HPV DNA positivity, GSK Biologicals or Delft DDL (Amended: 13 Nov 2008) Diagnostic Laboratories will use SPF 10 primers that amplify a 65 nucleotide region of the HPV L1 gene for most of the known HPV isolates.
## Efforts to maximize participation/retention
- We will invite 20,000 women to participate, but our power estimates (see [fig_ref] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations... [/fig_ref] Section 4.7) indicate the need for 7,500 12,000 -15,000 women (Amended: 13 Nov 2008). Only women who, in principle, agree to comply with the full protocol will be enrolled.
## Power to determine vaccine efficacy
Total accrual into the Costa Rica Trial (HPV-009) was lower than initial estimates. For the outcome of histological CIN2+ (expected # events = 24), the final accrual of 7,466 provides 91% power to detect a vaccine efficacy of 80% under the following assumptions: 1) cumulative incidence of histopathologically confirmed CIN2+ associated with HPV-16 or HPV-18 of 0.79% percent of non-virginal women with histologically confirmed CIN2+ at enrolment of 1.5%
## 3) proportion of women who are sexually inexperienced at entry of 23% 4) proportion of virginal women who initiate sexual activity during follow-up of 43% (constant rate of initiation assumed during the 4-year follow-up period) 5) proportion of sexually experienced women who are both hpv-16 and hpv-18 dna positive at entry of 0.6% 6) proportion of sexually experienced women who acquire hpv-16/18 during the vaccination phase of 1.5% (includes women who develop incident cin2+ associated with hpv-16/18 infection during this period) 7) proportion of women who receive 1+ vaccine doses outside the prescribed protocol window of 20% 8) proportion of women who drop out during the study period of 10% (constant dropout rate assumed during the 4-year follow-up period) 9) a two-sided alpha of 0.05. (amended: 13 nov 2008)
death, at any time after a subject has been discharged from the study, and considers the event reasonably related to the investigational product, the investigator will notify the sponsor. promptly notify Westat's Regulatory Associate. (Amended: 13 Nov 2008)
## Pregnancy
The PES, PO (Amended: 13 Nov 2008), PED (Amended: 29 Mar 2005) and CIF forms contain questions about pregnancies and pregnancy outcomes that have occurred since the last study visit. The principal investigator can also receive information about new pregnancies and pregnancy outcomes through other direct or indirect contacts with participants in the study (e.g., at home reminder visits). All new pregnancy and pregnancy outcome information for study participants, whether or not an SAE is involved, will be transmitted to the Westat Regulatory Associate via a read-only electronic mail report within 3-5 business days (i.e., no longer than 120 business hours) (Amended: 13 Nov 2008) of receipt by the principal investigator or designee. The Westat Regulatory Associate will, in turn, transmit the electronic mail to the NCI Medical Monitor and GSK Global Clinical Safety and Pharmacovigilance within one business day of receipt of the electronic mail from Costa Rica, documenting the receipt date, file name and transmittal date. While pregnancy itself is not considered an AE or SAE, all pregnancies will be followed to term.
A complication of pregnancy or delivery will be recorded as an AE or an SAE, as described in Section 1 and Section 2, and will be followed as described in Section 0 8. (Amended: 13 Nov 2008) A miscarriage is always considered an SAE and will be reported as outlined in Section 0 8. (Amended: 13 Nov 2008) Furthermore, SAEs occurring as a result of a post-study pregnancy AND which the investigator considers reasonably related in time to receipt of the investigational product, will be reported to the Westat Regulatory Associate as in Section 0 8. (Amended: 13 Nov 2008) While the investigator is not obligated to actively seek this information from former study participants, he/she may learn of a pregnancy through spontaneous reporting. Information on pregnancies identified during the screening phase/prior to first vaccine administration does not need to be communicated as a safety issue.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
A SAP for initial analyses that focuses on virological endpoints has been developed and submitted to the FDA. To account for this SAP and for the fact that one of its objectives overlaps with the 5 th secondary objective listed in the protocol, an alpha spending plan has been incorporated into the protocol.
- To include mention that pathologists responsible for histology review are NCIdesignates, and not exclusively NCI pathologists in section 3.19.2.
With the replacement of Dr. Diane Solomon with Dr. Mary Sidawy in protocol amendment #4, our pathology team is composed of both NCI pathologists and NCI designated pathologists. We have added wording to clarify this point in the protocol.
- To include the use of Twinrix® as a crossover vaccination vaccine.
In order to reduce clinic visits, women who are eligible to recieve both the hepatitis A vaccine and hepatitis B vaccine during the crossover vaccination will be offered vaccination with Twinrix® [Hepatitis A Inactivated & Hepatitis B (Recombinant) Vaccine].
- To clarify close-out activities at study completion.
To avoid unnecessary delays in offering crossover vaccination to women who have completed their 4 years of participation in the study, we will conduct data clean-up and data freeze after completion of study activities in a staggered/batched fashion to enable a more timely unblinding for crossover.
- Other minor revisions have been made and are noted throughout the protocol in bold, italic text.
Details of the revision and identification of the affected protocol sections are provided below. Amended text has been included in bold, italic text and deleted text is shown with strikethrough.
Footnote forTimeline of Costa Rican HPV16/18 VLP Vaccine Trial Activities ***This blood draw will be performed on a subset of women, and at selected study visits, as follows: A) A random sample of women at 0, 12, and 36 months; B) All women referred to accelerated screening visits per management algorithm, at yearly intervals before March 31, 2009 (Amended: 10 September 2009); C) All women referred for colposcopic evaluation who require a biopsy or LEEP, at the time of the colposcopy visit; D) women enrolled in the immunogenicity subcohort after March 1, 2006.
## Elimination criteria from the atp analyses during the study
For the primary according to protocol (ATP) analysis planned for our study, the following elimination criteria may apply. These are also discussed in the statistical analysis section of this protocol (Section 4.2).
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
10 September 2009 -159 -- Women who do not receive all three doses of the vaccine within the protocolspecified intervals (see [fig_ref] Table 4: Intervals Between Study Visits [/fig_ref].
- Women who are found to be HPV16 and HPV18 DNA positive (for the corresponding type considered in the analysis) during the vaccination phase, by clinician and/or self-administered collection (months 0 or 6) (Amended: 10 September 2009).
## Trial arms and randomization
Since knowledge of vaccine arm assignment would not affect treatment of participants who have an SAE, no provision is made to allow for unblinding on site in Costa Rica. Unblinding of individual participants, at the request of the DSMB, the IRBs, the GSK Global Clinical Safety Department or the NCI Medical Monitor as part of reporting requirements to the FDA (e.g., for rapid reporting of an unexpected SAE associated with vaccination) will be allowed. This individual unblinding will be performed in a manner that assures that the overall study blinding is maintained, that staff involved in the conduct, analysis, or reporting of the study remain blinded except where absolutely impractical, and that any unblinding be appropriately documented (see also Section 3.16.3, Exiting Women from the Trial) (Amended: 10 September 2009). Treatment codes will be kept at NCI and GSK under controlled/secured access.
## Study vaccines
It should be noted that after completion of this trial and under a separate protocol, participants will be offered cross-over vaccination. This will include offering the licensed formulation of Havrix (
## Activities at study completion (amended: 10 september 2009) the data and safety monitoring board (dsmb) for our trial has recommended crossover immunization of both treatment and control recipients. the participants who received the hpv vaccine will be offered vaccination against hepatitis a and hepatitis b and participants who received the control vaccine will be offered vaccination against hpv-16/18 and hepatitis b, as appropriate. vaccines to be used during cross-over immunization are the formulations licensed for use in costa rica. cross-over immunization will be implemented by nci and costa rican investigators under a separate protocol they have developed.
CONFIDENTIAL NCI Protocol # 04-C-N191; GSK Protocol # 580299/009 (HPV-009) Amendment 5
[fig] Figure 1 23, Figure 2 32, Figure 3: to Determine Vaccine Efficacy in the Costa Rican Trial ............... 75 LIST OF FIGURES PAGE The Cervical Cancer Continuum ............................................................ Diagrammatic Overview of the Costa Rican HPV16/18 VLP Vaccine Trial .......................................................................................... Expedited Reporting -Personnel Communication Flow Chart (Amended: 13 Nov 2008) ..................................................................... 107 [/fig]
[fig] 3 •: (Amended: 13 Nov 2008). The expected enrollment period is 18 (Amended: 31 Oct 2005) months. Treatment allocation: Randomized (1:1); randomization via sequential numbering of vaccines Blinding: Double-blind Treatment groups: Two groups • Vaccine: 40 µg HPV16/18 VLPs/50 µg MPL®/500 µg aluminum as Al(OH) Control: Investigational formulation of Havrix® (Hepatitis A Vaccine, Inactivated; Manufactured by GlaxoSmithKline Biologicals, Rixensart, Belgium) [720 ELISA units of antigen; 500 µg Al(OH) 3 ] Study visits: Average of seven scheduled study visits per subject (8 scheduled visits per subject in the immunogenicity cohort) Vaccination schedule: Three doses of control or vaccine administered on a 0, 1, 6-month schedule Cytologic assessments: Cervical liquid-based cytology evaluations at Month 0, 12, 24, 36 and 48 [/fig]
[fig] First,: Havrix 720 EL.U./0.5 mg Alum in 1.0 mL has been licensed for use in adults in several countries on a 0, 1, 6 month schedule. Between 1992 and 1997, Havrix 720 EL.U. adsorbed on 0.5 mg Alum (1.0 mL) was registered for use on a 0, 1, 6 month schedule in 68 countries including Canada, Australia, the United Kingdom, Brazil and Sweden. Data from trials which included 1,025 adults who were given Havrix 720 EL.U./0.5 mg Alum (1.0 mL) on a 0, 1, 6 month schedule were also included in the original product license application(PLA) (Ref. No. 92-0465) submitted for Havrix and supported licensure in the US. In most of the countries where Havrix 720 EL.U. (1.0 mL) was registered for use in adults on a 0, 1, 6 month schedule, this formulation has now been replaced by a formulation containing 1440 EL.U. of viral antigen adsorbed on 0.5 mg Alum per 1.0 mL dose, since the higher antigen content formulation can be given on a 2-dose schedule. Second, Havrix 720 EL.U./0.25 mg Alum in 0.5 mL (pediatric formulation) is currently licensed in the US for use in children aged 12 months (Amended: 31 Oct 2005) to 18 years on a 0, 6-12 month schedule. The 720 EL.U. dose has been demonstrated to be safe for use in children in numerous clinical trials and through post-marketing surveillance. This suggests that the formulation can be safely administered to adults. Finally, the formulation of Twinrix® (Hepatitis A Inactivated & Hepatitis B [Recombinant] Vaccine; Manufactured by GlaxoSmithKline Biologicals, Rixensart, Belgium) currently licensed in the US for use in adults contains 720 EL.U. of hepatitis A virus antigen (and 20 µg HBsAg adsorbed on 0.45 mg Alum) per 1.0 mL and is administered on a 0, 1, 6 month schedule. Data from 11 clinical trials involving a total of 1812 subjects submitted as part of the Twinrix PLA (BB-IND 6986) and post-marketing [/fig]
[fig] •: Oral, anal and vulvar specimen collection at the 48-month visit is being added to allow for the evaluation of the effect of vaccination on oral, anal, and vulvar HPV infection rates.(Amended: 13 Nov 2008) [/fig]
[table] Table 1 Table 2 Table 3 Table 4 Table 5: Expected Trial Sample Size (Amended: 13 Nov 2008) ........................... 31 Timeline of Costa Rican HPV16/18 VLP Vaccine Trial Activities (Amended: 13 Nov 2008) ....................................................................... 33 Vaccine Dosage and Administration ....................................................... 46 Intervals Between Study Visits (Amended: 13 Nov 2008) ....................... 47 Types of Biological Specimens Proposed and Expected Storage Requirements (Amended: 13 Nov 2008) ................................................ 60 [/table]
[table] Table 3: Vaccine Dosage and Administration [/table]
[table] Table 4: Intervals Between Study Visits (Amended: 13 Nov 2008)This interval applies only to the 600 women enrolled in the immunogenicity subcohort (Amended: 19 May 2006), i.e., the women in the immunogenicity subset. **Date of the Month 6 visit serves as the reference date. ***This schedule is to be followed by women detected as having evidence of LSIL (regardless of HPV testing results) or ASC-US concomitant with detection of an oncogenic HPV type. N.B. Except for Month 7 where the date of the Month 6 visit serves as the reference date, the date of the first [/table]
[table] Table 5 Types: participants (repository estimation purposes; does not take drop-outs into account) [/table]
[table] Table 6: Power to Determine Vaccine Efficacy in the Costa Rican Trial Power calculations to exclude efficacy of 30% were performed using an NCI Matrix Laboratory (MATLAB®, Mathworks, Inc.)-based program. [/table]
[table] Table 7: Timeline of Costa Rican HPV16/18 VLP Vaccine Trial Activities (Amended: 19 May 2006) [/table]
[table] Table 1: Expected Trial Sample Size (Amended: 13 Nov 2008) [/table]
[table] Table 2: Timeline of Costa Rican HPV16/18 VLP Vaccine Trial Activities (Amended: 13 Nov 2008) 1st Fup Interim Fup 2nd Fup Interim Fup 3rd Fup Interim Fup 4th Fup Visit potential subjects at their homes and provide them with a copy of the informed consent form To date At the time of study start, five Phase I-IIb clinical trials have been were conducted that and have evaluated vaccine safety. (Amended: 13 Nov 2008) Please refer to the Investigator Brochure (IB) (3rd edition, March 2004) for a review of the HPV16/18 VLP vaccine pre-clinical and clinical studies. (Amended: 13 Nov 2008) 3.11. Clinic Visits During Vaccination Phase (0, 1, 6, & 7 months) [/table]
[table] Table 5: Types of Biological Specimens Proposed and Expected StorageRequirements (Amended: 13 Nov 2008) Assuming 15,000 7,500 participants (repository estimation purposes; does not take drop-outs into account) [/table]
[table] 1440: ELISA units/0.5 mL) hepatitis A vaccine to women who received the HPV16/18 VLP vaccine and the licensed HPV16/18 VLP vaccine to women who received the investigational hepatitis A vaccine (if the HPV16/18 VLP vaccine is found to be effective and the DSMB and Costa Rica IRB approve) (Amended: 10 September 2009). In addition, all women will be offered vaccination with Engerix-B® (Hepatitis B Vaccine [Recombinant]; Manufactured by GlaxoSmithKline Biologicals, Rixensart, Belgium). In order to reduce clinic visits, women who are eligible to recieve both the hepatitis A vaccine and hepatitis B vaccine will be offered vaccination with Twinrix® [Hepatitis A Inactivated & Hepatitis B (Recombinant) Vaccine] (Amended: 10 September 2009).3.16.3.Exiting Women from the Trial (Amended: 13 Nov 2008) [/table]
|
Recent Advances in Multifunctional Mechanical–Chemical Superhydrophobic Materials
In recent years, biology-inspired superhydrophobic technology has attracted extensive attention and has been widely used in self-cleaning, anti-icing, oil-water separation, and other fields. However, the poor durability restricts its application in practice; thus, it is urgent to systematically summarize it so that scientists can guide the future development of this field. Here, in this review, we first elucidated five kinds of typical superhydrophobic models, namely, Young's equation, Wenzel, Cassie-Baxter, Wenzel-Cassie, "Lotus," and "Gecko" models. Then, we summarized the improvement in mechanical stability and chemical stability of superhydrophobic surface. Later, the durability test methods such as mechanical test methods and chemical test methods are discussed. Afterwards, we displayed the applications of multifunctional mechanical-chemical superhydrophobic materials, namely, anti-fogging, self-cleaning, oil-water separation, antibacterial, membrane distillation, battery, and anti-icing. Finally, the outlook and challenge of mechanical-chemical superhydrophobic materials are highlighted.
# Introduction
Nature has incubated many sophisticated superhydrophobic creatures during long-term evolution and natural selection [bib_ref] Biomimetism and Bioinspiration as Tools for the Design of Innovative Materials and..., Sanchez [/bib_ref] [bib_ref] Recent Developments in Bio-Inspired Special Wettability, Liu [/bib_ref]. Water droplets are spherical on the lotus leaf surface and can roll away the pollution form the surface, which is caused by the chemical composition and special structure of the surface of the lotus leaf. The waterproof composition and microscopic rough structure on the surface of the lotus leaf cause the superhydrophobic phenomenon. This is known as the "Lotus Effect" confirmed by W. Barthlott and C. Neihuis. In addition, many fascinating superhydrophobic phenomena in nature have been uncovered, such as low-adhesion water striders, water-collecting beetles, high-adhesion rose petals, and gecko feet. Inspired by these natural superhydrophobic phenomena, lots of superhydrophobic materials have been developed and used in many fields, self-cleaning [bib_ref] Mechanically Durable Carbon Nanotube−Composite Hierarchical Structures with Superhydrophobicity, Self-Cleaning, and Low-Drag, Jung [/bib_ref] [bib_ref] External Stimuli Responsive Liquid-Infused Surfaces Switching between Slippery and Nonslippery States: Fabrications..., Lou [/bib_ref] , anti-icing [bib_ref] Bio-Inspired Strategies for Anti-icing, Lv [/bib_ref] [bib_ref] Anti-Icing Potential of Superhydrophobic Coatings, Boinovich [/bib_ref] [bib_ref] Robust Anti-icing Superhydrophobic Aluminum Alloy Surfaces by Grafting Fluorocarbon Molecular Chains, Rico [/bib_ref] [bib_ref] Non-Fluorinated and Durable Photothermal Superhydrophobic Coatings Based on Attapulgite Nanorods for Efficient..., Xie [/bib_ref] [bib_ref] A Durable and Photothermal Superhydrophobic Coating with Entwinned Cnts-Sio2 Hybrids for Anti-icing..., Zhang [/bib_ref] [bib_ref] Design and Preparation of a Multi-Fluorination Organic Superhydrophobic Coating with High Mechanical..., Liu [/bib_ref] , anti-fogging [bib_ref] Wet-Style Superhydrophobic Antifogging Coatings for Optical Sensors, Yoon [/bib_ref] [bib_ref] Biomimetic Slippery Pdms Film with Papillae-like Microstructures for Antifogging and Self-Cleaning, Feng [/bib_ref] [bib_ref] Antifogging and Icing-Delay Properties of Composite Micro-and Nanostructured Surfaces, Wen [/bib_ref] , antibacterial [bib_ref] Silver-Nanoparticle-Colored Cotton Fabrics with Tunable Colors and Durable Antibacterial and Self-Healing Superhydrophobic..., Wu [/bib_ref] [bib_ref] A Superhydrophobic Coating to Create Multi-Functional Materials with Mechanical/Chemical/Physical Robustness, Wang [/bib_ref] [bib_ref] Nature-Inspired Chemistry toward Hierarchical Superhydrophobic, Antibacterial and Biocompatible Nanofibrous Membranes for Effective..., Ma [/bib_ref] [bib_ref] Textile Coatings Configured by Double-Nanoparticles to Optimally Couple Superhydrophobic and Antibacterial Properties, Ye [/bib_ref] , fluid drag reduction [bib_ref] Design and Preparation of Biomimetic Polydimethylsiloxane (Pdms) Films with Superhydrophobic, Self-Healing and..., Liu [/bib_ref] , liquid separation [bib_ref] 3D Printing of a Mechanically Durable Superhydrophobic Porous Membrane for Oil-Water Separation, Lv [/bib_ref] [bib_ref] A Study on the Fabrication of Porous Pvdf Membranes by In-Situ Elimination..., Chen [/bib_ref] [bib_ref] Performance Prediction of Magnetorheological Fluid-Based Liquid Gating Membrane by Kriging Machine Learning..., Zhang [/bib_ref] , membrane distillation [bib_ref] Development of Robust and Superhydrophobic Membranes to Mitigate Membrane Scaling and Fouling..., Liao [/bib_ref] [bib_ref] Fabrication of Robust Green Superhydrophobic Hybrid Nanofiber-Nanosphere Membrane for Membrane Distillation, Guo [/bib_ref] [bib_ref] Design of Firm-Pore Superhydrophobic Fibrous Membrane for Advancing the Durability of Membrane..., Ji [/bib_ref] , fog harvest [bib_ref] Understanding How Surface Chemistry and Topography Enhance Fog Harvesting Based on the..., Zhong [/bib_ref] , etc.
The construction of superhydrophobic materials is based on the combination of micro/nano structures and low surface energy chemicals [bib_ref] Superhydrophobic Foams with Chemical-and Mechanical-Damage-Healing Abilities Enabled by Self-Healing Polymers, Fu [/bib_ref] [bib_ref] A Superhydrophobic Coating to Create Multi-Functional Materials with Mechanical/Chemical/Physical Robustness, Wang [/bib_ref]. The micro/nano structures are vulnerable to mechanical wear and chemical corrosion in practical application [bib_ref] Mechanically Durable Superhydrophobic Surfaces, Verho [/bib_ref] [bib_ref] Recent Advances in the Mechanical Durability of Superhydrophobic Materials, Milionis [/bib_ref] [bib_ref] Moving Superhydrophobic Surfaces toward Real-World Applications, Tian [/bib_ref]. Once the superhydrophobic surface is worn or impacted by external pressure, the structure collapses and the chemical substances are worn off, causing the hydrophobic properties to be partially or completely lost immediately and cannot be recovered. In addition, the superhydrophobic materials suffer from the degradation induced by UV exposure and chemical reactions with solvents. Therefore, the development of superhydrophobic materials with excellent mechanical durability and chemical stability are highly desired.
In this review, we illustrated the recent development of multifunctional mechanical-chemical superhydrophobic materials. At first, the theories about superhydrophobic surfaces including Young's equation, , "lotus" model, "gecko" model are elucidated. Then, we summarized the improvement in mechanical stability and chemical stability of superhydrophobic surface. Later, the durability test methods such as mechanical test methods (sandpaper abrasion, tape-peeling, knife-scratch, finger wiping, Taber abrasion, impact test) and chemical test methods (solution immersion, UV irradiation, electrochemical) are discussed. Afterwards, the applications of multifunctional mechanical-chemical superhydrophobic materials are elaborated. Finally, conclusion and prospects of multifunctional mechanical-chemical superhydrophobic materials were discussed.
## Theory of superhydrophobicity
## Wetting definitions
If the interaction between liquid molecules and solid molecules is stronger than that between liquid molecules, the liquid will spread on the solid surface, which is called wetting phenomenon. Wettability is generally characterized by the contact angle of liquid on the solid surface. [bib_ref] Designing Superoleophobic Surfaces, Tuteja [/bib_ref] [bib_ref] Bio-Inspired, Smart, Multiscale Interfacial Materials, Xia [/bib_ref] [bib_ref] Apparent Contact Angles for Reactive Wetting of Smooth, Rough, and Heterogeneous Surfaces..., Bormashenko [/bib_ref]. When water contact angle (WCA) is lower than 10°, the surface is superhydrophilic. And the hydrophilicity is called at 10°-65°, hydrophobicity is denominated at 65°< CA < 150°. Especially, when the WCA is greater than 150°, the sample exhibits superhydrophobicity. Recently, through Jiang's theoretical research and experimental operation [bib_ref] Bio-Inspired, Smart, Multiscale Interfacial Materials, Xia [/bib_ref] , it is proved that CA of 65 defines non-wetting and wetting.
## Young's equation
In 1805, Thomas Young carried out force analysis on the threephase interface and proposed a force analysis model called Young's equation [bib_ref] Iii. An Essay on the Cohesion of Fluids, Young [/bib_ref] , which was only applicable to the contact angle value of water droplets with ideal smooth surface when they reached equilibrium state on the surface.
[formula] γ SV γ SL + γ LV cosθ, [/formula]
where θ is the static water contact angle; γ SV , γ SL, and γ LV represent surface tension of solid-vapor, solid-liquid, and liquid-vapor, respectively.
## Wenzel model
Based on Young's equation, Wenzel linked the roughness factor of the surface with the water contact angle by calculating the adhesion force balance in the surface wetting process [bib_ref] Resistance of Solid Surfaces to Wetting by Water, Wenzel [/bib_ref] , and the linear relationship between Young's contact angle and apparent contact angle are acquired:
[formula] cos θ w rcosθ , [/formula]
where r is the roughness factor, which is determined by the ratio of the actual surface area to the projected surface area, and θ w and θ represent the water CA in respective apparent and original states.
According to the Wenzel model , r can be regarded as the amplification factor in a mathematical relationship, which will make the hydrophilic surface more hydrophilic; on the contrary, for a hydrophobic surface, it will make the surface more hydrophobic.
## Cassie-baxter model
Cassie-Baxter model [bib_ref] Wettability of Porous Surfaces, Cassie [/bib_ref] can be used to analyze the wettability of porous hydrophobic fabric surface. On the basis of Young's equation, it is concluded that the apparent contact angle is the sum of the contributions of each contact phase (fabric and air (pore)):
[formula] cos θ CB f SL cosθ + f LV cos θ′ , [/formula]
where f SL and f LV , respectively, show the fraction between the solid-liquid and liquid-vapor interface at the contacted area and air (f SL +f LV = 1). θ CB and θ′ are the apparent contact angle of liquid droplets on rough surface and the contact angle of liquid on ideal air surface (θ' = 180°), respectively. The wetting state described by Cassie is shown in . The droplet is suspended on the convex surface, and the contact area between the surface and the droplet is very small.
## Wenzel-cassie state
The research of [bib_ref] Superhydrophobic States, Lafuma [/bib_ref] shows that Wenzel-Cassie model is an intermediate state between Wenzel model and Cassie model where water droplets are semi-filled on solid surface. The Cassie state will transform to the Wenzel state under the stimulation of external energy such as droplet impact, mechanical vibration, and droplet evaporation.
"Lotus" Model "Lotus" model [bib_ref] Explained: Two Reasons Why Two Length Scales of Topography Are Important, Gao [/bib_ref] is a special Cassie model, lotus leaf surface microscale mastoid and surface wax to give it a repellent ability, These structures reduce the contact area between solid surface and liquid, and water droplets are in a semi-suspended state, so pollutants can be rolled away by the falling water droplets, which gives a self-cleaning performance on lotus leaf.
## "gecko" model
The "gecko" model [bib_ref] Superhydrophobic Aligned Polystyrene Nanotube Films with High Adhesive Force, Jin [/bib_ref] comes from the classical superhydrophobic nanotube structure, and has good adhesion performance. It is similar to Wenzel model. One is in direct contact with the external atmosphere, and the other is trapped in the nanotube. Due to the change of air volume in the nanotubes, the negative pressure in the nanotubes increases, resulting in high CA, which makes the nanotubes have high adhesion to water .
## Improvement in the mechanical stability
## Self-hardness
Cement [bib_ref] Super-Robust Superhydrophobic Concrete, Song [/bib_ref] , diamond [bib_ref] Robust Diamond Meshes with Unique Wettability Properties, Yang [/bib_ref] [bib_ref] Robust Superhydrophobic Diamond Microspheres for No-Loss Transport of Corrosive Liquid Microdroplets, Wang [/bib_ref] [bib_ref] Robust Biomimetic Hierarchical Diamond Architecture with a Self-Cleaning, Antibacterial, and Antibiofouling Surface, Wang [/bib_ref] , and alloys [bib_ref] A Superhard Superhydrophobic Fe-Based Amorphous Alloy Coating, Qiao [/bib_ref] [bib_ref] A Robust and Repairable Superhydrophobic Co5zn21 Alloy Surface on a Zinc Substrate, Wu [/bib_ref] have inherently high hardness and are thus ideal materials to develop superhydrophobic surfaces with an enhanced mechanical robustness. A superhydrophobic concrete [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref] was prepared by combining metal mesh covering and fluoroalkylsilane modification [bib_ref] Super-Robust Superhydrophobic Concrete, Song [/bib_ref].
The obtained concrete can retain its superhydrophobic property after a sandpaper wear test (a pressure of 1100 Pa, standard sandpaper of 360#, and abrasion distance of 8 m). In addition, the superhydrophobic concrete is able to endure the knife-scratch and the hammer blow tests. This effectively demonstrates the remarkable mechanical strength of as-prepared superhydrophobic concrete. For its own hard materials, his preparation method is simple and easy to obtain, but because of the lack of materials, it is not suitable for large-scale production.
# Porous materials
Sponges [bib_ref] Robust Superhydrophobic Polyurethane Sponge as a Highly Reusable Oil-Absorption Material, Zhu [/bib_ref] [bib_ref] Fabrication of a Robust Superhydrophobic Polyurethane Sponge for Oil-Water Separation, Cheng [/bib_ref] [bib_ref] Facile Fabrication of Marine Algae-Based Robust Superhydrophobic Sponges for Efficient Oil Removal..., Dong [/bib_ref] , textiles [bib_ref] Superhydrophobic and Breathable Smart Mxene-Based Textile for Multifunctional Wearable Sensing Electronics, Luo [/bib_ref] [bib_ref] Fabrication of Robust and Self-Healing Superhydrophobic Pet Fabrics Based on Profiled Fiber..., Zhou [/bib_ref] , foamed nickel [bib_ref] Superhydrophobic and Superoleophilic Nickel Foam for Oil/Water Separation, Eum [/bib_ref] [bib_ref] Multi-Layer Superhydrophobic Nickel Foam (Nf) Composite for Highly Efficient Water-In-Oil Emulsion Separation, Wang [/bib_ref] , and other materials [bib_ref] Facile Fabrication of Robust Superhydrophobic Porous Materials and Their Application in Oil/Water..., Hou [/bib_ref] with multiple layers and porous [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref] , due to their large specific surface area, even if part of the material surface is rubbed off, the material still remains, so it has abrasion resistance and is an excellent superhydrophobic material. Superhydrophobic textiles [bib_ref] Superhydrophobic and Breathable Smart Mxene-Based Textile for Multifunctional Wearable Sensing Electronics, Luo [/bib_ref] are manufactured by decorating the textiles modified by polydopamine (PDA) with MXene (Ti 3 C 2 T x ) and then coating with polydimethylsiloxane (PDMS). The obtained superhydrophobic breathable textiles still maintain superhydrophobic properties in the sandpaper wear test (moving 2 cm with traction under the weight of 50 g), which demonstrates the robustness of the superhydrophobic textiles. Porous material is one of the recent research hotspots, which has the advantages of simple operation, low production cost, and suitable for large-scale production, while at the same time, porous materials have been widely used in separation, catalysis, and other fields.
## "paint + adhesive" method
In order to reduce the dependence of superhydrophobic surface on substrate and strengthen the interface bonding force, a strategy of "Paint + adhesive" was developed to prepare superhydrophobic surface. The surface superhydrophobic layer is connected with the substrate by an intermediate layer, which can not only anchor the micro-nano structure on the surface, but also serve as a shielding layer to provide additional protection for the substrate, thus obviously improving the mechanical properties of superhydrophobic surface and preparing durable superhydrophobic surfaces on various substrates. Lu et al. [bib_ref] Robust Self-Cleaning Surfaces that Function when Exposed to Either Air or Oil, Lu [/bib_ref] proposed a "paint + adhesive" strategy to build a durable superhydrophobic surface for the first time. TiO 2 nanoparticles modified by fluorosilane, was dispersed in ethanol solution and sprayed on the adhesive-coated substrate. The adhesive can firmly adhere the TiO 2 nanoparticles (superhydrophobic layer) to the substrates that the obtained superhydrophobic surface shows a water CA of >160°even after wiping with fingers, impacting with water droplets, and 40 cycles of sandpaper abrasion (standard glasspaper, grit no. 240, and moved for 10 cm). Based on the above "paint + adhesive" method, many organic/inorganic adhesives and superhydrophobic materials are used to develop superhydrophobic surfaces with good durability [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref] [bib_ref] Facile Fabrication of Robust, Biomimetic and Superhydrophobic Polymer/Graphene-Based Coatings with Self-Cleaning, Oil-Water..., Zheng [/bib_ref]. The method can improve the binding force between the substrate and the superhydrophobic material, and can be produced on a large scale which has wide selectivity to the substrate. However, the superhydrophobic layer is affected by external mechanical friction or chemical corrosion, and its service life is greatly reduced.
## "armor"
Armoring strategy is to use materials with excellent mechanical properties to protect the surface micro-nano structures, which is similar to the function of armor. At present, nano-scale armor and microscale armor are mainly used. In 2020, Wang and coworkers [bib_ref] Design of Robust Superhydrophobic Surfaces, Wang [/bib_ref] fabricated a robust superhydrophobic surface via constructing surface texture at two different length scales, including superhydrophobic nanostructures and a microstructure frame [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref]. The microstructure frame is made up of an array of microscale inverted-pyramidal cavities, which can house the superhydrophobic nanostructure and act as a protective "armor" to avoid the destruction of the superhydrophobic nanostructure by abradants. The combination of superhydrophobic nanostructures and the protective microstructure frame ensures that the obtained superhydrophobic surface could tolerate more than 1000 abrasion cycles and even under tape-peeling tests, Taber abrasion tests, and scratch tests. The armor model provides a new idea for the preparation of durable superhydrophobic materials, but it is still in the exploratory stage because of its complex preparation method.
## Improvement in the chemical stability
Improving the chemical stability of superhydrophobic surface is also a research hotspot in recent years. At present, the common preparation methods to improve the chemical stability of superhydrophobic surface include chemical etching, spraying, electrochemical deposition, sol-gel method and electrostatic spinning. However, they have their own advantages and disadvantages. [fig_ref] TABLE 1 |: Advantages and disadvantages of different methods [/fig_ref].
## Chemical etching
Chemical etching method refers to the preparation of superhydrophobic surface by using the strong corrosiveness of strong acid/alkali solution to construct a micro/nano composite structure on the substrate, which is simple to operate and fast to react. Xu et al. used nitric acid solutions with different concentrations to etch the nickel mold, discussed the importance of etching time and chemical solution concentration, and then copied the surface pattern of the chemical etching template to obtain a large-area micro/nano-structured polydimethylsiloxane (PDMS) film with superhydrophobicity. The film shows superhydrophobicity even under high-strength friction, and also has excellent acid and alkali resistance (excellent liquid repellency even after contacting with 1 M HCl, 1 M NaOH and 1 M NaCl solutions for 96 h), ultraviolet resistance, and optical transparency.
## Spraying
The spraying method uniformly disperses and overlays the raw materials of micro/nanoparticles on the surface of the base material to mode a uniform coating with a certain structure, which is not limited to the shape and size of the base material, simple and convenient to operate, low in cost, and high in coating efficiency. Yokoi et al. [bib_ref] Optically Transparent Superhydrophobic Surfaces with Enhanced Mechanical Abrasion Resistance Enabled by Mesh..., Yokoi [/bib_ref] deposited perfluorodecyl trichlorosilane on the surface of alkali-treated polyester, and then sprayed silica modified by fluorosilane on the surface of modified polyester to acquire a transparent superhydrophobic surface. The contact angle of the sample remained above 150°after 100 wear cycles under the pressure of 10 kPa, and the sample had strong repulsion to strong acid and alkali (the contact angle and sliding angle of acidic and alkaline aqueous solutions with pH values ranging from 2 to 14 were measured. The contact angle of all solutions was over 150°, and the sliding angle was less than 15°), which indicates that the prepared superhydrophobic polyester mesh not only had high mechanical strength, but also had good acid and alkali resistance. Frontiers in Bioengineering and Biotechnology | www.frontiersin.org
July 2022 | Volume 10 | Article 947327
## Electrochemical deposition
Electrochemical deposition [bib_ref] Nanoscale Electrodeposition: Dimension Control and 3d, Lee [/bib_ref] method refers to the preparation technology of depositing one or more materials on the workpiece surface of the anode, while the cathode undergoes a reduction reaction. She et al. [bib_ref] Highly Anticorrosion, Self-Cleaning Superhydrophobic Ni-Co Surface Fabricated on AZ91D Magnesium Alloy, She [/bib_ref] performed electroless nickel plating on the pre-treated AZ91D magnesium alloy and then electrodeposited the nickel-cobalt alloy coating, obtaining a superhydrophobic surface with a contact angle of 167.3 ± 1.3°and a rolling angle of about 1°, and the corrosion current density is three orders of magnitude lower than that of the blank sample, the corrosion rate is about 0.06% of the blank sample, which shows it has better corrosion resistance and pH stability.
# Sol-gel method
Sol-gel method refers to the use of highly chemically active compounds as precursors, hydrolysis, and condensation reaction in the liquid phase to form a stable transparent sol system, after polymerization, gel is formed, and then by drying, sintering curing treatment to prepare micro and nano pore structure, so as to give the surface of the material hydrophobic properties. Su et al. [bib_ref] Vapor-Liquid Sol-Gel Approach to Fabricating Highly Durable and Robust Superhydrophobic Polydimethylsiloxane@Silica Surface..., Su [/bib_ref] prepared hydrophobic sol by teosilicate ethyl ester and polydimethylsiloxane according to a certain mass ratio. Polyester fabric absorbed sol by immersion and reacted with acid to prepare superhydrophobic polyester surface with good mechanical stability. The prepared superhydrophobic textiles have excellent durability in deionized water, various solvents (the CAs were almost unchanged and still above 150°immersed in deionized water, hexane, hexane and toluene hexane for 168 h), strong acid/alkali solutions (the superhydrophobic textiles still had water repellency after being immersed in HCl solution for 60 h or an aqueous NaOH solution for 48 h) and boiling water/ice water.
## Electrostatic spinning
Electrospinning [bib_ref] Emerging Polymeric Electrospun Fibers: From Structural Diversity to Application in Flexible Bioelectronics..., Wan [/bib_ref] is a kind of method in which polymer solution forms a jet under the action of high-voltage electrostatic force, and finally one-dimensional nanofibers are prepared. The superhydrophobic surface can be obtained by covering the surface of the substrate with nanofiber membrane and then modifying it with low surface energy substances. It has the advantages of low spinning cost, simple manufacturing device, various kinds of spinnable substances, controllable process, etc. Cui et al. prepared superhydrophobic anticorrosive coating on aluminum substrate by electrospinning. Polyvinylidenefluoride (PVDF)/ stearic acid nanofibers are used to construct micron/ nanometer superhydrophobic structures to provide long-term corrosion protection. After corrosion in 3.5% NaCl solution for 30 days, it still had excellent corrosion resistance.
## Durability test
## Mechanical durability test
Inspired by lotus leaves, superhydrophobic surfaces have huge potential applications. However, their practical application is limited by poor durability. When exposed to harsh mechanical or chemical conditions, they can easily lose their functions.
Scientists also try to adopt various methods to improve the durability of materials, so we need to establish a test method for superhydrophobic durability. At present, there are many testing methods of superhydrophobic durability, which can be summarized into two aspects: one is mechanical durability test, such as sandpaper abrasion, tape-peeling, knife-scratch, finger wipe, Taber abrasion, and impact test, the other is chemical durability test, such as acid-base test, solution immersion, UV irradiation, and electrochemical corrosion.
## Sandpaper abrasion test
The sandpaper abrasion test is a common method to test the wear resistance of superhydrophobic surface at present. During the sandpaper abrasion test [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] [bib_ref] Fabrication of Durable Superhydrophobic Coatings Based on a Novel Branched Fluorinated Epoxy, Zhu [/bib_ref] [bib_ref] Robust Silicon Dioxide @ Epoxy Resin Micronanosheet Superhydrophobic Omnipotent Protective Coating for..., Wang [/bib_ref] , a certain load is applied on the superhydrophobic material, and the material is rubbed on the sandpaper. The surface between the superhydrophobic material and the sandpaper acts as a wear surface. Sandpaper abrasion test is the most common evaluation method, which has good practicability. However, at present, the test standards are not uniform and the test error is relatively large. Li et al. studied the effects of superhydrophobic coatings prepared with different filler particle sizes on surface morphology and hydrophobic properties under the same load, different abrasive particle sizes and friction distances. The results show that with the same filler content, the larger the filler particle size, the greater the wear resistance.
## Tape-peeling test
Tape peeling [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] [bib_ref] A Facile and Novel Emulsion for Efficient and Convenient Fabrication of Durable..., Wu [/bib_ref] [bib_ref] Fabrication of Durable Superhydrophobic Coating on Fabrics Surface for Oil/Water Separation, Ji [/bib_ref] is one of the easiest ways to determine the surface abrasion resistance of superhydrophobic materials, which is to fully contact the tape with the surface of the tested material under a certain pressure, and then peel off at a certain angle and speed. This method is mainly used to test the adhesion strength of superhydrophobic coating and its rough structure to substrate. However, this method can only evaluate the firmness of coating and substrate, but not the strength of superhydrophobic surface, which has certain limitations. By observing the SEM diagram, Zhao et al. compared the number of nanoparticles per unit area before and after peeling, evaluated the binding strength of silica particles with different sizes and epoxy resin substrate, and optimized the superhydrophobic surface durability by adjusting the ratio of different particle sizes to fillers.
## Knife-scratch test
Considering that the superhydrophobic surfaces are often subjected to scratches in practical application, such as car scratches, knife scratch is selected as a typical test to evaluate the mechanical wear resistance of superhydrophobic surfaces. This method is suitable for fields with high requirements for mechanical stability, but the current testing standards are not uniform [bib_ref] Robust Self-Cleaning Surfaces that Function when Exposed to Either Air or Oil, Lu [/bib_ref]. As shown in [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] , the knife is used to scrape the superhydrophobic surface, resulting in a dense array of wide and deep scars on the surface. Wu et al. [bib_ref] A Facile and Novel Emulsion for Efficient and Convenient Fabrication of Durable..., Wu [/bib_ref] used knives to form wide and deep lattice marks on the superhydrophobic wood, however, water droplets can easily roll down from it without leaving any traces, indicating that the superhydrophobicity still exists.
## Finger wiping test
As shown in [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] , the finger wipe test [bib_ref] Robust Self-Cleaning Surfaces that Function when Exposed to Either Air or Oil, Lu [/bib_ref] [bib_ref] A Facile and Novel Emulsion for Efficient and Convenient Fabrication of Durable..., Wu [/bib_ref] is to wipe the surface of the superhydrophobic material repeatedly with the finger in the same direction, and then test the change of the contact angle of the material surface. Finger wiping test can preliminarily evaluate the durability of superhydrophobic surface, and the experimental operation is convenient and easy. Liu et al. designed and prepared a new type of polyfluorinated organic superhydrophobic coating based on mercaptan-olefin click reaction. The coating has excellent superhydrophobicity and self-cleaning properties, and has good adhesion to the substrate, which still maintains excellent superhydrophobicity after finger wiping.
## Taber abrasion test
Taber friction [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] [bib_ref] Simple Spray Deposition of a Water-Based Superhydrophobic Coating with High Stability for..., Ye [/bib_ref] [bib_ref] Fabrication of Durable Superhydrophobic Coatings Based on a Novel Branched Fluorinated Epoxy, Zhu [/bib_ref] is also a kind of friction test, which is carried out in a special Taber friction testing machine. The machine consists of three parts: a turntable that clamps the sample, a friction wheel and a load. During the experiment, the superhydrophobic material is clamped on the turntable. Then, load a certain weight of the friction wheel for rotating friction, and take out the test piece after the specified number of revolutions to test its superhydrophobic performance. This method has certain evaluation criteria, the experimental operation is convenient and the data is accurate. Peng et al. [bib_ref] All-Organic Superhydrophobic Coatings with Mechanochemical Robustness and Liquid Impalement Resistance, Peng [/bib_ref] observed the variation of coating contact angle and coating thickness with Taber abrasion cycles under three different loads (150, 200 and 250 g). After 100 wear cycles, the CA of PTFE coating remained above 150°under 150 and 200 g loads and decreased to 146°u nder 250 g loads.
## Impact test
There are two types of impact tests [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref]. One is the water impact test [bib_ref] Fabrication of Durable Superhydrophobic Coatings Based on a Novel Branched Fluorinated Epoxy, Zhu [/bib_ref] , and the other is the sand impact test [bib_ref] Biomimetic, Dopamine-Modified Superhydrophobic Cotton Fabric for Oil-Water Separation, Yan [/bib_ref]. It is mainly a method to tilt the superhydrophobic surface at a certain angle, impact the surface with sand or water drops at a certain height, and evaluate the change of surface hydrophobicity. This method can effectively evaluate the outdoor durability of superhydrophobic materials. Deng et al. [bib_ref] Candle Soot as a Template for a Transparent Robust Superamphiphobic Coating, Deng [/bib_ref] used candle soot and silica to prepare superhydrophobic coating. To explore the mechanical properties of the coating, water drop impact and sand wear tests were carried out. Sand particles with a diameter of 100-300 mm hit the surface from a height of 10-40 cm. Although the coating surface is impacted by sand to form a cave, its microstructure has little change.
## Chemical durability test
## Solution immersion
At present, superhydrophobic materials have been used in various industries; however, their low corrosion resistance hinders their wider application. Therefore, there is a need to, at a relatively low-cost technology, improve the corrosion resistance of these materials. At the same time, scientists used a chemical solution immersion method to test the chemical resistance of materials. In acidic solution [bib_ref] A Multifunctional Transparent Superhydrophobic Gel Nanocoating with Self-Healing Properties, Si [/bib_ref] [bib_ref] Fabrication of Durable Superhydrophobic Coatings Based on a Novel Branched Fluorinated Epoxy, Zhu [/bib_ref] , high concentration of H + will hydrogenate with superhydrophobic materials, which will destroy their original properties and make them lose superhydrophobic properties. In alkali solutions, the chemical properties of strong base are relatively active, with strong reducibility, easy to react with other substances, so as to achieve corrosion. In chloride-containing solutions, because the radius is small and it has strong penetration ability, chloride ions are most likely to pass through the tiny voids in the oxidation film to get to the metal surface, interact with the metal to get soluble compounds, which changes the structure of the oxide film and causes corrosion of the metal. In aqua regia, aqua regia is a very corrosive liquid that can corrode the surface of the material. However, polytetrafluoroethylene (PTFE), the king of organic plastics, is not corroded by aqua regia, so researchers immersed a superhydrophobic material made of polytetrafluoroethylene in aqua regia to test its corrosion resistance.
## Ultraviolet light irradiation
Ultraviolet light irradiation [bib_ref] Fabrication of Durable Superhydrophobic Coatings Based on a Novel Branched Fluorinated Epoxy, Zhu [/bib_ref] is one of the common methods for testing the aging of materials, which is mainly tested by putting superhydrophobic materials under a certain wavelength and power ultraviolet lamp, evaluate the attenuation degree of the surface contact angle with the extension of irradiation time. This method is mainly used for evaluating and testing the outdoor durability of superhydrophobic materials. Huang et al. used polytetrafluoroethylene (PTFE) particles to prepare powder coatings without solvent and chemical modification. Due to the high bond energy and chemical inertia of PTFE, the surface contact angle of the coating remained above 160°after UV irradiation for 84 h, showing excellent chemical durability.
## Electrochemical corrosion
Electrochemical corrosionmeans the corrosion of metal due to electrochemical action in a conductive liquid medium, and current is generated during the corrosion process. When metal is placed in an aqueous solution or in a moist atmosphere, a microcell, also known as a corrosive cell, forms on the surface of the metal, oxidation reaction happens on the anode, so that the anode is dissolved, reduction reaction happens on the cathode, generally only play the role of electron transfer. This method can effectively evaluate the outdoor durability of metallic superhydrophobic materials.
## Applications
## Anti-fogging
Changing the wettability of the surface is a common method of anti-fogging, and two extreme cases are usually paid attention to: superhydrophilicity and superhydrophobicity. The hydrophilic anti-fogging method, which makes the surface of the substrate highly hydrophilic, the contact angle between the surface of the material and water approaches zero, and makes the water vapor quickly spread on the surface of the substrate after condensation to constitute a transparent water film, which has been deeply studied. Generally, superhydrophobic materials are able to firmly bond with the surfaces of other materials, and water droplets are easy to roll on the superhydrophobic surface. Therefore, it can be inferred that the droplets formed by condensation of water vapor on the surface can also roll off the surface of hydrophobic materials quickly, thus having anti-fogging pe rformance. Medical endoscopes have promoted the development of medical careers, but endoscopes are prone to mirror fogging due to liquid adsorption and high humidity, which reduces visibility. Lee et al. [bib_ref] Lubricant-Infused Directly Engraved Nano-Microstructures for Mechanically Durable Endoscope Lens with Anti-biofouling and..., Lee [/bib_ref] applied a laser to construct a lubricant-infused directly engraved nano/micro structured surface (LIDENS) on the lens,, which can repel various liquids after chemical modification of the LIDENS lens. Among them, the injection of lubricant can smoothen the rough surface structure and improve the transmittance. The low cost of LIDENS Nuclear density and dynamic coalescence can remove droplets under gravity, thereby preventing fogging. At the same time, the mechanical durability of the LIDENS directly etched on the surface morphology was tested, after 30 times of tape peeling, the SEM images in [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] shows that the dentate wrapped by F-SAM has no obvious topological changes, which proves it has good mechanical properties.
Yoon et al. [bib_ref] Wet㏒Tyle Superhydrophobic Antifogging Coatings for Optical Sensors, Yoon [/bib_ref] prepared a wet superhydrophobic coating, which maintained excellent antifogging performance. The top of the coating is a PDMS micro-well with low surface energy, which shows superhydrophobicity, and the bottom is a sacrificial oil (silicone oil) embedded polymer-silica nanocomposite as hydrophilic part, which guides the upper layer of water vapor condensation to the lower layer. The coating can prevent the formation of fog and maintain optical transparency during condensation.
## Self-cleaning
The lotus leaves that "come out of silt but do not dye" are typically natural self-cleaning surfaces. In addition, many animals and plants in nature have a superhydrophobic surface with selfcleaning property, such as rice leaves [bib_ref] Bioinspired Rice Leaf and Butterfly Wing Surface Structures Combining Shark Skin and..., Bixler [/bib_ref] [bib_ref] Bioinspired Self-Cleaning Surfaces with Superhydrophobicity, Superoleophobicity, and Superhydrophilicity, Nishimoto [/bib_ref] [bib_ref] Fabrication of an Anisotropic Superhydrophobic Polymer Surface Using Compression Molding and Dip..., Lee [/bib_ref] [bib_ref] Bio-Inspired Design of Bi/Tridirectionally Anisotropic Sliding Superhydrophobic Titanium Alloy Surfaces, Xu [/bib_ref] , pitcher plants [bib_ref] Water-Repellent and Corrosion-Resistance Properties of Superhydrophobic and Lubricant-Infused Super Slippery Surfaces, Song [/bib_ref] [bib_ref] A Switchable Cross-Species Liquid Repellent Surface, Huang [/bib_ref] [bib_ref] A Bio-Inspired Superhydrophobic Surface for Fog Collection and Directional Water Transport, Li [/bib_ref] , cicada wings [bib_ref] Exploring the Role of Habitat on the Wettability of Cicada Wings, Oh [/bib_ref] , butterfly wings [bib_ref] Bioinspired Self-Cleaning Surfaces with Superhydrophobicity, Superoleophobicity, and Superhydrophilicity, Nishimoto [/bib_ref] , gecko feet [bib_ref] Superhydrophobicity of the Gecko Toe Pad: Biological Optimization versus Laboratory Maximization, Stark [/bib_ref] , snail shells [bib_ref] Bioinspired Self-Cleaning Surfaces with Superhydrophobicity, Superoleophobicity, and Superhydrophilicity, Nishimoto [/bib_ref] , fish scales, shark skin [bib_ref] Rice-and Butterfly-Wing Effect Inspired Self-Cleaning and Low Drag Micro/Nanopatterned Surfaces in Water,..., Bixler [/bib_ref]. Water droplets can capture dust particles and roll away easily when arriving at the superhydrophobic surface, which offers the superhydrophobic surface its self-cleaning property.
Wu et al. [bib_ref] Efficient Fabrication of Lightweight Polyethylene Foam with Robust and Durable Superhydrophobicity for..., Wu [/bib_ref] proposed an efficient solution modification method to prepare superhydrophobic F-PE/SiO 2 foam materials, which shows a water CA of 158 ± 2°. The polyethylene foam has an interconnected threedimensional skeleton, which is composed of a polyethylene skeleton and irregular pores. The interconnected three-dimensional skeleton results in an enhanced wear resistance for the polyethylene foam. The polyethylene foam still exhibits superhydrophobic property even after sandpaper friction and water impact. In addition, F-PE/SiO 2 foam also shows excellent self-cleaning performance.
Photocatalysis [bib_ref] Elastic Superhydrophobic and Photocatalytic Active Films Used as Blood Repellent Dressing, Liu [/bib_ref] [bib_ref] Superhydrophobic Coating Using TiO 2 NPs/PMHS Composite for Self-Cleaning Application, Sutar [/bib_ref] [bib_ref] A Universal, Multifunctional, High-Practicability Superhydrophobic Paint for Waterproofing Grass Houses, Zhu [/bib_ref] can produce self-cleaning effects [bib_ref] An Anti-uv Superhydrophobic Material with Photocatalysis, Self-Cleaning, Self-Healing and Oil/Water Separation Functions, Zhu [/bib_ref]. Superhydrophobic materials with photocatalytic performance can convert light energy into chemical energy to decompose organic pollutants. During this process, the decomposed organic pollutants leave the surface of superhydrophobic material in the form of gas, and the residual solid particles will be taken away with the spreading of water film.
Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327
Our team [bib_ref] A Universal, Multifunctional, High-Practicability Superhydrophobic Paint for Waterproofing Grass Houses, Zhu [/bib_ref] mixed TiO 2 NPs, epoxy resin and 1H,1H,2H,2H-perfluorooctyltriethoxysilane (FAS) through stirring and ultrasonic treatment to compose an inorganic organic superhydrophobic coating (IOS-PA). The presence of TiO 2 NPs enables the degradation of Nile red. The superhydrophobicity of IOS-PA is preserved after sandpaper abrasionand sand impact, indicating the excellent mechanical durability. At the same time, after being stored in acidic (pH = 1) solution for 4 h and saline (pH = 7) and alkaline (pH = 14) solutions for 8 h, the high WCA and low RA remained on the coating samples. Moreover, the layer we studied had multifunctional self-cleaning ability, which can not only remove stains by gravity rolling of water, but also decompose organic dyes by ultraviolet.
## Oil-water separation
Frequent oil spills cause serious global water pollution [bib_ref] Environmentally Friendly Superhydrophobic Osmanthus Flowers for Oil Spill Cleanup, Zhu [/bib_ref] [bib_ref] Oil Pollution of Beaches, Huettel [/bib_ref] , which poses an urgent need for efficient solutions to oil-water separation. The traditional methods for oil-water separation include gravity separation [bib_ref] An Overview of Oil-Water Separation Using Gas Flotation Systems, Saththasivam [/bib_ref] , filtration, centrifugation , flotation (Rocha e Silva et al., 2018) and electrochemical methods [bib_ref] Investigation of Water Separation from Water-In-Oil Emulsion Using Electric Field, Kwon [/bib_ref]. However, most of them have low separation efficiency and complicated operation . Superhydrophobic material has high separation speed and efficiency and is a promising way to solve this serious matter [bib_ref] Low-Temperature Carbonization Synthesis of Carbon-Based Super-hydrophobic Foam for Efficient Multi-State Oil/Water Separation, Kong [/bib_ref].
Shang and his team [bib_ref] Sustainable and Robust Superhydrophobic Cotton Fabrics Coated with Castor Oil-Based Nanocomposites for..., Shang [/bib_ref] have prepared an environmentally friendly and sustainable superhydrophobic or superoleophilic castor oil-based nanocomposite on cotton fabric using a thiol-ene chemical method initiated by ultraviolet light. The cotton fabric has a rough surface and possesses a water CA of~160°and a water SA of 7.5°. The water droplets can penetrate into the pristine fabric immediately because of the capillary effect which is caused by the porosity and abundant hydroxyl groups on the fabric. In addition, high-strength superhydrophobic cotton fabrics can withstand at least 30 sandpaper wear cycles without losing their superhydrophobicity. At the same time, the functional cotton fabric can separate kinds oil and water mixtures and emulsions with high separation efficiency. Tang et al. [bib_ref] Fabrication of Calcium Carbonate Coated-Stainless Steel Mesh for Efficient Oil-Water Separation via..., Tang [/bib_ref] proposed a cheap, environmentally friendly and pollution-free method to prepare superhydrophobic calcium carbonate (CaCO 3 ) which coated stainless steel mesh (SSM). In the experiment, the superhydrophilic CaCO 3 -SSM was firstly prepared by using the biomineralization method induced by bacteria, and immersed in stearic acid (SA) to obtain a superhydrophobic SA/CaCO 3 -SSM. This has regular and large-size micro-pores, and thus shows high oil flux to various oil/water mixtures (0.2-9.12 × 104 L m −2 ·h −1 ) and high efficiency in separation (>94.8%). In addition, the SA/CaCO 3 -SSM also exhibits outstanding wear resistance. [bib_ref] One-pot Synthesis of Robust Superhydrophobic, Functionalized Graphene/polyurethane Sponge for Effective Continuous Oil-Water..., Zhou [/bib_ref] modified the interior of the PU sponge using (3-mercaptopropyl) trimethoxysilane and graphite oxide by solvent heat treatment, resulting in a graphene layer resembling a crater that was firmly attached to the polyurethane skeleton. Graphene/PU sponges are superhydrophobic with a WCA of over 160°and can effectively separate oil and water.
The recent development of superhydrophobic materials provides a simple and inexpensive solution for oil-water separation. For example, [bib_ref] Robust and Durable Superhydrophobic Steel and Copper Meshes for Separation of Oil-Water..., Tudu [/bib_ref] use TiO 2 nanoparticles and perfluorodecyltriethoxysilane (PFDTS) to make superhydrophobic steel and copper mesh. Yan's group [bib_ref] Biomimetic, Dopamine-Modified Superhydrophobic Cotton Fabric for Oil-Water Separation, Yan [/bib_ref] prepared superhydrophobic cotton fabric by combining micro-nano-binary structure of polydopamine (PDA) with grafting of octadecyylamine (ODA).
## Antibacterial action
The adhesion and proliferation of bacteria on the surface of objects will lead to the formation of biofilms, which poses huge challenges for medical, health, and industrial applications [bib_ref] Exposure and Health Effects of Bacteria in Healthcare Units: An Overview, Monteiro [/bib_ref]. The antibacterial material based on superhydrophobicity is an emerging method recently [bib_ref] Hydrothermally Structured Superhydrophobic Surface with Superior Anti-corrosion, Antibacterial and Anti-icing Behaviors, Lan [/bib_ref]. The information of bacterial biofilm involves transportation, adhesion, firmness, and reproduction. The strategies to remove biofilms on the surface of substrates mainly include preventing bacteria from adhesion [bib_ref] Silver-Perfluorodecanethiolate Complexes Having Superhydrophobic, Antifouling, Antibacterial Properties, Chung [/bib_ref] and killing bacteria that have attached.
Ye et al. [bib_ref] Textile Coatings Configured by Double-Nanoparticles to Optimally Couple Superhydrophobic and Antibacterial Properties, Ye [/bib_ref] used PDMS as the adhesive to attach fluorinated mesoporous silica nanoparticles (F-MSNS) and quaternary ammonium functionalized microporous silica nanoparticles (Q-MSNS)to the surface of various fabrics, and the resulting textiles showed obvious synergistic antibacterial effects against Escherichia coli and Staphylococcus aureus by "repellent", which is mainly because the superhydrophobicity can repel most bacteria, and Q-MSNS on the surface of cotton fabric can effectively kill some bacteria. At the same time, due to the surface of F/Q-MSNS coating being rough, even after 600 times of friction, the surface of the coating is still superhydrophobic.
Ou et al. [bib_ref] Washable and Antibacterial Superhydrophbic Fabric, Ou [/bib_ref] selected polydopamine as an adhesive to prepare a superhydrphobic cotton coated with silver nanoparticles. The polydopamine can increase the binding between silver cotton fibers and nanoparticles, so as to prevent silver nanoparticles from dropping from the surface of cotton fibers. At the same time, the fabric composite has obvious antibacterial effect on Staphylococcus aureus and Escherichia coli.
Zhu et al.prepared a superhydrophobic coating solution by dispersing hydrophobic silica nanoparticles (Aerosil ® gaseous silica) in ethanol at a concentration of 2.5 w/w %. Compared with the bare surface, the attachment amount of SARS-CoV-2 on the superhydrophobic (SHPB) surface was significantly reduced, up to 99.99995%. This suggests that the as-prepared coating can effectively resist the adhesion of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by repelling virus-carrying droplets.
## Membrane distillation
Membrane distillation (MD) [bib_ref] Fouling in Membrane Distillation, Osmotic Distillation and Osmotic Membrane Distillation, Laqbaqbi [/bib_ref] [bib_ref] A Pore-Size Tunable Superhydrophobic Membrane for High-Flux Membrane Distillation, Hong [/bib_ref] is a bright desalination technology because it is capable of treating highly saline water. Deng et al. [bib_ref] Robust Superhydrophobic Dual Layer Nanofibrous Composite Membranes with a Hierarchically Structured Amorphous..., Deng [/bib_ref] created a unique bilayer composite membrane using the superhydrophobic selective skin of amorphous polypropylene (APP) and the support composition of electrospun The WCAs of the paint-coated surfaces were tested after each abrasion cycle, and stable superhydrophobicity was obtained with almost all WCAs larger than 150°. (D) After sand impact for 50 cycles, the WCAs of the coatings remained high, also showing super water repellency. When placed in pH = 1 (E), pH = 7 (F), and pH = 14 (G) solutions for 2, 4, and 8 h, respectively, the coating still manifested super water repellency with high WCAs and low RCAs. (H) Multifunctional self-cleaning was shown on the coating, where sand particles could be removed by rolling water, and organic dye could be decomposed by UV light [bib_ref] A Universal, Multifunctional, High-Practicability Superhydrophobic Paint for Waterproofing Grass Houses, Zhu [/bib_ref].
Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 10 polyvinylidene fluoride (PVDF) nanofibers. The permeable vapor flux of the superhydrophobic APP/PVDF membrane is 53.1 kg/ (m 2 - h), and the permeable conductivity is stable. At the same time, it has great applicability in MD desalination. [bib_ref] Tri-Bore Pvdf Hollow Fibers with a Super-hydrophobic Coating for Membrane Distillation, Lu [/bib_ref] developed a porous polyvinylidene fluoride (PVDF) three-porous hollow fiber membrane with superhydrophobicity. The three-pored hollow fiber has greater mechanical strength than traditional single-pored fibers. Under the supreme coating conditions (0.025 wt% Teflon ® AF 2400, 30 s), a superhydrophobic surface was obtained which contact angle is 151°. At the same time, Teflon ® AF 2400-coated membrane has higher stability, Distilled water is produced by the differential partial pressure of steam due to the different temperatures between hot brine and cold deionized water, which drives the transfer of steam from the feed stream to the distillate stream. Su et al. [bib_ref] Robust Superhydrophobic Membrane for Membrane Distillation with Excellent Scaling Resistance, Su [/bib_ref] used electronic co-spinning/spraying (ES2) with chemical vapor welding to produce superhydrophobic films with mechanical strength, high porosity and robustness, which also has outstanding vapor permeability. The prepared superhydrophobic film WCA is bigger than 150°and SA is lower than 10°. Compared with the superhydrophobic film deposited on the surface of fluorinated nanoparticles, the superhydrophobic film has stronger wettability and wear resistance on MD, the surface of WCA and SA has little change after different ultrasonic treatment time, and the surface morphology of the solid superhydrophobic film does not change greatly after observation on SEM.
## Battery
Solar cells [bib_ref] Effect of Dust Accumulation on Solar Transmittance through Glass Covers of Plate-type..., Hegazy [/bib_ref] [bib_ref] Fabrication of Transparent, Durable and Self-Cleaning Superhydrophobic Coatings for Solar Cells, Liang [/bib_ref] are popular because of their low-cost, friendly environment, and renewable characteristics [bib_ref] Advances in Approaches and Methods for Self-Cleaning of Solar Photovoltaic Panels, Syafiq [/bib_ref]. However, in practical application, the solar cells will affect the efficiency due to the influence of environmental temperature, dust, and wind speed. Therefore, we need to develop a solar cell board which can resist pollution. Superhydrophobic materials can be used in batteries on account of their low surface energy and surface roughness, and they have the characteristics of self-cleaning. [bib_ref] A Super-hydrophobic Quasi-Solid Electrolyte for Li-O2Battery with Improved Safety and Cycle Life..., Wu [/bib_ref] developed a viable lithium-O 2 battery with lithium metal negative electrode in a humid environment (relative humidity of 45%), which prevents H 2 O by constructing a superhydrophobic quasi-solid electrolyte (SHQSE). In, the water contact angle is larger than 150°, which indicates that the SHQSE membrane is superhydrophobic and the SHQSE membrane has mechanical stability due to the porous substrate of nonwoven fabrics. From, it displays the classic discharge and charge profiles during cycles, which shows that the hydrophobic effects may take a vital part in the achievement of safe and permanent Li-air battery. [bib_ref] Fabrication of Transparent, Durable and Self-Cleaning Superhydrophobic Coatings for Solar Cells, Liang [/bib_ref] used plasma-improved chemical vapor deposition (PECVD) to prepare SiO 2 as the bottom layer, and then hydrolyzed and condensed epoxy propylpropyltrimethoxysilane (KH560) at both ends to shape a network structure as an intermediate connecting layer. The hydrophilic SiO 2 modified by hexamethyldisilazane (HMDS) to obtain the top superhydrophobic layer. The structure of the superhydrophobic surface is like the double layer structure of phospholipid in the cell membrane. Compared with the bare glass panel, the glass cover plate used in solar cells greatly improves the efficiency of utilization. Zhi et al.first formed threedimensional nanopores crosslinked network by the volatilization of pore-forming agents during calcination, then make the silica nanoparticles attached on the pore structure is formed on the double scale structure, thus forming a kind of superhydrophobic coating, a coating made of surface display WCA is 157.9°, which method is simple, and low coating can be applied in the solar cell cover glass.
## Others
The principle of superhydrophobic anti-icing [bib_ref] On the Nanoengineering of Superhydrophobic and Impalement Resistant Surface Textures below the..., Maitra [/bib_ref] [bib_ref] Effect of Decanol Vapors on the Delay in Water Droplet Crystallization on..., Boinovich [/bib_ref] [bib_ref] Mechanism Analysis and Durability Evaluation of Anti-icing Property of Superhydrophobic Surface, Liu [/bib_ref] is to cut down the contact Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 area between water drop and the superhydrophobic surface, and postpone the frozen time of water droplets on the surface. Meanwhile, before freezing, water droplets slide down with the help of gravity, reducing the possibility surface icing. Chen et al.structed a superhydrophobic composite coating on the basis of MOF (ZIF-8) nanoparticles and organic resins, which shows superhydrophobicity and the water contact angle is 168.2°because of the rough structure of ZIF-8 nanoparticles and the low surface energy. After being rubbed with sandpaper or immersed in different pH value, the superhydrophobicity can still be maintained, showing that the coating has excellent wear resistance and chemical stability.shows the freezing process of the coating surface after dripping 0°C water and the results reveal that the ZIF-8/POTS/EP superhydrophobic coating exhibits great anti-icing properties.
A superhydrophobic surface with a low rolling angle helps to reduce the resistance of the water surface, and the existence of the surface microstructure can make the liquid flow through the superhydrophobic surface to form a gas-liquid two-phase flow, resulting in a slip flow phenomenon, reducing the velocity gradient on the boundary surface, thereby reducing the resistance of the liquid flowing through the solid surface [bib_ref] Failure Pressures and Drag Reduction Benefits of Superhydrophobic Wire Screens, Venkateshan [/bib_ref]. Luo et al. [bib_ref] Robust and Durable Fluorinated 8-Maposs-Based Superamphiphobic Fabrics with Buoyancy Boost and Drag..., Luo [/bib_ref] prepared a sturdy and durable fluorinated 8-Methacryl polyhedral oligomeric silsesquioxane Cage Mixture-based superamphiphobic fabric (Fabrics-S-MAPOSS-F) [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref] , which could easily float on the surface of water or mixed oil, and could resist high temperature and acid corrosion [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref]. The navigation speed of Fabrics-S-MAPOSS-F in water and mixed oil is increased by 2.5 times, and the drag reduction rate is up to 154.7%. As shown in [fig_ref] FIGURE 2 |: Mechanical superhydrophobic models [/fig_ref] -E, the mechanical stability of the superamphiphobic fabric is evaluated through knifescratching, finger hand touch, hand twisting, and turbulent water flow impact, the results show that Fabrics-S-MAPOSS-F is still superhydrophobic.
The beetle [bib_ref] Beetle-Inspired Wettable Materials: From Fabrications to Applications, Zhu [/bib_ref] uses the special structure of the shell to collect water to provide itself with water resources. The cactus spines have a round cone-shaped wedge structure with Laplace pressure and surface energy gradient on the surface to achieve water collection . Inspired by natural creatures, lots of superhydrophobic materials are developed for water collection [bib_ref] Integration of Water Collection and Purification on Cactus-and Beetle-Inspired Eco-Friendly Superwettable Materials, Zhu [/bib_ref]. used copper particles and titanium dioxide particles to prepare coatings with superhydrophobic properties which can be used for water collection [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref]. As shown in [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] , when the molar ratio of the prepared sample precursor is 9:1, the water collection rate is the biggest water collection rate of 1309.9 mg h −1 cm −2 , and showed an approximate WCA and RA of 155.11, 4.51, respectively. After sandpaper friction [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] , it is observed that there is no great change in WCA and RA [fig_ref] FIGURE 3 |: Wear resistance test [/fig_ref] due to the excellent adhesion of epoxy resin is helpful to improve the surface firmness, indicating that the coating has excellent mechanical wear resistance.
# Conclusion
Superhydrophobic materials with outstanding mechanical and chemical stability are highly vital in practical application. This review elaborates the progress of mechanical-chemical superhydrophobic materials in recent years. Firstly, the typical superwetting models are introduced, such as "Young's contact," "Wenzel," "Cassie," "Wenzel-Cassie," "Lotus," and "Gecko" model. Secondly, some mechanical-chemical superhydrophobic models and corresponding tests to evaluate mechanical and chemical durability are discussed. Finally, the application of these mechanical-chemical superhydrophobic materials is described. Although great scientific progress has been made in the research of durable superhydrophobic surfaces, up to now, almost no superhydrophobic surface can withstand all types of wear required by strict industrial requirements and commercial standards. Therefore, the following are some of our views and opinions:
(1) There are a great many studies to increase the mechanical properties of superhydrophobic materials, and there are many differences in the durability tests carried out. However, unified standards to measure the durability of superhydrophobic materials are lacking and should be formulated.
(2) At present, the durable superhydrophobic surface has not been widely employed in practical application, which indicates that the development of durable superhydrophobic surface should take practical application into consideration.
(3) In the preparation of superhydrophobic materials, many used organic materials are harmful to the human body and environment. Environment-friendly materials and green preparation technology are highly recommended.
We believe that a comprehensive and depth review will provide strategic guidance for the development of multifunctional durable superhydrophobic materials, and that the most challenging aspect is to create a durable superhydrophobic material without affecting wettability. We believe that a comprehensive review can provide new ideas for the development and application of superhydrophobic materials. The research of durable superhydrophobic materials is constantly developing and innovating, and its research will become a hot development direction in the next few years.
# Author contributions
QL was responsible for literature retrieval and manuscript writing. HZ and SJ reviewed and edited the manuscript. JP, XC, HZ, and XD revised the manuscript.
# Funding
This project was supported by the Fundamental Research Funds for the Central Universities (CZY19005).
[fig] FIGURE 2 |: Mechanical superhydrophobic models: (A) Self-hardness: the surface of cement(Song et al., 2017a). (B) Porous materials: the surface of textile and sponge(Shang et al., 2020; Ozkan et al., 2020). (C) "Paint + adhesive" method(Qing et al., 2019;. (D) Schematic diagram of a strategy to enhance the mechanical robustness of superhydrophobic surfaces by containing hydrophobic nanostructures in protective microstructures "armor"(Wang et al., 2020c). [/fig]
[fig] FIGURE 3 |: Wear resistance test: (A) Sandpaper abrasion. (B) Tape-peel test. (C) Knife-scratch test. (D) Finger wiping test. (E) Taber abrasion test (Ye et al., 2017). (F) Sand or water impact test.Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 [/fig]
[fig] FIGURE 4 |: (A) Manufacturing process diagram of anti-fogging endoscope. (B) Picture of various liquids on the LIDENS (scale bar: 1 cm). (C) Schematic diagram of tape-peeling. (D) SEM images after 10, 20, 30 tape-peel experiment cycles (scale bars: 20 μm).(E) Continuous photographic images after exposed glass (left) and LIDENS (right) are placed on distilled water (~80°C, 100% relative humidity) for about 3 cm and 60 s(Lee et al., 2020). [/fig]
[fig] FIGURE 5 |: (A) PE foam and F-PE/SiO 2 foam schematic diagram of foam plastic preparation process. (B) Illustration of sandpaper abrasion for the foam surface. (C) The SEM image of F-PE/SiO 2 foam. (D) Water on the surface of F-PE/SiO 2 foam. (E) Picture of 30°inclined F-PE/SiO 2 foam polluted by sands before and after water drop washing (Wu et al., 2021).Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 [/fig]
[fig] FIGURE 6 |: (A) Schematic illustration of fabrication of IOS-PA. (B) UV-Vis spectra of Nile red solution showing decomposition by F-ER-TiO 2 NPs every 1 h. The insets are optical photos of the color variations. (C) [/fig]
[fig] FIGURE 7 |: (A) Schematic diagram of superhydrophobic cotton fabrics prepared by spray deposition of the thiol−ene resin and UV curing. (B) SEM images of the superhydrophobic CO/POSS/SiO 2 coated cotton fabric and the insets are the corresponding WCA and SA. (C) Schematic illustration of the sandpaper abrasion and CA and SA changes after different separation cycles. (D) Photos of different liquids on coated fabrics. (E) Schematic illustration of the separation process of the oil/water mixture and separation efficiency and flux of petroleum ether/water mixture after different separation cycles(Shang et al., 2020). [/fig]
[fig] FIGURE 8 |: (A) Schematic illustration of the configuring process of functionalized textiles. (B) Bacterial shielding experiments of cotton fabrics. (C) SEM images of the textiles. (D) Picture of a water drop (10 μl) on the treated cotton fabrics surface before and after 600 abrasion cycles. (E) The schematic diagram of anti-bacterial action (Ye et al., 2021); Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 11 which average flux is 21 kg m −2 h −1 and rejection rate is 99.99% in 60°C desalination applications. [/fig]
[fig] FIGURE 9 |: (A) Schematic diagram of the ES2 process for fabricating robust superhydrophobic membrane. (B) The WCA and SA of robust superhydrophobic membrane. (C) WCA and SA of the r-SH membranes after different durations of ultrasonication. (D) SEM surface morphology of ES2-derived robust superhydrophobic membrane 270 min before (left) and after (right) ultrasound. (E) The mechanism of membrane distillation. (F) Vapor flux and conductivity of superhydrophobic electrospun fiber membrane (Su et al., 2019).Frontiers in Bioengineering and Biotechnology | www.frontiersin.org July 2022 | Volume 10 | Article 947327 [/fig]
[fig] FIGURE 10 |: (A) Schematic diagram of solid Li-O 2 battery in humid atmosphere on basis of the superhydrophobic quasi-solid electrolyte (SHQSE). (B) SEM image of the original nonwoven fabric and the insets are the corresponding water CA. (C) The typical discharge-charge profiles of Li-O 2 batteries when relative humidity is 45% (Wu et al., 2017b). [/fig]
[fig] FIGURE 11 |: (A) The manufacturing process of EP coating, superhydrophobic ZIF-8/POTS coating and ZIF-8/POTS/EP coating. The pictures of (B-D) Q235 steel sheet and (E-G) ZIF-8/POTS/EP coating after (B,D) 2 h in −20°C refrigerator, and after (C,F) 0.1s and (D,G) 2 s of dripping 0°C water droplets on their surfaces. (H) The schematic of sandpaper abrasion test, and (I) the change of abrasion length on the CA. (J) The change of pH values of water droplet on the CA of ZIF-8/POTS/EP coating, inset picture is the photograph of litmus colored water droplets with different pH value on ZIF-8/POTS/EP coating(Chen et al., 2021). [/fig]
[fig] FIGURE 12 |: (A) Schematic diagram of manufacturing process of Fabric-S-MAPOSS-F; Durability tests through (B) knife-scratching, (C) hand twisting, (D) finger hand touch, and (E) turbulent water flow impact. (F) Common droplets (kerisine, dyed with oil red dye; glycol, colorless; blended oil, yellow; water; vinegar, brown; milk, lacte) on fabric, and liquid repellency of Fabric-S-MAPOSS-F after immersion in 98% H 2 SO 4 for 30 min and 300°C heating for 2 h(Luo et al., 2020). [/fig]
[fig] FIGURE 13 |: (A) WCA and RA on the surface after abrasion test. (B) Water collection rate changed with the precursor of Cu and TiO 2 . (C) Schematic of the abrasion test. (D) Schematic diagram of self-made fog collection system, H and T represent the humidity thermometer. [/fig]
[table] TABLE 1 |: Advantages and disadvantages of different methods. [/table]
|
Effect of Interactions on the Nutrient Status of a Tropical Soil Treated with Green Manures and Inorganic Phosphate Fertilizers
Integrated nutrient management systems using plant residues and inorganic P fertilizers have high potential for increasing crop production and ensuring sustainability in the tropics, but their adoption requires in-depth understanding of nutrient dynamics in such systems. This was examined in a highly weathered tropical soil treated with green manures (GMs) and P fertilizers in two experiments conducted in the laboratory and glasshouse. The treatments were factorial combinations of the GMs (Calopogonium caeruleum, Gliricidia sepium, and Imperata cylindrica) and P fertilizers (phosphate rocks [PRs] from North Carolina, China, and Algeria, and triple superphosphate) replicated thrice. Olsen P, mineral N, pH, and exchangeable K, Ca, and Mg were monitored in a laboratory incubation study for 16 months. The change in soil P fractions and available P was also determined at the end of the study. Phosphorus available from the amendments was quantified at monthly intervals for 5 months by 33 P-32 P double isotopic labeling in the glasshouse using Setaria sphacelata as test crop. The GMs were labeled with 33 P to determine their contribution to P taken up by Setaria, while that from the P fertilizers was indirectly measured by labeling the soil with 32 P. The P fertilizers hardly changed Olsen P and exchangeable cations during 16 months of incubation. The legume GMs and legume GM+P did not change Olsen P, lowered exchangeable Ca, and increased exchangeable K about threefold (4.5 cmol[+]kg -1 soil) in the first 4 months, even as large amounts of NH 4 -N accumulated (~1000 mg kg soil -1 ) and soil pH increased to more than 6.5. Afterwards, Olsen P and exchangeable Ca and Mg increased (threefold) as NH 4 + -N and soil pH declined. The legume GMs also augmented reversibly sorbed P in Al-P and Fe-P fractions resulting in high residual effect in the soil, while fertilizer-P was irreversibly retained. The GMs increased PR-P utilization by 40 to over 80%, mobilized soil P, and markedly enhanced uptake of N, K, Ca, and Mg. Thus GMs+PRs is an appropriate combination for correcting nutrient deficiencies in tropical soils.
# Introduction
Increasing crop production in many areas of the tropics (especially sub-Saharan Africa) is faced with major challenges because of the predominance of highly weathered and inherently infertile Ultisols and Oxisols [bib_ref] Myth and Science of Soils of the Tropics, Sanchez [/bib_ref]. These soils have severe chemical constraints associated with low pH including low cation saturation, toxic concentrations of Al, as well as deficiency in N and P [bib_ref] Management considerations for acid soils with high phosphorus fixation capacity, Sanchez [/bib_ref]. Other factors include declining arable land area due to soil degradation and demographic pressure [bib_ref] Soil management in the developing countries, Lal [/bib_ref] , and the inadequate use of high analysis fertilizers because of severe socioeconomic constraints on the farmers. Intensifying crop production in these circumstances often results in depletion of organic matter [bib_ref] World cropland soils as a source of sink for atmospheric carbon, Lal [/bib_ref] , high nutrient losses [bib_ref] Estimation rates of nutrient depletion in soils of agricultural lands of Africa, Henao [/bib_ref] , and ultimately severe land degradation [bib_ref] Soil fertility replenishment in Africa: an investment in natural resource capital, Sanchez [/bib_ref]. Consequently, soil fertility restoration and maintenance programs in the tropics need to emphasize the judicious use of chemical fertilizers coupled with improved management of soil organic matter to enhance nutrient cycling, mitigate losses, and hence maintain environmental quality. This could be through improved fallow management with N-fixing trees, hedgerow intercropping with legumes, biomass transfer, and the use of organic inputs like green manures (GMs) and animal manures [bib_ref] Soil fertility and hunger in Africa, Sanchez [/bib_ref].
Decomposing organic materials generally release N, P, and basic cations concomitantly with organic anions, which can detoxify heavy metal ions through complex formation [bib_ref] Amelioration of Al toxicity and P deficiency in acid soils by additions..., Haynes [/bib_ref]. Therefore, they have high potential for improving tropical soils. However, while crop N requirements could be satisfied easily by GM applications, that of P are fraught with major problems because of the generally low P content of organic matter [bib_ref] Combined use of organic and inorganic nutrient sources for soil fertility maintenance..., Palm [/bib_ref]. This implies enormous labor requirements for the management of P using organic materials because of their scarcity and bulky nature.
A better alternative for improving productivity in tropical soils would be to integrate organic materials with cheaper inorganic P sources such as phosphate rocks (PRs) [bib_ref] Agronomic and economic evaluation of organic and inorganic sources of phosphorus in..., Jama [/bib_ref] [bib_ref] Appropriate farm management practices for alleviating N and P deficiencies in low-nutrient..., Smithson [/bib_ref]. This is a more attractive and sustainable option for resource-poor farmers who cannot afford costly, manufactured fertilizers. The combined use of GMs and PRs has been shown to enhance the utilization of PR-P in acid soils [bib_ref] Agronomic and economic evaluation of organic and inorganic sources of phosphorus in..., Jama [/bib_ref] [bib_ref] Effect of green manures on P solubilization and uptake from phosphate rocks, Zaharah [/bib_ref]. In principle, the degrading GMs and PRs in the combined system release nutrients that can create conditions that would either promote or adversely affect soil nutrient status. For instance, both materials release Ca, which in the absence of a sink can retard further P release from PRs [bib_ref] Chemistry of phosphorus transformation in soil, Sanyal [/bib_ref]. Similarly, GM decomposition increases soil pH [bib_ref] Amelioration of Al toxicity and P deficiency in acid soils by additions..., Haynes [/bib_ref] , which at moderate levels facilitates P desorption [bib_ref] Organic amendments and phosphorus sorption by soils, Iyamuremye [/bib_ref] , but can exacerbate P retention at higher levels [bib_ref] Effect of liming on phosphate sorption by acid soils, Naidu [/bib_ref]. At the same time, soil microbial activity is promoted and significant amounts of nutrients can be immobilized in the microbial biomass [bib_ref] Biological immobilization and chemical fixation of native and fertilizer phosphorus in soil, Ghoshal [/bib_ref]. All these processes would have significant impact on plant-available nutrients in the soil. Therefore, there is need for better understanding of nutrient interactions and secondary transformations that occur when PRs and GMs are incorporated together in tropical soils. This is important for adapting such integrated nutrient management systems to various cropping systems in the tropics. As a result, this study was carried out to evaluate the effect of PRs and GMs on nutrient dynamics and soil nutrient status, and to demonstrate the potential of such a system for enhancing productivity in a tropical soil. Some emphasis was made on available P dynamics because P is the most limiting as well as the most difficult nutrient to replenish in tropical soils. Thus the influence of GMs on the efficacy of reactive and poorly reactive PRs in increasing plantavailable P in this soil was investigated using P tracer techniques. In this regard, the contribution of the GMs was assessed by directly labeling each of them with 33 P radioisotope. Labeling PRs without altering their characteristics is difficult, so the contribution of the PRs to P availability was indirectly determined by the 32 P isotope dilution technique. The latter is based on the principle that P tracers undergo exchange reactions with chemically identical, but isotopically different forms of P in a soil system. This results in the dilution of the P tracer, so available P contribution from PRs is determined by assuming that when a plant has more than one nutrient source, the quantity taken up from each source will be in direct proportion to its availability. Thus their contributions to plant-available P can be obtained without disturbing the soil system.
# Materials and methods
Laboratory incubation and glasshouse studies were carried out with a tropical soil treated with factorial combinations of three GMs and four P fertilizers. The soil was topsoil (0-20 cm) from the Bungor series (Typic Paleudult, kaolinitic, isohyperthermic) with pH water (1:2.5) of 4.8 and 1 kg soil has 18 g organic C, 120 mg N, 10.9 mg Olsen P, 8 g Al 2 O 3 , 24 g Fe 2 O 3 , 275 g clay, and 559 g sand. Also, cation exchange capacity (CEC) is 5.4 cmol (+) kg -1 with 0.37, 0.92, 0.44, and 1.8 cmol (+) kg -1 exchangeable K, Ca, Mg, and Al, respectively. Soil pH was measured in distilled water (1:2.5 w/v) using a glass electrode, organic C by the dichromate wet oxidation method [bib_ref] An examination of the Degtjareff method for determining soil organic matter and..., Walkley [/bib_ref] , total N by the Kjeldahl method [bib_ref] Nutrient status of a tropical soil treated with green manures and phosphate..., Bremner [/bib_ref] , and available P by extraction with 0.5M NaHCO 3 at pH 8.2and then measuring P in the extract by the method of Murphy and Riley [bib_ref] A modified single solution method for the determination of phosphate in natural..., Murphy [/bib_ref]. Active Al and Fe were analyzed by dithionite extraction [bib_ref] Iron oxide removal from soils and clays by a dithionite-citrate system buffered..., Mehra [/bib_ref] followed by measurement with an atomic absorption spectrophotometer. Particle size analysis was carried out by the pipette method after removing organic matter with H 2 O 2 and dispersing soil particles with sodium hexametaphosphate. Also, exchangeable acidity was assessed by extraction with 1M KCl (1:5, w/v) followed by titration with NaOH. Exchangeable Al in this extract was determined by atomic absorption spectrophotometry. CEC and exchangeable K, Ca, and Mg were measured by first leaching the soil with neutral molar NH 4 OAc to saturate exchange sites with NH 4 + and leach out adsorbed cations, then eluting the adsorbed NH 4 + with 0.5M K 2 SO 4 . The exchangeable cations were analyzed in the first leachate by atomic absorption spectrophotometry following dilution of an aliquot of the leachate with 1000 µg ml -1 SrCl 2 . The CEC was estimated in the second leachate via steam distillation wherein the distillate was collected in 2% H 3 BO 3 and titrated with 10 mM HCl. The GMs were green leaves of Gliricidia sepium (Gliricidia) and Imperata cylindirica (Imperata), and the leaves plus green stems of Calopogonium caeruleum (Calopogonium) [fig_ref] TABLE 1 Biochemical: Characteristics of the GMs UsedParameter [/fig_ref]. Chemical analysis of the GMs was conducted on oven-dried (70°C), ground (<1 mm) samples. Total P, K, Ca, and Mg were measured after ashing at 550°C for 4 h. The ash was dissolved in 2M HCl and evaporated to dryness. The residue was then dissolved in 20% HNO 3 and analysis of the elements carried out as described above. Total N was also measured as in the soil. Total polyphenols was determined by the Folin-Denis procedure and lignin by the acid detergent fiber method. For the glasshouse study, the GMs were labeled with 33 P by applying a solution of 33 P (activity 3.33 MBq) and 18 µg P (KH 2 PO 4 ) carrier to 15-month-old plants grown in Al containers. The plants were allowed to grow for another 3 weeks before harvesting the GMs. The P fertilizers were triple superphosphate (TSP) and PRs from Algeria (APR), China (CPR), and North Carolina (NCPR) [fig_ref] TABLE 2 Chemical: Properties of the P Fertilizers Used [/fig_ref]. The P fertilizers were characterized in terms of both elemental content and their solubility in citric acid. Two grams of each P fertilizer was digested by adding subsamples to a mixture of HNO [bib_ref] Soil management in the developing countries, Lal [/bib_ref] and HCl (4 ml each), in 50 ml H 2 O and boiling for 30 min. After filtration, determinations were carried out for P, Ca, Fe, and Al as described above. The solubility test was carried out with 2% citric acid (1:100, w/v).
## Incubation procedure
One kilogram of air-dried soil was sieved (<2 mm) and the moisture content increased to 90% of field capacity with distilled water in plastic bags to stimulate microbial activity. These were put aside for 14 d for equilibrium to be established. Factorial combinations of the fresh GMs (5% w/w, dry weight basis) and P fertilizers (66. 7 mg P kg -1 soil) were then mixed thoroughly with the soil in three replications and completely randomized. An untreated control was also added. Imperata and Calopogonium were cut into smaller pieces (1-2 cm) to minimize size differences among the GMs. Soil samples were collected at weekly intervals in the first month and thereafter at 2, 4, 8, and 16 months after treatment application, to measure NH 4 -N and NO 3 -N [bib_ref] Nitrogen -inorganic forms, Keeney [/bib_ref] , pH, exchangeable K, Ca, and Mg by the neutral molar NH 4 OAc method and bicarbonate extractable P. The final samples were also sequentially extracted for P fractions [bib_ref] Fractionation of soil phosphorus, Chang [/bib_ref] and organic P was also determined [bib_ref] Observations on the determination of total organic phosphorus in soils, Saunders [/bib_ref]. Plant-available P was estimated by the iron oxide impregnated paper strip (Pi) method [bib_ref] Determination of plant available phosphorus by the iron hydroxide-impregnated filter paper (Pi)..., Menon [/bib_ref].
## Glasshouse studies
The same treatments were used here except that the soil was labeled with a mixture of 32 P (1.85 MBq) and 2 mg P carrier (KH 2 PO 4 ) before being set aside for equilibrium to be established. The pool of available P in the soil becomes labeled with 32 P through exchange reactions. This is "diluted" by the P contributed by the applied materials (GMs and P fertilizers) after treatment application. Thus plants grown in pots without treatment would show maximum uptake of 32 P, which is reduced by the proportion of P contributed from the GMs and/or P fertilizers in the treated pots. In addition, the GMs were labeled with 33 P and this would give the amount of P directly contributed by the GMs. Two stem cuttings (12-15 cm) of Setaria sphacelata were planted in each pot and supplementary N and K (33.3 mg pot -1 ) were added to treatments without GMs. The plants were manually irrigated with distilled water daily and harvested at 4, 8, 15, and 20 weeks after treatment application. Dry weight was recorded after oven drying to constant weight (70°C) and manually crushed subsamples were ashed at 550°C for 4 h. The ash was dissolved in 2M HCl, filtered, and the activities of 33 P and 32 P determined by liquid scintillation counting using the external standard channels ratio method to determine efficiency [bib_ref] Liquid scintillation counting, L'annunziata [/bib_ref] , and total P was measured by the Scheel [bib_ref] Colorimetric determination of phosphoric acid in fertilizers with the pulfrich photometer, Scheel [/bib_ref] procedure. Total K, Ca, and Mg were also measured as described earlier. Percent P taken up from the GMs, %PdfGM, was then calculated from the following relationship: where SA trt is the specific activity of plants in the treated soil and SA gm is the specific activity of the applied GM.
[formula] %PdfGM = (SA trt /SA gm ) × 100 (1) [/formula]
Total GM-P (mg pot -1 ) = %PdfGM × PYLD/100 (2) wherein PYLD is total P uptake (mg pot -1 ), and PYLD = %P × DMY × 10 (3) DMY= dry matter yield (g pot -1 ). Percent P derived from the fertilizer treatments (%PdfT) is given by the following equation (4) where SA amend = specific activity of plants in the amended soil and SA control = specific activity of plants in the control.
[formula] %PdfT = {1 -(SA amend /SA control )} × 100 [/formula]
# Data analysis
Changes in the parameters monitored were calculated as differences between the treated soil and the control. Main effects of treatments were ascertained by analysis of variance (ANOVA) using the PROC GLM procedure of the SAS statistical package. In cases where GM×P interactions were significant, the PROC SORT procedure was used to ascertain the effect of GMs on the P fertilizers for the parameters. Means separation was carried out by the least significant difference (LSD) method at α = 0.05. Pearson's correlation analysis was used to ascertain relationships between some of the properties monitored and the contributions of soil P fractions to plant available P was determined by step-wise multiple regression.
# Results and discussion
## Effect of gms and p fertilizers on soil nutrients
The P fertilizers alone hardly affected soil mineral N, pH, and exchangeable cations due to the absence of N and low net alkalinity of these materials, as well as the buffering action of the soil. Also, less than 35% of applied P was available even from water soluble TSP and reactive APR due to retention of released P (large concentration of Al and Fe sesquioxides). Similarly, the legume GMs, alone or combined with inorganic P, had very little effect on available P, exchangeable Ca and Mg cations during the first 16 weeks (4 months), but markedly increased NH 4 -N as well as soil pH during this period. Calopogonium had the greatest effect, 600 µg NH 4 -N g -1 (≈40% of added GM-N) was accumulated and pH increased from 4.8 to more than 6.5 in the first 8 weeks of the incubation . This is consistent with the rapid release of N by high-quality GMs (low C/N, low lignin and polyphenols) [bib_ref] Mineralization of nitrogen from decomposing leaves of multipurpose trees as affected by..., Mafongoya [/bib_ref] [bib_ref] Nitrogen release from the leaves of some tropical legumes as affected by..., Palm [/bib_ref]. Therefore, the very small buildup of NH 4 -N in the Gliricidia treatment (<2% of applied GM-N) is rather surprising (C:N = 10; [lig + poly]:N = 3.5). This was likely due to volatilization of NH 3 owing to the low soil CEC, and perhaps the 5% application rate was too high for this GM [bib_ref] N transformations in tropical soils I. Mineralization of N rich organic materials..., Cornforth [/bib_ref]. As anticipated from its poor quality (C:N = 32; [lig + poly]:N = 12.5), Imperata caused little change in soil NH 4 -N. The pH increases caused by GMs is generally attributed to ammonification, ligand exchange reactions between hydroxyl groups on soil surfaces and organic anions, as well as dissociation/association reactions of various organic acids and decarboxylation of organic anions by microbial action [bib_ref] Amelioration of Al toxicity and P deficiency in acid soils by additions..., Haynes [/bib_ref] [bib_ref] Soil acidification, its measurement and the processes involved, Helyar [/bib_ref] [bib_ref] Effect of short-term legume residue decomposition on soil acidity, Tang [/bib_ref]. Ammonification was likely the overriding influence in our study as evidenced in the highly significant positive correlation between NH 4 -N and soil pH in the early stages of the incubation [fig_ref] TABLE 3 Correlation: Between pH and Some Soil Properties Monitored During Incubation of a Tropical... [/fig_ref]. The other mechanisms could have been important in the case of Gliricidia, considering that NH 4 -N accumulation was very low in this treatment. The correlation diminished noticeably from 16 weeks onwards in line with the decline in soil NH 4 -N probably due to loss through volatilization, nitrification, and depletion of organic N in the GMs during decomposition [bib_ref] Mineralization of nitrogen from decomposing leaves of multipurpose trees as affected by..., Mafongoya [/bib_ref] [bib_ref] Patterns of decomposition and nutrient release by fresh Gliricidia (Gliricidia sepium) leaves..., Zaharah [/bib_ref]. This coincided with the lowering of soil pH and remarkable increases in available P as well as exchangeable Ca and Mg [fig_ref] FIGURE 3: **, and *** denote significance at 5%, 1%, and 0 [/fig_ref] , and 5).
## Exchangeable cations
Exchangeable Ca was generally low during the first 8-16 weeks of the incubation, while exchangeable K and Mg declined throughout the study period in soils treated with inorganic P fertilizers alone (data not shown). On the other hand, the decomposing GMs markedly increased exchangeable Ca after 16 weeks, and for K and Mg after the 8th week of incubation . The extent of the effect was dependent on GM chemical characteristics (concentration of N, lignin and polyphenols, C/N, C/P; and concentration of K, Ca, and Mg), which govern their rate of decomposition as well as nutrient release patterns. Thus, the legume GMs generally caused much higher increases in exchangeable K and Mg than Imperata. As anticipated, Gliricidia (being the GM with the highest Ca content [ [fig_ref] TABLE 1 Biochemical: Characteristics of the GMs UsedParameter [/fig_ref] caused the most significant buildup of exchangeable Ca beyond the 16 th week of incubation, while the effect of Calopogonium was comparable with that of Imperata . The pattern of change was different among the nutrients; exchangeable K increased to a maximum in 4 months (16 weeks) and there was little change thereafter. Calcium showed depressed levels during this period, but an increasing trend afterwards. Magnesium, on the other hand, increased gradually throughout the incubation study . Beyond the 16 th week, Gliricidia gave much higher Ca and Mg levels, while Calopogonium gave the highest soil K after 8 months. Imperata had the least effect on soil K and Mg, but showed comparable buildup of Ca with the Calopogonium treatment after the 4 th month.
## Available p
The pattern of P availability was generally influenced by the release of other nutrients in the GM and GM+P treated soils. Ammonification resulted in the elevation of pH to more than 6.5 [fig_ref] FIGURE 3: **, and *** denote significance at 5%, 1%, and 0 [/fig_ref] , which increased P retention due to the precipitation of amorphous hydroxy Al polymers that acted as additional sorption sites [bib_ref] Effect of liming on phosphate sorption by acid soils, Naidu [/bib_ref] [bib_ref] Effect of liming and air-drying on the adsorption of phosphate by some..., Haynes [/bib_ref]. Therefore, Olsen P was depressed in the GM-and GM+P-treated soil in the first 16 weeks of the incubation, and on average, less than 10% of the GM-P applied was available during this period (lag phase) (Figs. 5 and 6). This was most likely due to the high P retention capacity of the soil because of high sesquioxides, high Al saturation, and its 1:1 clay mineralogy [bib_ref] Chemistry of phosphorus transformation in soil, Sanyal [/bib_ref]. Phosphorus retention in these conditions is via precipitation by Al and Fe ions, as well as adsorption on oxides and/or hydroxides of Al and Fe in the soil. This is evidenced by the significant increase in the Al-P and Fe-P fractions in the soil . These P fractions have been associated with P sorption sites in weathered soils [bib_ref] Amelioration of Al toxicity and P deficiency in acid soils by additions..., Haynes [/bib_ref] [bib_ref] Organic amendments and phosphorus sorption by soils, Iyamuremye [/bib_ref] , which implies that the released P mainly saturated sorption sites. This confirms the central role of PO 4 ions in mitigating P retention. Also divalent cations such as Ca 2+ and Mg 2+ were released during GM decomposition [bib_ref] Solubility of soil phosphorus as influenced by urea, Hartikainen [/bib_ref]. These cations reduced the mobilization of adsorbed PO 4 3by converting it to less soluble phosphate compounds of Ca and Mg [bib_ref] Dissolution of phosphate rock during composting of poultry manure: an incubation experiment, Mahimairajah [/bib_ref] , and probably also immobilized available P in complex compounds like struvite, NH 4 MgPO 4 .6H 2 O [bib_ref] Phosphate minerals, Lindsay [/bib_ref]. The Ca (from degrading PRs and GMs) could also suppress further P release from the PRs [bib_ref] Chemistry of phosphorus transformation in soil, Sanyal [/bib_ref]. However, the low exchangeable Ca observed during the early phase of the incubation suggests a sink for the released Ca, probably through the formation of soluble complexes with organic anions. Perhaps this partly explains the lower exchangeable Ca values observed in this phase . The precipitation of calcium phosphates could not have been one of the pathways for lowering exchangeable Ca and available P because the pH increase was not greater than 7 [bib_ref] The relative contribution of plant residues and fertiliser to the phosphorus nutrition..., Mclaughlin [/bib_ref]. Another pathway could have been through microbial action. It is known that the supply of soluble C by the high quality GMs fuels microbial activity, which results in the immobilization of considerable quantities of nutrients (including P) in the microbial biomass [bib_ref] Exchangeable and non-exchangeable phosphate sorption in Portuguese soils, Fernandes [/bib_ref]. In the second phase (>4 months), Olsen P increased significantly and more than 20% of GM-P from the legume GMs was available after 8 months. This increased further at the end of the study to about 40-60% of legume GM-P added . The change in available P in the second phase was related to GM quality (C-to-P ratio). Thus Calopogonium (C:P = 149) and Gliricidia (C:P = 225) increased available P by a much greater proportion of their organic P content than the poor-quality Imperata GM (0.08% P; C:P = 500). Similarly, the overall effect of the legume GMs was more considerable than that of the P fertilizer (3-4 times after 16 months). In the GM+P mixtures, Calopogonium made much greater improvements in available P than Gliricidia, irrespective of the fertilizer. This was likely due to high Ca release from the latter . Combinations of NCPR and TSP had similar effect as those of APR (data not shown).
## Soil nutrient status
The legume GMs alone and their combinations with P markedly improved the nutrient status of the tropical soil at the end of the incubation . Even though the pH in the end was less than the initial value, exchangeable K, Ca, and Mg as well as plant available P (Pi) were significantly increased. Gliricidia more than doubled exchangeable Ca, increased exchangeable Mg about 7 times, and K more than 70 times. On the other hand, Calopogonium had little effect on exchangeable Ca in the P treatments, but made significant improvements in the other cation levels (>4 times for Mg and >70 times for K) . The effect of Imperata was also significant, but much lower compared to the legume GMs. Plantavailable P was also augmented at the end of the incubation study because the P released in the earlier stages of the incubation saturated P sorption sites in the soil. This translated into marked increases in the Al-P and Fe-P fractions . As these two fractions made the most important contributions to plant available P , it suggests that that there was very little occlusion of the sorbed P in the presence of the legume GMs [bib_ref] Leaf litter decomposition of Piper aduncum, Gliricidia sepium and Imperata cylindrica in..., Hartemink [/bib_ref]. The P fertilizers were less effective in increasing the availability of P than the GMs, which suggests that much of the P released from these materials was irreversibly adsorbed. Nevertheless, the GMs increased the effectiveness of the P fertilizers by 24-30% (Imperata), 40-50% (Gliricidia), and 150-440% (Calopogonium) . Hence the GMs enhanced the residual effect of the P fertilizers.
The nutrient uptake pattern of Setaria sphacelata in the glasshouse experiment also reflected the improved nutrient status of the soil following treatment with the GMs and their mixtures with P fertilizers. These materials markedly increased the uptake of N, P, K, Ca, and Mg. [fig_ref] FIGURE 11: Uptake of K and Mg by Setaria sphacelata after 20 weeks growth... [/fig_ref]. The independent effect of the GMs was much greater than that of the P fertilizers and there was significant positive interaction between the GMs and P fertilizers in the uptake of P, K, and Mg (Figs. 9 and 11). The GMs considerably improved uptake of P as well as that of K and Mg in the fertilizers treatments [fig_ref] FIGURE 11: Uptake of K and Mg by Setaria sphacelata after 20 weeks growth... [/fig_ref]. Surprisingly, the low-quality Imperata made comparable or much greater improvements in P uptake from TSP and APR as the legume GMs . Perhaps the release of high levels of Ca limited the effectiveness of the legume GMs in the GM+P mixtures (especially Gliricidia) by suppressing the dissolution of the PRs. Moreover, the lower content of N, Ca, and Mg in Imperata resulted in net production of acidity during decomposition and this could have promoted the dissolution of the PR [bib_ref] Chemistry of phosphorus transformation in soil, Sanyal [/bib_ref]. However, the action of Imperata was more to maintain P in available forms rather than the promotion of PR degradation. This may have been achieved in this case by the action of organic acids and possibly the conservation of soil moisture because of the low rate of decomposition of Imperata [bib_ref] Leaf litter decomposition of Piper aduncum, Gliricidia sepium and Imperata cylindrica in..., Hartemink [/bib_ref]. This is evidenced by the fact that Imperata enhanced P uptake from highly reactive APR and water-soluble TSP, but not from the poorly reactive CPR . The highly reactive APR and TSP released more P which was maintained in available forms by Imperata, while poorly reactive CPR released very little P in the first place. On the other hand, the enhancing effect of the legume GMs on CPR was possibly due to the immobilization of Al and Fe in this PR since pH was elevated by the release of NH 4 -N from these GMs. However, the concomitant release of divalent cations from legume GMs suppressed further PR dissolution and reduced mobilization of adsorbed P. manures and phosphate fertilizers . Nutrient levels in a tropical soil treated with GMs and inorganic P fertilizers after incubation for 64 weeks.
## Table 4 stepwise multiple regressions between soil p fractions and available p after a 64-week incubation of a tropical soil treated with gms and p fertilizers
## The potential for nutrient management in tropical soils
The studies reported here were conducted under controlled conditions, so extrapolation of the results to field situations should be done with utmost caution because environmental factors such as temperature and rainfall can alter the trends described above. Higher temperatures increase GM decomposition rate and hence N mineralization, which could result in greater loss of the released NH 4 -N through volatilization. Similarly, high rainfall could lead to excessive leaching of released NO 3 -N. Nevertheless, it is evident that the integration of GMs and PRs offers potential for ameliorating deficiency in N, P, and exchangeable cations as well as eliminating Al toxicity (at least temporarily) in tropical soils. Between 50-60% of N applied as Calopogonium was released as NH 4 -N. This increased soil ammonium N over a thousand fold and elevated soil pH from 4.5-6.8 within 8 weeks (2 months). This increased P retention initially, but this was reversed after 4 months to increase available P as much as 6 times by the end of the study. Also, the effectiveness of CPR for increasing available P was increased 4-5 times and that of APR 5-7 times by Calopogonium. But the effect of Gliricidia on the PRs was limited by its high Ca content and the low soil CEC. Values within a column followed by the same letter are not significantly different (P< 0.05).
The positive influence of the GMs was also shown in the improvement of the utilization of fertilizer-P. The direct contribution of the GMs to P uptake, measured using 33 P tracer, was less than 5% [fig_ref] TABLE 5: Percent of total P uptake derived from 33 P labeled GMs by... [/fig_ref]. However, indirect measurements with 32 P radioisotope showed that the GMs markedly increased the uptake of fertilizer-P by 40 to over 80% irrespective of GM quality. Calopogonium made the greatest improvement on the uptake of P from NCPR, CPR, and TSP and Imperata had a greater effect on the PR treatments than Gliricidia [fig_ref] Figure 1 FIGURE 2: NH 4 -N accumulation in a tropical soil treated with GMs [/fig_ref]. Despite these positive interactions between GMs and P fertilizers, nutrient accumulation patterns in this study suggest that management of these materials may have to be timely and targeted to specific soil and climatic conditions for optimum benefits to be realized. Soils with higher CEC and leaching in high rainfall areas could provide sinks for Ca and increase available P in GM+PR mixtures even in early stages. Since the legume GMs decompose rapidly and can lose large amounts of N through volatilization, management techniques such as split application of the GMs or mixing legume and nonlegume GMs should be considered for minimizing these losses. Other studies have shown that incorporating high-quality organic materials (legume GMs) with those of a lower quality (materials with wide C:N, high lignin, and polyphenol) results in greater immobilization of released N, and hence minimize N loss. Also, applying the GMs after the inorganic P fertilizers would probably improve fertilizer-P availability at later stages. Even though significant levels of fertilizer-P could still be adsorbed initially, much of this was shown to be exchangeable adsorption wherein adsorbed P resided in the Al-P and Fe-P soil fractions ; . Therefore, the adsorbed P would be made available later on through the action of organic anions released by decomposing GMs added after the P fertilizers [bib_ref] Amelioration of Al toxicity and P deficiency in acid soils by additions..., Haynes [/bib_ref] [bib_ref] Organic amendments and phosphorus sorption by soils, Iyamuremye [/bib_ref]. This process is continued in the succeeding applications of the remaining GM portions. Such a system could fit in very well with the initial massive input of fertilizer P, designed to saturate sorption sites and enhance availability, in the P capitalization strategy proposed for highly weathered soils in sub-Saharan Africa [bib_ref] Soil fertility replenishment in Africa: an investment in natural resource capital, Sanchez [/bib_ref] [bib_ref] Building soil phosphorus capital in Africa, Buresh [/bib_ref]. Thus the legume GMs would serve to improve the availability of immobilized P in this case. However, careful consideration should be given to the choice of compatible GMs and PRs in these integrated systems. The GMs with high Ca contents like Gliricidia in our study may be targeted for soils with high CEC or in high rainfall areas where Ca could be leached out or possibly include plant genotypes with greater affinity for this nutrient. All this would further enhance plant-available P in the treated soils. In addition to N and P, this system can also improve the level of other macronutrients such as K, Ca, and Mg.
## Figure 12.
Total fertilizer-P uptake (indirectly quantified with 32 P) by Setaria sphacelata grown for 20 weeks in a tropical soil treated with inorganic P fertilizers and GMs.
# Conclusions
The benefit of integrating GMs and inorganic P fertilizers for combating acid soil infertility in the tropics was demonstrated in this investigation. The legume GMs increased soil pH to over 6.5 (albeit for only 8-16 weeks) and significantly enhanced soil nutrient status (N, P, K, Ca, and Mg). However, the 33 P direct labeling technique showed that the GMs provided very little plant available P (<5%). On the other hand, the indirect method wherein the soil was labeled with 32 P proved that the GMs significantly enhanced the availability of fertilizer P by as much as 40 to over 80%, irrespective of GM quality. These results have important implications for the sustainable management of highly weathered tropical soils. However, further field evaluation of this system is required where other factors such as leaching and much higher NH 3 volatilization could alter nutrient dynamics dramatically. Also, the potential of Imperata, a ubiquitous weed in the tropics, in integrated GM-PR systems including socioeconomic factors needs more in-depth evaluation. This could provide great opportunities for resource poor farmers in these regions who must maximize natural resource use because of precarious economic circumstances and fragile environment.
[fig] Figure 1 FIGURE 2: NH 4 -N accumulation in a tropical soil treated with GMs. GLR = Gliricidia, IMP = Imperata and CAL = Calopogonium. Bars = LSD 0.05 .. Changes in soil pH during incubation of a tropical soil treated with GMs. [/fig]
[fig] FIGURE 3: **, and *** denote significance at 5%, 1%, and 0.1% levels; n = sample size.Bah et al.: Nutrient status of a tropical soil treated with NH 4 -N accumulation (A) and pH changes (B) in a tropical soil treated with GMs and PRs. CPR = China PR; bars = LSD 0.05 . [/fig]
[fig] FIGURE 4 FIGURE 5: Changes in (A) exchangeable K, (B) Mg, and (C) Ca in a tropical soil treated with GMs and PRs. Bars = LSD 0.05 . Bah et al.: Nutrient status of a tropical soil treated with Changes in Olsen P in a tropical soil treated with GMs and PRs. (A) NCPR = North Carolina PR; (B) CPR = China PR; (C) APR = Algerian PR. Bars = LSD 0.05 . Bah et al.: Nutrient status of a tropical soil treated with green manures and phosphate fertilizers TheScientificWorldJOURNAL (2004) 4, 393-414 [/fig]
[fig] FIGURE 6 FIGURE 7: Net changes in Olsen P (as %GM-P applied) during incubation of a tropical soil treated with GMs. Bars = LSD 0.05 . Fractions of P in a tropical soil treated with GMs and PRs after 16 months of incubation. [/fig]
[fig] FIGURE 9 FIGURE 10: Total P uptake by Setaria sphacelata after 20 weeks growth in a tropical soil as affected by GMs and P fertilizers. Horizontal lines = LSD 0.05 . Bah et al.: Nutrient status of a tropical soil treated with Uptake of N and Ca by Setaria sphacelata after 20 weeks growth in a tropical soil treated with GMs and P fertilizers. Horizontal lines = LSD 0.05 . [/fig]
[fig] FIGURE 11: Uptake of K and Mg by Setaria sphacelata after 20 weeks growth in a tropical soil as affected by GMs and P fertilizers. Horizontal lines = LSD 0.05 . [/fig]
[table] TABLE 1 Biochemical: Characteristics of the GMs UsedParameter (g kg -1 ) Gliricidia Imperata Calopogonium [/table]
[table] TABLE 2 Chemical: Properties of the P Fertilizers Used [/table]
[table] TABLE 3 Correlation: Between pH and Some Soil Properties Monitored During Incubation of a Tropical Soil Treated with GMs Alone or in Combination with Inorganic P Fertilizers [/table]
[table] TABLE 5: Percent of total P uptake derived from 33 P labeled GMs by Setaria sphacelata grown for 20 weeks in a tropical soil treated with GMs and P fertilizers. [/table]
|
Effect of FAT1 gene expression on the prognosis of medulloblastoma in children
Background: It was reported that cloning human adipose atypical cadherin 1 (FAT1) has an effect on the prognosis of medulloblastoma (MB), while the conclusion still needs to be further proved. Therefore, this study attempted to explore the effect of the high expression of FAT1 on the prognosis of MB children.Methods: The database was retrieved from China National Knowledge Infrastructure (CNKI), Chinese Biomedical literature Database (CBM), Chinese Scientific and Journal Database (VIP), Wan Fang database, PubMed, and EMBASE. Hazard ratios (HRs) and its 95% confidence intervals (CIs) were applied to assess the prognostic effect of FAT1 on overall survival (OS) and disease-free survival (DFS). RevMan 5.3 and STATA 16.0 software were used to perform the meta-analysis.Results: The results of the study would be published in peer-reviewed journals or at relevant meetings. Conclusion: Our findings revealed the effect of the high expression of FAT1 on the prognosis of MB children. Such studies may find a new prognostic marker for MB children and help clinicians and health professionals make clinical decisions. Abbreviations: CBM = Chinese Biomedical Literature Database, CIs = confidence intervals, CNKI = Chinese National Knowledge Infrastructure, DFS = disease-free survival, FAT1 = adipose atypical cadherin 1, HRs = hazard ratios, MB = medulloblastoma, NOS = Newcastle Ottawa Scale, OS = overall survival, PRISMA-P = Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols, VIP = Chinese Scientific and Journal Database.
# Introduction
Medulloblastoma (MB) is a highly invasive malignant tumor of posterior cranial fossa, and usually occurs in children.It is one of the most common malignant tumors in children, accounting for about 20% of central nervous system tumors.Epidemiological data displays that the annual incidence of MB is about from 0.20 to 0.58 per 100,000 people worldwide.The disease can occur in all age groups, and the peak of diagnosis is between 6 and 8 years old.As a cerebellar embryonal tumor, MB may evolve from primitive neural stem cells. However, the specific pathogenesis is not clear at present. It is the most malignant glioma, because the tumor cells are easy to fall off and spread along the cerebrospinal fluid circulation pathway, with about 30% of patients suffering from metastases at the time of diagnosis.At present, targeted treatment is the first choice of surgery. However, the effect of surgical resection on the prognosis is still inconclusive. Especially, the tumors with severe adhesion to the brainstem are difficult to achieve total resection. It is easy to plant and spread through cerebrospinal fluid, the overall prognosis is poor, the case fatality rate is high, and the 5-year average survival rate is 50% to 75%.The latest research illustrated that gene mutation and abnormal intracellular signal transduction are important reasons for the occurrence and development of MB.Therefore, the exploration of the key molecules in the occurrence and development of MB is helpful to find new treatment methods, and is of great clinical significance to improve the disease risk stratification, formulate a reasonable treatment plan and evaluate the prognosis of children.
Previous studies indicated that FAT1 is a tumor suppressor gene in gliomas, colon cancer and other tumors,and it is important in tumorigenesis and tumor cell proliferation, migration and invasion.However, the high expression of FAT1 in MB children is still controversial. In order to analyze the influence of high expression of FAT1 on childrens MB survival more accurately. This study comprehensively searched literatures on the relationship between the high expression of FAT1 and the prognosis of MB children, and the method of meta-analysis was adopted to evaluate the effect of the high expression of FAT1 on the prognosis of MB children.
# Methods
## Study registration
This meta-analysis protocol is based on the Preferred Reporting Items for Systematic Reviews and meta-analysis Protocols (PRISMA-P) statement guidelines. The PRISMA-P checklist for the protocol is provided in the PRISMAP-checklist.
The protocol of the systematic review was registered on Open Science Framework, and the registration number is DOI 10.17605/OSF.IO/5FN8M.
## Data sources and search strategy
We searched for China National Knowledge Infrastructure (CNKI), Chinese Biomedical literature Database (CBM), Chinese Scientific and Journal Database (VIP), Wan Fang database, PubMed, EMBASE, and all these electronic databases without language restrictions. The PubMed search strategy was illustrated in detail in . Other electronic databases applied similar search strategies.
## Inclusion criteria for study selection
1. According to pathology and histology, patients diagnosed as MB, and under the age of 18 2. FAT1 was expressed in tumor tissues. 3. Reported FAT1 survival-related data, including overall survival (OS) and disease-free survival (DFS) 4. Patients were divided into FAT1 positive and FAT1 negative. 5. Published as full-text articles.
Papers without sufficient data, non-peer-reviewed articles, meta-analyses, literature reviews, case reports, case series, conference summaries, animal studies, letters to the editor, reviews, editorials, and other unrelated studies were excluded from the analysis.
## Data collection and analysis 2.4.1. selection of studies.
The titles and abstracts were independently examined by 2 authors to select studies that meet the criteria. In this process, irrelevant literature, reviews, case reports or series, letters, recommendations or guidelines, and studies published only as summaries of meetings were excluded. Then, the full-text level of remaining articles is reviewed by the 2 same reviewers. Differences on the choice of research would be resolved by discussing with the third author. All excluded studies with detailed reasons would be recorded at different stages. The flowchart of the study selectioncould give a detailed description.
## 2.4.2.
Data extraction and management. The 2 authors collected information independently from each included studies. Any conflict was resolved by consensus with the help of a third experienced author. The extracted information included the title of the manuscript, the name of the first author, journal, year of publication, country, race, age, sex, sample size, FAT1 detection method, OS and DFS hazard ratio (HRs) and 95% confidence interval (CIs). We obtained HRs and 95% CIs from the Kaplan-Meier survival curves by referring to Engauge Digitizer version 4.1 (http://digitizer.sourceforge.net/). We contacted the lead authors to obtain any missing or ambiguous information from included studies.
## Assessment of quality in included studies
Newcastle-Ottawa quality Assessment scale (NOS), a bias risk assessment tool for observational studies recommended by Cochrane Collaboration, was adopted to evaluate the quality of the included studies.Any dispute was settled through discussion. NOS consists of the following 3 quality parameters: selection, comparability and result evaluation. According to these parameters,each study was scored from 0 to 9.
## Measures of prognosis
OS and DFS were taken as prognostic outcomes. The results were expressed as HRs with 95% CIs.
## Management of missing data
If there existed insufficient or missing data in the literature, we would only analyze the currently available data and discuss its potential value. Search strategy (PubMed).
## Number
Search terms , P < .1), a fixed effect model would be used. When there was statistical heterogeneity among included literatures (P < .1, I 2 ≥ 50%), the sources of heterogeneity could be analyzed. Clinical heterogeneity was treated by subgroup analysis. In the absence of significant clinical and methodological heterogeneity, statistical heterogeneity was considered and analyzed by adopting a random effect model. If the clinical heterogeneity of the subgroup analysis wad significantly high, there was only descriptive analysis, rather than meta-analysis.
2.9. Additional analysis 2.9.1. Subgroup analysis. According to the detection methods of FAT1, the source of survival data and the high expression threshold of FAT1, we analyzed the subgroup.
2.9.2. Sensitivity analysis. Sensitivity analysis was conducted to assess the impact of individual studies on the overall merger value. If the number of studies included in a certain outcome index was no less than 10, funnel chart could be used to evaluate publication bias.Besides, Egger and Begg test were adopted for the evaluation of potential publication bias.
## Ethics
Our research data was derived from published literature, because there was no patient recruitment and personal information collection. Therefore, ethical approval was not required.
# Discussion
Although advances in surgery, radiotherapy, and chemotherapy improved overall survival, the lifetime sequelae of these treatments was a major health care burden, thus leading to sustained efforts for effective targeted treatments.With the development of molecular biology and cytogenetics, it is agreed that gene mutation and abnormal intracellular signal transduction are important reasons for the occurrence and development of MB, and they are closely related to its prognosis.Previous studies proved that MB patients have a series of molecular genetic patterns, such as deletion of tumor suppressor gene PTEN, mutation of p53, and abnormality of SHH and WNT signal transduction pathway.MB contains a variety of molecular subtypes. According to the latest molecular typing study, MB can be divided into 4 subtypes, including WNT type, SHH type, Group3 type and Group4 type. Different molecular subtypes lead to different prognosis, so molecular typing is an important basis for disease risk stratification. However, new and more detailed molecular typing is still in-depth study, and the introduction of signal transduction pathway and molecular typing provides a new idea for individualized treatment and accurate prognosis of MB in children. At present, it is obvious that FAT1 is abnormally expressed in many kinds of malignant tumors, such as cholangiocarcinoma, breast cancer, oropharyngeal squamous cell carcinoma, diffuse astrocytoma, glioblastoma, acute lymphoblastic leukemia, and so on.RNAi technique was applied to knock down the expression of FAT1 gene in human MB cell line Daoy.It was clear that the proliferation ability of tumor cells is significantly enhanced, suggesting that FAT1 gene may play a role as a tumor suppressor gene in the pathogenesis of MB children. The change of FAT1 gene expression can affect the biological behavior of MB children, after which affect the prognosis of children. Therefore, we would like to conduct a meta-analysis in strict accordance with the PRASMA statement, so as to provide more evidence to support FAT1 as a prognostic marker of MB children. However, this study has some limitations. Our search does not include studies in languages other than Chinese and English, thus may result in certain selective bias. Most importantly, perhaps, there is some heterogeneity due to the application of different treatments in different studies.
# Author contributions
|
Model-based estimation of measures of association for time-to-event outcomes
Background: Hazard ratios are ubiquitously used in time to event applications to quantify adjusted covariate effects. Although hazard ratios are invaluable for hypothesis testing, other adjusted measures of association, both relative and absolute, should be provided to fully appreciate studies results. The corrected group prognosis method is generally used to estimate the absolute risk reduction and the number needed to be treated for categorical covariates. Methods: The goal of this paper is to present transformation models for time-to-event outcomes to obtain, directly from estimated coefficients, the measures of association widely used in biostatistics together with their confidence interval. Pseudo-values are used for a practical estimation of transformation models. Results: Using the regression model estimated through pseudo-values with suitable link functions, relative risks, risk differences and the number needed to treat, are obtained together with their confidence intervals. One example based on literature data and one original application to the study of prognostic factors in primary retroperitoneal soft tissue sarcomas are presented. A simulation study is used to show some properties of the different estimation methods. Conclusions: Clinically useful measures of treatment or exposure effect are widely available in epidemiology. When time to event outcomes are present, the analysis is performed generally resorting to predicted values from Cox regression model. It is now possible to resort to more general regression models, adopting suitable link functions and pseudo values for estimation, to obtain alternative measures of effect directly from regression coefficients together with their confidence interval. This may be especially useful when, in presence of time dependent covariate effects, it is not straightforward to specify the correct, if any, time dependent functional form. The method can easily be implemented with standard software.
# Background
Measures of disease frequency and measures of associations derived from them are among the basic building blocks of biostatistics and epidemiology. The appropriateness of the use of a specific measure of association may depend on the study objectives and design. Sometimes, however, the use of specific measures of association depends also on the statistical methods available for estimation. For example, in epidemiology, a debated subject concerns the use of odds ratios, estimated through logistic regression, in cohort studies of common outcomes [bib_ref] Model-based estimation of relative risks and other epidemiologic measures in studies of..., Greenland [/bib_ref] [bib_ref] Estimating the relative risk in cohort studies and clinical trials of common..., Mcnutt [/bib_ref]. When time-to-event outcomes are analyzed, the presence of censoring calls for specific methods of analysis [bib_ref] Analysing Survival Data from Clinical Trials and Observational Studies, Marubini [/bib_ref]. The evaluation of the effect of a treatment in a controlled trial can be performed through the graphical display and comparison of Kaplan-Meier curves at selected times, when adjustment is not required. Otherwise, the measure of effect generally considered is the adjusted hazard ratio estimated by means of Cox proportional hazard model, [bib_ref] Regression models and life-tables (with discussion), Cox [/bib_ref].
However, the clinical literature in randomized controlled trials suggests the use of absolute measures of effect to assess the effects of a treatment, such as risk difference or the number needed to be treated, which are better suited than relative measures of effect for clinical decision support, see [bib_ref] An assessment of clinically useful measures of the consequences of treatment, Laupacis [/bib_ref] [bib_ref] Clinically useful measures to assess the effectiveness of treatments: hints for a..., Boracchi [/bib_ref] [bib_ref] Basic statistics for clinicians: 3. Assessing the effects of treatment: measures of..., Jaeschke [/bib_ref] [bib_ref] Clinically useful measures of the effects of treatment, Dicenso [/bib_ref] [bib_ref] Interpreting measures of treatment effect in cancer clinical trials, Case [/bib_ref] among others. Schechtman highlights how relative measures are appropriate for http://www.biomedcentral.com/1471-2288/14/97 summarizing the evidence while absolute measures for the concrete application in a clinical setting,.
The need for alternatives to hazard ratios, a relative measure of effect based on (instantaneous) incidence rates, is increasing in medical/epidemiological literature. In particular, the possibility to provide absolute measures of association computed using adjusted survival curves was explored in literature [bib_ref] A tutorial on methods to estimating clinically and policy-meaningful measures of treatment..., Austin [/bib_ref] [bib_ref] Estimating adjusted risk difference (RD, and number needed to treat (NNT, measures..., Laubender [/bib_ref] [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref].
To precisely define the different measures of association in time-to-event applications, it is useful to distinguish between the risk of the event, F(t), i.e. the probability of a patient having the event over a defined follow-up time, and the event rate, λ(t), i.e. the number of events in a specified follow-up interval divided by the time at risk accumulated during the interval. The instantaneous hazard rate is obtained when the interval length approaches 0. The hazard rate at time t refers to the population survived until time t, while the risk refers to the whole population. The measures considered in this paper refer to ratios and differences between the risk of event of different groups of subjects. Let F 1 (t) and F 2 (t) be the event risk by t in two groups of subjects, (exposed and non exposed, standard treatment and new treatment), then we define the risk difference as RD(t) = F 1 (t) − F 2 (t). It is useful to translate RD(t) expressed as a percentage measure in a measure more sensible form a clinical perspective. To this end it is usual to use the number needed to be treated NNT(t) = 1/RD(t) which is interpreted as the expected number of patients needed to be treated to avoid one additional death compared to the untreated. The measure has its roots in clinical trial literature and was extended in an epidemiological framework as the number needed to be exposed, NNE(t), i.e. the expected number of subjects to be exposed to have one additional event compared to the unexposed. In observational studies, an alternative definition of NNE(t) is the exposure effect among the unexposed, while the exposure impact number, EIN(t), describes the effect of removing the exposure among the exposed [bib_ref] Methods to calculate relative risks, risk differences, and numbers needed to treat..., Bender [/bib_ref] [bib_ref] Estimating adjusted nnt measures in logistic regression analysis, Bender [/bib_ref] [bib_ref] Different measures of treatment effect for different research questions, Austin [/bib_ref]. It is also interesting to define a relative risk, RR(t) = F 1 (t) F 2 (t) , to be contrasted with the hazard ratio, HR(t) = λ 1 (t) λ 2 (t) . The different measures of effect are in general time-varying. In certain situations, however, they are estimated as constant through follow-up, as it happens for example with the Cox proportional hazard model for the hazard ratio. When the measure is assumed to be constant during follow-up the time dependence is omitted (i.e., HR(t) is written as HR).
The purpose of this paper is to provide an outline of the methods generally adopted to estimate adjusted summary measures of associations, different from the hazard ratio, in time-to-event studies and to present a new method based on transformation models. The focus of this paper is not to provide guidance about which association measure should be used in different situations, but simply to provide an estimation method. Moreover, particular attention will be given to the estimate of adjusted RD(t). In fact, absolute measures of association are particularly advocated in survival analysis, to be combined with the generally used hazard ratio. A small simulation study is provided to show a preliminary evaluation of the properties of the different estimation procedures. Two examples are then developed. The first concerns literature data on a clinical trial on 506 prostate cancer patients [bib_ref] Treatment effects in competing-risks analysis of prostate cancer data, Kay [/bib_ref]. The goal is to estimate the treatment RD(t) and NNT(t) with their confidence intervals. A comparison of the model based estimated RD(t) with that obtained with the classical corrected group prognosis method [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref] is provided. The second application concerns an observational study of prognostic factors in primary retroperitoneal soft tissue sarcomas [bib_ref] Histology-specific nomogram for primary retroperitoneal soft tissue sarcoma, Ardoino [/bib_ref].
# Methods
## Computation of association measures
In many situations a researcher is interested in providing adjusted estimates of covariate associations with the outcome. In observational studies (involving no randomization) the exposure effect has to be adjusted for known confounders. Also in randomized controlled trials (RCT) the use of adjusted estimates is suggested for example to account for potential covariate imbalances or since prognostically relevant covariates were considered for a stratified randomization [bib_ref] Should we adjust for covariates in nonlinear regression analyses of randomized trials?, Hauck [/bib_ref] [bib_ref] Covariate imbalance and random allocation in clinical trials, Senn [/bib_ref]. In these cases, Cox regression is widely used to adjust the estimated association between the covariate of interest and outcome for the other covariates.
For the purpose of illustration, let us consider a controlled trial, where the event of interest is death, investigating the efficacy of a new treatment (T = 1) in comparison to a standard treatment (T = 0). Two covariates such as age, A, and gender, G, are considered for adjustment. The multivariable proportional hazard Cox model can be specified as follows:
[formula] λ(t|T, A, G) = λ 0 (t) exp(αT + βA + γ G) [/formula]
where t is the time to event; λ(t|T, A, G) is the hazard function conditional to covariate values; α, β and γ are the regression coefficients; λ 0 (t) is the baseline hazard for a subject in the control group (T = 0), 0 years old and female (G = 0). The adjusted hazard ratio for the treatment, HR constant through follow-uptime, is simply obtained as exp(α). Using such a model, RD(t) or NNT(t) for the treatment can be obtained specifying a covariate pattern and the baseline risk. For example, the estimated NNT(t), conditional on being male 40 years old is:
[formula] 1 S(t|T = 1, A = 40, G = 1) −Ŝ(t|T = 0, A = 40, G = 1) . [/formula]
(1) http://www.biomedcentral.com/1471-2288/14/97 whereŜ(t|T = 1, A = 40, G = 1) is the estimated survival probability for a male 40 years old in the experimental treatment group, given by [Ŝ 0 (t)] exp(α+β 40+γ ) , whilê S(t|T = 0, A = 40, G = 1) is the estimated survival probability for a male 40 years old, in the control group, given by [Ŝ 0 (t)] exp(β 40+γ ) .Ŝ 0 (t) is the baseline survivor function from a Cox proportional hazards model estimated according to one of the available methods [bib_ref] Nonparametric estimation of the survival distribution in censored data, Fleming [/bib_ref]. In order to obtain adjusted measures of association, different from the hazard ratio, the Cox proportional hazard model is used to estimate adjusted survival curves [bib_ref] Comparison of 2 methods for calculating adjusted survival curves from proportional hazards..., Ghali [/bib_ref] as outlined in the following paragraphs.
# Average covariate method
The simplest approach for obtaining adjusted survival curves is the average covariate method. The mean values among the study patients of the covariates used for adjustment are plugged into the multivariable Cox proportional hazard model. Considering the above example, if the mean age of subjects in study is 45 and 30% males are included, the adjustedNNT(t) for the treatment will be
[formula] 1 S(t|T=1,A=45,G=0.3)−Ŝ(t|T=0,A=45,G=0.3) [/formula]
. The average covariate method was once popular and largely adopted, due to its simplicity, but it was severely criticized [bib_ref] Comparison of 2 methods for calculating adjusted survival curves from proportional hazards..., Ghali [/bib_ref] [bib_ref] Adjusting survival curves for confounders: a review and a new method, Nieto [/bib_ref].
In fact it involves the averaging of categorical covariates, such as gender, which is difficult to understand. Moreover the method provides an estimate of the measure of effect for an hypothetical average individual and not a population averaged estimate.
## Corrected group prognosis method and developments
An alternative idea is the corrected group prognosis method (CGPM), [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref] [bib_ref] Corrected group prognostic curves and summary statistics, Chang [/bib_ref] [bib_ref] Adjusted survival curve estimation using covariates, Makuch [/bib_ref]. In the following, the CGPM to estimate RD(t), as described by Austin [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref] , is outlined:
- A Multivariable Cox (or fully parametric) regression is used for the treatment and the covariates. - For each subject, the predicted survival probabilities, at the times of interest, are estimated using the multivariable model, assuming each subject is in the experimental treatment group; then the predictions are averaged; - the same predictions are obtained and averaged assuming each subject is in the control group. - the difference between the averaged predicted probabilities between experimental and control group is an estimate of the adjusted RD(t) for the experimental treatment at the specified times.
Pointwise confidence intervals of the obtained RD(t) estimates may be computed via bootstrap resampling [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref]. For each bootstrap sample, i.e. a sample of the same size of the original one and randomly drawn with replacement from it, the RD(t) is computed according to the procedure outlined. A non parametric bootstrap 95% pointwise confidence interval is obtained resorting to the 2.5 th and 97.5 th percentiles of the obtained RD(t) bootstrap distribution.
A simulated example of the estimation of RD(t) in presence of confounding is exemplified in [fig_ref] Figure 1: Simulated data [/fig_ref].
The CGPM can be applied in principle to whatever regression model and an adequate model must be chosen. Considering, for example, the Cox regression model, in presence of time dependent covariate effects, an interaction of the covariates with a pre-specified function of time should be specified, in order to estimate HR(t) varying during follow-up time. It is important to remark that it is not always easy to specify an adequate model in presence of time dependent covariate effects. In fact it is not always obvious how to model the time dependence itself. In general simple functions of time (linear or logarithm) or more flexible alternatives are used, [bib_ref] Assessing time-by-covariate interactions in proportional hazards regression models using cubic spline functions, Hess [/bib_ref].
To allow the estimation the data set must be augmented as it is done for true time-dependent covariates. It is to be remarked that, although the use of predicted values from regression models is simple from a practical point of view, the standard way to obtain summary measures of effect and their confidence interval is to use directly regression model coefficient estimates. The CGPM applied to the Cox model will be used for comparison with the method here proposed and described in the following section.
Laubender and Bender, [bib_ref] Estimating adjusted risk difference (RD, and number needed to treat (NNT, measures..., Laubender [/bib_ref] , proposed different averaging techniques to estimate relevant impact numbers in observational studies using Cox model. For the purpose of illustration, let us consider the same example as before simply considering an exposure (E) instead of treatment. To obtain an estimate of NNE(t) it is possible to average predictions considering the subjects as if they were unexposed and as if they were exposed and taking the difference. As the distributions of the covariates used for adjusting are in general different in the exposed and unexposed groups, two different measures should be considered. Specifically, the estimate of the NNE(t) is obtained considering the unexposed subjects only, while EIN(t) is obtained considering the exposed subjects only. A comparison of the model based estimated RD(t) with that obtained through different averaging techniques, namely NNE(t) and EIN(t) [bib_ref] Estimating adjusted risk difference (RD, and number needed to treat (NNT, measures..., Laubender [/bib_ref] [bib_ref] Methods to calculate relative risks, risk differences, and numbers needed to treat..., Bender [/bib_ref] , is provided in the second example. However, the focus of the paper is not the comparison of different averaging techniques which are provided only for illustrative purposes. In particular, only the estimates obtained through the averaging performed over the whole population are compared with those based on transformation models methods. http://www.biomedcentral.com/1471-2288/14/97
## Model-based estimates of association measures
Adjusted model-based estimates of measures of association can be obtained resorting to a general class of regression models used in Survival Analysis called transformation models.
## Pseudo values
Considering the previous example, the transformation model can be written as g(S(t|T, A, G)) = g(S 0 (t)) + αT + βA + γ G.
A possibility to estimate transformation models, using standard available software, is through pseudo-values [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref]. The pseudo value is defined for each subject i at any time t and is given bŷ
[formula] θ i (t) = nŜ(t) − (n − 1)Ŝ −i (t)(2) [/formula]
where n is the sample size,Ŝ(t) is the survival probability based on the Kaplan-Meier estimator using the whole sample andŜ −i (t) is the survival probability obtained by deleting the i subject from the sample. When no censoring is present in the data, the pseudo values for subject i at time t is simply 1 if the subject is alive at t, while it is 0 if the event happened by t. Suppose to have an exposed male, 40 years old, which dies after 30 months of followup. The pseudo values computed at 12, 24 and 36 months are equal to 0, 0 and 1 respectively. The times at which the pseudo-values are computed are called pseudo-times. When censoring is present in the data, pseudo-values are still defined for each subject (even those censored) and for each time, but the values may also be less than 0 or greater than 1 (See [bib_ref] Checking hazard regression models using pseudo-observations, Perme [/bib_ref] ; page 5310-11 for further details on the properties of pseudo-values).
In general, to allow inference on the entire survival curve, M (greater than 5) pseudo times are used, considering, for example, the quantiles of the unique failure time distribution. As M pseudo values are computed for each subject, an augmented data set is created with M observations for each subject.
## Transformation models and association measures
The pseudo-values are then used as responses in a regression model for longitudinal data, where time is a covariate. As no explicit likelihood is available for pseudo-values, generalized estimating equations (GEE), [bib_ref] Longitudinal data analysis using linear models, Liang [/bib_ref] , are used accounting for the correlation of the pseudo-values within each subject. The cluster robust variance-covariance is used for hypothesis testing using Wald tests. In general an independence working variance-covariance matrix can conveniently be used in the estimation process [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref].
In order to model g(S 0 (t)), the transformed baseline survival function, the standard procedure is to insert in the regression model indicator functions for each pseudotime. If all event times would be used to compute the pseudo-values, the insertion of indicator functions would result in a non parametric representation of the (transformed) baseline survival, as in the Cox model. In general only a small number of pseudo-times are used obtaining a parametric baseline representation. As an alternative, http://www.biomedcentral.com/1471-2288/14/97 spline functions can be inserted in the regression model, as did [bib_ref] Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application..., Royston [/bib_ref] in a non-pseudo-values framework.
Considering for simplicity only two spline bases, the regression model of the example can be written as follows:
[formula] g(θ i (t)) = φ 0 + φ 1 B 1 (t) + φ 2 B 2 (t) + αT i + βA i + γ G i (3) [/formula]
where B 1 (t) and B 2 (t) represent the first and second spline bases for time t. For example, if a restricted cubic spline basis is used with three knots at k 1 ,
[formula] k 2 , k 3 , then B 1 (t) = t and B 2 (t) = (t − k 1 ) 3 + − (t−k 2 ) 3 + (t 3 −t 1 ) (t 3 −t 2 ) + (t−k 3 ) 3 + (t 2 −t 1 ) (t 3 −t 2 ) , where, for example, (t − k 1 ) 3 + is equal to (t − k 1 ) 3 if t > k 1 , otherwise is 0. [/formula]
Knots are chosen at quantiles of the failure time distribution. In the case of 3 knots the quantiles commonly suggested are 0.1, 0.5 and 0.9,. To choose the complexity of the spline the QIC, [bib_ref] Akaike's information criterion in generalized estimating equations, Pan [/bib_ref] , an information criterion proposed for generalized estimating equations, can be used. A less formal strategy is the graphical comparison between the Kaplan-Meier marginal survival probability and the marginal probability obtained from the transformation model without covariates. Such a procedure will be used in the examples.
The first part of the model,
[formula] φ 0 + φ 1 B 1 (t) + φ 2 B 2 (t) [/formula]
, provides a parametric representation of the (transformed) baseline survival function, g(S 0 (t)), during follow-up time.
The coefficients α, β and γ represent the covariate effects expressed as differences in the Survival probability, transformed by g associated with a unit increase in the covariates. Let us consider such an issue in detail. When g is the logit link function, a proportional odds model is estimated. Accordingly, α, β and γ represent the logarithm of the ratio of the odds of surviving associated with the change of one unit in the covariates. Such an effect is constant through follow-up times. The exponentiation of the parameter estimates represent therefore the ratio of the odds of surviving. Similarly, the logarithmic link produces a proportional risks model and the exp(α), exp(β) and exp(γ ) represent the ratio of the survival probabilities (Relative Risks, RR). The identity link produces a constant survival difference model: α, β and γ represent the adjusted differences in survival probabilities (risk differences, RD). A constant difference model through follow-up is often not practical as a model such that at the beginning of the follow-up the survival curves start at 1 and then, eventually, become different. However, it is to be noted that the first pseudo-time is never placed at time 0, but later on the follow-up time scale. In [fig_ref] Figure 1: Simulated data [/fig_ref] an example of the model based RD estimate with pointwise confidence intervals, constant through time, is reported in the right panel. The constant model estimated RD can be used to obtain a constant estimate of NNT by inversion. In the case of treatment T:NNT =[α] −1 . The value of 1 indicates the largest possible effect of NNT, while in correspondence of no covariate effect (RD=0) the NNT value is ±∞. The largest possible harmful effect is −1. Positive and negative values of NNT represent the expected number of patients needed to be treated for one additional patient to benefit and to be harmed, respectively.
In the case of the log-log link, g = log(−log(- )), exp(α), exp(β) and exp(γ ) are the ratio between cumulative hazard functions associated with the change of one unit in the covariates. This ratio is equal to that of hazard functions, only in the proportional hazard case.
The method allows to estimate the measures of effect also for continuous covariates. For example, the evaluation of a biomarker effect measured on a continuous scale, without cutoffs, is still possible with this methodology.
The use of different link functions to obtain a particular measure of effect, is an established technique in binomial regression, where the use of non-canonical links, such as the logarithm, allows to obtain adjusted measures of impact different from the odds ratio, [bib_ref] Model-based estimation of relative risks and other epidemiologic measures in studies of..., Greenland [/bib_ref]. Wacholder, [bib_ref] Binomial regression in GLIM: estimating risk ratios and risk differences, Wacholder [/bib_ref] , is an excellent reference for deepen such aspects in the framework of logistic regression.
When there is evidence for time dependent effects of the covariates, the interaction between the covariates and the spline bases B 1 (t) and B 2 (t) are included in model (3).
[formula] g(θ i (t)) = φ 0 + φ 1 B 1 (t) + φ 2 B 2 (t) + αT i + βA i + γ G i + γ 1 B 1 (t)T i + γ 2 B 2 (t)T i + γ 3 B 1 (t)A i + γ 4 B 2 (t)A i + γ 5 B 1 (t)G i + γ 6 B 2 (t)G i [/formula]
In such a case, the estimated g-transformed survival probability differences change during follow-up time. In order to show the effect, varying in time, of a dichotomous covariate, for example treatment T, it is useful to adopt a graphical display, where the time is put on the horizontal axis while the function exp(α + γ 1 B 1 (t) + γ 2 B 2 (t)) is on the vertical axis (exponentiation is not used with the identity link;
[formula] RD(t) = α + γ 1 B 1 (t) + γ 2 B 2 (t)) [/formula]
. In this case the estimated NNT(t) is naturally varying through follow-up time and again obtain by inversion:
[formula] RD(t) −1 . [/formula]
For a continuous covariate, such as Age, A in the example, it is possible to use a surface plot, where Age and time are on the x and y axis, while the z axis reports the covariate effect with respect to a reference value. It would also be possible to model Age effect with spline bases. In this case, the interaction between Age and time is obtained through tensor product spline bases of Age and time.
When a large number of pseudo-times is used, spline functions allow to model parsimoniously the baseline risk compared to indicator functions. This is particularly important for the modelling of time-dependent effects in connection to the different link functions. In principle when a covariate effect is constant using a specific link, it should be time-varying with the other links. No http://www.biomedcentral.com/1471-2288/14/97 statistical evidence against a constant covariate effect for more than one link may only be due to lack of power. The problem can also be exacerbated by some multiple testing issue. Time-dependent effects selection depends therefore on the link transformation used. As a consequence, the adjusted effect of a covariate may be constant using a link function, but time-dependent using a different link.
Moreover, the fitted values of the different models selected for the different link functions are generally different, being equal only if the models are saturated. Traditionally, the strategy used in the application of transformation models such as (3) was to select the best fitting g transform, i.e., the transform where covariate effects are constant through time, see [bib_ref] Analysis of survival data by the proportional odds model, Bennett [/bib_ref] [bib_ref] Testing goodness of fit for proportional hazards model with censored observations, Wei [/bib_ref] as examples. The approach considered here is different. The interest is in using the g-transform which is the most informative for the clinical or biological counterpart. Generally the best fitting link function and the one selected by the researcher are not the same. Time dependent effects should therefore be expected in the model.
## Pointwise confidence intervals
Approximate pointwise 95% confidence intervals are calculated from model results as in standard GLM/GEE modelling. The computation is easy when covariate effects are constant on the g-transformed scale. The clusterrobust variance-covariance matrix must be used. Using model (3) as example, the 95% CI for treatment, on the transformed g scale, will be l lower , l upper =α ± 1.96 × st.error where st.error(α) is the estimated cluster-robust standard error for the model parameter α. When g is the log, logit or log − log link function, the 95% CI for the treatment effect With time-varying covariate effects, the variance of the sum of a linear combination of different parameter estimates must be computed for the each of the follow-up times. For example, for treatment T, the variance of interest, at a specific time t, is that of the linear combination α+γ 1 B 1 (t)+γ 2 B 2 (t). Written in matrix terms the variance at time t is given by:
[formula] 1 B 1 (t) B 2 (t) V (α) Cov(α, γ 1 ) Cov(α, γ 2 ) Cov(α, γ 1 ) V (γ 1 ) Cov(γ 1 , γ 2 ) Cov(α, γ 2 ) Cov(γ 1 , γ 2 ) V (γ 2 ) 1 B 1 (t) B 2 (t) [/formula]
where V (- ) stands for the cluster-robust variance, while Cov(- , - ) stands for the cluster-robust covariance of two random variables. When the variances at the different times are calculated, the pointwise 95% CI can be computed as before.
## Software implementation
The approach to censored data regression based on pseudo values was applied to regression models for the cumulative incidence functions in competing risks and for multi-state modeling [bib_ref] Generalised linear models for correlated pseudo-observations, with applications to multistate models, Andersen [/bib_ref] , for the restricted mean [bib_ref] Regression analysis of restricted mean survival time based on pseudo-observations, Andersen [/bib_ref] and for the survival function at a fixed point in time [bib_ref] Analyzing survival curves at a fixed point in time, Klein [/bib_ref]. Implementation details and software can be found in Klein et al., [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref] , and Andersen and Perme, [bib_ref] Pseudo-observations in survival analysis, Andersen [/bib_ref].
Software is available to compute pseudo values (macro %pseudosurv in SAS and function pseudosurv in R package pseudo [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref] Standard GEE tools available in SAS or R can be used for regression. In SAS the proc genmod allows to change link functions using the instructions FWDLINK and INVLINK. In R, the package geepack can be conveniently used, see [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref] for details.
As an example of the R software implementation, the identity link is used: geese( pseudo~rcs(tpseudo,3) + ...., family = "gaussian", mean.link = "identity", id = id, corstr="independence", data = b, scale.fix=TRUE, scale.value=1 )
where the variable pseudo contains the pseudo values and the variable tpseudo the pseudo-times according to the software reported in [bib_ref] SAS and R functions to compute pseudo-values for censored data regression, Klein [/bib_ref]. The R function rcs of the package rms,, is used to compute restricted cubic spline bases. Each subject is represented by multiple rows in the data, one for each pseudo time. The records for each subject are identified by means of the variable id which is used to estimate the robust standard error by the geese function. Using the identity link function, the estimated coefficients can be interpreted as the adjusted RD(t) estimates.
# Results
The estimation of RD(t) through pseudo-values is evaluated through a simple simulation study. Moreover, two real examples are presented. The first example concerns a prostate cancer trial, well known in competing risks literature, to show a situation where proportional hazard fail to model the treatment effect during the whole followup. The second example regards the analysis of prognostic http://www.biomedcentral.com/1471-2288/14/97 factors in an observational study of primary retroperitoneal soft tissue sarcoma patients. R software, [bib_ref] Vienna: R Foundation for Statistical Computing, R Core Team [/bib_ref] , was used for the simulation and both the examples presented.
## Simulation
Data are generated to simulate Cox regression model, according to [bib_ref] Estimating adjusted risk difference (RD, and number needed to treat (NNT, measures..., Laubender [/bib_ref]. Suppose 100 persons are exposed (Z = 1) and 100 unexposed (Z = 0). A confounder X is generated normally distributed with mean 40 for Z = 0 and 45 for Z = 1 (standard deviations is 8 for both). Event times are generated according to an exponential proportional hazard model λ(t|X, Z) = λe log(1.01)X+log(1.80) * Z . Censoring times are obtained in the same way adjusting the baseline hazard to have about 10% censoring. In such a situation the true exposure RD(t) may be calculated integrating out the covariate X:
[formula] ∞ −∞ (S(t|Z = 1, X) f (X|Z = 1) − S(t|Z = 0, X)f (X|Z = 0))dx(4) [/formula]
Pointwise confidence intervals for the CGPM are calculated using percentile bootstrap (200 bootstrap samples for each simulated data set). RD(t) is estimated by transformation models using pseudo values and the identity link. The baseline cumulative risk is modelled using a restricted cubic spline with 5 knots at 5%, 25%, 50%, 75% and 95% quantiles of the failure time distribution. In the pseudo value model the Z covariate is inserted either without and with an interaction with the baseline risk (i.e. estimating a constant RD, and a time-dependent RD(t) through follow-up time). The estimated RD and RD(t) at times 200, 400 and 600 are collected and compared with the true values. Such times are chosen to be sure that they are between the first and the last pseudo-times in all simulations. The results are analyzed in terms of bias, root mean squared error, average length and coverage of the 95% confidence intervals. The results are reported in [fig_ref] Table 1: Event times generated according to a Cox-exponential model with a confounder X... [/fig_ref].
RD(t) estimated through the CGPM using the Cox proportional hazard regression model (the model used to generate the simulation data) is used as the benchmark estimation method.
The method based on pseudo-values with Z time dependent appears effective especially in terms of bias. Confidence interval coverage is good, although the width of the confidence intervals with pseudo-values is fairly large. It is interesting to observe the results of pseudo-values with the covariate Z not time dependent. In this case the estimated risk difference is constant through time, namely RD, a situation which can result from lack of power to detect the time-dependence of the RD(t). The simulation results appear very interesting for late followup times. At time 600, results are very similar to that of the CGPM was adopted using the Cox proportional hazard regression model. PV model was adopted with identity link. Pseudo times were fixed at quantiles 5%, 25%, 50%, 75%, 95% of the failure time distribution. The Z covariate was inserted in the PV model with an interaction with the baseline risk function modelled with a restricted cubic spline with 3 knots (Z td) giving an estimated risk difference varying through follow-up time, RD(t), and without time-dependence to estimate a constant RD through follow-up time between the first and the last pseudo-times (Z const).
Cox proportional model. However, the PV method with identity link and without time dependence estimating a constant RD leads to a strong undercoverage demonstrating that the estimation of a constant RD may be misleading. The same simulation is performed generating times from a Cox model with a time-dependent effect for Z according to the following formula: λ(t|X, Z) = λe log(1.01)X−log(0.95) * Z * log(t) . In this second simulation the Cox model used with CGPM iss specified in two different ways. Specifically, the covariate Z is inserted with an interaction with log(t), the correct one, and with a restricted cubic spline of time with 3 knots. Namely the use of cubic splines for modelling time dependent effects was proposed by Hess, [bib_ref] Assessing time-by-covariate interactions in proportional hazards regression models using cubic spline functions, Hess [/bib_ref] , allowing the study of possible covariate-time interactions without having to specify a specific functional form, using a limited number of parameters. The results are reported in [fig_ref] Table 2: Event times generated according to a Cox-exponential model with a confounder X... [/fig_ref].
In this simulation the Cox model with the correct specification of the time dependent effect of Z, that is log(t), is used as benchmark estimation. When the time dependence of Z is modelled using the restricted cubic spline, the performance of the CGPM is less appealing, compared to the benchmark, regarding all the parameters considered into the simulation. The pseudo-value model is really a competitor in this situation. It is in particular interesting to observe the 95% confidence interval width. The transformation model using pseudo-values with identity link is a valuable alternative to the CGPM when http://www.biomedcentral.com/1471-2288/14/97 The effect of Z is multiplied by log(t). CGPM was adopted using the Cox proportional hazard regression model with the time dependent effect of Z specified in two ways: [bib_ref] Model-based estimation of relative risks and other epidemiologic measures in studies of..., Greenland [/bib_ref] with an interaction between Z and log(t), corresponding to the generating model; [bib_ref] Estimating the relative risk in cohort studies and clinical trials of common..., Mcnutt [/bib_ref] with and interaction between Z and a restricted cubic spline of time (rcs(t)). PV model was adopted with identity link. Pseudo times were fixed at quantiles 5%, 25%, 50%, 75%, 95% of the failure time distribution. The Z covariate was inserted with an interaction with the baseline risk function modelled with a restricted cubic spline with 3 knots, giving an estimated risk difference varying through follow-up time, RD(t).
the time dependent effect in the Cox model is unknown and modelled using a flexible method. Model checking is therefore very important and pseudo-values can be of help also in this case, see the work of Anderson and Perme, [bib_ref] Pseudo-observations in survival analysis, Andersen [/bib_ref].
## Prostate cancer
Literature data on 502 prostate cancer patients, publicly available at the web site http://biostat.mc.vanderbilt. edu/wiki/Main/DataSets, (Byar & Greene prostate cancer data), treated with different doses of diethylstilbestrol in a randomized clinical trial, [bib_ref] Treatment effects in competing-risks analysis of prostate cancer data, Kay [/bib_ref] , were used to estimate the adjusted treatment effect (high versus low dose) on overall mortality. Seven covariates were used for adjustment, namely: age (0, < 75 years; 1, 75 − 80 years; 2, ≥ 80 years), weight index (0, ≥ 100;1, 80 − 99; 2, < 80), performance rating (0, normal; 1, limitation of activity), history of cardiovascular disease (0, no; 1, yes), serum haemoglobin (0, ≥ 12 g/100 ml; 1, 9-12 g/100 ml; 2, < 9 g/100 ml), size of primary lesion (0, < 30 cm 2 ; 1, ≥ 30 cm 2 ), and Gleason stage category (0, ≤ 10; 1, > 10). 483 patients with complete information on the seven covariates available were considered. 344 patients died: 149 for cancer; 139 for cardiovascular causes; 56 for other causes. The estimated RD(t) according to the CGPM using a Cox proportional hazard model [bib_ref] Absolute risk reductions and numbers needed to treat can be obtained from..., Austin [/bib_ref] is reported in [fig_ref] Figure 2: Prostate Cancer Trial [/fig_ref]. The estimated RD(t) is increasing through follow-up time, from 1% to 5%. Correspondingly the NNT(t) is about 100 at the beginning of the followup time and about 17 after 6 years follow-up. The 95% confidence interval is obtained using 2.5 and 97.5 percentiles of the bootstrap distribution of the estimated RD(t) (1000 bootstrap resamples). The RD(t) confidence interval includes 0.
The estimated RD(t) is based on a proportional hazard model. In fact there is no evidence for time-dependent treatment effect in Cox model according to Schoenfeld residuals. However, Kay, [bib_ref] Treatment effects in competing-risks analysis of prostate cancer data, Kay [/bib_ref] , carefully investigated the fit of the Cox model, dividing the time axis into three time interval: [ 0−13]; ; . The log HR for treatment has a positive sign in the first period (0.09), then becomes negative (−0.40 and −0.31). Comparing by likelihood ratio the model fitted on three intervals with an overall survival model, evidence is found against the proportional HR assumption. In fact, cardiovascular deaths are more frequent than cancer deaths earlier during follow-up while, later on, cancer deaths are prevalent. Accordingly, the beneficial effect of treatment appears evident only after the first year of follow-up. The estimates obtained with the CGPM appear therefore distorted due to the use of a proportional hazard model.
As a model with jumps in covariate effects at specific times is not biologically plausible, a Cox regression model with an interaction between the treatment and a function of time is used (specifically a B-spline with 1 interior knot at the median of the failure time distribution). According to the work of Hess, [bib_ref] Assessing time-by-covariate interactions in proportional hazards regression models using cubic spline functions, Hess [/bib_ref] , splines are used to model flexibly the possible treatment time dependent effect. The estimated RD(t) according to this flexible model is reported in [fig_ref] Figure 2: Prostate Cancer Trial [/fig_ref]. The harmful initial treatment effect is not yet captured.
A transformation model based on pseudo-values with identity link is used to estimate a possibly time-dependent treatment RD(t). In order to have about 10 events between each pseudo consecutive times, 32 pseudo-times are considered at quantiles of the unique failure time distribution. In this case B-splines are used to model the baseline cumulative risk. One knot, placed at the median of the failure time distribution, seems to be sufficient to model the marginal survival function, see [fig_ref] Figure 3: Marginal Survival function [/fig_ref].
A backward selection procedure is used to select the time dependent effects for each covariate. The complete model has a total of 45 degrees of freedom, including the 8 covariates and their time dependent effects. The selected model exhibit a time dependent effect of history of cardiovascular disease, size of primary lesion and Gleason stage. There is no evidence for a time dependent treatment effect. The constant estimated RD is 2.2% with 95% confidence interval [−3.5%; 7.9%]. The corresponding constant estimate of NNT is about 45 patients to be treated for one patient to benefit with 95% confidence interval [−∞ to −28; 13 to ∞]. http://www.biomedcentral.com/1471-2288/14/97 It is interesting, however, to compare the RD(t) estimated by the CGPM, which is by definition time dependent, with the one estimated through the transformation model letting the treatment effect varying in time. The results are reported in [fig_ref] Figure 2: Prostate Cancer Trial [/fig_ref] and summarized in [fig_ref] Table 3: RD [/fig_ref]. The treatment effect is harmful during the first year, while the benefits appear afterwards in agreement with the analysis of Kay [bib_ref] Treatment effects in competing-risks analysis of prostate cancer data, Kay [/bib_ref] further refined using competing risks.
In the right panel of [fig_ref] Figure 2: Prostate Cancer Trial [/fig_ref] are also reported the averaged survival probabilities obtained by the CGPM using the Cox model and the transformation model both with the time-dependent effect of treatment. The plot of the averaged survival probabilities is an important completion of the RD(t) plot. In fact RD(t), as usual for the effect measures, is reducing two numbers to a single number.
The prostate cancer data outline a scenario in which the proportional hazard assumption for the treatment effect is not tenable during all follow-up times. Data on prostate cancer should be actually analyzed accounting for competing risks. When a non competing risks survival analysis is performed, CGPM applied to the Cox model without a time dependent treatment effect gives an estimate of the RD(t) increasing during follow-up until reaching a plateau. At the same time, the constant estimate RD obtained using PV provides a distorted estimate constant through follow-up. CGPM applied to the Cox model with a time dependent treatment effect provides an RD(t) estimate not yet capturing the initial harmful treatment effect. The time-dependent estimate obtained from the pseudo-value model is instead effective in describing the treatment effect during follow-up time: harmful at the beginning, when cardiovascular deaths are more frequent, beneficial later on when cancer deaths are more frequent
## Primary retroperitoneal soft tissue sarcoma
Retroperitoneal soft tissue sarcoma (STS) is an uncommon disease. Histologic grade and completeness of macroscopical resection are considered to be the major prognosticators for survival, while histologic type, size and age are more debated. One hundred and ninety two patients with retroperitoneal STS admitted to National Cancer Institute in Milan, consecutively, between 1985 and 2007 are considered. Patients are without evidence of locoregional recurrence or distant metastasis. The median follow-up is 54.8 moths (IQR: 25-104). Among the 192 patients included in the case series, 78 died during follow-up. For a complete description of patients characteristics see [bib_ref] Histology-specific nomogram for primary retroperitoneal soft tissue sarcoma, Ardoino [/bib_ref]. Kaplan-Meier Survival estimates for resection margin and grading are reported in [fig_ref] Figure 4: Cox prop [/fig_ref]. The covariate resection margins is strongly unbalanced in this case series. No evidence for time dependent covariate effects is found using Schoenfeld residuals. A restricted cubic spline with 3 knots, for a total of 2 bases, is used to model the marginal transformed baseline cumulative survival. The fit and the monotonicity of the estimated baseline cumulative risk is controlled by visual inspection [data not shown]. A backward procedure is used for timedependent covariate selection, fixing the significance at a conservative 0.01 level. Using the cloglog link function, no evidence for time dependent effects is found. Considering the log link, the effect of Grading appears time-varying, while using the identity link, Grading and resection margins are modelled varying in time. For the sake of completeness, the estimates of the association between resection margins and survival using the pseudo value model and a cloglog link are compared with the results from the Cox model. The cumulative constant HR is 2.4 (95% CI: 1.38 − 4.31) with pseudo-values. Such an association appears conservative compared to that estimated through Cox regression: HR = 4.19 (95% CI: 2.04 − 8.59). In general it is possible to note an attenuation of estimated effects using PV and the cloglog scale, compared to the Cox model: in the case of resection margins the attenuation is quite important. Clearly, the pseudovalue model with cloglog link is of no interest in this setting, while it has an interest for model checking in Cox regression, [bib_ref] Pseudo-observations in survival analysis, Andersen [/bib_ref].
Also the constant RR estimated using pseudo values and log link appears conservative: RR = 1.3 (95% CI: 1.03 − 1.71). A comparison of the estimated RD(t) obtained with pseudo values and identity link with the estimates obtained using the CGPM is reported in . Different averaging procedures are also reported. The estimated RD(t) and point-wise confidence intervals using the different methods are in close agreement. The different averaging procedures yield quite similar results in this case. According to pseudo-values, the RD(t) reaches a maximum value at 60 months and is about 40%. According to Cox regression the HR of grading 2 vs 1 is 2.56 (1.27 − 5.16). The estimate obtained through pseudo-values and cloglog link is 2.34 . There is still an attenuation but not as huge as in the case of margins. However, considering grading 3 vs 1 the attenuation effect is more evident: the HR from Cox regression is 6.49 (3.31-12.75), while using pseudo-values is 4.88 (2.54-9.39). The RR(t) and RD(t) are reported in . After 8 years the RR(t) is 1.6 and 3.1, and the RD(t) is 36% and 59%, for 2 vs 1 and 3 vs 1 contrasts, respectively. Also in the case of grading, the estimates obtained using pseudo-values and those obtained using different averaging procedures from the Cox regression model appear in good agreement. The quantification of effects through risk differences may be very important for the clinical management of the disease.
# Discussion
In survival analysis the adjusted measure of association everywhere adopted is the hazard ratio. Although the efficiency of the hazard ratio makes it attractive for hypothesis testing, it may not carry the most useful information for clinicians/biostatisticians. Schechtman,, suggests using also absolute measures in conjunction with relative measures of covariate-outcome association. To provide adjusted measures of association different from the hazard ratio, a simple strategy is going through the calculation of the predicted probabilities of event for an "average" subject, the so called "average covariate method". Such a procedure estimates the measures of effect for an hypothetical average subject and not population averaged estimates. An alternative idea to provide adjusted summary measures of effect is the corrected group prognosis method [bib_ref] Corrected group prognostic curves and summary statistics, Chang [/bib_ref] [bib_ref] Adjusted survival curve estimation using covariates, Makuch [/bib_ref]. Extending this idea, using the concept of counterfactuals, Laubender and Bender [bib_ref] Estimating adjusted risk difference (RD, and number needed to treat (NNT, measures..., Laubender [/bib_ref] proposed methods for computing such population measures taking into account the confounder distribution. Bootstrap is then used to obtain pointwise confidence intervals. Such approaches are particularly appealing as they may adopt the Cox regression model which is widely accepted in medical literature. However the Cox model may easily not be the best regression procedure to be applied, simply because of the assumption of proportional hazards. In fact, in presence of time dependent effects Cox regression may be less appealing. This is demonstrated here through a simple simulation. When time-dependent effects in the Cox model are specified using well-known flexible methods, [bib_ref] Assessing time-by-covariate interactions in proportional hazards regression models using cubic spline functions, Hess [/bib_ref] , without committing to a specific functional form, the estimates are not optimal, especially in terms of efficiency. In these circumstances the results of the simple simulation presented here, suggest that the use of the pseudo-value model may represent a valid alternative to the CGPM. The simulation is not exhaustive and more work is needed to fully understand the properties and the relationships among the different estimation methods.
Considering a similar problem in the context of logistic regression, Gehrmann and colleagues, [bib_ref] Logistic regression was preferred to estimate risk differences and numbers needed to..., Gehrmann [/bib_ref] , concluded that the CGPM applied to logistic regression is the preferred method to estimate RD and NNT adjusted for covariates compared to binomial, Poisson and linear regression methods that directly estimates the RD (similar http://www.biomedcentral.com/1471-2288/14/97 to pseudo-values with identity link) even if the fitted response function differs from the true response function. The context of time-to-event outcomes is more complex than that of logistic regression especially for the problem of time dependent effects. Whether similar results hold for the Cox model has therefore to be further explored thorough a series of simulation studies. In any case, when using the CGPM, the basic model used to obtain event probabilities during follow-up has to be adequate. This means, for example, that the proportional hazard Cox regression should not be applied if the proportional hazard assumption is not satisfied.
In clinical literature, results of statistical analysis are commonly reported in terms of regression coefficients and their confidence intervals. Applied survival analysis resorts entirely to the Cox model which is a particular case of transformation models. Transformation models include also the accelerated failure time models providing a variety of measures of effects to be considered.
It is to be noted that, additive and multiplicativeadditive hazard regression models,are not comprised in this class. The estimated coefficients are differences of hazard rates rather than ratios. These models were mainly proposed to improve the fitting where the proportional hazards model is not adequate or to check for proportional hazard assumptions. Moreover, the measures of impact provided are still based on hazards.
In this work a simple approach to obtain point and interval estimates of association measures, by using transformation models through suitable link functions, is presented. The general technique of estimation based on pseudo-values proposed by Andersen and colleagues [bib_ref] Generalised linear models for correlated pseudo-observations, with applications to multistate models, Andersen [/bib_ref] is used as it is simply implementable with standard software.
Other techniques could have been considered to estimate the transformation models. In this context, it is of particular interest the estimate of the baseline survival function to model time dependent effects. Therefore, semi-parametric techniques (see for example [bib_ref] Analysis of transformation models with censored data, Cheng [/bib_ref] are not of interest here. Maximum likelihood (ML) estimation can however be conveniently used. Royston and Parmar proposed ML for transformation models with cloglog and logit links, [bib_ref] Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application..., Royston [/bib_ref] , which are, however, the links of less interest here. Moreover as the pseudo-value is defined between the first and the last pseudo times (which are not the first and last event times) it still makes sense to have a constant RD model. In fact considering the whole follow-up time interval, and especially the beginning of the follow-up, RD(t) estimates should instead always be time-dependent.
From the methodological viewpoint, the only difference introduced in the presented examples with respect to standard applications of the pseudo-value model is the use of spline functions to estimate the transformed baseline survival. This modification is without practical efforts, considered the wide availability of software to compute spline bases. From the modeling point of view, care must be paid to the monotonicity of the estimated transformed baseline survival function. In general, provided the number of knots is limited, no problems of non monotonicity were observed.
Another issue concerns the possible simultaneous use of different association measures estimated through different link functions in a transformation model. In such a case, practitioners must be aware that, due to lack of power, not all time dependent effects can be correctly specified, and likely the different models cannot hold simultaneously. In such a case, if different measures are of interest, different models can be used simultaneously only if the results are in agreement with each other.
From a theoretical point of view, cloglog and logit links guarantee that the estimated probabilities are within the range 0-1 but this is not guaranteed if log or identity links are used with standard software. Research is in progress to face this issue.
# Conclusions
The use of different link functions in transformation models has been studied by several authors simply investigating the goodness of fit of the different links [bib_ref] Analysis of survival data by the proportional odds model, Bennett [/bib_ref] [bib_ref] Testing goodness of fit for proportional hazards model with censored observations, Wei [/bib_ref].
The alternative perspective considered here, evaluates the use of different links with the goal of providing suitable measures of association between covariates and outcome. As a consequence, when a specific link function is chosen, it should be expected the need to include time dependent covariate effects into the model.
Transformation models estimated through pseudovalues appear an easily implemented alternative to the available approaches mainly based on Cox proportional hazard model to obtain adjusted measures of association eventually time-dependent also for continuous covariates.
[fig] *: Correspondence: [email protected] 1 Department of Clinical Sciences and Community Health, University of Milan, Via Venezian 1, 20133 Milan, Italy Full list of author information is available at the end of the article [/fig]
[fig] Figure 1: Simulated data: RD(t) and Kaplan-Meier Curves. Estimation of RD(t) associated with an hypothetical experimental treatment using artificial data simulated from a proportional hazard model (Details on the simulation are reported in the manuscript). The treatment effect is confounded by two covariates. The unadjusted Kaplan-Meier survival probabilities (----) are reported together with the adjusted estimates (--) obtained with the corrected group prognosis methods (left panel). The corresponding RD(t) estimates are reported in the right panel together with the model based constant RD estimate (with confidence interval). [/fig]
[fig] Figure 2: Prostate Cancer Trial: Treatment RD(t). Left: Estimate of treatment RD(t) and NNT(t) in the prostate cancer diethylstilbestrol trial by different methods. Continuous line: CGPM with the Cox proportional hazard model. The estimated RD(t) is increasing through follow-up time. Broken line: CGPM with a Cox model with a time dependent treatment effect (interaction treatment by time modeled with a B-spline with 1 knot at the median of the failure time distribution). There is no treatment effect until month 20, then RD(t) increases until month 40. Dotted line: estimate obtained using the transformation model with identity link. The RD(t) estimate is negative at the beginning of follow-up (harmful treatment) then becomes positive after about 20 months (beneficial treatment) reflecting the differential impact on cardiovascular and cancer deaths. On the top: three period Cox model used by Kay. The log hazard ratio is positive in the first period (0.09) [ 0 − 13], then becomes negative in the second (-0.40), (13 − 32], and in the third (-0.31), (32 − ∞), periods. The three period model was preferred to the overall Cox model according to likelihood ratio, accounting for non-proportional hazards. Right: estimated population averaged survival probability for the two intervention groups of the prostate cancer trial. The estimate from the time-dependent Cox model is reported from left to the right. The estimate from the transformation model with identity link is reported from right to the left. It is possible to observe a crossing of the survival curves. [/fig]
[fig] Figure 3: Marginal Survival function. Estimate of the marginal survival function for the prostate cancer data using a transformation model estimated with pseudo-values. 32 pseudo-times were considered. Each subject is therefore replicated 32 times in the dataset. The typical estimation of the baseline risk function is through indicator variables. In this case 32 coefficients should be included in the model. A B-Spline was instead used with one knot at the median of the unique failure time distribution resulting in 4 bases plus the intercept for modeling the baseline risk. The figure shows the estimated marginal survival with superimposed the Kaplan-Meier estimate. [/fig]
[fig] Figure 4: Cox prop: CGPM with a Cox model without treatment time dependent effect; Cox TD: CGPM with a Cox model with treatment time dependent effect; PV identity Z const: Pseudo-value model with identity link and constant treatment effect RD; PV identity Z TD: Pseudo-value model with identity link and time dependent treatment effect. B-splines were used to model treatment time dependent effects http:Kaplan-Meier overall survival curves. Patients with retroperitoneal soft tissue sarcoma. Left panel: Surgical Margins. All patients underwent surgery but some of them (14), especially those with advanced disease status, failed to have complete resection. Right panel: Grading (1: 60; 2: 62; 3: 70). [/fig]
[fig] Figure 5, Figure 6: Retroperitoneal soft tissue sarcoma: Margins. RD(t) of extension of resection margins estimated using different methods. PV identity: estimated effect using pseudo values, identity link and interaction between time and margins. Five pseudo times were used at times 6, 11, 23, 51, 96, corresponding to quantiles 10%, 25%, 50%, 75%, 90% of the failure time distribution. CGPM: Corrected group prognosis method using a proportional hazard Cox regression model. The averaging is performed on the whole sample of patients. Other two measures are reported: (1) average over exposed: obtained averaging only on patients with macroscopic resection margins (EIN(t)); (2) average over unexposed: obtained averaging only on patients without microscopic resection margins (NNE(t)). Retroperitoneal soft tissue sarcoma: Grading. The effect of the different Grading categories on the RR(t) and RD(t) scales are reported, using different estimation methods. CGPM applied to a proportional hazard Cox model; PV with log and identiy links and interaction between time and grading. The estimates obtained with the different techniques appears quite in agreement. No evidence for time dependent effects in the Cox model was found. [/fig]
[table] Table 1: Event times generated according to a Cox-exponential model with a confounder X and an exposure status Z [/table]
[table] Table 2: Event times generated according to a Cox-exponential model with a confounder X and an exposure status Z with time-dependent effect [/table]
[table] Table 3: RD(t) estimated with different methods at months 13, 32 and 60 [/table]
|
Richter Syndrome: From Molecular Pathogenesis to Druggable Targets
## Definition of richter syndrome
Richter Syndrome (RS) was reported for the first time by Maurice N. Richter as "reticular cell sarcoma" in 1928 [bib_ref] Generalized Reticular Cell Sarcoma of Lymph Nodes Associated with Lymphatic Leukemia, Richter [/bib_ref]. Currently, according to the World Health Organization (WHO) classification of Tumours of Haematopoietic and Lymphoid Tissues, RS is defined as the occurrence of an aggressive lymphoma in patients with a previous or concomitant diagnosis of chronic lymphocytic leukemia (CLL) or small lymphocytic lymphoma (SLL). RS is currently divided into two recognized pathological variants: diffuse large B-cell
## Epidemiology
CLL is the most frequent leukemia in adults, with an incidence of 4.7/100,000 per year in the US. In the chemo-immunotherapy (CIT) era, data mainly based on retrospective studies showed an incidence of RS transformation ranging from 1 to 7%. Consistently, the CALGB 9011 clinical trial demonstrated a RS transformation rate of~7% in treatment-naïve CLL patients after at least 15 years after treatment with fludarabine or chlorambucil [bib_ref] The impact of initial fludarabine therapy on transformation to Richter syndrome or..., Solh [/bib_ref]. Remarkably, in the CLL8 trial evaluating the effect of fludarabine and cyclophosphamide with or without the anti-CD20 monoclonal antibody (mAb) rituximab, the use of rituximab proved to be a protective factor against RS transformation, leading to lower rates of progression to RS in patients receiving rituximab [bib_ref] Long-term remissions after FCR chemoimmunotherapy in previously untreated patients with CLL: Updated..., Fischer [/bib_ref]. A retrospective study suggested a potential role played by prolymphocytes in RS development, underlining that deaths due to RS were significantly more common in CLL patients who had ≥10% circulating prolymphocytes [bib_ref] The morphology of CLL revisited: The clinical significance of prolymphocytes and correlations..., Oscier [/bib_ref].
More recently, a pooled analysis of the German CLL Study Group (GCLLSG) considering frontline treatment trials with both CIT and pathway inhibitors, including Bruton tyrosine kinase (BTK) and BCL2 inhibitors, has documented a 3% prevalence of RS transformation among 2975 CLL patients monitored after their enrolment in clinical trials, recruited from 1999 to 2016, with a median observation time of 53 months [bib_ref] Richter transformation in chronic lymphocytic leukemia (CLL)-a pooled analysis of German CLL..., Al-Sawaf [/bib_ref]. Data from the Surveillance, Epidemiology and End Results (SEER) database of CLL patients diagnosed between 2000 and 2016 have documented that the incidence of RS transformation was 0.7% [bib_ref] Outcomes of Richter's transformation of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL): An..., Elnair [/bib_ref].
The issue of RS epidemiology in the era of novel agents has been partially answered by results collected from the first clinical trials with pathway inhibitors: in first-line treatment, novel agents showed RS transformation rates comparable to those of the CIT era, suggesting that pathway inhibitors are neither harmful nor fully protective [bib_ref] Ibrutinib plus obinutuzumab versus chlorambucil plus obinutuzumab in first-line treatment of chronic..., Moreno [/bib_ref] [bib_ref] Acalabrutinib with or without obinutuzumab versus chlorambucil and obinutuzmab for treatment-naive chronic..., Sharman [/bib_ref]. Among relapsed/refractory (R/R) CLL patients treated with novel agents, the RS incidence was higher than the overall incidence of RS in the CIT era, probably due to the biological behavior and genetic profile of R/R CLL [bib_ref] Venetoclax-Rituximab in Relapsed or Refractory Chronic Lymphocytic Leukemia, Seymour [/bib_ref] [bib_ref] Final analysis from RESONATE: Up to six years of follow-up on ibrutinib..., Munir [/bib_ref] [bib_ref] Extended follow-up and impact of high-risk prognostic factors from the phase 3..., Brown [/bib_ref]. Because pathway inhibitors have been approved for the treatment of CLL relatively recently (the first BTK inhibitor, ibrutinib, was approved in 2014), further investigation is needed to assess the RS transformation rate with these novel agents in a real-life setting. Currently available data on RS frequency in cases treated with pathway inhibitors are summarized in [fig_ref] Table 1: RS frequency in patients treated with therapeutic regimens based on pathway inhibitors [/fig_ref].
## Molecular pathways in rs
Several molecular alterations associated with DLBCL-type RS have been described, whereas the development of HL-type RS has been studied less extensively and is thought to be similar to that of de novo HL and possibly linked to EBV-mediated immunosuppression, thus favoring CLL progression to HL [bib_ref] Hodgkin lymphoma transformation of chronic lymphocytic leukemia: Cases report and discussion, Kaźmierczak [/bib_ref] [bib_ref] Richter syndrome: Biology, incidence, and therapeutic strategies, Tsimberidou [/bib_ref].
Genetic lesions of RS. The pathogenesis of DLBCL-type RS is linked to the dysregulation of intracellular pathways involved in DNA damage response, tumor suppression, apoptosis, modulation of the cell cycle and proliferation. The main genetic lesions associated with DLBCL-type RS are represented by somatic mutations or disruptions of the TP53, CDKN2A, NOTCH1 and c-MYC genes [fig_ref] Figure 1: Molecular pathway alterations in DLBCL-type RS [/fig_ref] [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia transformation to Richter syndrome, Fabbri [/bib_ref] [bib_ref] The genetics of Richter syndrome reveals disease heterogeneity and predicts survival after..., Rossi [/bib_ref] [bib_ref] Two main genetic pathways lead to the transformation of chronic lymphocytic leukemia..., Chigrinova [/bib_ref] [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref].
TP53 encodes for one of the main regulators of the DNA-damage-response pathway, and its disruption, generally acquired at the time of transformation, leads to the chemorefractoriness characteristic of RS and, therefore, favors the positive selection and expansion of mutated tumor cells [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] Mutant p53 as a target for cancer treatment, Duffy [/bib_ref]. Disruption of TP53 by mutation and/or deletion has been documented in a large fraction of DLBCL-type RS, including 60 to 80% of clonally related RS cases, which represent the overwhelming majority of RS events, and in 20% of clonally unrelated RS cases [bib_ref] The genetics of Richter syndrome reveals disease heterogeneity and predicts survival after..., Rossi [/bib_ref]. The high recurrence of TP53 disruption explains, at least in part, the frequency of chemorefractoriness in this condition.
CDKN2A deletion occurs in~30% of DLBCL-type RS and is commonly acquired at the time of transformation [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia transformation to Richter syndrome, Fabbri [/bib_ref]. The CDKN2A gene is responsible for the negative regulation of the G1 to S transition of the cell cycle and for the activation of the p53 transcriptional program through its transcripts p16INK4A and p14ARF, respectively, leading to tumor suppression [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia transformation to Richter syndrome, Fabbri [/bib_ref] [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] Ink4-Arf locus in cancer and aging, Sherr [/bib_ref]. Importantly, both positive and negative cell cycle regulators are induced by B-cell receptor (BCR) signaling in murine models. On these grounds, the concomitant loss of the negative cell cycle regulators TP53 and CDKN2A/B shifts BCRdependent signaling toward the promotion of positive cell cycle regulators, leading to an aggressive proliferation compatible with RS transformation [bib_ref] B-cell receptor signaling and genetic lesions in TP53 and CDKN2A/CDKN2B cooperate in..., Chakraborty [/bib_ref].
NOTCH1 encodes for a surface receptor that, after being triggered by a ligand belonging to the SERRATE/JAGGED or DELTA families, is cleaved by γ-secretase, migrates into the nucleus and activates the transcription of several genes involved in cell proliferation and survival. At the time of diagnosis, the frequency of NOTCH1 mutational activation in RS is significantly higher compared to the frequency in CLL (31% vs. 8.3%, respectively) [bib_ref] Analysis of the chronic lymphocytic leukemia coding genome: Role of NOTCH1 mutational..., Fabbri [/bib_ref]. These data, together with the noticeably higher reported risk of developing DLBCL-type RS in NOTCH1 mutated CLL (45% at 15 years) vs. NOTCH1 wild type CLL (4.6% at 15 years), indicate that mutations of NOTCH1 are a significant risk factor for developing RS transformation [bib_ref] Different impact of NOTCH1 and SF3B1 mutations on the risk of chronic..., Rossi [/bib_ref]. The pathogenesis of DLBCL-type RS is due to the dysregulation of multiple molecular pathways due to genetic lesions of proto-oncogenes and tumor suppressor genes, stereotyped B-cell receptor (BCR) configuration, and BCR signaling alterations; these lead to the enhanced cell survival and proliferation typical of DLBCL-type RS cells. Inhibition of apoptosis may also be involved. BTK, Bruton tyrosine kinase; PI3K, phosphatidylinositol-3 kinase; PLCγ2, phospholipase C gamma 2; XPO1, exportin 1. Image created with Biorender.com (accessed on 4 August 2022).
TP53 encodes for one of the main regulators of the DNA-damage-response pathway, and its disruption, generally acquired at the time of transformation, leads to the chemorefractoriness characteristic of RS and, therefore, favors the positive selection and expansion of mutated tumor cells [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] Mutant p53 as a target for cancer treatment, Duffy [/bib_ref]. Disruption of TP53 by mutation and/or deletion has been documented in a large fraction of DLBCL-type RS, including 60 to 80% of clonally related RS cases, which represent the overwhelming majority of RS events, and in 20% of clonally unrelated RS cases [bib_ref] The genetics of Richter syndrome reveals disease heterogeneity and predicts survival after..., Rossi [/bib_ref]. The high recurrence of TP53 disruption explains, at least in part, the frequency of chemorefractoriness in this condition.
CDKN2A deletion occurs in ~30% of DLBCL-type RS and is commonly acquired at the time of transformation [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia transformation to Richter syndrome, Fabbri [/bib_ref]. The CDKN2A gene is responsible for the negative regulation of the G1 to S transition of the cell cycle and for the activation of the p53 transcriptional program through its transcripts p16INK4A and p14ARF, respectively, leading to tumor suppression [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia transformation to Richter syndrome, Fabbri [/bib_ref] [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] Ink4-Arf locus in cancer and aging, Sherr [/bib_ref]. Importantly, both positive and negative cell cycle regulators are induced by B-cell receptor (BCR) signaling in murine models. On these grounds, the concomitant loss of the negative cell cycle regulators TP53 and CDKN2A/B shifts BCRdependent signaling toward the promotion of positive cell cycle regulators, leading to an aggressive proliferation compatible with RS transformation [bib_ref] B-cell receptor signaling and genetic lesions in TP53 and CDKN2A/CDKN2B cooperate in..., Chakraborty [/bib_ref]. The c-MYC proto-oncogene and its transcriptional product take part in many crucial cellular pathways, and a dysregulation of this network results in altered cell survival, proliferation, metabolism, self-renewal, and genomic instability [bib_ref] The MYC oncogene-The grand orchestrator of cancer growth and immune evasion, Dhanasekaran [/bib_ref]. The main genetic lesions deregulating c-MYC expression in DLBCL-type RS are represented by chromosomal translocations between the c-MYC locus and the IGHV regulatory regions, and gene amplifications as well as gain-of-function mutations of the c-MYC promoter. An alternative mechanism of c-MYC deregulation is represented by loss-of-function mutations of the MGA gene, which encodes for a protein inhibiting c-MYC heterodimerization with its partner MAX. Overall, by combining genetic lesions of c-MYC and MGA,~40% of DLBCL-type RS cases harbor c-MYC deregulation [bib_ref] MGA, a suppressor of MYC, is recurrently inactivated in high risk chronic..., De Paoli [/bib_ref].
The NF-κB pathway is also involved in RS pathogenesis [bib_ref] Genomic and transcriptomic correlates of Richter transformation in chronic lymphocytic leukemia, Klintman [/bib_ref]. TRAF3, a gene implicated in the negative regulation of signaling through the NF-κB and mitogen-activated protein kinase (MAPK) pathways, is disrupted by heterozygous deletions and frameshift mutations in a fraction of RS cases [bib_ref] Genomic and transcriptomic correlates of Richter transformation in chronic lymphocytic leukemia, Klintman [/bib_ref]. Inactivation of TRAF3 leads to NF-κB activation, promoting B cell survival and, in particular, enhancing the expression of c-MYC and PIM-2 [bib_ref] TRAF3 regulates the oncogenic proteins Pim2 and c-Myc to restrain survival in..., Whillock [/bib_ref]. PIM-2 maintains high levels of NF-κB, which are required for its antiapoptotic function.
The pathogenetic role of PIM-2 in B-cell neoplasia is documented by its overexpression, translocation, or amplification in a fraction of B-cell lymphomas [bib_ref] Genomic and transcriptomic correlates of Richter transformation in chronic lymphocytic leukemia, Klintman [/bib_ref] [bib_ref] TRAF3 regulates the oncogenic proteins Pim2 and c-Myc to restrain survival in..., Whillock [/bib_ref] [bib_ref] PIM2 inhibition as a rational therapeutic approach in B-cell lymphoma, Gómez-Abad [/bib_ref] [bib_ref] Pim kinases in hematological malignancies: Where are we now and where are..., Mondello [/bib_ref].
Other molecular alterations detected in DLBCL-type RS include (i) overexpression and amplification of PTPN11, a positive regulator of the MAPK-RAS-ERK signaling pathway; (ii) deletion of the SETD2 histone methyltransferase, which plays a major role in chromatin epigenetic remodeling; and (iii) disrupting mutations of the tumor suppressor gene PTPRD, which encodes for a receptor-type protein, tyrosine phosphatase, regulating cell growth and found to be inactivated also in other types of B-cell neoplasia and solid cancers [bib_ref] Genomic and transcriptomic correlates of Richter transformation in chronic lymphocytic leukemia, Klintman [/bib_ref] [bib_ref] Understanding the language of Lys36 methylation at histone H3, Wagner [/bib_ref] [bib_ref] The tyrosine phosphatase PTPRD is a tumor suppressor that is frequently inactivated..., Veeriah [/bib_ref] [bib_ref] The genetics of nodal marginal zone lymphoma, Spina [/bib_ref].
Modification of immune regulators in RS. Recent studies have underlined the importance of immune checkpoints and of the tumor microenvironment in lymphatic tissues of DLBCLtype RS [bib_ref] PD-1 Expression in Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL) and Large B-cell..., He [/bib_ref] [bib_ref] Distinct immune signatures in chronic lymphocytic leukemia and Richter syndrome, Wang [/bib_ref]. The main immune checkpoints involved in RS are PD-1, LAG3 and TIGIT.
PD-1 is a T cell surface molecule which stimulates effector T-cell apoptosis and Treg survival through its binding with the PD-L1 ligand, which is expressed mainly on the surface of antigen-presenting cells (APC), such as macrophages, B cells and dendritic cells (DC) [bib_ref] PD-1/PD-L1 pathway: Basic biology and role in cancer immunotherapy, Salmaninejad [/bib_ref]. Augmented levels of PD-1 in RS cells and enhanced expression of PD-L1 in histiocytes and dendritic cells of the RS microenvironment have been reported [bib_ref] PD-1 Expression in Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL) and Large B-cell..., He [/bib_ref] [bib_ref] Distinct immune signatures in chronic lymphocytic leukemia and Richter syndrome, Wang [/bib_ref]. Altered expression of the PD-1/PD-L1 axis leads to RS tumor-cell resistance to the cytotoxicity exerted by T cells [bib_ref] PD-1/PD-L1 pathway: Basic biology and role in cancer immunotherapy, Salmaninejad [/bib_ref].
LAG3 is a T cell surface protein with a structure similar to the T helper antigen CD4, and its main ligand is the major histocompatibility complex (MHC) class II, typically expressed by APC. Additional ligands have been identified, namely Galectin-3 (Gal-3), which is expressed by several cell types [bib_ref] Lymphocyte-activation gene 3 (LAG3): The next immune checkpoint receptor, Ruffo [/bib_ref]. The interaction between LAG3 and its ligands promotes tumor escape from apoptosis through the recruitment of tumor-specific CD4+ T cells (through the interaction with MHC class II) and the inhibition of CD8+ T cells' cytotoxic function by Gal-3 binding [bib_ref] Lymphocyte-activation gene 3 (LAG3): The next immune checkpoint receptor, Ruffo [/bib_ref]. Higher LAG3 levels have been observed in RS neoplastic and tumor-infiltrating lymphocytes, suggesting its potential role in promoting tumor immune escape and neoplastic cell survival [bib_ref] Characterisation of immune checkpoints in Richter syndrome identifies LAG3 as a potential..., Gould [/bib_ref].
TIGIT, expressed on normal T and NK cells and overexpressed in RS, is capable of immune suppression as a consequence of its binding with the CD155 ligand, exposed on the cell membrane of various cell types, such as dendritic cells, T cells, B cells, and macrophages [bib_ref] Characterisation of immune checkpoints in Richter syndrome identifies LAG3 as a potential..., Gould [/bib_ref] [bib_ref] TIGIT as an emerging immune checkpoint, Harjunpää [/bib_ref]. The mechanism of action of TIGIT is supposed to be linked to the transduction of immune-suppressive stimuli on T and NK cells and to the promotion of tolerogenic DC that downregulate T cell responses [bib_ref] TIGIT as an emerging immune checkpoint, Harjunpää [/bib_ref]. The finding that immune checkpoints are overexpressed in RS cells and tumor-infiltrating lymphocytes suggests the potential role of these molecules in the promotion of a permissive immune microenvironment, resulting in immune suppression and tumor escape.
The BCR pathway in RS. Several BCR pathway alterations related to RS transformation have been documented. The BCR is a transmembrane complex expressed in B cells, composed by a surface immunoglobulin linked to a signal transduction subunit and responsible for antigen recognition and B cell activation [fig_ref] Figure 1: Molecular pathway alterations in DLBCL-type RS [/fig_ref] [bib_ref] B-cell receptor: From resting state to activate, Treanor [/bib_ref].
The variable part of the BCR IGHV subunit is characterized by a molecular pattern typical of mature B cells, the VDJ rearrangement, which causes a considerable diversity across the BCR expressed by different B cell clones. Approximately thirty percent of CLL patients carry stereotyped BCR, which are characterized by almost identical VDJ rearrangement across patients and are groupable in well-codified subsets identified by progressive numbers [bib_ref] Distinctive Signaling Profiles With Distinct Biological and Clinical Implications in Aggressive CLL..., Gerousi [/bib_ref] [bib_ref] Over 20% of patients with chronic lymphocytic leukemia carry stereotyped receptors: Pathogenetic..., Stamatopoulos [/bib_ref].
CLL patients carrying BCR subset 8 (characterized by IGHV4-39/IGHD6-13/IGHJ5 rearrangement) display a significantly higher risk of developing DLBCL-type RS, especially in combination with NOTCH1 mutations. [bib_ref] Stereotyped B-cell receptor is an independent risk factor of chronic lymphocytic leukemia..., Rossi [/bib_ref] [bib_ref] Association between molecular lesions and specific B-cell receptor subsets in chronic lymphocytic..., Rossi [/bib_ref]. From a mechanistic perspective, CLL cells harboring BCR subset 8 tend to overreact to multiple autoantigens and immune stimuli derived from the microenvironment [fig_ref] Figure 1: Molecular pathway alterations in DLBCL-type RS [/fig_ref]. The propensity of these cells to undergo RS transformation can be explained by this promiscuous antigen reactivity [bib_ref] Excessive antigen reactivity may underlie the clinical aggressiveness of chronic lymphocytic leukemia..., Gounari [/bib_ref].
In the BCR signal transduction, a key role is played by Bruton tyrosine kinase (BTK) and phosphatidylinositol-3 kinase (PI3K) [bib_ref] Targeting B cell receptor signalling in cancer: Preclinical and clinical advances, Burger [/bib_ref]. BTK is phosphorylated subsequently to BCR stimulation and leads to the activation of phospholipase C gamma 2 (PLCγ2), causing calcium mobilization and activation of cell survival, proliferation and differentiation pathways, including MAPK and NF-κB signaling [fig_ref] Figure 1: Molecular pathway alterations in DLBCL-type RS [/fig_ref] [bib_ref] BTK inhibition targets in vivo CLL proliferation through its effects on B-cell..., Cheng [/bib_ref]. PI3K, which is responsible for the activation of the serine/threonine kinase Akt and for the delta isoform of protein kinase C (PKC), was observed to be constitutively active in CLL patients, resulting in an enhanced anti-apoptotic effect [bib_ref] Constitutively activated phosphatidylinositol-3 kinase (PI-3K) is involved in the defect of apoptosis..., Ringshausen [/bib_ref].
Akt signaling in RS. Akt takes part in cell-survival signaling through mTOR and is constitutively active in high-risk CLL (i.e., TP53 or NOTCH1 mutated CLL) and in >50% of cases of RS [bib_ref] Constitutively activated phosphatidylinositol-3 kinase (PI-3K) is involved in the defect of apoptosis..., Ringshausen [/bib_ref] [bib_ref] Active Akt signaling triggers CLL toward Richter transformation via overactivation of Notch1, Kohlhaas [/bib_ref]. In an Eµ-TCL1 murine model of CLL with constitutively active Akt alleles in B cells, the excessive Akt activation led to an aggressive DLBCL-type lymphoma with histological and biological features coherent with human RS [bib_ref] Active Akt signaling triggers CLL toward Richter transformation via overactivation of Notch1, Kohlhaas [/bib_ref]. Additionally, this murine model enlightened the correlation between hyperactivation of Akt and NOTCH1 signaling, since mice with constitutively active Akt alleles presented an expansion of CD4+ T cells expressing the NOTCH1 ligand DLL1 in the tumor microenvironment, implying a higher engagement of NOTCH1 by its ligands in neoplastic cells [bib_ref] Active Akt signaling triggers CLL toward Richter transformation via overactivation of Notch1, Kohlhaas [/bib_ref].
## Molecular history of clonally related dlbcl-type rs
Recently, an analysis performed on longitudinal tumor samples from CLL patients developing clonally related DLBCL-type RS has underlined the presence of dormant seeds of RS already at the time of CLL diagnosis [bib_ref] Detection of early seeding of Richter transformation in chronic lymphocytic leukemia, Nadeu [/bib_ref]. The samples were obtained before and after treatment with chemotherapy, immunotherapy, and novel agents (mainly BCR pathway inhibitors), and were analyzed with bulk whole-genome, epigenome and transcriptome sequencing. A very early diversification of CLL was shown to lead to the emergence of neoplastic cells with RS features up to 19 years before the clinical and histological evidence of transformation [bib_ref] Detection of early seeding of Richter transformation in chronic lymphocytic leukemia, Nadeu [/bib_ref]. These early clones displayed fully assembled genomic, immunogenetic and transcriptomic profiles of RS already at CLL diagnosis.
Moreover, RS tumor cells showed an activation of the oxidative phosphorylation (OXPHOS) pathway and, on the other side, a downregulation of BCR signaling [bib_ref] Detection of early seeding of Richter transformation in chronic lymphocytic leukemia, Nadeu [/bib_ref]. The OXPHOS high -BCR low transcriptional axis could partly explain the resistance and rapid selection of DLBCL-type RS cells observed after treatment with BCR inhibitors [bib_ref] Molecular profiling of diffuse large B-cell lymphoma identifies robust subtypes including one..., Monti [/bib_ref] [bib_ref] Metabolic signatures uncover distinct targets in molecular subsets of diffuse large B..., Caro [/bib_ref] [bib_ref] Differential contribution of the mitochondrial translation pathway to the survival of diffuse..., Norberg [/bib_ref]. In this regard, OXPHOS inhibition has been tested in vitro, demonstrating an activity against the proliferation of DLBCL-type RS cells, a finding worth exploring in future therapeutic strategies [bib_ref] Detection of early seeding of Richter transformation in chronic lymphocytic leukemia, Nadeu [/bib_ref].
## Molecular differences between clonally related and clonally unrelated dlbcl-type rs
At least 80% of DLBCL-type RS cases carry the same IGHV rearrangement in both CLL and RS cells and thus originate from the direct transformation of the pre-existent CLL clone [bib_ref] IgVH mutational status and clonality analysis of Richter's transformation: Diffuse large B-cell..., Mao [/bib_ref] [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref]. Conversely, a minority of DLBCL-type RS cases (approximately 20%) carry a different IGHV rearrangement compared to the CLL phase, and are therefore de novo DLBCL arising in the context of a condition, i.e., CLL, which predisposes to secondary lymphoid malignancies [bib_ref] IgVH mutational status and clonality analysis of Richter's transformation: Diffuse large B-cell..., Mao [/bib_ref] [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] Trends in the risk of second primary malignancies among survivors of chronic..., Kumar [/bib_ref]. The distinction between clonally related and clonally unrelated DLBCL-type RS is clinically relevant, since clonally unrelated cases display a better prognosis that is overall overlapping with that of DLBCL arising de novo in patients without CLL [bib_ref] The genetics of Richter syndrome reveals disease heterogeneity and predicts survival after..., Rossi [/bib_ref] [bib_ref] Diffuse large B-cell lymphoma (Richter syndrome) in patients with chronic lymphocytic leukaemia..., Parikh [/bib_ref].
The molecular pathogenesis of clonally unrelated DLBCL-type RS has been poorly understood for a long time. A very recent study exploited ultra-deep NGS to interrogate the timing of emergence and the genetic profile of this rare disorder [bib_ref] Clonally unrelated Richter syndrome are truly de novo diffuse large B-cell lymphomas..., Favini [/bib_ref]. The investigation of IGHV gene rearrangements of clonally unrelated DLBCL-type RS clarified that the RS clone was not present either at the time of CLL diagnosis or later in the clinical course of the indolent phase of the disease, even with the sensitivity of 10-6 allowed by ultra-deep NGS [bib_ref] Clonally unrelated Richter syndrome are truly de novo diffuse large B-cell lymphomas..., Favini [/bib_ref]. Therefore, clonally unrelated DLBCL-type RS unequivocally arises as a de novo DLBCL emerging as a secondary tumor whose development is favored by the deep immune suppression of the CLL host [bib_ref] Trends in the risk of second primary malignancies among survivors of chronic..., Kumar [/bib_ref].
Mutational analysis performed by CAPP-Seq with an ample panel of genes affected in B-cell neoplasia revealed that genetic lesions of clonally unrelated DLBCL-type RS predominantly involve tumor suppressor genes implicated in the DNA damage response (TP53, ATM) as well as genes regulating cell cycle and proliferation (NOTCH1, ID3, and MYC) [bib_ref] Clonally unrelated Richter syndrome are truly de novo diffuse large B-cell lymphomas..., Favini [/bib_ref]. Remarkably, genes that are recurrently affected in DLBCL without a previous CLL phase, namely, KMT2D, CREBBP, EP300 and TNFAIP3, and genes implicated in BCR signaling are not affected in clonally unrelated DLBCL-type RS [bib_ref] Clonally unrelated Richter syndrome are truly de novo diffuse large B-cell lymphomas..., Favini [/bib_ref]. On these grounds, the genetic profile of clonally unrelated DLBCL-type RS is reminiscent, at least in part, of the mutational spectrum associated with clonally related DLBCL-type RS. The fact that both clonally unrelated and clonally related DLBCL-type RS share the involvement of a similar panel of genes despite their different histogenetic origins may reflect commonalities in the degree of immune suppression and/or in the lymph node microenvironment in both conditions. Also, the previous exposure of the host to antineoplastic agents might exert a role favoring the accumulation of specific genetic lesions in both clonally unrelated and clonally related DLBCL-type RS.
From a clinical standpoint, this novel information on clonally unrelated DLBCL-type RS is of value. On the one side, in fact, the molecular vulnerabilities of both clonally related and clonally unrelated DLBCL-type RS have a certain degree of overlap and might be exploited as common targets with innovative agents [bib_ref] Clonally unrelated Richter syndrome are truly de novo diffuse large B-cell lymphomas..., Favini [/bib_ref]. On the other side, the high frequency of TP53 disruption in both clonally related and clonally unrelated DLBCL-type RS mandates the analysis of this genetic lesion in the RS sample, since TP53 disruption is a consolidated predictor of chemorefractoriness in the context of both CLL and DLBCL arising de novo [bib_ref] Prognostic impact of TP53 mutation in newly diagnosed diffuse large B-cell lymphoma..., Chiappella [/bib_ref] [bib_ref] Genetic lesions associated with chronic lymphocytic leukemia chemo-refractoriness, Messina [/bib_ref]. Finally, because the treatment of clonally related versus clonally unrelated DLBCL-type RS differs, at least for what concerns the indication to hematopoietic stem cell transplant, whenever possible, clinicians should implement strategies aimed at defining the clonal relationship between CLL and DLBCL-type RS [bib_ref] Biology and treatment of Richter syndrome, Rossi [/bib_ref] [bib_ref] How we treat Richter syndrome, Parikh [/bib_ref]. From a practical point of view, it might be cumbersome to have the availability of both samples for genomic analysis. However, interpolation between PD-1 expression and immunoglobulin genes analyses showed a 90% correlation of PD-1 expression with clonally related DLBCL-type RS defined in molecular terms; consequently, PD-1 expression could represent a surrogate for defining the clonal relation between CLL and DLBCL-type RS [bib_ref] PD-1 Expression in Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL) and Large B-cell..., He [/bib_ref].
## Druggable targets for the treatment of dlbcl-type rs
The understanding of the molecular pathways involved in RS pathogenesis has led to the identification of novel druggable targets, namely BTK, BCL2, PD-1 and PD-L1 [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref].
Basic concepts in the treatment of RS. DLBCL-type RS is characterized by an aggressive and particularly chemorefractory behavior, partly due to the genetic features of the disease. Furthermore, RS patients commonly suffer from poor performance status, impaired bone marrow function or immunodeficiency, displaying a frailty profile that represents a predictor of low survival rate after RS transformation [bib_ref] Richter syndrome: Molecular insights and clinical perspectives, Rossi [/bib_ref]. The combination between chemorefractoriness, on one side, and frailty, on the other side, may be a possible cause of the dismal prognosis observed in RS treated with CIT because of the limited possibility to tolerate high-dose chemotherapy typical of frail and heavily pretreated patients. In a phase II trial, the R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) regimen, widely used in de novo DLBCL, showed an overall response rate (ORR) of 67% in RS patients, with a complete response (CR) of only 7%, a median overall survival (OS) of 21 months and a median progression-free survival (PFS) of 10 months [bib_ref] Poor efficacy and tolerability of R-CHOP in relapsed/refractory chronic lymphocytic leukemia and..., Langerbeins [/bib_ref]. The most common severe adverse events were hematological toxicity (grade ≥3 anemia, neutropenia and thrombocytopenia) and infections. Regimens alternative to R-CHOP have also been tested for RS, although none of them has demonstrated a clear superiority. A retrospective study assessed first-line R-EPOCH (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, and doxorubicin) and resulted in a median OS of 5.9 months and underlined a poorer prognosis in patients with TP53 deletion or complex karyotype [bib_ref] A single-institution retrospective cohort study of first-line R-EPOCH chemoimmunotherapy for Richter syndrome..., Rogers [/bib_ref]. The hyper-CVXD (fractioned cyclophosphamide, vincristine, doxorubicin, and dexam-ethasone) regimen has been investigated in a phase II study and, despite the clinically significant CR rate of 38%, an important toxicity was reported, with an overall mortality rate of 20% [bib_ref] Fractionated cyclophosphamide, vincristine, liposomal daunorubicin (daunoXome), and dexamethasone (hyperCVXD) regimen in Richter's..., Dabaja [/bib_ref]. A phase I/II study assessing safety and efficacy of the OFAR2 (oxaliplatin, fludarabine, ara-C and rituximab) regimen reported an ORR of 38.7%, with a CR of 6.5% and a median survival duration of 6.6 months [bib_ref] Phase I-II clinical trial of oxaliplatin, fludarabine, cytarabine, and rituximab therapy in..., Tsimberidou [/bib_ref]. The CHOP-OR phase II trial evaluated the CHOP-O (cyclophosphamide, doxorubicin, vincristine, prednisolone, with induction and maintenance with the anti-CD20 mAb ofatumumab) regimen in newly diagnosed RS, reporting an ORR of 46%, with a CR rate of 27%. The median PFS of 6.2 months and the median OS of 11.4 months did not demonstrate superiority compared to R-CHOP in terms of outcome [bib_ref] NCRI phase II study of CHOP in combination with ofatumumab in induction..., Eyre [/bib_ref]. These findings highlight the need for alternative therapeutic strategies in RS treatment. practical point of view, it might be cumbersome to have the availability of both samples for genomic analysis. However, interpolation between PD-1 expression and immunoglobulin genes analyses showed a 90% correlation of PD-1 expression with clonally related DLBCL-type RS defined in molecular terms; consequently, PD-1 expression could represent a surrogate for defining the clonal relation between CLL and DLBCL-type RS [bib_ref] PD-1 Expression in Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL) and Large B-cell..., He [/bib_ref].
## Druggable targets for the treatment of dlbcl-type rs
The understanding of the molecular pathways involved in RS pathogenesis has led to the identification of novel druggable targets, namely BTK, BCL2, PD-1 and PD-L1 [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref]. BTK inhibitors in RS. Covalent BTK inhibitors challenge the Cys481 residue of BTK, which lies within the ATP-binding pocket of the kinase, resulting in a BCR pathway blockade [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] BTK Inhibitors in Chronic Lymphocytic Leukemia, Gaballa [/bib_ref]. Ibrutinib, the first-in-class BTK inhibitor that binds covalently to Cys481, has shown a limited efficacy profile in DLBCL-type RS [bib_ref] The efficacy of ibrutinib in the treatment of Richter syndrome, Tsang [/bib_ref] [bib_ref] Ibrutinib has some activity in Richter's syndrome, Giri [/bib_ref] [bib_ref] Successful Treatment of Richter Transformation with Ibrutinib in a Patient with Chronic..., Master [/bib_ref] [bib_ref] Ibrutinib-induced rapid response in chemotherapy-refractory Richter's syndrome, Fischer [/bib_ref]. An analysis of the ACE-CL-001 study led to similar outcomes with acalabrutinib, a second generation covalent BTK inhibitor, with a CR of 8% and a median PFS of only 3 months [bib_ref] Acalabrutinib monotherapy for treatment of chronic lymphocytic leukaemia (ACE-CL-001): Analysis of the..., Eyre [/bib_ref]. The poor efficacy profile displayed by covalent BTK inhibitors may be explained, at least in part, by the development of BTK inhibitor-resistant clones, which harbor mutations of BTK affecting Cys481 (the most common is C481S) and gain-of-function mutations of PLCγ2 [bib_ref] BTK Inhibitors in Chronic Lymphocytic Leukemia, Gaballa [/bib_ref] [bib_ref] Resistance to Acalabrutinib in CLL Is Mediated Primarily By BTK Mutations, Woyach [/bib_ref]. For this reason, a reversible noncovalent BTK inhibitor may be preferred. Initial results of the use of noncovalent inhibitors are coming from an ongoing phase I/II study (NCT03740529) evaluating the safety and efficacy of pirtobrutinib, a noncovalent BTK inhibitor, in previously treated DLBCL-type RS [bib_ref] LOXO-305), a next generation highly selective non-covalent BTK inhibitor in previously treated..., Jurczak [/bib_ref]. Pirtobrutinib showed encouraging results, with an ORR of 75% and few toxicity events in patients with a median number of two prior lines of therapy [bib_ref] LOXO-305), a next generation highly selective non-covalent BTK inhibitor in previously treated..., Jurczak [/bib_ref] [bib_ref] CLL-039: Pirtobrutinib (LOXO-305), a Next-Generation, Highly Selective, Non-Covalent BTK Inhibitor in Previously..., Coombs [/bib_ref]. A promising reversible noncovalent BTK inhibitor is nemtabrutinib (ARQ 531), tested in murine models of CLL and aggressive B-cell lymphoma [bib_ref] The BTK Inhibitor ARQ 531 Targets Ibrutinib-Resistant CLL and Richter Transformation, Reiff [/bib_ref]. Nemtabrutinib demonstrated a better efficacy profile compared with ibrutinib in vivo, and two clinical trials involving RS patients are ongoing (NCT03162536 and NCT04728893) [bib_ref] The BTK Inhibitor ARQ 531 Targets Ibrutinib-Resistant CLL and Richter Transformation, Reiff [/bib_ref].
BCL2 inhibitors in RS. The BCL-2 family of proteins plays a central role in inhibition of apoptosis, leading to an enhanced tumor cell survival if overexpressed [bib_ref] Targeting BCL2 in Chronic Lymphocytic Leukemia and Other Hematologic Malignancies, Yalniz [/bib_ref]. The BCL-2 inhibitor venetoclax has shown efficacy in CLL, independent of TP53 mutations and chromosome 17p deletion [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] Potential of BCL2 as a target for chronic lymphocytic leukemia treatment, Moia [/bib_ref] [bib_ref] Venetoclax for Patients with Chronic Lymphocytic Leukemia with 17p Deletion: Results from..., Stilgenbauer [/bib_ref]. DLBCL-type RS, as already mentioned, commonly harbors TP53 disruption, making venetoclax a potentially viable strategy for RS treatment. A phase 1 study (NCT01328626) attained a response rate of 43% in DLBCL-type RS treated with venetoclax monotherapy, but no cases of CR were reported [bib_ref] Phase I First-in-Human Study of Venetoclax in Patients With Relapsed or Refractory..., Davids [/bib_ref]. More positive results have come from a phase II trial (NCT03054896), testing the combination of venetoclax with R-EPOCH in DLBCL-type RS, that reached an ORR of 62%, with a remarkable CR rate of 50% (13/26, while 11 of them achieved undetectable bone marrow minimal residual disease for CLL) [bib_ref] Venetoclax plus dose-adjusted R-EPOCH for Richter syndrome, Davids [/bib_ref]. The median PFS and median OS were 10.1 months and 19.6 months, respectively, and the grade ≥3 AEs were neutropenia, thrombocytopenia, and febrile neutropenia [bib_ref] Venetoclax plus dose-adjusted R-EPOCH for Richter syndrome, Davids [/bib_ref].
Immune checkpoint inhibitors in RS. PD-1 signaling can be targeted with mAbs that interfere with the PD-1/PD-L1 interaction [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] Venetoclax for Patients with Chronic Lymphocytic Leukemia with 17p Deletion: Results from..., Stilgenbauer [/bib_ref]. Pembrolizumab, an anti-PD-1 mAb, was evaluated in a phase II clinical trial (NCT02332980), in which DLBCL-type RS patients achieved an ORR of 40% and a median OS of 11 months. The safety profile was acceptable, with only 20% of patients experiencing severe hematologic toxicity. These results were obtained in a non-homogeneous cohort of RS patients, in which 70% of cases were previously treated with ibrutinib, underlining the need for further studies in this treatment strategy. Pembrolizumab monotherapy was assessed also in 21 R/R RS patients in the KEYNOTE-170 phase II study, which underlined poor results for R/R DLBCL-type RS; in particular, the ORR was 4.8% (only 1 patient), with no CR [bib_ref] Pembrolizumab in relapsed or refractory Richter syndrome, Armand [/bib_ref].
Combination therapies with immune checkpoint inhibitors seem to grant more positive results compared to monotherapy. Nivolumab, a PD-1 inhibitor, in combination with ibrutinib displayed an ORR of 65%, with a 10% CR rate, in a DLBCL-type RS cohort of a phase I/II study [bib_ref] Safety and activity of ibrutinib in combination with nivolumab in patients with..., Younes [/bib_ref]. Encouraging results came from an ongoing phase II trial investigating the anti-PD-L1 mAb atezolizumab in combination with venetoclax and the anti-CD20 mAb obinutuzumab in treatment-naïve and R/R RS. The initial results of the study reported an ORR of 100%, with a 71% CR in the treatment-naïve cohort, whereas only one patient was recruited in the R/R group.
## Druggable targets for the treatment of hl-type rs
Due to its similarity to de novo cases, HL-type RS is mainly treated with the ABVD (Doxorubicin, Bleomycin, Vinblastine, Dacarbazine) chemotherapy regimen, with a reported response rate of 40-60% and a median OS of~4 years [bib_ref] Hodgkin transformation of chronic lymphocytic leukemia: The M. D. Anderson Cancer Center..., Tsimberidou [/bib_ref] [bib_ref] Hodgkin's variant of Richter transformation in chronic lymphocytic leukemia; a retrospective study..., Tadmor [/bib_ref] [bib_ref] Hodgkin lymphoma as Richter transformation in chronic lymphocytic leukaemia: A retrospective analysis..., Bockorny [/bib_ref]. The main possible adverse event is the severe pulmonary toxicity caused by the exposure to bleomycin, which can be omitted after two cycles of ABVD if interim PET shows remission with no significant outcome variations [bib_ref] Bleomycin pulmonary toxicity has a negative impact on the outcome of patients..., Martin [/bib_ref].
Immune checkpoint inhibitors and brentuximab vedotin, an anti-CD30 conjugated with the antimicrotubular agent monomethyl auristatin E (MMAE), have proved to be effective in de novo R/R HL, but data on the activity on R/R HL-type RS are insufficient [bib_ref] SGN-35), an antibody-drug conjugate for the treatment of CD30-positive malignancies, Gualberto [/bib_ref] [bib_ref] Novel Agents For Relapsed and Refractory Classical Hodgkin Lymphoma: A, Zhang [/bib_ref]. A case report showed encouraging results in R/R HL-type RS, with a successful outcome in a patient refractory to bendamustine, which at relapse was treated with brentuximab, vedotin and radiotherapy, and, later, with pembrolizumab [bib_ref] Report: Sequential Development of Three Mature Lymphoid Neoplasms in a Single Patient:..., Salvetti [/bib_ref]. Based on this evidence, investigations with these innovative drugs might be envisaged in HLtype RS.
## Perspectives of new potential molecular targets for rs
Recently, patient-derived xenograft (PDX) models for DLBCL-type RS have been developed in order to study human-like models of this rare disease, overcoming the limited number of patients [bib_ref] Novel Richter Syndrome Xenograft Models to Study Genetic Architecture, Biology, and Therapy..., Vaisitti [/bib_ref]. Thanks to PDX models, promising pre-clinical results are being reached, paving the way for further clinical trials.
The receptor tyrosine kinase-like orphan receptor 1 (ROR1) is a surface receptor tyrosine kinase that, through the binding to proteins of the Wnt family, primarily Wnt5a, activates non-canonical WNT signaling, promoting cell survival, proliferation and migration events [bib_ref] Tyrosine Kinase ROR1 as a Target for Anti-Cancer Therapies, Zhao [/bib_ref]. Despite its important role in embryogenesis, ROR1 is not expressed in the majority of normal adult tissues and cells, such as B cells, whereas it is overexpressed in many cancer cell types, including CLL and DLBCL-type RS [bib_ref] Gene expression profiling of B cell chronic lymphocytic leukemia reveals a homogeneous..., Klein [/bib_ref] [bib_ref] Unique cell surface expression of receptor tyrosine kinase ROR1 in human B-cell..., Baskar [/bib_ref]. The effect of the antibody drug conjugate (ADC) VLS-101, consisting of a mAb targeting ROR1 linked to the antimicrotubular agent MMAE, has been assessed in a xenograft model [bib_ref] ROR1 targeting with the antibody-drug conjugate VLS-101 is effective in Richter syndrome..., Vaisitti [/bib_ref]. When VLS-101 binds to ROR1 on DLBCL-type RS cells, the complex is internalized by the cell and included in lysosomes, where, via proteolytic cleavage, MMAE is released and can fulfill its cytotoxic function, blocking cell cycle progression and inducing apoptosis in the neoplastic cell [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] ROR1 targeting with the antibody-drug conjugate VLS-101 is effective in Richter syndrome..., Vaisitti [/bib_ref]. This cellular effect translates into reduction of tumor burden and into prolonged mice survival, with high selectivity and no toxicities recorded in treated mice [bib_ref] ROR1 targeting with the antibody-drug conjugate VLS-101 is effective in Richter syndrome..., Vaisitti [/bib_ref]. Based on these encouraging findings, a phase I clinical trial testing VLS-101 in hematological malignancies, including RS, has been recently started (NCT03833180). Other ADCs are giving early promising results in PDX models, such as immunoconjugates linked to the cytotoxic payload amanitin and directed against CD37, an antigen expressed on the cellular membrane of DLBCL-type RS cells [bib_ref] Anti-CD37 Alpha-Amanitin Conjugated Antibodies As Therapeutic Weapons for Richter's Syndrome, Vaisitti [/bib_ref].
Due to the constitutive activation of the PI3K/Akt pathway in DLBCL-type RS, a targeted therapeutic strategy might be represented by PI3K inhibition [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref]. Although the PI3Kδ inhibitor idelalisib showed some clinical effect in DLBCL-type RS in a retrospective study, the toxicities linked to this drug represent a major obstacle to treatment, as extensively reported in CLL patients [bib_ref] BCR kinase inhibitors, idelalisib and ibrutinib, are active and effective in Richter..., Visentin [/bib_ref] [bib_ref] Revolution of Chronic Lymphocytic Leukemia Therapy: The Chemo-Free Treatment Paradigm, Scheffold [/bib_ref]. More recently, in RS-PDX models, the PI3Kδ/γ inhibitor duvelisib, combined with venetoclax, showed encouraging efficacy and selectivity, leading to an increased apoptosis rate in RS cells and a significant reduction of tumor volume coupled to a prolonged survival in treated mice [bib_ref] Synergistic efficacy of the dual PI3K-δ/γ inhibitor duvelisib with the Bcl-2 inhibitor..., Iannello [/bib_ref]. This synergistic effect is probably obtained through the cross-talk between PI3K and BCL2 pathways, thanks to the activation of GSK3β by the inhibition of Akt promoted by duvelisib, resulting in reduced levels of the anti-apoptotic protein Mcl-1, which, in turn, leads to sensitization to venetoclax [bib_ref] Synergistic efficacy of the dual PI3K-δ/γ inhibitor duvelisib with the Bcl-2 inhibitor..., Iannello [/bib_ref] [bib_ref] Targeted inhibition of PI3Kα/δ is synergistic with BCL-2 blockade in genetically defined..., Bojarczuk [/bib_ref]. These encouraging results led to the design of a phase I trial with duvelisib and the anti-PD-1 mAb nivolumab, which is still ongoing (NCT03892044). A potential target can also be represented by Akt, tested only in CLL with the allosteric inhibitor MK-2206, in combination with a BR (rituximab plus bendamustine) regimen [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] Akt inhibitor MK-2206 in combination with bendamustine and rituximab in relapsed or..., Larsen [/bib_ref]. MK-2206 obtained promising results in terms of safety and efficacy, but, due to the decision of the sponsor to stop its development, further studies with alternative Akt inhibitors deserve to be undertaken in RS.
Another possible target for RS therapy could be represented by exportin 1 (XPO1), a key mediator of the nuclear export of macromolecules [bib_ref] Exportin 1 (Crm1p) is an essential nuclear export factor, Stade [/bib_ref]. Recently, the XPO1 gene has been found to be frequently overexpressed or mutated in CLL and affected by gain-offunction point mutations in 25-35% of RS, resulting in an unbalance between pro-oncogenic and antioncogenic proteins (i.e., p53) [fig_ref] Figure 1: Molecular pathway alterations in DLBCL-type RS [/fig_ref] [bib_ref] Recurrent XPO1 mutations alter pathogenesis of chronic lymphocytic leukemia, Walker [/bib_ref] [bib_ref] Gain in the short arm of chromosome 2 (2p+) induces gene overexpression..., Cosson [/bib_ref]. Selinexor, a XPO1 inhibitor, was evaluated in a phase I study that reported an ORR of 33% in R/R DLBCL-type RS, with an acceptable safety profile [bib_ref] The Oral Selective Inhibitor of Nuclear Export (SINE) Selinexor (KPT-330) Demonstrates Broad..., Kuruvilla [/bib_ref]. Unfortunately, the phase II trial (NCT02138786) assessing selinexor monotherapy in RS was closed due to lack of efficacy, but strategies involving combination therapies with selinexor or other XPO1 inhibitors might be explored.
Since TP53 is commonly disrupted in RS, an intriguing potential therapeutic approach might include TP53-pathway-restoring drugs under development [bib_ref] Targeting p53 in chronic lymphocytic leukemia, Moia [/bib_ref]. APR-017 and APR-246 are first-in-class drugs, capable of binding to cysteine residues in the DNA binding domain of p53, restoring its physiological conformation in TP53 mutated in vitro models of CLL [bib_ref] Targeting p53 in chronic lymphocytic leukemia, Moia [/bib_ref] [bib_ref] Effects of PRIMA-1 on chronic lymphocytic leukaemia cells with and without hemizygous..., Nahi [/bib_ref]. In TP53-deleted CLL, evidence of the effectiveness of RG7388, an inhibitor of the downregulating p53 ligand MDM2 (an E3 ubiquitin ligase), has been documented in vitro [fig_ref] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS [/fig_ref] [bib_ref] Non-genotoxic MDM2 inhibition selectively induces a pro-apoptotic p53 gene signature in chronic..., Ciardullo [/bib_ref]. Because these compounds are not yet included in clinical trials for RS patients, further investigation into this therapeutic strategy is needed.
Lastly, the inhibition of cyclin-dependent kinase 4 and 6 (CDK4/6), involved in the positive regulation of the cell cycle, might represent a way to overcome CDKN2A/B deletion in RS. The CDK4/6 inhibitor palbociclib showed activity both in vitro and in vivo in preclinical models of RS, with a synergistic effect with BCR inhibitors [bib_ref] B-cell receptor signaling and genetic lesions in TP53 and CDKN2A/CDKN2B cooperate in..., Chakraborty [/bib_ref] [bib_ref] Combined Genetic Lesions in TP53 and CDKN2A/CDKN2B Drive B Cell Receptor-Dependent/Costimulatory Signal-Independent..., Chakraborty [/bib_ref]. Ongoing clinical trials in RS are depicted in [fig_ref] Table 2: Summary of ongoing trials in RS [/fig_ref] (https://clinicaltrials.gov/, last access on 3 August 2022). Abbreviations: VR-EPOCH, venetoclax + rituximab + etoposide + prednisone + vincristine + cyclophosphamide + doxorubicin; R-CHOP, rituximab + cyclophosphamide + doxorubicin + vincristine + prednisone; VR-CHOP, venetoclax + R-CHOP; R-EPCH, rituximab + etoposide + prednisone + cyclophosphamide + hydroxydaunorubicin; NAE, NEDD8-activating enzyme; BTK, Bruton tyrosine kinase; PI3K, phosphatidylinositol-3 kinase; PD-1, programmed death 1; PD-L1, programmed death ligand 1; CBL-B, Casitas B-lineage lymphoma.
# Conclusions
In clinical practice, RS still represents a major unmet medical need, and most patients die because of a lack of dedicated treatments and/or because of frailty due to previous exposure to multiple lines of ineffective therapy. The molecular pathogenesis of RS, in particular of DLBCL-type RS, has been elucidated to a certain extent, revealing druggable vulnerabilities both in the tumor clone and in the RS microenvironment. The availability of pathway inhibitors, namely BCR and BCL2 inhibitors, coupled with the development of innovative mAbs with different modes of action, is substantially changing the therapeutic landscape of RS. These novel medicines are currently being tested in clinical trials with or without chemoimmunotherapy, and a few promising results are emerging. The development of PDX models of RS may represent a further tool for testing novel drug combinations that, if successful in these pre-clinical models, may then be exported to clinical trials in RS patients. A precision medicine approach to RS may thus be considered for the future and will require a detailed characterization of the clonal relationship, of the genetic profile and of the expression pattern of RS cells in individual patients.
[fig] Figure 1: Molecular pathway alterations in DLBCL-type RS. [/fig]
[fig] Figure 2: Drugs under development for the targeted therapy of DLBCL-type RS. The figure represents the main drugs and targets under development for the molecular treatment of DLBCL-type RS. The agents depicted here can be divided into small molecules and monoclonal antibodies (mAb). Small molecules target BTK, PI3K, Akt, CDK4/6, BCL2, XPO-1, MDM2 and p53. mAb include immune checkpoint inhibitors and antibody drug conjugates (ADC). The anti-ROR1 ADC VLS-101 interrupts the microtubule polymerization through the release of its payload monomethyl auristatin E (MMAE) in the cytoplasm of the RS cell. APC, antigen presenting cell; BCR, B cell receptor; BTK, Bruton tyrosine kinase; PI3K, phosphatidylinositol-3 kinase; XPO1, exportin 1; CDK4/6, cyclin-dependent kinase 4 and 6; ROR1, receptor tyrosine kinase-like orphan receptor 1. Image created with Biorender.com (accessed on 4 August 2022). [/fig]
[fig] Author: Contributions: S.M. and G.G. have contributed equally to the conception and drafting of the manuscript and have read and agreed to the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.Funding: Work by the authors described in this review has been funded by Molecular bases of disease dissemination in lymphoid malignancies to optimize curative therapeutic strategies, (5 × 1000 No. 21198), Associazione Italiana per la Ricerca sul Cancro Foundation Milan, Italy; Progetti di Rilevante Interesse Nazionale (PRIN; 2015ZMRFEA), Rome, Italy; the AGING Project-Department of Excellence-DIMET, Università del Piemonte Orientale, Novara, Italy; and Ricerca Finalizzata 2018 (project RF-2018-12365790), MoH, Rome, Italy; and AIL Novara Onlus, Novara, Italy. Conflicts of Interest: S.M. declares no conflict of interest for this specific work. G.G. declares advisory board and speaker's bureau honoraria from Abbvie, Astra-Zeneca, BeiGene, Incyte, Janssen and Roche. [/fig]
[table] Table 1: RS frequency in patients treated with therapeutic regimens based on pathway inhibitors. [/table]
[table] Table 2: Summary of ongoing trials in RS. [/table]
|
Sentinel Listeriosis Surveillance in Selected Hospitals, China, 2013–2017
During 2013-2017, a total of 211 cases of listeriosis were reported by 64 sentinel hospitals in China to a national foodborne disease surveillance network. The average case-fatality rate was 31.2% for perinatal cases and 16.4% for nonperinatal cases. Sequence types 87 and 8 were the most prevalent types.
L isteriosis is caused by the gram-positive bacterium Listeria monocytogenes, which is ubiquitous in the environment and a foodborne pathogen of importance to public health. Listeriosis occurs sporadically and mainly in highrisk groups, such as pregnant women, neonates, and immunocompromised and elderly persons. Although listeriosis occurs rarely in humans, it has a high case-fatality rate of 20%-50%. Nearly all reported listeriosis cases are transmitted to humans via food, and L. monocytogenes can grow at refrigeration temperatures, which makes it particularly challenging to control.
In China, surveillance of L. monocytogenes in food products was launched in 2000; however, as yet, listeriosis is not a notifiable disease in China. The National Foodborne Disease Surveillance Plan was implemented in 2011. Human listeriosis surveillance was included as a special pilot project in 2013. We provide an overview of the listeriosis sentinel surveillance data for the period 2013-2017. We summarize the demographic and clinical characteristics of patients with listeriosis and analyze the prevalent sequence types (STs) of all identified isolates.
## The study
In 2013, listeriosis surveillance started in 6 selected provinces in China. The target was to detect whether human listeriosis existed in China and to determine illness and death rates for listeriosis. In 2017, this pilot surveillance had expanded to 12 provinces with the additional objectives to investigate high-risk factors and detect potential outbreaks . A total of 78 sentinel hospitals were selected using convenient sampling: 40 general hospitals, 28 maternity hospitals, and 10 children's hospitals.
We defined invasive listeriosis as the isolation of L. monocytogenes from a normally sterile site (e.g., blood or cerebrospinal fluid) or products of conception (e.g., placental or fetal tissue). Pregnancy-associated patients were considered perinatal case-patients, including pregnant women, fetuses, or infants <28 days of age; maternal-fetal infections were counted as a single case. We defined stillbirths and miscarriages as deaths, which were tallied in case-fatality rates. Pregnant women and neonates were the focused population groups, with immunocompromised and older adults also included. All demographic data, clinical manifestations, and laboratory tests were submitted to the China National Center for Food Safety Risk Assessment (CFSA) through the National Foodborne Disease Reporting System. All confirmed isolates were finally referred to CFSA for pulsed-field gel electrophoresis and wholegenome sequencing analysis through the National Molecular Tracing Network for Foodborne Disease Surveillance (TraNet).
During 2013-2017, a total of 211 listeriosis cases were diagnosed and reported by 64 sentinel hospitals, 138 (65.4%) perinatal cases and 73 (34.6%) nonperinatal cases. All case-patients were hospitalized; 55 deaths or fetal losses (case-fatality rate 26.1%) were reported, and 43 (78.2%) fatal cases occurred among fetuses and neonates. The average case-fatality rates were 31.2% (43/138) for perinatal and 16.4% (12/73) for nonperinatal cases. No maternal death was reported. Seventy-four (35.1%) case-patients acquired listeriosis in the summer (June-August).
Of the 138 perinatal infections, the median age of the mother was 29 years (range 20-41 years), and the median gestational age was 32 weeks (range 8-40 weeks). Preterm labor (<37 weeks gestational age) was reported in 63 (45.7%) pregnant women with listeriosis. Clinical signs in pregnant women included intrauterine
## Sentinel listeriosis surveillance in selected hospitals, china, 2013-2017
infection, abortion, preterm labor, and influenza-like symptoms. Clinical manifestations and outcomes of infection in neonates included neonatal sepsis, asphyxia, pneumonia, meningitis, aspiration of amniotic fluid, meconium syndrome, and death.
Of the 73 nonperinatal infections, 45 (61.6%) cases were bloodstream infections such as septicemia and bacteremia, 20 (27.4%) were central nervous system infections, 6 (8.2%) were acute gastroenteritis, and 2 (2.7%) were focal infections. The median age of nonperinatal casepatients was 53 years (range 2 months-102 years); 22.9% were >65 years of age. The sex ratio was 1:1. Fifty-seven (78.1%) patients had positive blood samples, 11 (15.1%) had positive cerebrospinal fluid, and 15 (6.9%) were positive in other specimens, such as pleural effusion, cystic liquid, bone marrow, and feces. The all-cause immunosuppression rate was 28.8% (21/73 cases). We detected the following underlying immunosuppression conditions: hematologic malignancy, systemic lupus erythematosus, chronic obstructive pulmonary disease, chronic kidney disease, liver disease, organ tumor, lung transplantation, and tuberculosis.
Of the reported listeriosis cases, 28.9% (61/211) were followed up with epidemiologic investigation, and 18.0% (11/61) yielded positive results for L. monocytogenes in suspicious food, chopping boards, refrigerators, or kitchen sinks. However, the pulsed-field gel electrophoresis patterns were not identical to those of clinical isolates, and >100 allele differences were found by using the core genome multilocus sequence typing (MLST) profile of 1,748 loci (9). These results showed no links between food, environmental, and clinical isolates. A total of 116 isolates isolated during 2013-2016 were submitted to CFSA for whole-genome sequencing analysis: 108 from human listeriosis and 8 from the environment and suspicious food. The distribution of these 108 clinical L. monocytogenes clones was determined by MLST. A previous study reported that clonal complex (CC) 1, CC2, CC121, and CC155 were frequent clones in eastern Asia. We found that sequence type (ST) 87 (lineage I) and ST8 (lineage II) were the predominant STs; 15.7% of isolates were ST87 and 13.9% were ST8. The prevalences of ST87 in clinical isolatesand in domestic food products were also reported previously. ST87 was seldom linked to human listeriosis in other countries; only 2 outbreaks (both in Spain) were associated with ST87 strains. The most common PCR serogroups were IIb and IIa. A total of 89 different core genome MLST types were identified as groups that differ by up to 7 allelic mismatches among the clinical isolates.
# Conclusions
Our study describes epidemiologic characteristics of listeriosis from sentinel surveillance in China. An estimated 1,662 cases of listeriosis occur each year in the United States (3); a detailed analysis should be expedited in China to estimate incidence. The Universal Two-Child Policy was proposed and passed in 2015, which likely will increase the number of pregnancies and births in China and might therefore increase the incidence of listeriosis.
This study has limitations (1). All cases came from sentinel hospitals but were not a complete picture of listeriosis occurrence because of the gradual increase of provinces included in surveillance (from 6 to 12 provinces), which meant the population served by selected hospitals could not be estimated accurately. All case-patients might be the most ill patients; cases might have been missed because those patients with milder illness might not go to the hospital and therefore will not be reflected in the data. The number of perinatal cases was nearly twice the number of nonperinatal cases, which cannot represent the actual illness and death rates because perinatal infection is given more attention in some sentinel hospitals. The case-fatality rates might be underestimated because all live-born infants, premature infants, and case-patients who did not complete follow-up surveillance were assumed to survive unless they were reported to have died.
In summary, health education and reasonable diet advice regarding listeriosis prevention should be provided to high-risk groups in China, and a focus on L. monocytogenes infection should be strengthened in hospitals. Moreover, L. monocytogenes is common in domestic food products in mainland China, especially in meat, poultry, seafood, and Chinese salad. An urgent need exists for improving surveillance of food and humans, exploring the mechanisms of pathogenesis, determining higher-risk foods, detecting potential outbreaks, and implementing control measures to protect vulnerable populations. |
Circadian rhythms of hemostatic factors in tetraplegia: a double-blind, randomized, placebo-controlled cross-over study of melatonin
Study design: This is a double-blind, randomized, placebo-controlled cross-over study of melatonin in complete tetraplegia. Objectives: Tetraplegic patients have an increased risk of venous thrombosis despite prophylaxis, blunted variations in melatonin and altered circadian variation of several hemostatic markers. To examine whether melatonin could modify the regulation of hemostasis, we measured plasma melatonin and several markers of hemostasis in tetraplegic subjects with or without melatonin supplement. Setting: The study was conducted in the Section for Spinal Cord Injury, Sunnaas Hospital, Nesoddtangen, Norway. Methods: Six subjects with long-standing complete tetraplegia were included in this cross-over study with 2 mg of melatonin or placebo given 4 days before sampling. We also included six able-bodied men without any intervention. Plasma samples were then collected frequently during a 24-h awake/sleep cycle. The plasma concentrations of melatonin and the various markers were analyzed using linear mixed models.Results:The 24-h profiles of prothrombin fragment 1+2 and von Willebrand factor, but not D-dimer, activated FVII, tissue factor pathway inhibitor and plasminogen activator inhibitor type 1, differed (Po0.05) between tetraplegic patients and able-bodied subjects. The absolute plasma concentration of activated FVII was higher (Po0.05) among the able-bodied compared with the tetraplegic groups. Supplementation of melatonin had no impact on these findings. Conclusions: We found differences in circadian variation of several hemostatic markers between able-bodied and tetraplegics. These differences were apparently unrelated to fluctuations in the melatonin concentrations, suggesting little or no role of melatonin in the regulation of hemostasis in tetraplegia. Sponsorship: Financial support was provided from the Throne Holst Foundation.
# Introduction
Spinal cord injury (SCI) results in a reduction or lack of autonomic control that is directly related to the level of injury, which affects most of the body organs including the cardiovascular system, 1 giving rise to acute cardiovascular complications, 2 as well as hypotension and altered vascular regulation. Patients with SCI currently have a life expectancy that is approximately the same as able-bodied subjects, and they are therefore susceptible to the same chronic conditions across the life span. In fact, cardiovascular disease is one of the leading causes of mortality in both able-bodied subjects and in patients with SCI. [bib_ref] Mortality after spinal cord injury in Norway, Lidal [/bib_ref] Whereas in able-bodied subjects there is an increased cardiovascular disease risk in the few hours after waking up in the morning, postulated to be owing to the early-morning surge in blood pressure, this is apparently not the case with SCI patients. [bib_ref] Diurnal blood pressure variation in quadriplegic chronic spinal cord injury patients, Krum [/bib_ref] Moreover, in ablebodied subjects, the incidence of both venous and arterial thrombotic disorders seems to peak in the morning. [bib_ref] Circadian variations within the hemostatic and fibrinolytic systems with emphasis on venous..., Iversen [/bib_ref] [bib_ref] Circadian variation within hemostasis: an underrecognized link between biology and disease?, Montagnana [/bib_ref] In line with this, the plasma concentrations of several hemostatic factors vary according to specific circadian patterns, including certain coagulation factors and markers of coagulation activation, the coagulation inhibitors protein C, protein S and tissue factor pathway inhibitor (TFPI), as well as the fibrinolytic inhibitor plasminogen activator inhibitor type 1 (PAI-1). [bib_ref] Impaired circadian variations of haemostatic and fibrinolytic parameters in tetraplegia, Iversen [/bib_ref] [bib_ref] Morning hypercoagulability and hypofibrinolysis. Diurnal variations in circulating activated factor VII, prothrombin..., Kapiotis [/bib_ref] [bib_ref] Circadian variations in natural coagulation inhibitors protein C, protein S and antithrombin..., Undar [/bib_ref] The biological mechanisms causing these circadian variations are poorly understood. Experiments in transgenic mice and in vitro studies demonstrated that variations of coagulation factor VII were driven at the transcriptional level through the recruitment of circadian transcription factors. [bib_ref] Daily and circadian rhythms of tissue factor pathway inhibitor and factor VII..., Pinotti [/bib_ref] [bib_ref] Regulation of the PAI-1 promoter by circadian clock components: differential activation by..., Schoenhard [/bib_ref] Furthermore, deletion or mutation of circadian transcription factors in mice changed the time to thrombotic vascular occlusion. [bib_ref] Genetic components of the circadian clock regulate thrombogenesis in vivo, Westgate [/bib_ref] In addition, circadian variations of mRNA expression of coagulation and fibrinolytic factors have been demonstrated in several murine organs. [bib_ref] Circadian mRNA expression of coagulation and fibrinolytic factors is organ-dependently disrupted in..., Oishi [/bib_ref] These studies indicate that blood coagulation is influenced by endogenous biological clocks and daylight.
Despite recommendation-based preventive use of antithrombotic medication, the incidence of venous thrombosis (VT) is unacceptably high in patients suffering from paraplegia or tetraplegia, with about 1 in 4 being affected. [bib_ref] Heparin for venous thromboembolism prophylaxis in patients with acute spinal cord injury:..., Chen [/bib_ref] [bib_ref] The short-and long-term risk of venous thromboembolism in patients with acute spinal..., Giorgi [/bib_ref] It is therefore of interest that in subjects with tetraplegia the circadian rhythm of several hemostatic factors varies either more (for example, PAI-1) or less (for example, D-dimer and TFPI) than in able-bodied subjects. [bib_ref] Impaired circadian variations of haemostatic and fibrinolytic parameters in tetraplegia, Iversen [/bib_ref] Moreover, the mean plasma concentrations differ between able-bodied and tetraplegic subjects (for example, elevated fibrinogen and PAI-1 and reduced von Willebrand factor plasma levels in tetraplegia). [bib_ref] Impaired circadian variations of haemostatic and fibrinolytic parameters in tetraplegia, Iversen [/bib_ref] Melatonin is an important signaling molecule for circadian rhythms of many biological processes, but its circadian variation is almost totally blunted in tetraplegic subjects. [bib_ref] Absence of detectable melatonin and preservation of cortisol and thyrotropin rhythms in..., Zeitzer [/bib_ref] Notably, in able-bodied subjects, we previously found that the plasma concentrations of free TFPI and melatonin varied with seemingly opposite rhythms: when plasma melatonin reached its nadir values, plasma TFPI free antigen peaked and vice versa. [bib_ref] Opposite circadian rhythms in melatonin and tissue factor pathway inhibitor type 1:..., Dahm [/bib_ref] In tetraplegia, however, the plasma concentrations of neither free TFPI nor melatonin displayed any circadian rhythm. [bib_ref] Opposite circadian rhythms in melatonin and tissue factor pathway inhibitor type 1:..., Dahm [/bib_ref] We next showed that melatonin stimulated vascular endothelial cells grown in vitro to selectively secrete TFPI without altering transcription of the TFPI gene. [bib_ref] Melatonin stimulates release of tissue factor pathway inhibitor from the vascular endothelium, Kostovski [/bib_ref] If melatonin increases TFPI release in a similar manner in vivo as in vitro, this could have potential clinical implications in both prophylaxis and treatment of thromboembolic events. Hence, we have now conducted a double-blind, randomized placebo-controlled cross-over study and measured various hemostatic factors in blood sampled repeatedly throughout a 24-h cycle in tetraplegic subjects with or without supplementation of melatonin.
# Materials and methods
## Subjects and design of study
The study was approved by the Regional Committee for Medical Health Research Ethics in Norway (no. 2010/295) and the Norwegian Medicines Agency (EUDRACT no. 2010-021212-24), and it is registered with Clinicaltrials.gov unique identifier NCT 01741389.
The study was designed as a double-blind, randomized cross-over study of six tetraplegic men conducted at Sunnaas Hospital in December 2012. We also included a control group of six able-bodied men. Hence, we studied three groups: tetraplegic men with placebo, tetraplegic men with melatonin and ablebodied men [fig_ref] Table 1: Characteristics of the study subjects [/fig_ref]. The intervention in the tetraplegic group was capsules with placebo (Kragerø Tablettproduksjon AS, Kragerø, Norway) or 2 mg melatonin (Circadin; Neurim Pharmaceuticals, Zug, Switzerland). The ablebodied group received no intervention. The tetraplegic men were randomized to receive melatonin or placebo (daily at 22:00 h) for 4 days before they were subjected to a 24-h period of blood sampling. Randomization was performed by a person not involved in the study and using the software program www.randomization.com. and using sealed envelopes. Blood was drawn 10 times during a 24-h cycle (at 7, 12, 16, 18, 20, 22, 24, 2, 4 and 7 h). Then, the tetraplegic men were crossed over, and thus the three tetraplegic men who first received melatonin received placebo for 4 days and the three tetraplegic men who first received placebo received melatonin for 4 days before both groups went through a new period of 24-h sampling. The values measured at 07:00 h on day 1 did not differ significantly from those measured at 07:00 h on day 2. The 'wash-out' period lasted for 4 days before the cross-over, which is assumed to be sufficient because the half-life of melatonin is about 35-50 min, thus ensuring minimal, if any, carry-over effect. The able-bodied men were subjected to a similar 24-h period of blood sampling without any intervention and with normal sleep. All the participants received standardized meals at regular time points. The only restriction was no alcohol intake, and a maximum of two cups of coffee were required by the participants during the study period. They slept only between 23:00 and 06:00 h (daylight from about 09:00 till 15:00 h).
## Blood sampling
Venous blood samples were collected in 5-ml Vacutainer vacuum tubes containing 0.5 ml of buffered sodium citrate (0.129 M) (Becton-Dickinson, Plymouth, UK), and in 4.9-ml Monovette tubes containing potassiumethylenediaminetetraacetic acid (Sarstedt AG, Nümbrecht, Germany). Citrated blood was kept at room temperature and immediately centrifuged at 2000 g for 15 min. Platelet-poor plasma aliquots and EDTA-blood were stored at − 70°C until they were assayed.
## Assays
Activated factor VII (factor VIIa), free TFPI antigen, von Willbrand factor (VWF) and D-dimer were analyzed using commercial ELISA assays (Staclot VIIa-rTF, Asserachrom Free TFPI, AsserachromVWF:Ag and Asserachrom D-DI) from Diagnostica Stago (Asnières, France), PAI-1 activity was analyzed with the Zymutest PAI-1 activity kit from Aniara Corporation (West Chester, OH, USA) and prothrombin fragment 1+2 (F 1+2 ) was analyzed with a kit from Siemens Healthcare (Erlangen, Germany). The lowest detection values are 6 mU ml − 1 for FVIIa, 3 ng ml − 1 for free TFPI, 5% for VWF, 60 ng ml − 1 for D-dimer, 0.6 ng ml − 1 for PAI-1 activity and 20 pmol ml − 1 for F 1+2 . The intra-assay coefficients of variation were o5.2% for all the assays, whereas the inter-assay coefficients of variation were between 2.9 and 11.2%.
Melatonin was first extracted from the blood plasma samples using a vacuum manifold procedure, as described in the Direct Saliva Melatonin ELISA kit obtained from Buhlmann Laboratories AG (Basel, Switzerland). The extracted melatonin concentrations were then quantified using the Buhlmann Laboratories ELISA kit, which was based on a melatonin biotin conjugate, an enzyme label and a tetramethyl benzidine substrate. The product of the substrate was measured spectrophotometrically at 450 nm. The assay sensitivity range was 1-60.6 pg ml − 1 .
All analyses were performed examiner-blind, and the samples were run in-batch using a balanced setup with equal number of cases and controls in each run.
## Statistics
Formal power calculation was considered impossible owing to lack of relevant information. We reasoned that the tetraplegic subjects with and without melatonin supplementation would differ with respect to the circadian variation of the studied hemostatic markers. The statistical analyses were performed with SPSS version 18.0 (IBM Corporation, Chicago, IL, USA) and the MedCalc Software (Ostend, Belgium). We considered P-values ⩽ 0.05 to indicate statistical significance. Values are given both as mean absolute plasma concentrations with s.e.m. and as mean percent change (s.e.m.) from the starting point at 07:00 h on day 1. Circadian variations, as well as differences in the absolute plasma concentrations of the various coagulation parameters between the three study groups, were evaluated with linear mixed models. [fig_ref] Figure 1: Plasma concentrations of melatonin [/fig_ref] shows that, as expected, the plasma concentrations of melatonin remained quite low and with virtually no significant circadian pattern among the tetraplegic subjects who received placebo. When they were given melatonin, a circadian pattern of plasma melatonin appeared that resembled that of the reference group of ablebodied subjects, although with a peak concentration considerably above that for the able-bodied subjects, probably as a result of the 4-day treatment with melatonin before the blood-sampling period.
# Results
## Hemostatic factors and melatonin in tetraplegia e kostovski et al
The tetraplegic subjects reported unchanged sleeping duration and sleeping quality irrespective of whether they received placebo or melatonin.
Having established that supplementation of melatonin in tetraplegia could mimic its endogeneous circadian variation, we next examined the 24-h plasma profiles of various hemostatic markers. The circadian patterns of the F 1+2 , a marker of thrombin generation, and D-dimer, a marker of activated coagulation and fibrinolysis, were apparently similar (P40.05) between the tetraplegic groups, that is, irrespective of whether they were given melatonin [fig_ref] Figure 2: Plasma concentrations of markers of coagulation [/fig_ref]. Compared with the able-bodied group, the tetraplegic groups with or without melatonin supplementation showed an apparent increase in the circadian variation of F 1+2 (P = 0.01 and Po0.001, respectively), whereas the circadian pattern for D-dimer was similar (P40.05) across the three study groups. The absolute values for the plasma concentrations of F 1+2 and D-dimer [fig_ref] Figure 2: Plasma concentrations of markers of coagulation [/fig_ref] were similar (P40.05) in all three study groups.
For FVIIa, no significant changes were observed for the circadian variation [fig_ref] Figure 2: Plasma concentrations of markers of coagulation [/fig_ref] among the three study groups. The absolute plasma concentration was higher among the able-bodied subjects compared with the tetraplegic groups with (P = 0.050) or without (P = 0.039) melatonin [fig_ref] Figure 2: Plasma concentrations of markers of coagulation [/fig_ref]. With regard to VWF, which is a marker of endothelial function, the able-bodied subjects had a different circadian variation characterized by decreasing (Po0.001) VWF concentrations during the night [fig_ref] Figure 3: Plasma concentrations of various hemostatic markers [/fig_ref]. The absolute plasma concentration for VWF was, however, not different (P40.05) in the able-bodied reference group compared with tetraplegic subjects with or without supplementation of melatonin [fig_ref] Figure 3: Plasma concentrations of various hemostatic markers [/fig_ref].
The plasma concentrations of free TFPI, the only endogenous inhibitor of tissue factor induced coagulation, did not differ significantly with regard to circadian variations [fig_ref] Figure 3: Plasma concentrations of various hemostatic markers [/fig_ref] or absolute plasma concentration [fig_ref] Figure 3: Plasma concentrations of various hemostatic markers [/fig_ref] between the able-bodied and the tetraplegic subjects with or without supplementation of melatonin. Similarly, the plasma concentrations of PAI-1, a fibrinolytic inhibitor,
# Discussion
We here show that melatonin, but not placebo, supplementation for 4 days to patients with stable, complete tetraplegia could nearly restore the 24-h profile of this major biological clock regulator. However, melatonin supplementation did not induce any major alterations in the circadian variation of a wide range of hemostatic markers when their plasma concentrations were compared with those in tetraplegic patients given placebo. Yet, we did demonstrate differences in the circadian variation of several hemostatic markers between tetraplegic and able-bodied subjects, suggesting that these differences might be unrelated to fluctuations in the melatonin concentrations.
The absolute peak concentrations of melatonin were considerably higher among the tetraplegic patients receiving melatonin compared with the endogeneous concentrations in the able-bodied reference subjects. Importantly, these concentrations apparently did not induce any adverse effects among the study patients, and the dose used (2 mg) is the one recommended for treating sleeping disorders, the most common indication for melatonin supplementation.
Even with recommended prophylaxis, for example, the use of lowmolecular-weight heparins, patients suffering from SC are prone to VT. In line with this, over a median follow-up period of 36 months, Giorgi et al. [bib_ref] The short-and long-term risk of venous thromboembolism in patients with acute spinal..., Giorgi [/bib_ref] recently found that VT was diagnosed in 23.4% SCI patients despite thromboprophylaxis. Similarly, Chung et al. [bib_ref] Increased risk of deep vein thrombosis and pulmonary thromboembolism in patients with..., Chung [/bib_ref] reported adjusted hazard ratios of 2.46 and 1.57, respectively, for the risk of deep vein thrombosis and pulmonary embolism in SCI patients compared with the general population in a nationwide survey in Taiwan. Moreover, the risk for VT is highest in the acute phase following the SCI, suggesting that immobilization alone cannot explain the increased risk of venous thromboembolism. Collectively, the time-dependent decline in the risk of VT and the inadequate effect of current antithrombotic prophylaxis emphasize the need for a better understanding of the regulation of hemostasis in SCI patients.
The altered circadian variations in the concentrations of several hemostatic markers coupled with a dysregulated 24-h profile of melatonin concentrations in tetraplegia and a putative regulatory role of melatonin on vascular endothelial secretion of free TFPI in vitro 5,17,18 prompted us to examine the effect of melatonin supplementation on hemostatic markers in these patients. Our present findings indicate similar effects regarding the absolute plasma concentrations of F 1+2 and D-dimer, as well both the 24-h profiles and absolute plasma concentrations of D-dimer, TFPI and PAI-1. On the other hand, the 24-h profile of the concentrations of F 1+2 and VWF were different among the able-bodied subjects compared with those in the tetraplegic patients; however, melatonin supplementation had no apparent effect on the circadian pattern of F 1+2 and VWF. With regard to FVIIa, the able-bodied subjects had apparently higher absolute plasma concentrations than both the tetraplegic groups, but similar 24-h profiles.
We cannot exclude the possibility that our repeated blood sampling may have been a stressor on the hemostatic factors. Saliva sampling might have overcome this possible confounder; however, this sampling technique was not available to us. Despite the low number of participants, our study was robustly designed and the study subjects were carefully monitored under standardized conditions during the 24-h study periods. Moreover, the tetraplegic group was rather uniform, as only men were included and all had a complete and stable, long-standing injury (43 years). The tetraplegic and able-bodied subjects were well matched with regard to gender, age and body mass index. In addition, the level of injury, a factor that may affect the risk of venous thromboembolism, showed little variation (Cervical 5-Cervical 8). [bib_ref] Risk of venous thromboembolism after spinal cord injury: not all levels are..., Maung [/bib_ref] In conclusion, it appears that there are differences in the circadian variations of the coagulation system between tetraplegic and ablebodied subjects, as illustrated by the 24-h profiles of F 1+2 , FVIIa and VWF; however, they are probably unrelated to melatonin. To gain further insight into the regulation of hemostasis in SCI, detailed studies of clotting formation and fibrin production should probably be conducted.
## Data archiving
There were no data to deposit.
[fig] Figure 1: Plasma concentrations of melatonin. Values (mean+s.e.m.) are expressed as percentages (a) or as absolute plasma concentrations (b). Data are from tetraplegic subjects receiving either placebo (blue line) or melatonin (red line), and from a reference group of healthy able-bodied subjects (green line). The lines are somewhat mutually offset for illustration purposes. [/fig]
[fig] Figure 2: Plasma concentrations of markers of coagulation. Values (mean+s.e.m.) are expressed as percentages (a, c, e) or as absolute plasma concentrations (b, d, f). The panels show data for F 1+2 (a, b), D-dimer (c, d) and FVIIa (e, f). Data are from tetraplegic subjects receiving either placebo (blue line) or melatonin (red line), and from a reference group of healthy able-bodied subjects (green line). The lines are somewhat mutually offset for illustration purposes. [/fig]
[fig] Figure 3: Plasma concentrations of various hemostatic markers. Values (mean+s.e.m.) are expressed as percentages (a, c, e) or as absolute plasma concentrations (b, d, f). The panels show data for VWF (a, b), free TFPI (c, d) and PAI-1 (e, f). Data are from tetraplegic subjects receiving either placebo (blue line) or melatonin (red line), and from a reference group of healthy able-bodied (green line). The lines are somewhat mutually offset for illustration purposes.Hemostatic factors and melatonin in tetraplegia E Kostovski et al did not differ (P40.05) among the three study groups, with regard to circadian variations(e) or absolute plasma concentration(f). [/fig]
[table] Table 1: Characteristics of the study subjects [/table]
|
Electrocorticography reveals thalamic control of cortical dynamics following traumatic brain injury
The return of consciousness after traumatic brain injury (TBI) is associated with restoring complex cortical dynamics; however, it is unclear what interactions govern these complex dynamics. Here, we set out to uncover the mechanism underlying the return of consciousness by measuring local field potentials (LFP) using invasive electrophysiological recordings in patients recovering from TBI. We found that injury to the thalamus, and its efferent projections, on MRI were associated with repetitive and low complexity LFP signals from a highly structured phase space, resembling a low-dimensional ring attractor. But why do thalamic injuries in TBI patients result in a cortical attractor? We built a simplified thalamocortical model, which connotes that thalamic input facilitates the formation of cortical ensembles required for the return of cognitive function and the content of consciousness. These observations collectively support the view that thalamic input to the cortex enables rich cortical dynamics associated with consciousness.
F undamental mechanisms of consciousness are mysterious, despite significant recent theoretical and experimental progress. Thus, the return of consciousness following TBI is uncertain and hard to predict [bib_ref] Modeling the return to consciousness after severe traumatic brain injury at a..., Winans [/bib_ref] [bib_ref] Agitation following severe traumatic brain injury is a clinical sign of recovery..., Wang [/bib_ref]. While corticothalamic circuits are believed to be crucial, few direct observations of their function are available. Most studies of cortical signals have used electroencephalography (EEG) [bib_ref] Preserved feedforward but impaired top-down processes in the vegetative state, Boly [/bib_ref] [bib_ref] A theoretically based index of consciousness independent of sensory processing and behavior, Casali [/bib_ref] [bib_ref] Information sharing in the brain indexes consciousness in noncommunicative patients, King [/bib_ref] [bib_ref] Large scale screening of neural signatures of consciousness in patients in a..., Sitt [/bib_ref] , which is hampered by limited spatial and frequency resolution. However, the features of EEG associated with consciousness are now reasonably well-characterized; EEG in conscious subjects has greater spectral power in higher frequencies [bib_ref] Prognostic value of standard EEG in traumatic and nontraumatic disorders of consciousness..., Bagnato [/bib_ref] [bib_ref] Prognostically important EEG coma patterns in diffuse anoxic and traumatic encephalopathies in..., Synek [/bib_ref] , increased complexity [bib_ref] A theoretically based index of consciousness independent of sensory processing and behavior, Casali [/bib_ref] [bib_ref] Consciousness and complexity during unresponsiveness induced by propofol, xenon, and ketamine, Sarasso [/bib_ref] [bib_ref] Quantifying cortical EEG responses to TMS in (un) consciousness, Sarasso [/bib_ref] , and higher information content 5 relative to EEG recorded in unconscious subjects. However, how these features arise is not known with certainty due to a lack of direct observations from involved regions and a lack of lesion-type experiments needed to make causal inferences.
Studies of severe traumatic brain injury (sTBI) represent an opportunity to address both gaps and a clinical need. Injury to the cortex, subcortical white matter, and thalamus characterize sTBI [bib_ref] Lesion volume, injury severity, and thalamic integrity following head injury, Anderson [/bib_ref] [bib_ref] Traumatic axonal injury: the prognostic value of lesion load in corpus callosum,..., Moen [/bib_ref] , and serial MRI studies have reported extensive brain atrophy after sTBI [bib_ref] Brain atrophy in mild or moderate traumatic brain injury: a longitudinal quantitative..., Mackenzie [/bib_ref] [bib_ref] Magnetic resonance imaging evidence of progression of subacute brain atrophy in moderate..., Ng [/bib_ref] [bib_ref] Long-term global and regional brain volume changes following severe traumatic brain injury:..., Sidaros [/bib_ref]. Thalamic atrophy is strongly correlated with prolonged unconsciousness, leading to the question of how thalamic function facilitates consciousness [bib_ref] Neuropathological findings in the brain of Karen Ann Quinlan-the role of the..., Kinney [/bib_ref] [bib_ref] Thalamic atrophy in antero-medial and dorsal nuclei correlates with six-month outcome after..., Lutkenhoff [/bib_ref]. Moreover, the integrity of frontothalamic circuitry also correlates with the level of consciousness [bib_ref] Diffusion weighted imaging distinguishes the vegetative state from the minimally conscious state, Fernández-Espejo [/bib_ref] [bib_ref] Thalamo-frontal connectivity mediates top-down cognitive functions in disorders of consciousness, Monti [/bib_ref] , which highlights the critical need to understand the role of the thalamus in consciousness.
Early studies in brain slices and intact animals have led to the longstanding view that the firing mode of the thalamus correlates with the level of consciousness [bib_ref] Thalamic burst patterns in the naturally sleeping cat: a comparison between cortically..., Domich [/bib_ref] [bib_ref] Discharge rate and excitability of cortically projecting intralaminar thalamic neurons during waking..., Glenn [/bib_ref]. It has been argued that rhythmic oscillations in the somatosensory thalamus are not conducive to the transmission of complex information [bib_ref] Burst and tonic firing in thalamic cells of unanesthetized, behaving monkeys, Ramcharan [/bib_ref]. This observation led to the proposal that thalamic neurons control the cortical transition from sleep to wakeful states and facilitate information transmission [bib_ref] Bursting of thalamic neurons and states of vigilance, Llinás [/bib_ref]. Investigations in the 2010s have revealed functional significance to this transition; thalamic input has been shown to synchronize cortical oscillations facilitating cortico−cortical information transmission in the visual system [bib_ref] The pulvinar regulates information transmission between cortical areas based on attention demands, Saalmann [/bib_ref]. In the frontal lobe, the mediodorsal thalamus functions as a content-free amplifier of cortical representations that boosts synaptic gain and thus enables neuronal ensemble formation in the prefrontal cortex (PFC) [bib_ref] Thalamic regulation of switching between cortical representations enables cognitive flexibility, Rikhye [/bib_ref] [bib_ref] Thalamic amplification of cortical connectivity sustains attentional control, Schmitt [/bib_ref]. These data are consistent with a model in which the thalamus capacitates cortical representations needed for consciousness.
However, it is not clear whether the thalamus's key function in the recovery of brain injury is the maintenance of arousal (especially from central thalamic efferents [bib_ref] Recovery of consciousness after brain injury: a mesocircuit hypothesis, Schiff [/bib_ref] and/or facilitation of cortical function leading to complex behavior [bib_ref] Thalamic and extrathalamic mechanisms of consciousness after severe brain injury, Lutkenhoff [/bib_ref]. According to two recent reports, thalamic injury was not strongly associated with unconsciousness [bib_ref] Thalamic strokes that severely impair arousal extend into the brainstem, Hindman [/bib_ref] [bib_ref] Deep structural brain lesions associated with consciousness impairment early after hemorrhagic stroke, Rohaut [/bib_ref] , and there is no apparent impairment of arousal with pure thalamic injuries [bib_ref] Thalamic strokes that severely impair arousal extend into the brainstem, Hindman [/bib_ref]. However, stimulation of the central thalamic nuclei does appear to increase the level of consciousness in several reports, in both chronic [bib_ref] Behavioural improvements with thalamic stimulation after severe traumatic brain injury, Schiff [/bib_ref] and subacute 32 traumatic brain injury. Given the massive burden of sTBI worldwide, a detailed understanding of the contribution of thalamic activity to the recovery of consciousness is critical.
Here, we use direct recordings of local field potentials (LFP) from frontal areas implicated in consciousness to understand how neural signals associated with consciousness arise. We had a rare and unique clinical opportunity: we recorded and stimulated depth electrodes implanted in the PFC and anterior cingulate cortex (ACC) for seizure monitoring after sTBI. Due to the current limitations for direct recording from the thalamus in sTBI patients, we used in silico modeling to gain insight into the role of the thalamus in shaping the functional state of the cortex in the context of recovery of consciousness. Our model suggests that thalamocortical projections to the frontoparietal network (FPN) facilitate the complex dynamics needed for consciousness. By contrast, injury to these connections results in a dysfunctional state of cortical networks, incapable of maintaining neuronal ensembles required for consciousness.
# Results
Study population. Five sTBI patients were enrolled in the study, and we obtained combined LFP/EEG and recorded responses to single-pulse electrical stimulation from all five (see "Methods"). We also enrolled a single control patient with epilepsy. There were no surgical complications in the patients, and the study procedures were tolerated well. Ultimately, four of the five patients returned to consciousness over days to weeks [fig_ref] Table 1: Patient summaries [/fig_ref]. In one case, a patient was comatose at implant and almost fully conscious by the time the study procedures were conducted (Subject 4, CRS 10, [fig_ref] Table 1: Patient summaries [/fig_ref] ; this patient was explanted rapidly). We implanted a single depth electrode spanning ACC and PFC based on prior data supporting frontal depth electrodes in comatose patients for seizure monitoring [bib_ref] Intracortical electroencephalography in acute brain injury, Waziri [/bib_ref]. Complete patient histories are found in Supplementary Note 1. See for a detailed timeline of injury, implantation, and measures of consciousness.
Higher frequencies in LFP are associated with consciousness. We recorded depth LFP from PFC and ACC in all study subjects. shows these recordings in a recovering comatose patient (blue) and a patient who never recovered consciousness (red). Consistent with prior reports, we found that unconsciousness was associated with delta-band oscillations in ACC and PFC . The recovery of consciousness is associated with the development of faster theta, alpha, and beta frequencies , blue). We observed increases in theta-band oscillations in ACC , left) during the return of consciousness and beta-band oscillations in PFC .
Power spectra for all subjects over days of recordings are included in .
Complexity and variability of cortico−cortical evoked potentials are associated with consciousness. Consciousness has previously been associated with the complexity of the brain's response to perturbation reflective of its capacity for information integration (measured as the "perturbational complexity index"
[PCI] in prior reports) [bib_ref] A theoretically based index of consciousness independent of sensory processing and behavior, Casali [/bib_ref] [bib_ref] Stratification of unresponsive patients by an independently validated index of brain complexity, Casarotto [/bib_ref]. Thus, we applied single pulses of electrical stimulation and recorded cortico−cortical evoked potentials (CCEP; . In general, ACC stimulation resulted in a broader spatial response than PFC stimulation [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref] , suggesting that ACC has widespread connectivity to other cortical areas. Thus, most single-pulse stimulation in this study was administered to ACC. Patients with strictly cortical and/or white matter injuries exhibited characteristic multiphasic responses to stimulation with a high level of complexity reflected in increased zero crossings , left panel, P < 0.01, two-tailed t-Test) and Lempel−Ziv complexity , right panel, P < 0.01, two-tailed t-Test) as well as increased variability of the CCEP across trials. However, the most gravely injured patient with a bilateral thalamic injury had a stereotypical monophasic response , red line at bottom), which was broadly detectable throughout the cortex . These features were similar at longer timescales (tens of seconds). Single pulses of stimulation revealed a high degree of susceptibility to external perturbations in two patients with thalamic injuries. This observation was reproducible over days. The most severely injured patient with bilateral thalamic injury exhibited a high degree of susceptibility to single pulses of stimulation. In this patient (Subject 5) and the patient (Subject 4) with unilateral thalamic injury, trains of single pulses of stimulation produced highly deterministic cortical responses . However, trains of stimulation failed to produce deterministic responses in patients with strictly cortical injuries who quickly recovered consciousness. These observations suggest a modulatory role for the thalamic input to the cortex.
We note that one patient who rapidly recovered consciousness (Subject 4) exhibited features similar to the most gravely injured patient. Both patients had evidence of injury to the thalamus and/ or its efferent fibers; in the case of the patient who rapidly recovered, the injury was unilateral . These observations led us to assess the mechanistic role of the thalamus.
Thalamic injury leads to a low-dimensional, ring attractor-like state in vivo. Phase space analysis is a parsimonious representation of how complex systems change over time, especially when these systems are nonstationary or otherwise difficult to represent in analytic terms. Previous dynamical analysis of wakeful EEG has identified a transition between two attractors driven by thalamic input [bib_ref] Biophysical mechanisms of multistability in resting-state cortical rhythms, Freyer [/bib_ref]. We, therefore, hypothesized that a lack of thalamic input would constrain cortical dynamics during coma. To understand the underlying dynamics of coma, we reconstructed the phase space of PFC recordings using delay embedding. We used autocorrelation and false nearest neighbors (FNN) metrics to estimate the appropriate time lag and the embedding dimension [bib_ref] Determining embedding dimension for phase-space reconstruction using a geometrical construction, Kennel [/bib_ref].
We reconstructed the two-dimensional phase space of ten seconds of PFC dynamics (taken from five minutes of LFP recording) for all five patients [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref]. The phase space of cortical dynamics for patients with thalamic injuries is structured LFP activity distinguishes coma, wakefulness, and the transition from one state to another. a Left: a depth electrode is placed to provide direct recordings throughout PFC and ACC. The depth electrode location is shown on a T1-weighted coronal MRI. All electrodes were right-sided, except for Subject 2, whose electrode was on the left. Right: locations of electrodes from external and internal sagittal slices for all the patients. b, c Representative LFP recordings and their associated power spectral density recorded from PFC and ACC from the patient who recovered (Subject 1, CRS-R~2−10) versus the deeply comatose patient who did not recover (Subject 5, CRS-R~0-2). Note that the wakefulness exhibits spatially distinctive peaks in the ACC (theta and alpha) and PFC (delta and beta). d A patient who recovered consciousness over days of recordings (Subject 2) showed a gradual shift of the ACC theta peak towards higher frequencies and increased beta power in the PFC (black arrows).
with a small repertoire of available states that are periodically visited, resembling a limit cycle attractor in the phase space [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref] , top right panels, Subjects 4 and 5). Also, PCA on these recordings from depth contacts and scalp EEG contacts showed that fewer dimensions were required to represent their data . In contrast, patients without thalamic injuries required a larger number of dimensions to explain the variance in the data (Supplementary and no attractor-like structure was evident in the phase space [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref].
We examined the robustness of the attractor structure observed in the phase space by perturbation (single-pulse electrical stimulation at 1 Hz). [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref] shows how a ten-second train of single pulses of stimulation affects the temporal evolution of the visited states in phase space. The gray trajectories are associated with ten seconds of PFC LFP before stimulation. The red trajectory represents the ten-second period when single pulses of stimulation at 1 Hz were applied. In patients with thalamic injury [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref] , right panels), the system repetitively and predictably visits the same states (notice similar-appearing straight red lines). When stimulation was applied to the patients without thalamic injury, the response to stimulation was much less predictable.
To summarize, we found that thalamic injuries put the cortical networks in a dysfunctional state with simple, low variability and repetitive dynamics. This functionally passive regime has a limited number of allowed states with periodic trajectories between these allowed states, which resembles a low-dimensional attractor stable to external perturbation/ stimulation. We then sought to identify the mechanism underlying the attractor-like dynamics in patients with thalamic injuries.
Thalamic injury recapitulates delta-band, low-complexity LFP activity in silico. To evaluate how the thalamic input to FPN could explain the LFP features we observed in coma, we have built an abstract model of FPN structures in silico using a simplified network of leaky-integrate-and-fire (LIF) neurons [bib_ref] Measuring predictability of autonomous network transitions into bursting dynamics, Mofakham [/bib_ref] [bib_ref] Lapicque's introduction of the integrate-and-fire model neuron, Abbott [/bib_ref]. Although this abstract model lacks many of the biological details of the thalamo-cortical circuit, it still replicates many experimental observations and makes valuable predictions. This model includes two populations representing cortical cells: one in the frontal cortex and another in the parietal cortex. Each receives inputs from a thalamic population. Each group of cortical neurons is recurrently connected, and the two populations also communicate with each other via long-distance connections. Thalamic cells do not directly interact with each other but instead project to the prefrontal and parietal cortices . We posited two linked roles for the thalamus in this network: (1) driving: providing excitatory input to their post-synaptic neurons, and (2) Cortical evoked potentials exhibit distinctive features in patients with thalamic injury. a The evoked responses to ACC stimulation in a recovering comatose patient with no thalamic injury and a patient with a bilateral thalamic injury who never recovered consciousness are overlaid in blue and red, respectively. b Propagation of these evoked responses across scalp and depth channels are shown. c A representative scalp evoked response to ACC stimulation in each sTBI patient is shown. d Evoked responses across patients were classified based on their complexity using zero-crossing (left panel) and LZ complexity (right panel). Each circle represents the average of evoked recordings across trials (90−100 trials per patient) from each scalp contact. Horizontal lines represent the average across all scalp contacts. Evoked responses from patients with thalamic injury exhibited significantly diminished complexity with both measures (n = 3564 trials recorded from 18 scalp contacts in patients with thalamic injury and 5220 trials for patients without thalamic injury, P < 0.001, two-tailed t-Test). In this figure, "fast recovery" and "delayed recovery" refer to Subjects 2 and 3, who followed commands at 38 and 51 days, respectively. modulatory: adjusting the cortico−cortical connectivity in which thalamic activity amplifies the strength of connectivity across (β) and within (α) cortical networks [bib_ref] Thalamic regulation of switching between cortical representations enables cognitive flexibility, Rikhye [/bib_ref] [bib_ref] Thalamic amplification of cortical connectivity sustains attentional control, Schmitt [/bib_ref]. Here, we simulated coma as a state where the thalamic activity is reduced and phasic, similar to sleep 20 . In the awake state, the thalamus exhibits a tonically high firing rate in diverse spatiotemporal patterns [bib_ref] Discharge rate and excitability of cortically projecting intralaminar thalamic neurons during waking..., Glenn [/bib_ref].
When we reduced the thalamic activity in the model, we found that the cortical activity is also reduced, and the overall simulated cortical baseline LFP activity shifted towards the delta-band . Moreover, single-pulse electrical stimulation of the model in a coma state (reduced thalamic activity) resulted in simple, broad, and monophasic responses similar to what was experimentally observed when the thalamic input was missing .
Thalamic injury leads to a low-dimensional ring attractor-like state in silico. We reconstructed the phase space of the simulated coma and awake states using a similar approach as in [fig_ref] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor... [/fig_ref]. We used FNN and autocorrelation to estimate the appropriate delay and embedding dimension and plotted the state trajectories over time. The phase space was much more complex and random when thalamic input was fully present (awake state), , bottom row). When thalamic input was reduced (coma state), the phase space was simple and structured , bottom row), resembling a ring attractor, in which cortical networks are only allowed to visit a limited number of states repetitively. Together, these data support the view that the thalamic input allows the cortical networks to visit a larger repertoire of available states with a higher degree of freedom of transitioning from one to the other.
Thalamic injury impairs neuronal ensemble formation in silico. Recovery of consciousness requires both arousal and goaldirected behavior [bib_ref] The JFK Coma Recovery Scale-Revised: measurement characteristics and diagnostic utility, Giacino [/bib_ref]. Goal-directed behavior, in turn, requires activation of neural ensembles encoding specific behaviors [bib_ref] The organization of behavior: a neuropsychological theory, Hebb [/bib_ref]. With our in silico model, we set out to understand how activation of these neural ensembles is impaired in coma. We embedded a neuronal ensemble in the model by strengthening synaptic weights among a subpopulation of PFC neurons to form a clique of neurons that could support a thought, memory, or another functional subcomponent of behavior. When there is no thalamic amplification of cortical connections (as in the simulated coma state), the activity of such a cluster of neurons follows the ongoing cortical "up" and "down" states similar to the rest of the neurons that are not in the cluster , right raster plot). However, in the awake state, an embedded pattern reemerges due to the thalamic strengthening of neuronal functional connectivity. This pattern exhibits distinct and stable firing patterns that differentiate this neural subpopulation from the rest of the network . These data are consistent with the hypothesis that thalamic input needs to increase excitability and both local and long-range connectivity to optimize cortical function as measured by ensemble activation.
# Discussion
The return of consciousness after TBI is associated with a welldescribed transition from delta-band to alpha, beta, and theta oscillations. We extend these findings by reporting that thalamic injury is associated with an attractor-like dysfunctional cortical state characterized by simple and repetitive ECoG activity in vivo and in silico. Single-pulse electrical stimulation of the cortex generated monophasic, repetitive, deterministic evoked potentials when thalamic injury was present. We replicated these findings in silico; our hypothesis-driven model had strong homology to the recorded electrocorticography. Moreover, in our model, thalamic input facilitated the formation of neuronal ensembles required for the content of consciousness; neuronal ensembles did not form when the thalamic input was reduced. We suggest that simple and repetitive cortical dynamics associated with the coma attractor (when thalamic input is absent) are inadequate to engage the vast repertoire of ensembles needed for goal-directed behavior consistent with other recent models [bib_ref] The thalamus integrates the macrosystems of the brain to facilitate complex, adaptive..., Shine [/bib_ref]. This leads to a limited behavioral repertoire of the comatose sTBI patient. Critically, the thalamus has to adjust excitability and connectivity to bring about complex dynamics, consistent with recently proposed theoretical models of thalamic function [bib_ref] Thalamocortical circuit motifs: a general framework, Halassa [/bib_ref].
To the best of our knowledge, this study represents the first report combining electrocorticography with neuroimaging to understand how thalamic injury affects cortical function. These data have important implications for coma therapeutics. Specifically, this model furnishes an explanation for the mixed results of central thalamic deep brain stimulation in TBI patients [bib_ref] Deep brain stimulation for recovery of consciousness in minimally conscious patients after..., Haddad [/bib_ref]. Our data imply that when higher-order thalamic projections are injured, interventions to increase cortical excitability alone are unlikely to restore consciousness because the full repertoire of human behavior requires dynamic adjustment of connectivity. Further work will be required to understand how central thalamic stimulation affects connectivity; it has clear effects on cortical excitability [bib_ref] Thalamus modulates consciousness via layer-specific control of cortex, Redinbaugh [/bib_ref] , but deficits in synaptic connectivity may explain why it has mixed results in the clinical treatment of consciousness disorders [bib_ref] Behavioural improvements with thalamic stimulation after severe traumatic brain injury, Schiff [/bib_ref] [bib_ref] Deep brain stimulation for recovery of consciousness in minimally conscious patients after..., Haddad [/bib_ref]. It is possible that stimulation of mediodorsal or other thalamic nuclei could increase synaptic connectivity in the required fashion. Finally, our data provide a possible explanation for recent observations that thalamic injuries do not totally prevent consciousness [bib_ref] Thalamic strokes that severely impair arousal extend into the brainstem, Hindman [/bib_ref]. Based on our model, we propose that The computational model of thalamocortical networks in consciousness predicts content of consciousness requires thalamic modulation of cortical networks. a The thalamocortical model consists of two interconnected cortical populations that receive thalamic input. Each cortical region includes interacting excitatory and inhibitory cells. In addition to the excitatory input, the thalamus modulates the local (α) and long-distance (β) connectivity of cortical networks. b Depending on the activity mode of thalamus (tonic vs. phasic) cortical cells switch between rhythmic (coma) and asynchronous (awake) dynamics. The reconstruction of phase space for awake and coma shows a structured phase space for coma and much more complex phase space for the awake state; similar features are seen in the experimental data. c The evoked responses for 100 simulated trials of coma and awake states are shown in heat maps for a one-second period where the color intensity denotes the deflection from the baseline. Underneath the heat map, the average of these responses across one hundred trials is shown; the stimulation time is marked by the yellow lightning bolt. Note that stimulation in the simulated awake state triggers multiple waves of excitation. Stimulation in coma results in a monophasic, and stereotypical response. d In our model, an embedded cortical cluster (i.e., an ensemble, orange bracket on raster plots) only becomes activated when it receives both excitatory thalamic input and amplification of its cortico−cortical connections (left versus right raster plots). This ensemble represents a low-level "building block" of consciousness.
thalamic injury leads to impairment of neuronal ensemble formation in the cortex, and thus, the content of consciousness; thus, even extensive thalamic injuries could allow for a limited repertoire of conscious content (i.e., the cortical ensembles that can be activated by the intact thalamic projections). In the case of Subject 4, he was conscious, despite an extensive but partial thalamic injury. A detailed neuropsychological examination may have revealed deficits related to right hemisphere dysfunction, but the language-based control needed to assay consciousness was evidently intact. When there is a bilateral thalamic injury, there is likely to be less functioning cortex needed to support consciousness. New therapies for brain injury should focus on restoring the thalamic role in modulating cortical networks and thus the content of consciousness, in addition to its role in augmenting arousal.
Conscious behavior exhibits similarities with the lowpredictability dynamics we identified in the conscious LFP recordings. William James described that thalamic input to the cortex in frogs leads to unpredictable behavior; in contrast to a frog with the thalamus ablated, "...[the frog's] conduct has become incalculable. We can no longer foretell it exactly. The effort to escape is his dominant reaction, but he may do anything else…". After a severe brain injury, the return of consciousness is similarly characterized by a gradual return of increasingly complex goal-directed behavior. While increased complexity of EEG in conscious patients has been previously reported 4,6 , our data also show that the return of consciousness implies decreased predictability as well as increased complexity. This lack of predictability is a consequence of thalamic input to the cortex, which is in turn intimately linked with the cortex's ability to form neuronal ensembles.
Further research is needed to define the dynamics of consciousness exactly; a tantalizing possibility is that the dynamics are formally chaotic [bib_ref] More than synchrony: EEG chaoticity may be necessary for conscious brain functioning, Fell [/bib_ref] , consistent with the low predictability of animal behavior observed by James. However, it is currently not possible to distinguish this possibility from simpler kinds of nonlinear interactions [bib_ref] Dimensional analysis of resting human EEG II: Surrogate-data testing indicates nonlinearity but..., Pritchard [/bib_ref]. We can currently only conclude that the conscious cortex exhibits complex dynamics with low predictability and a large number of possible states.
Our sample size is relatively small (n = 5) and limited to a single disease process. Other conditions such as cardiac arrest may lead to coma in a manner distinct from TBI. We also did not directly observe electrical activity in the thalamus because we do not routinely place electrodes in this structure for seizure monitoring. Even so, some groups have advocated thalamic monitoring for seizures in refractory epilepsy cases [bib_ref] Responsive neurostimulation targeting the anterior nucleus of the thalamus in 3 patients..., Elder [/bib_ref]. However, in the case of critically-ill, comatose sTBI patients, it could be beneficial to implant thalamic electrodes-especially if preclinical data suggest that stimulation could be therapeutic. Animal models will likely be only modestly informative about the role of the thalamus in disorders of consciousness because of the challenges associated with studying consciousness in animals as well as the difficulty that accompanies modeling traumatic coma.
In addition, although our computational model is a simplified model and lacks many of the biological details of the thalamocortical circuit, it replicates many of our in vivo observations and makes important predictions. Further experiments are needed to refine the model in ways that reflect emerging evidence about the organization of these circuits, and will shed additional light on how they recover from injury.
# Methods
Patient selection. One control subject (aged 24) and five patients (aged 25−78) who became comatose (GCS < 8) after suffering severe traumatic brain injury were recruited for this study after admission to Stony Brook University Hospital from 2016-2018. Patient clinical histories are summarized in Supplementary Note 1.
Candidate sTBI patients were identified during neurosurgical consultation in the emergency department. A seizure monitoring stereotactic depth electrode was placed to detect seizures; in one recent series, seizures or periodic discharges occurred in 61% of sTBI patients, and 43% of these seizures were observable only with depth recordings [bib_ref] Metabolic crisis occurs with seizures and periodic discharges after brain trauma, Vespa [/bib_ref]. Informed consent to enroll in the study and implant depth electrodes was obtained from the patient's Legally Authorized Representative (LAR) according to NIH, Declaration of Helsinki, and institutional guidelines, and was monitored by the Stony Brook University Hospital Committee on Research Involving Human Subjects, as well as a separate Data Safety and Monitoring Board (DSMB). Pregnant females and patients with major brain structural abnormalities (excepting hemicraniectomy) that would preclude between-subject comparisons were excluded from the study (see [fig_ref] Table 1: Patient summaries [/fig_ref] for full inclusion and exclusion criteria). describes the exact timeline of injury, depth electrode implantation, and CRS-R scoring. Supplementary Note 1 provides additional clinical information.
Electrophysiological recordings and stimulation. Patients were implanted with a ten-contact stereotactic depth electrode (Ad-Tech, Spencer Depth™) spanning the dorsolateral prefrontal cortex (DLPFC) to ACC. Electrodes were right-sided for all patients except for Subject 2 (see Supplementary Note 1). These areas were selected for the study because it is technically simple to place depth electrodes there for clinical seizure monitoring. FPN is well-known to be critical to goal-directed behavior [bib_ref] Distinct brain networks for adaptive and stable task control in humans, Dosenbach [/bib_ref] ; ACC is also a well-defined "hub" region that has broad cortical connectivity. We thus expected to probe cortical function broadly from ACC as a single point of access. PFC was identified as the caudal part of the middle frontal gyrus at the level of the coronal suture, and the anterior cingulate was typically targeted 20 mm from the tip of the frontal horn. Each contact can be used for ECoG recordings and stimulation. We also attached scalp EEG leads in an 18-contact standard 10−20 montage. Impedances for scalp contacts were within 5−10 kΩ. Impedance was not directly measured on intracranial contacts, but recordings were visually inspected for quality. Artifacts were rejected by visual inspection as well as by the Fieldtrip Toolbox 53 . After implantation, patients underwent detailed behavioral testing, stimulation, and recording. After implantation of the depth electrode, continuous scalp (EEG) and depth (ECoG) monitoring was initiated to detect seizure activity and to provide antiepileptic treatment and prophylaxis as needed. The sedative medication for each patient is listed in Supplementary .
Single-pulse stimulation (10 mA, 100 μA, 100 stimuli) was applied to contacts within the ACC and DLPFC using a Nicolet cortical stimulator (Natus Medical: Pleasanton, CA) while recording all other contacts including scalp and depth electrodes [bib_ref] Mapping human brain networks with cortico− cortical evoked potentials, Keller [/bib_ref]. Stimulation pulses were biphasic and bipolar and applied between two adjacent contacts. The maximum parameters were 10 mA at ≥1 Hz (biphasic square-wave pulse, 100 μS per phase); 10 mA was selected based on prior reports [bib_ref] Functional connectivity in the human language system: a cortico−cortical evoked potential study, Matsumoto [/bib_ref]. No adverse effects were noted from stimulation. There were no obvious changes to the level of consciousness or ongoing EEG activity as a result of single-pulse stimulation.
Behavioral testing. After enrollment, we measured the CRS-R score on a daily basis before delivering single pulses of stimulation in each patient to assess the level of consciousness [bib_ref] The JFK Coma Recovery Scale-Revised: measurement characteristics and diagnostic utility, Giacino [/bib_ref]. Most patients gradually returned to consciousness over a period of days or weeks [fig_ref] Table 1: Patient summaries [/fig_ref].
Imaging. Imaging was obtained at multiple time points to verify proper electrode placement. CT was obtained post-operatively after implantation to localize the electrode and to identify potential surgical complications. MRI was performed after electrode removal and is shown in . MRIs were obtained for clinical (not research) purposes and represent the best images that could be obtained under the circumstances. Volumetric Siemens FLAIR (fluid-attenuated inversion recovery) sequences were used for all patients. The matrix size was 256 × 256. Slice thickness/gap was 1/0 mm. Flip angle = 120, TE/TR = 335/5000 ms (exception: TR = 4500 for subject 1), TI = 1800 ms, pixel size = 0.9375 mm. Scans were typically acquired in the sagittal plane and resampled to axial. For Subject 4, diffusion tensor images were acquired using an EPI sequence with flip angle = 90, TE/TR = 90/5400 ms, slice thickness/gap = 4/0 mm, 32 diffusion direction, single shell scheme with b = 1000. For Subject 5, a diffusion-weighted EPI image is also shown (TE/TR = 90/10300 ms; flip angle = 90, b = acquisition matrix = 200 × 200 and slice thickness/gap was 4/0 mm). Field strength was for 3T for Subjects 1, 2, and 4 and 1.5T for Subjects 3 and 5.
Human subject safety. All study procedures were conducted under the supervision of a Data Safety and Monitoring Board and the Stony Brook University Committee on Research in Human Subjects (CORIHS). There were no complications from depth electrode placement or stimulation in this study. In all cases, MRI after implantation did not show evidence of injured tissue in the area of the electrode. Notably, the used current amplitudes and charge densities were below the manufacturer's recommendations and are comparable to stimulation parameters used for therapeutic cortical stimulation (in some cases, ≤8 mA) chronically [bib_ref] Dosing of electrical parameters in deep brain stimulation (DBS) for intractable depression:..., Ramasubbu [/bib_ref] , with bursts up to 12 mA 57 . We had no study-related adverse events. Data analysis. Data analysis was performed using MATLAB. Scalp and depth channel recordings were subject to mean subtraction, bipolar re-referencing, and band-pass filtering (0.5−30 Hz).
Frequency analysis. Baseline power spectral density (PSD) estimates shown in , were created using Welch's overlapped segment averaging estimator with a Hamming window and 50% overlap. Frequency analysis was performed over the 0.5−30 Hz range.
Cortico−cortical evoked potentials (CCEPs). CCEPs were created by cutting the recorded EEG and ECoG signals into one-second epochs (from 250 ms prestimulation to 750 ms post-stimulation). These were time-locked to the depth electrode stimulation pulse. The CCEPs recordings were band-pass filtered with cutoff frequencies of 0.5−30 Hz and then were demeaned and averaged.
Complexity analysis. The zero-crossing (ZC) score was used to estimate the complexity of the CCEP. This measure represents the number of times that the demeaned and band-pass filtered (0.5−10 Hz, 4th order FIR filter) CCEP signal crossed zero (changed sign) per second. This measure was computed for each trial, and the averaged ZC for each patient across all channels was computed.
The Lempel−Ziv complexity measure (LZ) was used to quantify the complexity of CCEP evoked responses [bib_ref] On the complexity of finite sequences, Lempel [/bib_ref]. The LZ method is a robust information-theoretic measure based on universal lossless data compression, and the most popular algorithmic complexity estimator of the Kolmogorov class. The complexity of a one-dimensional signal is estimated by quantifying the number of unique patterns contained within that signal.
CCEPs were band-passed from 0.5 to 10 Hz using a 4th order Butterworth notch filter. Scalp CCEPs were converted to one-dimensional binary signals using the median method. The LZ complexity method was then calculated [bib_ref] On the complexity of finite sequences, Lempel [/bib_ref]. The output of the method for computing LZ complexity was then normalized based on the signal sequence length according to:
[formula] C norm ¼ C raw ½ n ðnÞð1Þ [/formula]
where, C norm is the normalized LZ complexity, C raw is the non-normalized complexity, and n is the signal length.
Susceptibility index. We hypothesize that the susceptibility of cortical networks to external perturbations is due to a lack of strong internal dynamics. Thus, the susceptibility is an indication of the poor functional state of the cortical networks. We measured the susceptibility (δ) of the ongoing cortical LFP to external stimulation by quantifying the similarity of the LFP waveform in response to single pulses of stimulation and trains of stimulation , 30 s, low-passed LFP signals, <0.1 Hz) across trials [bib_ref] Clustering in globally coupled inhibitory neurons, Golomb [/bib_ref] [bib_ref] Interplay between excitability type and distributions of neuronal connectivity determines neuronal network..., Mofakham [/bib_ref]. This index quantifies the similarity of LFP responses by comparing the variance of the mean of the responses across trials normalized by the average variances of each trial [bib_ref] Clustering in globally coupled inhibitory neurons, Golomb [/bib_ref]. In , the susceptibility index was used to quantify the temporal stability of 30 s LFPs (initial ten seconds that include the stimulation block followed by twenty seconds with no stimulation) across all trials. The susceptibility index is bounded from 0 to 1, and larger values (~1) indicate a highly stable and deterministic response to external stimulation.
Phase space reconstruction. We reconstructed the underlying dynamical systems of the cortical networks following sTBI in the phase space using delay embedding methodology. The phase space represents the trajectory of states that the system visits at any time point. Delay embedding theorem enables us to reconstruct this phase space via a single time-series (X j = 1, .., N) 62 . Here, we used depth cortical recordings from the prefrontal cortex to reconstruct the phase space. The onedimensional time series can be mapped into the higher dimensional space by plotting it against its delayed version as follows:
[formula] Y j ¼ ðX j ; X jþτ ; X jþ2τ ; ; X jþðdÀ1Þτ Þ j ¼ 1; 2; ; N À ðd À 1Þτð2Þ [/formula]
Here, d is the embedding dimension and τ the delay. The time delay was estimated based on the signal autocorrelation; the threshold was drawn when the autocorrelation dropped under zero. The minimum sufficient embedding dimension was determined based on the FNN algorithm [bib_ref] Determining embedding dimension for phase-space reconstruction using a geometrical construction, Kennel [/bib_ref] Principal component analysis (PCA). We performed PCA, a widely used dimensionality reduction methodology, to correlate the dimensionality of the recordings to the level of consciousness. We used a MATLAB built-in PCA function to compute the principal components, which are the eigenvectors of the covariance matrix of the dataset. Where the eigenvalues describe how much of the variance in the data is explained by the respective eigenvectors. We performed PCA analysis on the combined scalp (n = 17) and depth electrode channels (n = 8) on a 5 min stimulation-free period (n = 25, . These recordings were subject to bipolar re-referencing, normalization, and band-pass filtering (0.5−30 Hz).
Statistics and reproducibility. This study included five TBI patients with depth electrophysiological recordings and one control. Statistical analyses were performed on complexity levels across patients using IBM SPSS. CCEPs were delivered in trains of ten pulses separated by one second; typically, each session included ten stimulation trains. Depending on the availability of patients, we recorded between 2 and 6 sessions for each patient. Thus, each patient has between CCEP 200−600 trials. The examples of average CCEP over 100 trials (one session) in the patients with and without thalamic injury are shown in . To examine statistical differences between the CCEPs waveforms, statistical analyses were performed on complexity levels across patients using IBM SPSS. First, we calculated zero crossing and LZ complexity measures at all the EEG contacts for each trial across patients. Then, we grouped these complexity measures into two categories based on whether patients had a thalamic injury or not. Finally, we used a two-tailed t-test to determine if there was a significant difference in the mean values of these groups.
Our results show that the group with thalamic injury had a significantly lower complexity measured with zero-crossing and LZ complexity (P < 0.01, two-tailed T-test).
Network structure for modeling FPN. A simplified and reduced thalamocortical model has been constructed consisting of three populations of LIF neurons representing frontal and parietal cortical populations and a thalamic population. The cortical populations each include 50 excitatory and inhibitory neurons. The excitatory neurons in each cortical population are connected to all other neurons within their respective cortical population. Further, the excitatory cells in each cortical population project to and receive feedback from their local inhibitory cells, and all the cells within the two cortical populations receive excitatory thalamic inputs. Finally, the synaptic gain of the local and long-distance cortical connections is additionally controlled by the state of thalamic activity via the parameters α and β, respectively .
Leaky-integrate-and-fire. The simulated neuronal network is constructed using the LIF model. The LIF neuron integrates all the incoming signals and spikes if the voltage across its membrane reaches a certain threshold (V th = −50 mV). After the spike, the membrane voltage is reset to the resting potential (−65 mV) and it cannot spike for a refractory period of τ ref~1 ms. The voltage across the LIF cell membrane evolves as follows:
[formula] C dV j dt ¼ ÀgðV j À E L Þ þ I j ext: þ I j syn: þ I j stim:ð3Þ [/formula]
Here, C is the membrane capacitance (1 nF), E L is the resting potential of the neuron, V j is the voltage across the jth neuron membrane, and g is the membrane conductance (1/R, where R = 10 [10 6 Ω]). The neuron's time constant, τ, is equal to τ = RC. Here, I ext is the driving current for each population of cells (cortical, thalamic), and I j stim: is the stimulation pulse that resembles the stimulation delivered by the depth electrode to the jth neuron. The synaptic current received by the jth cortical neuron is represented by I j syn: and is calculated using the following equation, Eq. (4):
[formula] I j syn: ¼ αW L ∑ j A L i;j ðHðt À t s À t d Þ À Hðt À t s À t d À TÞÞ þ βW D ∑ k A D k;j ðHðt À t s À t d Þ À Hðt À t s À t d À TÞÞ þ W Th ∑ h A Th h;j ðHðt À t s À t d Þ À Hðt À t s À t d À TÞÞ À W inh ∑ n A inh n;j e 1ÀðtÀts Àt d Þ Qð4Þ [/formula]
where the meaning of the symbols is as follows. According to the equation, each cortical cell receives three different synaptic currents: (1) local excitatory and inhibitory, (2) long-distance cortical, and (3) thalamic. The first and the second terms in the equation correspond to the first two synaptic currents: where A L i;j and A D k;j matrices are local and long-distance connectivity matrices representing each synaptic connection in the cortical network. Zero entries of these matrices stand for no connection between neurons i and j, and entries of one represent the existence of a connection. We assumed all-to-all connectivity for connections within each cortical population and a partial connectivity across populations (each neuron establishes long-distance connections within the radius of fifteen). The strength of these connections is determined with W L (0.001) and W D (0.004), which are the local and long-distance synaptic weights, respectively. We assumed 3 ms synaptic delay for all connections (t d ). The H operator indicates the Heaviside function, which controls the duration of the incoming square pulse (T) following the presynaptic cell spike (t s ).
In this model, the thalamic activity controls the gain of local and long-distance connections through the α and β coefficients, respectively. Here, the coma state was characterized by reduced thalamic activity. The terms α and β are greater in the awake state in comparison to the coma state (α awake /α coma = 2, β awake /β coma = 1.5). The third term corresponds to the thalamic excitatory input to the cortical cells where W Th is the synaptic weight for the driving thalamic inputs to the cortex W Th (0.01), and A Th is the connectivity matrix between thalamic and cortical cells (we assumed all-to-all connectivity, A Th is a unity matrix). In the simulated coma state, the firing rate of thalamic cells and thus driving input from the thalamus to cortical networks was reduced. This was implemented by: (1) increasing the time constant of the thalamic cells in coma (τ = 100 ms) compared to the awake state (τ = 10 ms), (2) reducing the noisiness of external current (lowering probability [*0.5] and magnitude [*0.025] of the excitatory noise input) relative to the deterministic external current to the thalamic cells in coma (I j ext: ¼ 1:5 nA, in awake I j ext: ¼ 1:49 nA). Together, in this model, thalamic activity controls cortical dynamics in coma and consciousness by adjusting the excitatory driving input to the cortex, and the gain of local and long-distance cortical connections.
In addition to the aforementioned excitatory currents, each cortical cell also receives inhibitory input from local inhibitory cells-the fourth term in Eq. (4). The inhibitory cells control the high excitation level in this model and only spike when they receive synchronous excitatory inputs. After spiking, the inhibitory cells inject the local cortical cells with a long-lasting negative current with a synaptic delay of t d = 3 ms that decays exponentially with time. The LFP was obtained by the summation of the excitatory and inhibitory synaptic currents at each cortical area.
Formation of functional local and long-distance cell assemblies. In this model, we also investigated the functional state of cell assemblies in the coma and awake states. As mentioned previously, each simulated neuron in the cortical areas is locally connected to the rest of the local cells with W L . To investigate the activity of the cortical representations in the coma and awake states, we embed a cluster of neurons by increasing the W L to W L' (*4) within a subpopulation of these neurons (neurons 1−20). Hypothetically, this cluster could potentially encode certain aspects of a task or memory. shows that this cluster can be functionally activated when receiving further amplification from the thalamus. In the coma state, without thalamic amplification, the connections within this cluster are too weak to be functionally distinguished. However, in the awake state, the general thalamic amplification of connections and selective amplification within this cluster (*3) turns the cluster ON. As a result, the activity within the cluster becomes remarkably synchronous, distinguishing it from the rest of the cells.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.
## Data availability
All source data for our figures are included as Supplementary Data files. Original recordings and images are available upon reasonable request.
[fig] Figure 3: Reconstructed phase space of depth cortical recordings reveals a low dimensional attractor for patients with thalamic injury. a We reconstructed the phase space of recordings obtained from depth PFC recording for all the patients. Patients 4 and 5 with unilateral and bilateral thalamic injury, respectively, show a structured phase space that repeatedly visits similar trajectories reflecting a limited number of available states in the repertoire of cortical networks. b Phase space of 20 s of cortical recordings consisted of a ten-second baseline (gray) followed by a ten-second train of single pulses of stimulation (red). The timing of single pulses is shown in red on the recorded LFP. Responses are consistent with a ring attractor in patients 4 and 5. [/fig]
[table] Table 1: Patient summaries.Abbreviations: GCS Glasgow Coma Scale, CRS-R Coma Recovery Scale-Revised. *Patient initially presented with a higher GCS (14) but declined prior to enrollment. [/table]
|
Plasma Amino Acid Profile in Patients with Aortic Dissection
[fig] Fig: S2 Permutation test plots (200 permutations) for all the PLS-DA score plots. A: Acute AD vs. [/fig]
[table] Table S1: Data for quality control samples. [/table]
|
Microsatellite instability in colorectal cancer
Colorectal cancer (CRC) is a heterogeneous disease that is caused by the interaction of genetic and environmental factors. Although it is one of the most common cancers worldwide, CRC would be one of the most curable cancers if it is detected in the early stages. Molecular changes that occur in colorectal cancer may be categorized into three main groups: 1) Chromosomal Instability (CIN), 2) Microsatellite Instability (MSI), and 3) CpG Island Methylator phenotype (CIMP). Microsatellites, also known as Short Tandem Repeats (STRs) are small (1-6 base pairs) repeating stretches of DNA scattered throughout the entire genome and account for approximately 3 % of the human genome. Due to their repeated structure, microsatellites are prone to high mutation rate. Microsatellite instability (MSI) is a unique molecular alteration and hyper-mutable phenotype, which is the result of a defective DNA mismatch repair (MMR) system, and can be defined as the presence of alternate sized repetitive DNA sequences which are not present in the corresponding germ line DNA. The presence of MSI is found in sporadic colon, gastric, sporadic endometrial and the majority of other cancers. Approximately, 15-20 % of colorectal cancers display MSI. Determination of MSI status in CRC has prognostic and therapeutic implications. As well, detecting MSI is used diagnostically for tumor detection and classification. For these reasons, microsatellite instability analysis is becoming more and more important in colorectal cancer patients. The objective of this review is to provide the comprehensive summary of the update knowledge of colorectal cancer classification and diagnostic features of microsatellite instability.
# Introduction
Colorectal cancer (CRC) is the third most prevalent cancer in humans and the third most common cause of cancer related deaths in both males and females that contributes to a significant public health problem worldwide [bib_ref] Cancer statistics, Siegel [/bib_ref] [bib_ref] Forman D. Global cancer statistics, Jemal [/bib_ref]. The majority of colorectal cancer cases are sporadic (about 75 %) that display no apparent evidence of having inherited disorders, suggesting contribution of genetic and environmental factors, whereas only 25 % of the patients have family histories of the disease [bib_ref] Hereditary and familial colon cancer, Jasperson [/bib_ref]. Only 5-6 % of patients with colorectal cancer with a family background are due to inherited mutations in major CRC genes, while the rest are the result of accumulation of both genetic mutations and epigenetic modifications of several genes [bib_ref] Genetics, cytogenetics, and epigenetics of colorectal cancer, Migliore [/bib_ref]. According to the current knowledge, CRC mainly develops through a gradual accumulation of genetic and epigenetic alterations of the genome [bib_ref] A genetic model for colorectal tumorigenesis, Fearon [/bib_ref]. Molecular changes that occur in colorectal cancer may be categorized into three main groups: 1) Chromosomal Instability (CIN), 2) Microsatellite Instability (MSI), and 3) CpG Island Methylator phenotype (CIMP) that silences gene function with aberrant hypermethylation [bib_ref] Colorectal cancer: molecular features and clinical opportunities, Worthley [/bib_ref]. CIN-positive CRCs are featured with alterations in the structure and number of chromosome besides increased mutation rates in both tumor suppressor genes and oncogenes. Actually, mutation rates in single nucleotides are higher in MSI-positive CRCs than in CIN-positive CRCs . On the other hand, 15-20 % of CRCs display MSI caused by defective DNA mismatch repair (MMR) system [bib_ref] Microsatellite instability in colorectal cancer-the stable evidence, Vilar [/bib_ref]. In this review, we provide the comprehensive summary of the current knowledge in the field of colorectal cancer classification and diagnostic features of microsatellite instability.
## Types of crc and their
GENETIC BASIS According to etiology and genetics of the disease, CRC is usually classified into three distinct groups: sporadic, familial, and hereditary [fig_ref] Table 1: Types of colorectal cancer and their genes [/fig_ref] [bib_ref] Colorectal cancer: Molecular mutations and polymorphisms, Sameer [/bib_ref].
## Sporadic crc
Sporadic colorectal cancer is the most common type of CRC and includes approximately 75 % of cases that display no apparent evidence of having inheritance of disorder. However, this is unclear, since genetic factors seem to affect the likelihood of cancer even in the absence of specific mutations. Sporadic colorectal cancer is common among elder people, probably as a result of environmental factors, dietary, and aging. MSI-H sporadic colorectal cancers are the most often (about 70 %-95 %) caused [bib_ref] Clinical significance and prognostic relevance of microsatellite instability in sporadic colorectal cancer..., Copija [/bib_ref].
## Familial type of crc
This type of CRC is often considered sporadic and not any associated gene has been identified yet. People with a history of colorectal cancer in a first-degree relative are at increased risk of two to three times higher than the normal population [bib_ref] Colorectal cancer screening in patients at moderately increased risk due to family..., Lin [/bib_ref].
## Hereditary type of crc
The hereditary type of CRC is divided into five subtypes:
## Familial adenomatous polyposis (fap)
Familial adenomatous polyposis is the most common hereditary polyposis syndrome, which is an autosomal dominant disease caused by a mutation in the APC gene on chromosome 5q21. The APC gene is a tumor suppressor gene that produces APC protein, a multifunction protein which controls how quickly cells grow and prevent the development of tumors. The normal function of APC protein is the regulation of β-catenin by its degradation. β-catenin plays an important role in cell communication, Wnt signalling pathway, and growth by acting as a transcription factor for proliferation genes. Mutations in the APC gene lead to loss of APC function and result in an accumulation of β-catenin. Many different mutations (e.g. insertions, deletions, nonsense mutations) of the APC gene are described as a cause of FAP [bib_ref] Molecular genetics of colorectal cancer, Bogaert [/bib_ref]. In addition to APC mutation, other mutations, such as K-RAS, DCC, P53, COX-2, BCL-2 and etc. are required to develop cancer [bib_ref] A de novo germline APC mutation (3927del5) in a patient with familial..., Zeichner [/bib_ref]. In FAP patients, polyps are mainly found in the proximal colon and rarely in the rectum. The average age for people with FAP to develop polyps is 35 years and if FAP is not recognized and treated, most likely it will develop colorectal cancer [bib_ref] Familial adenomatous polyposis: clinical presentation, detection and surveillance, Laurent [/bib_ref]. There are four subtypes of FAP that are caused by different germ line mutations in the APC gene: Attenuated FAP (AFAP), Gardner syndrome, Turcot syndrome, and Gastric Adenocarcinoma and Proximal Polyposis of the Stomach (GAPPS).
## Mutyh-associated polyposis (map)
MUTYH-associated polyposis is one of the hereditary polyposis syndromes which is an autosomal recessive disease and is caused by a biallelic germline mutation in MUTYH gene on chromosome 1p34.1. MUTYH gene encodes an enzyme called MYH glycosylase that is involved in a DNA repair system called Base Excision Repair (BER). The number of polyps in MAP disease is less than FAP and it is phenotypically similar to attenuated FAP [bib_ref] Lynch syndrome and MYH-associated polyposis: review and testing strategy, Goodenberger [/bib_ref]. The average age for diagnosis of MAP is between 40-60 years old and if MAP is not recognized and treated, patients will have an 80 % risk of developing CRC [bib_ref] A large-scale metaanalysis to refine colorectal cancer risk estimates associated with MUTYH..., Theodoratou [/bib_ref]. The majority of studied CRCs in people with MAP were microsatellite stable; although the MSI-H (high-level microsatellite instability) phenotype is reported in the minority of CRCs of persons with MAP [bib_ref] MUTYHassociated polyposis (MAP), Nielsen [/bib_ref] [bib_ref] Prevalence of germline MUTYH mutations among Lynch-like syndrome patients, Castillejo [/bib_ref].
## Peutz-jeghers syndrome (pjs)
Peutz-Jeghers syndrome is a rare autosomal dominant disorder, which is one of the hereditary polyposis syndromes and is characterized by multiple benign hamartomatous polyps in the gastrointestinal tract, most often found in the small intestine. The number of polyps in PJS is less than MAP syndrome and those polyps are present from childhood [bib_ref] Peutz-Jeghers syndrome and management recommendations, Giardiello [/bib_ref]. The major cause of this disease is germ line mutations in the STK11 (serine threonine kinase 11) gene, also known as the LKB1, which is a tumor suppressor gene and is located on chromosome 19p13.3 [bib_ref] Peutz-Jeghers syndrome with germline mutation of STK11, Chae [/bib_ref]. Mutations in this gene change the structure and/or function of the STK11 protein, disrupting its ability to restrain cell division with the loss of kinase activity. Microsatellite instability, LOH nearby the APC gene, and KRAS mutations have been identified in some tumors [bib_ref] Lower gastrointestinal tract cancer predisposition syndromes, Shah [/bib_ref].
## Serrated polyposis syndrome (sps)
Serrated polyposis syndrome is a relatively rare syndrome characterized by multiple serrated polyps of the colon, which was previously known as the hyperplastic polyposis syndrome [bib_ref] Serrated colon polyps as precursors to colorectal cancer, Sweetser [/bib_ref]. This syndrome is usually considered sporadic and underlying genetic causes are related to germline mutations of oncogene-induced senescence pathway genes [bib_ref] Germline mutations in oncogeneinduced senescence pathways are associated with multiple sessile serrated..., Gala [/bib_ref]. These tumors will be more often considered MSI-low or MSS (Microsatellite Stable) [bib_ref] Serrated polyposis syndrome: molecular, pathological and clinical aspects, Guarinos [/bib_ref]. There are three criteria for diagnosis of SPS and an individual would be considered affected if have at least one of them; A) at least five serrated polyps proximal to sigmoid with at least two of them being greater than 1 cm; B) Any number of serrated polyps occurring proximal to the sigmoid colon in an individual who has a firstdegree relative with SPS; and C) more than 20 serrated polyps distributed throughout the colon. Furthermore, it would be noteworthy to indicate that SPS has been associated with an increased risk of developing CRC considering previous studies [bib_ref] Serrated lesions of the colorectum: review and recommendations from an expert panel, Rex [/bib_ref].
## Lynch syndrome (ls)
Lynch syndrome (LS), also was known as Hereditary non-polyposis colorectal cancer, is the most common hereditary colon cancer syndrome which is an autosomal dominant disease and is caused by germline mutations in one of several DNA mismatch repair (MMR) genes, including MSH2 on chromosome 2p16, MLH1 on chromosome 3p21, MSH6 on chromosome 2p16, and PSM2 on chromosome 7p22 [bib_ref] Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and..., Lynch [/bib_ref]. MSH2 and MLH1 mutations account for the majority of Lynch syndrome cases [bib_ref] Practical genetics of colorectal cancer, Lynch [/bib_ref]. MMR genes encode proteins that are critical to the suitable repair of DNA sequence mismatch and correct base mismatches or small deletions or insertions [bib_ref] Molecular testing in colorectal cancer: diagnosis of Lynch syndrome and personalized cancer..., Shi [/bib_ref]. Inactivation of these genes interrupts DNA repair and causes an alteration in the short-tandem DNA repetitive sequences or microsatellites, resulting in the development of a phenotype known as microsatellite instability which is a hallmark of Lynch syndrome. This syndrome is accounting for approximately 2-3 % of the total CRC cases [bib_ref] Screening for the Lynch syndrome (hereditary nonpolyposis colorectal cancer), Hampel [/bib_ref]. High-level microsatellite instability is observed in approximately 90 % of LS-associated CRCs [bib_ref] Report from the Jerusalem workshop on Lynch syndrome-hereditary nonpolyposis colorectal cancer, Boland [/bib_ref].
## Consensus molecular sub-types of crc
To resolve inconsistencies between CRC classifications based on gene expressions reported and simplify clinical translation four Consensus Molecular Subtypes (CMSs) with distinct features have been reported. The CMS1 are (microsatellite instability immune, 14 %), hypermutated, microsatellite unstable and are immunogenic. The CMS2 are (canonical, 37 %), epithelial, marked WNT and MYC signalling activation and have the highest overall survival. The CMS3 are (metabolic, 13 %), epithelial and evident metabolic cancer phenotype; and the CMS4 (mesenchymal, 23 %), prominent transforming growth factor-β activation, stromal invasion and angiogenesis that have a worst survival. The Consensus Molecular Subtypes of CRC with clear biological interpretability may have a better prognosis, therapeutic response, and potential new treatment strategies [bib_ref] The consensus molecular subtypes of colorectal cancer, Guinney [/bib_ref] [bib_ref] Consensus molecular subtypes of colorectal cancer and their clinical implications, Thanki [/bib_ref].
## Molecular basis of dna mismatch repair system
DNA mismatch repair system (MMR) corrects erroneous insertion, deletion, and base-base mismatches generated during DNA replication and recombination that have escaped the proofreading process [bib_ref] The multifaceted mismatch-repair system, Jiricny [/bib_ref]. This repair pathway is highly conserved from bacteria to humans and safeguards the integrity of the genome [bib_ref] DNA mismatch repair: molecular mechanism, cancer, and ageing, Hsieh [/bib_ref]. MutS and MutL are the main proteins involved in prokaryote MMR system that function as homodimers whereas, in eukaryotes, MSH2, MSH3, and MSH6 are homologs for MutS; MLH1, MLH2, MLH3 are MutL homologs. There are also other homologs for MutL (post-meiotic segregation) named PMS1 and PMS2 which interact as heterodimers [bib_ref] DNA mismatch repair in eukaryotes and bacteria, Fukui [/bib_ref]. When a mismatch is detected in the eukaryotic genome, DNA mismatch repair system functions through a series of steps: MSH2 associates with MSH6 or MSH3 causing the formation of MutSα and MutSβ heterodimers, respectively. MutSα recognizes single base mismatches and small insertion/deletion loops (IDLs), while MutSβ recognizes larger loops. MutSα or MutSβ can recruit MutLα, MutLβ or MutLγ heterodimers (if MLH1 couples with PMS2, PMS1 or MLH3, respectively) by means of exchanging adenosine triphosphate (ATP) to adenosine diphosphate (ADP). This complex (MutS-MutL) creates a sliding clamp around the DNA. The proteins in sliding clamp interact with exonuclease-1 and proliferating cell nuclear antigen (PCNA). This complex excises the daughter strand back to the site of the mismatch. Finally, resynthesize and re-ligation are performed by DNA polymerase and DNA ligase, respectively. The correction occurs [fig_ref] Figure 1: Mechanism of mismatch repair system [/fig_ref] [bib_ref] Mechanisms and functions of DNA mismatch repair, Li [/bib_ref] [bib_ref] The mechanism of mismatch repair and the functional analysis of mismatch repair..., Martín-López [/bib_ref].
## Microsatellite instability (msi)
Microsatellites, also known as Short Tandem Repeats (STRs) are small (1-6 base pairs) repeating stretches of DNA scattered throughout the entire genome (both in coding and non-coding regions) and account for approximately 3 % of the human genome. Due to their repeated structure, microsatellites are prone to high mutation rate [bib_ref] Microsatellites: simple sequences with complex evolution, Ellegren [/bib_ref]. Microsatellite instability in tumor DNA is defined as the presence of alternate sized repetitive DNA sequences that are not present in the corresponding germline DNA. Microsatellite instability (MSI) is a molecular phenotype due to a defective DNA mismatch repair system.
The presence of MSI is found in the sporadic colon, gastric, sporadic endometrial and the majority of other cancers [bib_ref] Microsatellite instability: An update, Yamamoto [/bib_ref]. Determination of MSI status in CRC has prognostic and therapeutic implications. As well, MSI can be used diagnostically for tumor detection and classification [bib_ref] Microsatellite instability in colorectal cancer: clinicopathological significance, Setaffy [/bib_ref]. MSI has always been associated with an improved prognosis, stage for stage. Recently the reason for this has been discovered, a reason that has changed our approach to advanced MSI-high disease. The unstable microsatellites are highly immunogenic, so that therapy that activates the immune system can have almost miraculous effects on unstable tumors. This has led to proposals to develop tumor vaccines and to turn MSS tumors into MSI to make them more immunogenic.
## Detection of msi
MSI is detected indirectly by analysis of MMR protein expression by Immunohistochemical (IHC) staining, or directly by PCRbased amplification of specific microsatellite repeats, which is the most common method to detect MSI [bib_ref] Role of microsatellite instability in the management of colorectal cancers, Buecher [/bib_ref].
# Ihc method
Immunohistochemical analysis can determine loss of expression of one or more of MMR proteins. Actually, IHC is correlates with MSI but it is not a perfect test for MSI determination. It is a test of expression of mismatch repair proteins in cells. In this method, antibodies against MMR proteins such as MLH1, MSH2, PMS2 and MSH6 provide information of the MMR system functionality. IHC analysis with PMS2 and MSH6 antibodies is able to detect most abnormalities in the corresponding encoding genes as well as mutations in MLH1 and MSH2; however, IHC assay with MLH1/MSH2 antibodies can detect a fraction of MLH1 or MSH2 abnormalities but not all of them. Therefore, IHC analysis with MSH6 and PMS2 antibodies has more diagnostic potential than analysis with MLH1 and MSH2 antibodies [bib_ref] Immunohistochemistry versus microsatellite instability testing for screening colorectal cancer patients at risk..., Shia [/bib_ref].
The main relevance of MSI and IHC is as a screening test for Lynch Syndrome. Universal MSI/IHC on tumors is increasingly performed throughout the world.
# Pcr-based method
For MSI analysis by the fluorescent multiplex PCR-based method, we need DNA from tumor tissues and normal tissues, a series of primers one of which is fluorescently end labeled (the sense strand or antisense strand of each primer), a sequencer, and appropriate software. The principle of this method is to measure the presence of different lengths of specific microsatellite markers in tumor cells comparing to normal cells [bib_ref] Microsatellite instability in colorectal cancer: clinicopathological significance, Setaffy [/bib_ref]. (2) The proteins in sliding clamp interact with exonuclease-1 (EXO1) and proliferating cell nuclear antigen (PCNA). This complex excises the daughter strand back to the site of the mismatch. Finally, re-synthesize and re-ligation are performed by DNA polymerase and DNA ligase, respectively. The correction occurs .
In the first attempt to the diagnosis of MSI in CRC, a consensus conference recommended a panel of microsatellite markers included three dinucleotide repeats (D5S346, D2S123, and D17S250) and 2 mononucleotide repeats (BAT25 and BAT26). Three distinct MSI phenotypes have been described. If two or more microsatellite markers are mutated, the tumor is considered MSI-high (MSI-H); if only one is mutated, the tumor is defined as MSI-low (MSI-L); and if none of the examined loci demonstrate instability, the tumor will be considered Microsatellite Stable (MSS). This panel was known as the Bethesda panel [bib_ref] A National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome: meeting..., Rodriguez-Bigas [/bib_ref].
A few years later, it was found that mononucleotide markers have a better specificity and sensitivity than dinucleotide repeats (dinucleotide markers have a polymorphic nature) [bib_ref] Evaluation of tumor microsatellite instability using five quasimonomorphic mononucleotide repeats and pentaplex..., Suraweera [/bib_ref] and hence Bethesda guideline criteria were revised by NCI (National Cancer Institute) at the following conference in 2004 [bib_ref] Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite..., Umar [/bib_ref]. After that, the uses of panels containing more mononucleotide markers have been increased due to their higher sensitivity and specificity in the diagnosis of MSI in CRCs [fig_ref] Table 2: Microsatellite markers used to detect of MSI in CRC [/fig_ref] [bib_ref] Performance of different microsatellite marker panels for detection of mismatch repairdeficient colorectal..., Xicola [/bib_ref] [bib_ref] An optimized pentaplex PCR for detecting DNA mismatch repair-deficient colorectal cancers, Goel [/bib_ref] [bib_ref] Tumours with loss of MSH6 expression are MSI-H when screened with a..., You [/bib_ref] [bib_ref] A ten markers panel provides a more accurate and complete microsatellite instability..., Agostini [/bib_ref] [bib_ref] Quality assessment and correlation of microsatellite instability and immunohistochemical markers among population-and..., Cicek [/bib_ref].
## Msi in treatment
The uses of MSI status in the prediction of response to adjuvant chemotherapy is controversial although it has been confirmed that colorectal tumors displaying MSI have a better prognosis compared with MSS tumors. Antimetabolites (5-flourouracil), Alkylating agents and Topoisomerase Inhibitors are the three categories of chemotherapeutic agents that are used in CRC treatment . The chemotherapeutic treatment is effective in some certain patients, but it can cause many adverse effects, nonetheless [bib_ref] Mortality associated with Irinotecan plus bolus Flourouracil/Leucovorin. Summary findings of an independent..., Rothenberg [/bib_ref] [bib_ref] Prediction of the response of colorectal cancer to systemic therapy, Adlard [/bib_ref]. MSI-H is one of the potential predictive points to the chemotherapeutic treatment efficacy and to the level of adverse effects in a patient; therefore, several clinical trials have been conducted regarding this opinion (De la Chapelle and [bib_ref] Clinical relevance of microsatellite instability in colorectal cancer, De La Chapelle [/bib_ref]. There are different therapeutic responses in MSI-H CRCs depending on type of adjuvant chemotherapy. When a tumor is to be MSI-high, for diagnosis it is either Lynch or Methylated. If IHC is done and the unexpressed protein is MSH2, PMS2 or MSH6 then it is Lynch. Germline testing is indicated. If the unexpressed protein is MLH1 it could be a CIMP tumor with hypermethylation of MLH1 promoter, or Lynch. To tell which is which, BRAF mutation testing or methylation assay on the tumor are helpful. For treatment it is a candidate for immune activation therapy if it is advanced. If not advanced it has a good prognosis and will not respond to 5 FU based therapy.
## Conclusion and concluding remark
Since CRC is one of the most prevalent cancers in humans and causes a remarkable public health problem worldwide, identifying the ways of diagnosis and treatment of CRC is of most importance. MSI is a significant genetic marker in CRC that can be useful in diagnosis, prognosis, and prediction of chemotherapeutic treatment efficacy. Nowadays, molecular techniques have been developed for detection of MSI and drug development strategies are focused on specific tumor molecular characteristics.
[fig] Figure 1: Mechanism of mismatch repair system: (1) MutSα or MutSβ has recognized the mismatched DNA base pairs during replication that the DNA polymerase has matched the mistake base G (guanosine) in daughter strand with the T (thymidine) on the template. MutSα or MutSβ can recruit MutLα, MutLβ or MutLγ heterodimers by means of exchanging ATP to ADP. This complex (MutS-MutL) creates a sliding clamp around the DNA and moves along the new DNA chain when it encounters the DNA polymerase complex. [/fig]
[table] Table 1: Types of colorectal cancer and their genes [/table]
[table] Table 2: Microsatellite markers used to detect of MSI in CRC [/table]
|
The scalability of common paradigms for assessment of cognitive function: A functional transcranial Doppler study
Cognitive paradigms induce changes in cerebral blood flow (CBF) associated with increased metabolic demand, namely neurovascular coupling (NVC). We tested the hypothesis that the effect of complexity and duration of cognitive paradigms will either enhance or inhibit the NVC response. Bilateral CBF velocity (CBFV) in the middle cerebral arteries (MCAs) via transcranial Doppler ultrasound (TCD), blood pressure (BP), electrocardiogram (ECG) and end-tidal CO 2 (EtCO 2 ) of 16 healthy participants (aged 21-71 years) were simultaneously recorded at rest and during randomized paradigms of different complexities (naming words beginning with P-,R-,V-words and serial subtractions of 100-2,100-7,1000-17), and durations (5s, 30s and 60s). CBFV responses were population mean normalized from a 30-s baseline period prior to task initiation. A significant increase in bilateral CBFV response was observed at the start of all paradigms and provided a similar pattern in most responses, irrespective of complexity or duration. Although significant inter-hemispherical differences were found during performance of R-word and all serial subtraction paradigms, no lateralisation was observed in more complex naming word tasks. Also, the effect of duration was manifested at late stages of 100-7, but not for other paradigms. CBFV responses could not distinguish different levels of complexity or duration with a single presentation of the cognitive paradigm. Further studies of the ordinal scalability of the NVC response are needed with more advanced modelling techniques, or different types of neural stimulation.
# Introduction
Neurovascular coupling (NVC) describes the mechanism of cerebral blood flow (CBF) regulation through changes in arteriolar and capillary diameter to match increasing metabolic demand during neural activation [bib_ref] Neurovascular coupling in the normal brain and in hypertension, stroke, and Alzheimer..., Girouard [/bib_ref] [bib_ref] Regulation of cerebral blood flow, Peterson [/bib_ref]. Action potentials, resulting from neural stimulation, release neurotransmitters to the neurovascular unit (neurons, astrocytes, pericytes, endothelial cells and extracellular matrix components), where neuronal mediators, such as K + , NO and adenosine, control vascular smooth muscle contraction or relaxation [bib_ref] Blood-brain barrier, brain metabolism and cerebral blood flow, Paulson [/bib_ref] [bib_ref] Neurovascular regulation in the normal brain and in Alzheimer's disease, Iadecola [/bib_ref] [bib_ref] Vascular and haemodynamic issues of brain ageing, Beishon [/bib_ref]. response to cognitive task activations can be studied with a number of non-invasive methods, including magnetic resonance imaging (MRI) [bib_ref] Neurophysiological investigation of the basis of the fMRI signal, Logothetis [/bib_ref] , positron emission tomography (PET) [bib_ref] Decrease in parietal cerebral hemoglobin oxygenation during performance of a verbal fluency..., Hock [/bib_ref] , single-photon emission computed tomography (SPECT) [bib_ref] Regional cerebral blood flow patterns in visual imagery, Goldenberg [/bib_ref] , near infrared spectroscopy (NIRS) [bib_ref] Differences in the hemodynamic response to event-related motor and visual paradigms as..., Jasdzewski [/bib_ref] , laser Doppler flowmetry [bib_ref] Coupling of neural activation to blood flow in the somatosensory cortex of..., Ances [/bib_ref] and transcranial Doppler ultrasound (TCD) [bib_ref] Visually evoked dynamic blood flow response of the human cerebral circulation, Aaslid [/bib_ref]. MRI, PET and SPECT may cause patient discomfort, and are not suitable for screening and follow-up studies. However, TCD can measure CBF velocity (CBFV) at rest, or during cognitive tasks, usually referred to as functional TCD (fTCD) [bib_ref] AVERAGE: a Windows program for automated analysis of event related cerebral blood..., Deppe [/bib_ref]. fTCD has multiple advantages in terms of low-cost, high temporal resolution, portability to the bedside and non-harmful repetitive studies. Therefore, current TCD applications have been adopted to study a wide range of conditions such as sickle cell disease [bib_ref] TCD in sickle cell disease: an important and useful test, Adams [/bib_ref] , ischaemic [bib_ref] Effects of cerebral ischemia on human neurovascular coupling, CO2 reactivity, and dynamic..., Salinet [/bib_ref] or haemorrhagic [bib_ref] Transcranial Doppler monitoring in subarachnoid hemorrhage: a critical tool in critical care, Rigamonti [/bib_ref] stroke, brain stem death [bib_ref] Brain death and transcranial Doppler: experience in 130 cases of brain dead..., Ducrocq [/bib_ref] , intraoperative monitoring [bib_ref] The potential benefits and the role of cerebral monitoring in carotid endarterectomy, Pennekamp [/bib_ref] , cerebral microembolism in right-to-left cardiac shunt [bib_ref] Patent foramen ovale in older patients with cryptogenic stroke or transient ischaemic..., Thaler [/bib_ref] , and cognitive impairment [bib_ref] The Assessment of Cerebrovascular Response to a Language Task from the Addenbrooke's..., Beishon [/bib_ref] , including cerebral pressure autoregulation and neuroactivation . The quantification of the NVC response has been assessed by CBFV changes (ΔCBFV), induced by neural stimulation [bib_ref] AVERAGE: a Windows program for automated analysis of event related cerebral blood..., Deppe [/bib_ref]. Moreover, fTCD has been extensively used for assessing NVC in the lateralization of functional brain regions with a range of tasks, such as cognitive [23, 26-28], visual [bib_ref] Visually evoked dynamic blood flow response of the human cerebral circulation, Aaslid [/bib_ref] and sensorimotor paradigms.
In addition to detecting differences in lateralisation, a diagnostic tool to assess cognitive function needs to satisfy the condition of scalability, that is, to demonstrate the sensitivity to reflect differences in cognitive load, or its complexity. In the case of fTCD, this has been shown to be the case with visual stimulation , with changes in ΔCBFV detected in the posterior cerebral artery (PCA). With cognitive paradigms though, evidence of scalability is limited, and far from generalizable. In a few studies, fTCD has shown sensitivity to detect differences in difficulty and cognitive load, with the use of n-back paradigms . The nback paradigm stimulates specific aspects of cognition, involving attention and short-memory [36], but does not involve other aspects of cognitive function, such as language, fluency and visuospatial domains, which are part of a more comprehensive assessment of cognitive function . Previous studies have shown that fTCD can detect the NVC response to all five main cognitive domains of the Addenbrooke's Cognitive Examination (ACE-III) battery of tests, an important step to demonstrate the way fTCD might contribute to improving the diagnostic accuracy of routine cognitive assessment . Given that the ACE-III battery involves 20 different tasks, assessing the scalability of each of these paradigms is beyond the capacity of a single study. As a first step, we have chosen to study the scalability of naming words (NW) and serial subtraction (SS) as these tasks have been shown to induce consistent CBFV responses [26, 41] and good reproducibility . The relevance of NW and SS is that they are part, not only of the ACE-III, but also of other current methods of cognitive assessment such as the Mini Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) .
In summary, we tested the hypotheses that the complexity and duration of stimulation of NW and SS tasks can: 1) induce either larger or smaller magnitude of NVC response, and 2) lead to inter-hemispherical differences that could be the consequence of either enhanced or inhibitory mechanisms.
# Materials and methods
## Subjects and measurements
Sixteen healthy participants (eight women) without history of cardiovascular, neurological, or respiratory conditions, including metabolic abnormalities, inflammatory abnormalities, and severe somatic (e.g., cancer) or psychiatric (e.g., psychotic or bipolar), were recruited from the University of Leicester as students or staff. The Research Ethics committee approved the study (ref: 19452-rp9-ls:cardiovascularsciences), and fully informed, written consent was signed by each participant. Only adults 18 years old or older, with good insonation of bilateral temporal windows were suitable for study inclusion. The Edinburgh inventory was used for assessment of handedness .
All participants abstained from caffeine, alcohol, vigorous exercise, heavy meal, and smoking for at least 4 hrs before attending a temperature (22-24˚C) and light controlled laboratory. All measurements were undertaken with participants in the seated position. To avoid any artefacts due to movement, participants were asked to sit comfortably and avoid any movement, and all signals were checked prior to recording. The 2MHz TCD probes (Viasys Companion III device) were securely placed over the temporal windows using a head frame, and bilateral CBFV was monitored continuously in the middle cerebral artery (MCA) with the depth of insonation at 45-55 mm. Arterial volume clamping of the left middle finger digital artery was used for continuous beat-to-beat BP measurement (Finometer, Finapres Medical Systems; Amsterdam, the Netherlands), and automated brachial BP was also intermittently recorded in the right arm to calibrate the Finometer before each recording, using a validated OMRON (UA767) device. Continuous measurements were recorded for heart rate (HR), with a threelead electrocardiography (ECG), respiratory rate and end-tidal CO 2 (EtCO 2 ), using small nasal cannula (Salter Labs, ref 4000) with a capnograph (Capnocheck Plus).
## Paradigms of different complexity and duration
Prior to the tests, the paradigms were described in detail, and participants were encouraged to provide their best mental effort when providing their answers in a low voice. Each paradigm was recorded during 60s with subjects asked to keep an 'empty mind', Tasks were given verbally to participants during recording before task initiation. Thereafter, paradigms were performed over one of three different periods of duration stimulation, with a 60s recovery phase. Naming words (e.g. P-words) and serial number subtraction (subtract sequentially 7 from 100) paradigms have been applied routinely for diagnosing mild cognitive impairment (MCI) and dementia, and have been increasingly adopted as equivalent cognitive paradigms for assessment of NVC and also functional cerebral lateralization . Hence, both types of paradigms were modified to determine the effect of complexity and duration in this study. For modulating intensity of neuronal activations, naming words (NW) beginning with P-, R-and V-were adopted to represent increasing levels of difficulty corresponding to easy, moderate and hard, respectively . Also, three levels of difficulty in sequential subtractions (SS) were created with SS of (100-2), (100-7) and (1000-17), respectively .
Subjects rested for 15 min and a resting baseline recording was performed for 5 min . Two separate continuous recordings of cognitive paradigms were then obtained for duration and complexity data. R-words and subtraction (100-7) were used as a reference for representing word-naming (NW REF ) and serial subtraction (SS REF ), respectively. Paradigms of different complexity were randomized and presented during 60s. For changing the duration of presentation, the NW REF and SS REF paradigms were also randomised and then interrupted after 5, 30 or 60s, without participants' prior knowledge. The 60 s has been the traditional duration of NW and SS tasks , but a plateau is usually observed from 30 s onwards and using 5 s would only detect the potential effects of central command, before the metabolic deficit starts to act as a feedback error to correct the need for additional O 2 supply. The beginning and end of stimulation were marked with an on/off electrical switch to identify the duration of the paradigm and allow synchronization for coherent averaging of CBFV responses. All physiological signals were sampled at 500 Hz and recorded in the PHYSIDAS data acquisition system (Department of Medical Physics, University Hospitals of Leicester NHS Trust, Leicester, UK) for offline analysis.
# Data analysis
BP was calibrated at the beginning of each recording, and all signals were visually inspected to identify artefacts, with narrow (<0.1s) spikes, which were linearly interpolated. Bilateral CBFV signals were passed through a median filter with a window width of five samples to remove random noise. All signals were low-pass filtered by a zero-phase Butterworth filter with a cutoff frequency of 20 Hz. RR interval was automatically marked from the ECG, and HR was continuously plotted against time and manually corrected under visual inspection in case of missed markings. Mean values of CBFV and BP were calculated for each cardiac cycle. Linear interpolation was applied to obtain estimates of EtCO 2 synchronized to the end of each cardiac cycle. Beat-to-beat data were spline-interpolated with a third order polynomial and resampled at 5 Hz to obtain signals with a uniform time base.
Each variable was synchronized by the start of each paradigm, to obtain the coherent average of CBFV for each subject. Bilateral CBFV were expressed in percent (%) by normalization by the 30s of baseline before the paradigm activation. The synchronization placed the beginning of each paradigm at 60s, within a window of 180s duration. Only bilateral CBFV signals were averaged for 10s during five time segments, with the beginning of stimulation starting at 60s: baseline (T1) at 40-50s, (T2) at 60-70s, (T3) at 110-120s, (T4) at 120-130s, and (T5) at 160-170s. Therefore, T1 represents baseline values before stimulation, T2 the immediate response to stimulation, T3 the sustained response, T4 near the end of the response (for 60 s duration of stimulation) and T5 during recovery. All timepoints were fixed for all paradigms, independently of the duration of presentation.
# Statistical analysis
All data were tested for normality using Shapiro-Wilk test. Continuous data with a Gaussian distribution are presented as mean (SD). Outliers of parameters were determined by using the boxplot with interquartile limits between 25 th and 75 th percentile (interquartile range, IQR) for indicating extreme values beyond 1.5 times (above or below) of the IQR as outliers which were modified by replacing the second smallest or largest values in observations to the threshold boundaries as recommended in Winsorization . Paired t-test (parametric data) or Wilcoxon test (nonparametric data) were used to detect differences in hemispherical comparisons of CBFV response to each paradigm. Two-way ANOVA was adopted to test for differences between time segments and paradigms for interaction in each bilateral hemisphere. Post hoc analysis with Tukey's was applied for multiple comparisons. Analyses were conducted using SPSS version 25 for Windows or Graphpad Prism version 9.2 for Windows. Statistical significance was adopted at a value of p < 0.05.
# Results
Sixteen healthy participants (eight women) were recruited (mean age 31.6 years, [range 21-71 years] and mean BMI 22.6 ± 3.0 kg/cm 2 ), and completed all the NW and SS paradigms of variable duration and complexity. A majority of participants (n = 14) were right-handed and two subjects were ambidextrous, following assessment with the Edinburgh handedness inventory. Half of all participants had Asian background, six were Caucasian, and two subjects had Caribbean and African ancestry, respectively. Eight people abstained from alcohol and two subjects were current smokers. Absolute mean (SD) values of baseline CBFV did not show a significant difference between right hemisphere (RH) and left hemisphere (LH) at 66.8 (11.4) and 64.8 (11.4) cm/s, respectively. Baseline values for mean BP, HR and EtCO 2 were 89.7 (9.4) mmHg, 70.8 (8.1) bpm and 38.5 (3.1) mmHg, respectively.
## Temporal cbfv responses during cognitive paradigms
The overall paradigm-synchronized patterns of bilateral cerebral haemodynamic coherent averages for the whole population with increasing complexity of NW [fig_ref] Fig 2: p < 0 [/fig_ref] and SS [fig_ref] Fig 2: p < 0 [/fig_ref] showed considerable similarity. A rapid increase in bilateral CBFV was observed at the start of stimulation (grey bar), with the largest peak around T2 and reducing rapidly towards baseline values before T3, around 25s after stimulation. There was a further small rise in CBFV around T3, except for SS (100-2), and a steady return to baseline after the paradigm was completed at approximately 20s. The 2-way ANOVA confirmed a highly significant effect of time for NW and SS [fig_ref] Fig 2: p < 0 [/fig_ref] [fig_ref] Table 1: Mean [/fig_ref] steep increase to the highest peak at T2 for all durations was observed bilaterally. Thereafter, there was a reduction to baseline, though a rebound increase in bilateral CBFV to a second peak at T3 was seen for the 60s duration paradigm. The 2-way ANOVA corroborated the [fig_ref] Table 1: Mean [/fig_ref] 100-2 -0.6(1.2) 6.9(4.0) a 3.0(7.4) 2.6(7.5) � -0. [bib_ref] Blood-brain barrier, brain metabolism and cerebral blood flow, Paulson [/bib_ref] , p < 0.001).
[formula] R-word -1.3(1.7) 6.1(3.7) a -2.2(5.9) b -0.6(6.3) b -2.9(4.3) b V-word -0.3(1.1) 8.8(5.3) a 1.3(8.0) b 1.1(6.4) b -1.8(3.9) b SSR-word -1.2(1.5) 6.7(2.9) a 0.8(7.9) b 0.1(6.1) b -3.2(6.1) b V-word -0.5(1.1) 8.5(3.4) a 1.8(6.5) b 0.5(5.6) b -2.6(2.9) b SS [/formula]
## Effect of task complexity on cerebral haemodynamic responses
Increases in the complexity of NW and SS tasks induced similar temporal CBFV changes during neural activation as no obvious difference was seen in its effect on CBFV responses within each time period of stimulation. [fig_ref] Table 1: Mean [/fig_ref] presents values of bilateral CBFV changes at each time segment for both NW and SS tasks, in both RH and LH. There was no significant interaction effect on bilateral CBFV responses between complexity paradigms and time segments on two-way ANOVA [fig_ref] Table 1: Mean [/fig_ref]. There were no significant differences between CBFV changes for NW and SS complexities at each time-point.
## Effect of task duration on cerebral haemodynamic responses
Differences in the duration of NW REF and SS REF tasks induced a similar pattern of temporal CBFV responses [fig_ref] Fig 3: Average population responses [/fig_ref] , with a lack of significance in its effect on CBFV changes at initial NVC activation (
## Interhemispheric differences
Overall, there were contradictory responses between RH and LH regarding the dominance of CBFV responses during SS and NW tasks, respectively. Significant interhemispheric differences were observed for RH with SS and LH for NW tasks but only for the timepoints T3-T5 [fig_ref] Table 1: Mean [/fig_ref]. In paradigms of different complexity, CBFV response for LH was significantly greater than RH (p = 0.028) during R-words at T3. On the other hand, for SS, RH had significantly higher CBFV response than LH at T4 (100-2, p = 0.041 and 100-7, p = 0.043) and T5 (1000-17, p = 0.007). However, hemispherical differences in both NW REF and SS REF were not seen at each time segment during 05s and 30s durations.
# Discussion
## Main findings
As a non-invasive tool, fTCD has often been used to detect changes in cerebral haemodynamic responses during neural activation using cognitive paradigms [bib_ref] Visually evoked dynamic blood flow response of the human cerebral circulation, Aaslid [/bib_ref]. In this novel study of the effects of paradigm complexity and duration, we focused on the coherent averages of CBFV, as the dominant technique that has been used for assessment of NVC with fTCD. Our main finding was that in healthy participants, the temporal CBFV response failed to demonstrate sensitivity to NW and SS cognitive paradigms of varying complexity and duration. By showing similar patterns of otherwise highly significant alterations in the CBFV response at the onset of all tasks, these changes, and the subsequent temporal evolution of the CBFV response was not consistently sensitive to detect any effect of complexity or duration with both the NW and SS paradigms. However, different CBFV responses were observed during the sustained response and at the end of stimulation, depending on the type of paradigm undertaken. More work is therefore needed to obtain a better understanding of the NVC response to these commonly used cognitive tasks and the poor sensitivity they express to changes in cognitive load.
## Temporal pattern characteristics of ftcd responses to cognitive tasks
The highly significant bilateral CBFV differences observed across time [fig_ref] Table 1: Mean [/fig_ref]. Neurogenic regulation may therefore nourish a rapid adaptive response with uncoupling of the cortical metabolic rate during initial stimulation. For the second peak, a gradual CBFV increase was seen as the sustained response (T3), with differences related to the complexity of NW and SS, except for SS (100-2). This slow phase of the CBFV response could reflect metabolic regulation that induced decreases in local microvascular resistance depending on the actual level of tissue CO 2 concentration [49]. However, in SS (100-2), the absence of a second peak might be due to this being the easiest paradigm [52], combined with a desire to quickly complete the task with rapid answers and associated hyperventilation. Therefore, an attenuated CBFV response might be particularly associated with retrieving arithmetic facts in simple calculations
## Influence of paradigm complexity on ftcd responses
Paradigms of different complexity induced CBFV responses with similar temporal patterns, and without significant differences in the amplitude of the CBFV changes, or evidence of lateralization [fig_ref] Table 1: Mean [/fig_ref]. Although our cognitive-challenging tasks did not provide CBFV responses with consistent interhemispheric differences, our findings were in line with other studies showing reduced lateralisation with increasing task difficulty, possibly due to recruitment of additional neural networks compensating for increased cognitive demands . Also, most paradigms used in common cognitive assessments have not resulted in significant lateralisation . On the other hand, there is evidence in the imaging literature that other techniques for NVC assessment might be more responsive to differences in cognitive effort. In fMRI investigations, the consequences of increasing complexity in verbal fluency performance showed greater activation in the anterior cingulate region [54], with activation predominantly in the right cerebellar hemisphere [55], and with laterality increased in the posterior part of the inferior frontal gyrus during mentally complex calculation [bib_ref] Laterality of brain areas associated with arithmetic calculations revealed by functional magnetic..., Zhang [/bib_ref]. This might help to explain our results with increasing complexity of NW and SS tasks, respectively. However, only the inferior frontal gyrus is part of the territory perfused by the MCA [bib_ref] Probabilistic anatomic mapping of cerebral blood flow distribution of the middle cerebral..., Kim [/bib_ref]. Thus, changes in CBF, resulting from paradigms of different complexity, might occur in regions other than those perfused by the MCA. As reported by [bib_ref] Where's Waldo? The utility of a complicated visual search paradigm for transcranial..., Smirl [/bib_ref] , using different levels of challenge in visual search paradigms, it was possible to discriminate ΔCBFV responses in the posterior cerebral artery (PCA), but ΔCBFV changes to visual paradigms of different complexity could not be detected in the MCA [bib_ref] Where's Waldo? The utility of a complicated visual search paradigm for transcranial..., Smirl [/bib_ref]. Similarly, bilateral CBFV changes showed a larger response during the Stroop test with cognitive incongruent (difficult) than congruent (easy) stimuli . With more demanding tasks, larger ΔCBFV was seen bilaterally during n-back paradigms compared to less demanding tasks . Also, the differences between interhemispheric activations [49, 59, 60] could be discriminated by task difficulty. These pronounced differences probably depend on duration of stimulation, type and complexity of paradigms .
Noteworthy, the majority of previous studies used repeated presentations of the same paradigm to improve signal-to-noise ratio (SNR) of intra-subject averages of the ΔCBFV response. As seen in previous studies, presentation of complex tasks was repeated a variable number of times, including < 5 [bib_ref] Lateralization of cerebral blood flow velocity changes during cognitive tasks. A simultaneous..., Vingerhoets [/bib_ref] , 5 to or > 10 times [bib_ref] Inferring common cognitive mechanisms from brain bloodflow lateralization data: a new methodology..., Meyer [/bib_ref]. Studies with repeated task presentation were likely to provide more consistent responses across individuals, as compared to the present study where tasks were presented only once.
There are potential additional explanations for the lack of significance in our results to the effects of cognitive load with NW and SS paradigms. First, the differences in complexity of paradigms in the current study might not have been sufficient to induce differential CBFV responses, as proposed in terms of the compensation-related utilization of neural circuits hypothesis (CRUNCH) to recruit more neural activations matching with load increases [bib_ref] The adaptive brain: aging and neurocognitive scaffolding, Park [/bib_ref] [bib_ref] The CRUNCH model does not account for load-dependent changes in visuospatial working..., Jamadar [/bib_ref]. As reported for other neuroimaging techniques (fMRI or SPECT), it provided greater activation in some specific brain areas, with increased task difficulty with verbal [54], arithmetic [bib_ref] Laterality of brain areas associated with arithmetic calculations revealed by functional magnetic..., Zhang [/bib_ref] or visual [bib_ref] Non-invasive assessment of cerebral blood flow changes during complex activation, Lisak [/bib_ref] paradigms. Although NVC assessment with fTCD has been rated as equivalent to that obtained with fMRI [bib_ref] Assessment of hemispheric language lateralization: a comparison between fMRI and fTCD, Deppe [/bib_ref] , this does not seem to be the case in our study, with respect to the ability of fTCD to reflect differences in complexity. We can speculate that increasing tasks difficulty cannot induce the CRUNCH effect on TCD-measured CBFV changes.
Secondly, the phenomenon of 'central command' anticipation of the brain when preparing to receive an increased demand for oxygen may be important. To our knowledge, the role of the 'central command' during cognitive paradigms in fTCD studies has received very little attention in the human NVC literature [bib_ref] What is the key mediator of the neurovascular coupling response?, Hosford [/bib_ref]. Previous studies suggested that the preparatory process of blood flow adjustment involves alertness of attentional function [bib_ref] Cerebral blood flow modulations during antisaccade preparation in chronic hypotension, Duschek [/bib_ref] , correlated with increasing CBF response before the onset of exercise that is likely governed by the 'central command' network [bib_ref] Influence of central command on cerebral blood flow at the onset of..., Sato [/bib_ref]. Therefore, increased CBFV at the onset of all paradigms, independently of complexity or duration, was probably determined by the effects of 'central command' on the neurovascular unit.
Thirdly, cognitive stimulation of some brain areas leads to inhibition of other cortical centres [bib_ref] The neural basis of cognitive control: response selection and inhibition, Goghari [/bib_ref]. Despite the cerebral cortex circulation supporting the local metabolism during cognitive activation [bib_ref] Functional magnetic resonance imaging of human brain activity in a verbal fluency..., Schlosser [/bib_ref] , there is evidence in the literature showing a decreased activation in the medial prefrontal, precuneus, posterior cingulate and left angular gyrus, near to Wernicke's area (language function), following increased cognitive load of NW tasks [bib_ref] The neurodevelopmental differences of increasing verbal working memory demand in children and..., Vogan [/bib_ref]. Similarly, [bib_ref] Developmental cognitive neuroscience of arithmetic: implications for learning and education, Menon [/bib_ref] highlighted a shift in the process of brain response from prefrontal cortex to the bilateral posterior parietal cortices, including additional brain areas (caudate nucleus and cerebellum) during complex mental arithmetic [bib_ref] Developmental cognitive neuroscience of arithmetic: implications for learning and education, Menon [/bib_ref]. It is plausible that the impact of increasing cognitive complexity might be a true decrease in some local brain areas [bib_ref] The neurodevelopmental differences of increasing verbal working memory demand in children and..., Vogan [/bib_ref] [bib_ref] Developmental cognitive neuroscience of arithmetic: implications for learning and education, Menon [/bib_ref]. The expected increased response in CBFV following stimulation might therefore be overshadowed by the increase in inhibition with greater cognitive demand. Finally, increases in the amplitude of the CBFV response, following cognitive paradigms of increasing complexity, could have been confounded by concomitant changes in systemic covariates. The influence of sympathetic activation inducing a rise of BP and HR during cognitive paradigms has previously been reported, and changes in EtCO 2 with hypocapnia-induced hyperventilation or hypercapnia-induced breath holding can also accompany CBFV alterations during cognitive tasks [bib_ref] Bilateral simultaneous assessment of cerebral flow velocity during mental activity, Silvestrini [/bib_ref]. Furthermore, increases in HR were reported at the beginning of more complex cognitive tasks [48], resulting from arousal or stress. The contribution of BP and PaCO 2 to CBFV changes following cognitive tasks has been confirmed in previous fTCD studies [bib_ref] Contribution of arterial blood pressure and PaCO2 to the cerebrovascular responses to..., Panerai [/bib_ref]. However, whether these co-variates can obscure the changes in CBFV response that could be expected with increasing complexity needs to be investigated with the use of more advanced multivariate dynamic modelling techniques.
## Influence of paradigm duration on ftcd responses
For our hypothesis regarding paradigms of different durations, it would be expected that shorter or extended duration paradigms would result in different magnitude changes of CBFV following cognitive activation. With different durations for both NW REF and SS REF , temporal CBFV responses showed similar patterns at the first peak of initial stimulation (T2), without significant differences in the amplitude of CBFV or lateralized dominance. Only CBFV responses to the longer (60s) stimulations led to a second peak for SS REF. Despite the 60s stimulation being more likely to induce greater metabolic demand, than 5s and 30s durations, the magnitude of CBFV changes was lower than for stimulation with 30s duration [fig_ref] Fig 3: Average population responses [/fig_ref]. These results were consistent with previous studies of different durations of stimulation, which demonstrated that the longer duration stimulation would not necessarily maximize the CBFV magnitude change [bib_ref] Age and stimulus dependency of visually evoked cerebral blood flow responses, Panczel [/bib_ref].
To our knowledge, there have been no previous fTCD studies of the effects of changing cognitive paradigm duration on the induced changes in CBFV. However there have been previous TCD studies using visual stimulation of different durations [bib_ref] Influence of stimulus duration on the neurovascular coupling response, Rosengarten [/bib_ref] [bib_ref] Age and stimulus dependency of visually evoked cerebral blood flow responses, Panczel [/bib_ref] [bib_ref] Temporal patterns of evoked cerebral blood flow during reading, Tiecks [/bib_ref]. The results of these studies were consistent, showing an identical curve of CBFV response at the initial phase, independent of the duration of stimulation, as seen in other imaging literature (e.g. BOLD) [bib_ref] Coupling of neural activation to blood flow in the somatosensory cortex of..., Ances [/bib_ref] [bib_ref] Nonlinear temporal dynamics of the cerebral blood flow response, Miller [/bib_ref] and our paradigms. According to the 'central command' hypothesis, the lack of marked differences in CBFV responses with varying durations could be explained by the anticipatory response to cognitive demand in the attentional process [bib_ref] Cerebral blood flow modulations during antisaccade preparation in chronic hypotension, Duschek [/bib_ref] , governing an upsurge of CBFV around 5s after onset stimulation, without further influence of the duration of stimulation on the NVC mechanism.
In tasks with 60s duration, a slow CBFV peak was found in both hemispheres before T3 in SS REF [fig_ref] Fig 3: Average population responses [/fig_ref]. This is in contrast to CBFV patterns seen in previous studies with stimuli of varying duration, that demonstrated a plateau phase with longer duration stimulation, using both fTCD [bib_ref] Influence of stimulus duration on the neurovascular coupling response, Rosengarten [/bib_ref] and BOLD [bib_ref] Nonlinear temporal dynamics of the cerebral blood flow response, Miller [/bib_ref]. The difference in CBFV responses might depend on the type of paradigm. The plateau curve might relate to an adaptative brain response as a compensatory function [bib_ref] BOLD correlates of trial-by-trial reaction time variability in gray and white matter:..., Yarkoni [/bib_ref] , but the two-peaked pattern in CBFV response seen in our and other studies [46, 48] might be confounded from involuntary hyperventilation-induced hypocapnia after the first peak [bib_ref] Transcranial Doppler study of the cerebral hemodynamic changes during breath-holding and hyperventilation..., Settakis [/bib_ref] which affects approximately 30s of the magnitude of the CBFV response. Accordingly, a second peak will be missing in paradigms of shorter (5s and 30s) duration.
## Clinical implications and limitations
This is the first systematic study to determine fTCD-measured NVC responses to NW and SS paradigms of varying complexity (easy, moderate and hard) and duration (shorter, intermediate and longer) by modifying well known paradigms used for clinical assessment of cognitive performance [37-39, 41]. The results suggest that fTCD, based on a single presentation of a paradigm, provides questionable results, regarding its feasibility to detect alterations in mild cognitive impairment (MCI) and other conditions. From previous studies in our group, demonstrated a significant decrease of CBFV at sustained response during a language task in Alzheimer's disease but not in MCI when compared with healthy controls [bib_ref] The Assessment of Cerebrovascular Response to a Language Task from the Addenbrooke's..., Beishon [/bib_ref]. Although the mean of repeated task presentations to each patient might provide more robust estimates of the CBFV changes [bib_ref] A method for the automated assessment of temporal characteristics of functional hemispheric..., Deppe [/bib_ref] [bib_ref] Reproducibility of functional transcranial Doppler sonography in determining hemispheric language lateralization, Knecht [/bib_ref] , combining the use of fTCD with other modalities (EEG or near infrared spectroscopy) may enhance discriminatory power in clinical NVC studies [bib_ref] Simultaneous acquisition of EEG and NIRS during cognitive tasks for an open..., Shin [/bib_ref]. These and other alternatives need to be explored further to provide greater confidence in the use of fTCD in clinical studies of NVC.
There are a number of limitations to our study. First, TCD reflecting CBFV changes in the MCA, can be directly proportional to changes in absolute CBF with the assumption that the diameter of insonated MCA remains constant despite small changes in CO 2 and BP [bib_ref] MRI measures of middle cerebral artery diameter in conscious humans during simulated..., Serrador [/bib_ref] [bib_ref] Cerebral arterial diameters during changes in blood pressure and carbon dioxide during..., Giller [/bib_ref]. Second, fTCD measurements limited to the MCA might not have the same sensitivity to cognitive stimulation as techniques with greater spatial resolution, such as PET or fMRI . One interesting possibility would be to perform simultaneous measurements in the MCA and PCA as an attempt to improve sensitivity. Third, language task difficulty has been reported to have reduced sensitivity [bib_ref] Using functional transcranial Doppler ultrasonography to assess language lateralisation: Influence of task..., Badcock [/bib_ref] [bib_ref] Lateralised visual attention is unrelated to language lateralisation, and not influenced by..., Rosch [/bib_ref] , although [bib_ref] Stimulus rate increases lateralisation in linguistic and non-linguistic tasks measured by functional..., Payne [/bib_ref] showed differences in fTCD responses to the joint fast paced presentation of rhyme and line array judgement tasks [bib_ref] Stimulus rate increases lateralisation in linguistic and non-linguistic tasks measured by functional..., Payne [/bib_ref]. It is possible that the level of difficulty used in our NW paradigms was insufficient to induce differential changes in CBFV responses in the MCA. In addition, the motivation level of participants to engage and perform the tasks to the best of their ability could have affected their CBFV responses, despite our encouragement prior to each test. The role of personal motivation, and how it could be individually assessed, is complex and warrants further investigation [bib_ref] A Brief Review of Research Using Near-Infrared Spectroscopy to Measure Activation of..., Bendall [/bib_ref]. Fourth, the education level could be an important co-variate of cognitive performance. We have not taken this aspect into consideration, given the homogeneity of educational background in our sample, with all participants having had between 16 and 22 years of education. A recent review has highlighted inconsistent findings in previous studies about the effect of ageing on cognitive function [bib_ref] Vascular and haemodynamic issues of brain ageing, Beishon [/bib_ref]. For this reason, further studies are needed to understand the influence of ageing in healthy older adults, and its interaction with the NVC response to tasks with varying complexities and durations. Finally, another limitation was the relatively small sample size (n = 16). However, previous sample size calculations showed that 14 participants could detect a difference of 2% in CBFV at 80% power with p = 0.05, for the case of the Pwords task of 60s duration . Although our sample size was approximately the same, it is possible that much larger numbers are needed to detect the subtle differences resulting from NW tasks of greater or lesser complexity in comparison with naming P-words.
# Conclusions
We have demonstrated that complexity and duration of NW and SS paradigms provided similar patterns and amplitudes of CBFV responses, without the possibility of making a scalable metric for assessment of NVC in healthy subjects. Nevertheless, some significant changes were seen in comparison with SS REF for the effects of different task durations, indicating a complex interaction between NVC response and duration of stimulation. Further studies, on the sensitivity of the fTCD response to stimulation of varying complexity and duration, is needed by performing similar studies in the influence of ageing, patients with early signs of MCI, Alzheimer's disease or other types of dementia, exploring alternative cognitive paradigms. Also, a more integrated approach as described previously [bib_ref] Cerebral blood flow response to neural activation after acute ischemic stroke: a..., Salinet [/bib_ref] is needed to take into consideration the potential contribution of other determinants of CBFV, such as BP and EtCO 2 , to identify parameters of the NVC response that could be scalable, ideally in a linear fashion.
[fig] Fig 2: p < 0.001). Changes in paradigm duration, also led to similar patterns of the CBFV bilateral coherent averages, for both the NW REF (R-word) (Fig 3A-3C) and SS REF (100-7) tasks (Fig 3D-3F). A Average population responses (n = 16) in the right (continuous line) and left (dashed line) hemispheres for the complexity of NW (A, B, C) and SS (D, E, F) paradigms of CBFV, respectively. Error bar represents the largest ±SD at the point of occurrence. Changes are shown before, during (grey bar) and after 60s of P-word (A), R-word (B), V-word (C), 100-2 (D), 100-7 (E) and 1000-17 (F). Responses were synchronized by the beginning of stimulation, set at 60s. For statistical analyses, shorter segments (T1-T5) were averaged from 40-50s (T1), 60-70s (T2), 110-120s (T3), 120-130s (T4) and 160-170s (T5), respectively. https://doi.org/10.1371/journal.pone.0266048.g002 [/fig]
[fig] Fig 3: Average population responses (n = 16) in the right (continuous line) and left (dashed line) hemispheres for the duration of NW REF ; R-word (A, B, C) and SS REF ; 100-7 (D, E, F) paradigms of CBFV, respectively. Error bar represents the largest ±SD at the point of occurrence. Changes are shown before, during (grey bar) and after each duration of 05s of NW REF (A), 30s of NW REF (B), 60s of NW REF (C), 05s of SS REF (D), 30s of SS REF (E) and 60s of SS REF (F). Responses were synchronized by the beginning of stimulation, set at 60s. For statistical analyses, shorter segments (T1-T4) were averaged from 40-50s (T1), 60-70s (T2), 110-120s (T3) and 120-130s (T4), respectively. https://doi.org/10.1371/journal.pone.0266048.g003 change in bilateral CBFV response to different duration stimulations, with a highly significant effect of time for NW REF and SS REF ( [/fig]
[table] Table 1: Mean (SD) values of bilateral cerebral hemodynamic changes at each timepoint for increasing complexity of NW and SS paradigms. [/table]
[table] Table 2: Mean (SD) values of bilateral cerebral hemodynamic changes at each timepoint for different duration of NW REF and SS REF paradigms.CBFV was bilaterally normalized at baseline values. Sample size = 16, T1-T4 represents each extracted timepoint for statistical analyses. Interaction showed p-value of paradigm x timepoint effect by two-way repeated measures ANOVA followed by post-hoc analysis with Tukey's multiple comparison. CBFV; cerebral blood flow velocity, RH; right hemisphere, LH; left hemisphere, s; second, NW REF ; reference word-naming and SS REF ; reference serial subtraction. a p < 0.05 for difference compared with baseline (after Tukey correction) b p < 0.05 for difference compared other timepoints with T2 (after Tukey correction) c p < 0.05 for difference between paradigm comparisons (after Tukey correction) [/table]
|
Improvements in productivity at paid work and within the household, and increased participation in daily activities after 24 weeks of certolizumab pegol treatment of patients with psoriatic arthritis: results of a phase 3 double-blind randomised placebo-controlled study
# Introduction
Psoriatic arthritis (PsA) is a chronic inflammatory arthritis which affects up to 30% of patients with psoriasis. [bib_ref] Psoriatic arthritis: a critical review, Dhir [/bib_ref] [bib_ref] High prevalence of psoriatic arthritis in patients with severe psoriasis with suboptimal..., Haroon [/bib_ref] [bib_ref] Prevalence of rheumatologist-diagnosed psoriatic arthritis in patients with psoriasis in European/North American..., Mease [/bib_ref] Over half of the patients with PsA will go on to exhibit a more severe, erosive form of the disease that can be associated with rapidly progressive joint damage and substantial functional impairment. [bib_ref] Psoriatic arthritis (PA): a clinical, immunological and radiological study of 180 patients, Alonso [/bib_ref] [bib_ref] A prospective, clinical and radiological study of early psoriatic arthritis: an early..., Kane [/bib_ref] [bib_ref] Progression of peripheral joint disease in psoriatic arthritis: a 5-yr prospective study, Mchugh [/bib_ref] The complex nature of PsA can also result in a considerable negative impact on multiple physical and emotional aspects of patients' lives as well as severe work disability and loss of productivity. [bib_ref] Work disability in psoriatic arthritis: a systematic review, Tillett [/bib_ref] This has a significant impact on the quality of life as well as financial status of individuals.
While there is published evidence on the burden of the disease on work disability in related rheumatic diseases such as rheumatoid arthritis (RA) and ankylosing spondylitis (AS), [bib_ref] The economic burden of disease: comparison between rheumatoid arthritis and ankylosing spondylitis, Boonen [/bib_ref] [bib_ref] Work productivity in rheumatoid arthritis: relationship with clinical and radiological features, Del Moral [/bib_ref] [bib_ref] The economic burden of rheumatoid arthritis: beyond health care costs, Zhang [/bib_ref] [bib_ref] Review of the costs of illness of ankylosing spondylitis and methodologic notes, Boonen [/bib_ref] to date there are few data on work disability in PsA. Some evidence indicates that the employment rate in patients with PsA is significantly lower than that seen in the general population, slightly lower than in the AS population but higher than in the RA population. [bib_ref] Employment across chronic inflammatory rheumatic diseases and comparison with the general population, Mau [/bib_ref] In contrast to RA, very little is known about the indirect costs associated with PsA and its treatment. The age of onset of PsA is in the 30-40s for men and women, an age at which many men and women would be in the midst of working careers. Additionally, two separate studies to identify issues of concern among PsA patients reported the reduced ability to work or volunteer [bib_ref] Willingness-to-pay stated preferences for 8 health-related quality-of-life domains in psoriatic arthritis: a..., Hu [/bib_ref] and the reduced ability to fulfil personal roles, social life and work, as common concerns. [bib_ref] Development of the PsAQoL: a quality of life instrument specific to psoriatic..., Mckenna [/bib_ref] Reports from several studies of tumour necrosis factor (TNF) inhibitor in PsA have suggested that these biologic agents can have a positive impact on productivity in PsA patients, improving employment status and work productivity compared to placebo. [bib_ref] Effectiveness and safety of etanercept in patients with psoriatic arthritis in a..., Gladman [/bib_ref] [bib_ref] Effect of infliximab therapy on employment, time lost from work, and productivity..., Kavanaugh [/bib_ref] [bib_ref] Patient-reported outcomes and the association with clinical response in patients with active..., Kavanaugh [/bib_ref] However, further research is required to better understand the impact of PsA and TNF inhibitor therapies on workplace productivity, as well as on daily activities within the household and participation in social or leisure activities.
The efficacy and safety of certolizumab pegol (CZP), a PEGylated Fc-free anti-TNF agent, in PsA patients has been investigated and reported in the RAPID-PsA trial (NCT01087788). [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] Here we report on the impact of CZP on productivity in paid and household work, and on participation in family, social and leisure activities, in patients with active PsA.
# Methods patients
The RAPID-PsA trial (NCT01087788) randomised 409 patients aged 18 years or older, with a diagnosis of adult-onset active PsA of ≥6 months' duration Open Access Scan to access more free content as defined by the ClASsification criteria for Psoriatic ARthritis (CASPAR). [bib_ref] Classification criteria for psoriatic arthritis: development of new criteria from a large..., Taylor [/bib_ref] All recruited patients had to have active psoriatic skin lesions or a documented history of psoriasis. In addition, erythrocyte sedimentation rate ≥28 mm/h (Westergren) or C-reactive protein levels above the upper limit of normal (ULN; 7.9 mg/L) was required. Patients must have previously failed treatment with, or been resistant to, at least one disease modifying antirheumatic drug and up to 40% of patients could have previously been treated with a TNF inhibitor. Patients were excluded from the study if they had a form of inflammatory arthritis other than PsA or had a secondary, noninflammatory condition symptomatic enough to interfere with evaluation of CZP for PsA. Patients were also excluded if they had received previous treatment for PsA or psoriasis with more than two biologics or at least two TNF inhibitors, or had primary failure to a prior TNF inhibitor.
## Study design
RAPID-PsA is an ongoing, phase 3, multicentre study in PsA patients. Details of the RAPID-PsA trial design are reported elsewhere. [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] In brief, the study was double-blind and placebocontrolled to week 24, dose-blind to week 48 and is open-label to week 216. Patients were randomised 1:1:1 to placebo, or 400 mg CZP at weeks 0, 2 and 4 (loading dose) followed by either 200 mg CZP every 2 weeks (Q2W) or 400 mg CZP every 4 weeks (Q4W), administered subcutaneously by investigators using a blinded prefilled syringe. [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] Randomisation was stratified according to investigator site and prior TNF inhibitor exposure according to an interactive voice response system.
Placebo patients who failed to achieve a 10% improvement from baseline in tender and swollen joint counts at weeks 14 and 16 underwent mandatory escape and were re-randomised at week 16 to receive CZP 200 mg Q2W or CZP 400 mg Q4W following loading dose at weeks 16, 18 and 20. The clinical primary endpoints were ACR20 response at week 12 and change from baseline in modified Total Sharp Score (mTSS) at week 24. CZP safety was also investigated and is reported elsewhere. [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref]
## Work productivity survey
The Work Productivity Survey (WPS) is a novel validated questionnaire assessing the impact of arthritis-related symptoms on patient productivity in the workplace and at home, and on participation in family, social and leisure activities. [bib_ref] Discriminant validity, responsiveness and reliability of the rheumatoid arthritis-specific Work Productivity Survey..., Osterhaus [/bib_ref] The WPS has demonstrated good psychometric properties (ie, discriminant validity, responsiveness to clinical changes and reliability) in RA, [bib_ref] Validity, responsiveness and reliability of the arthritis-specific work productivity survey assessing work..., Osterhaus [/bib_ref] PsA 26 and axial spondyloarthritis (axSpA) (unpublished data).
The WPS is self-reported, but interviewer-administered and has a recall period of 1 month. The first question addresses employment status and provides additional information on job type for employed patients (non-manual, manual with no supervisory duties, or mixed) and on the status of those not employed. For employed patients, three questions relate to paid work productivity outside the home. These questions assess over the past month: absenteeism (full days of work missed due to arthritis); presenteeism (days with work productivity reduced by ≥50% due to arthritis); and the level of interference of arthritis on work productivity (on a 0-10 scale; where 0=no interference and 10=complete interference).
All patients, regardless of their employment status, answer five questions which are related to household work and daily activities. These questions specifically assess over the previous month: the number of days with no household work performed due to arthritis; days with household productivity reduced by ≥50% due to arthritis; days with outside help hired due to arthritis; days with family, social or leisure activities missed due to arthritis; and the level of arthritis interference with household productivity (on a 0-10 scale; where 0=no interference and 10=complete interference). The WPS was completed at the baseline visit of the RAPID-PsA trial and every subsequent 4 weeks.
# Statistical analysis
All analyses were conducted in the randomised set (RS), which included all patients randomised to the study. The RS has also been used to report clinical and patient-reported efficacy outcomes from the RAPID-PsA study in previous publications. [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] [bib_ref] Effect of certolizumab pegol on multiple facets of psoriatic arthritis as reported..., Gladman [/bib_ref] The WPS questions relating to workplace productivity were analysed for employed patients only, whereas the questions assessing household productivity were analysed for all patients, in the entire RS. The burden of PsA on productivity in the workplace and within the home was assessed through the WPS responses at study baseline in the entire RS population, as well as by treatment groups. Mean WPS responses were compared between CZP and placebo treatment arms at specific study visits using a non-parametric bootstrap-t test (a 5% statistical significance level was used). [bib_ref] Analysis of cost data in randomized trials: an application of the non-parametric..., Barber [/bib_ref] Last observation carried forward (LOCF) was used to impute missing data following patient withdrawal (for any reason) or for any patient with a missing measurement. For placebo patients who escaped to active treatment arms, the last observation prior to escape was carried forward.
Additional post-hoc analyses were conducted to calculate the frequency distribution of the WPS scores at specific visits and the cumulative gains over 24 weeks. Cumulative productivity scores were derived for each individual treatment group by summing the productivity scores over the 24 weeks of the double-blind phase, starting at week 4. The cumulative gains were derived as the total difference in gains between the CZP treatment and placebo groups and are presented over the 6-month period.
# Results
## Baseline characteristics
A total of 409 patients were randomised and received at least one dose of study medication. Of these, 136 patients were randomised to placebo, 138 patients to CZP 200 mg Q2W and 135 patients to CZP 400 mg Q4W. A total of 368 patients (90.0%) completed the 24-week placebo-controlled period; 59 placebo patients (43.4%) underwent mandatory escape and were re-randomised to CZP treatment at week 16. Additionally, 18 (13.0%) CZP 200 mg Q2W and 21 (15.6%) CZP 400 mg Q4W patients fulfilled escape criteria, but continued randomised treatment with the same dose regimen as per protocol. [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] Baseline demographics and characteristics, including measures of disease severity, were generally well balanced between treatment groups [fig_ref] Table 1: RAPID-PsA baseline demographics and disease characteristics at study baseline [/fig_ref].
A total of 56.6%, 60.1% and 61.5% of patients in the placebo, CZP 200 mg Q2W and CZP 400 mg Q4W treatment Days with no household work due to arthritis 5.6 1. groups, respectively, were employed at study baseline. Among the RS population, 13.0-15.6% of the patients were unable to work due to PsA, 12.6-14.2% were retired and 3.7-8.2% were homemakers [fig_ref] Figure 1: RAPID-PsA [/fig_ref].
## Burden of disease on workplace and household productivity at baseline
Overall, the burden of disease at study baseline was high in PsA patients. Employed patients reported on average more than 1 week of workplace productivity (in terms of both absenteeism and presenteeism) affected by PsA over the previous month [fig_ref] Table 2: RAPID-PsA [/fig_ref]. The burden of PsA on household productivity was even greater than that seen in the workplace, with on average more than 2 weeks of household work and social activities affected by the disease over the previous month. Baseline productivity was comparable across treatment arms [fig_ref] Table 2: RAPID-PsA [/fig_ref].
## Improvements in workplace productivity
At baseline, all patients answered at least one of the WPS questions; there were no patients with a completely missing WPS questionnaire. The beneficial effects of CZP treatment on workplace productivity compared with placebo were observed as early as week 4 and continued up to week 24.
At week 4, employed patients in both CZP treatment arms had an average of 0.6-0.9 fewer days of absenteeism per month, had 2.2-2.4 fewer days per month with reduced productivity and a reduced monthly rate of PsA interference with work productivity of 0.8-1.4 (on a 0-10 scale), compared with an average 0.5 fewer days of absenteeism per month, 1.2 fewer days with At week 24, the number of CZP-treated patients who did not report workplace absenteeism increased by 12.2-19.3%, compared with 1.0% in the placebo group. Similar frequency distribution patterns were seen for the number of paid work days with reduced productivity and the level of PsA interference with paid work productivity. These monthly improvements in workplace productivity reported by CZP-treated patients resulted in cumulative gains over the 24 weeks equivalent to a mean of 6.7 or 8.3 full days gained of paid work and an additional mean of 8.8 or 5.2 days with increased productivity for CZP 200 mg Q2W or CZP 400 mg Q4W, respectively, over placebo.
## Improvements in productivity within the home and increased participation in family, social and leisure activities
Regardless of employment status, CZP-treated patients reported reduced loss of productivity within the home and greater participation in daily activities compared to placebo patients. Improvements in household work productivity in CZP-treated patients were observed as early as week 4 and continued up to week 24 (figure 3).
At week 4, CZP-treated patients reported on average 1.9-2.2 fewer days with no household productivity per month, 1.8-1.9 fewer days per month with reduced household productivity, and a reduced monthly rate of PsA interference with work productivity of 1.4-1.6 (on a 0-10 scale), compared with 1.2 days gained per month, 1.9 fewer days with reduced productivity, and a reduced rate of PsA interference of 0.4 for patients in the placebo group [fig_ref] Figure 3: RAPID-PsA [/fig_ref]. Further improvements were reported in CZP-treated patients up to week 24 with an average of 3.0-3.5 fewer days with no household work, 3.6-4.2 fewer days with reduced household productivity, and PsA interference with productivity within the home reduced by 2.3-3.0, compared with an average of 1.0 less day with no household work per month (p<0.05; either CZP arm vs placebo), 1.8 fewer days per month with reduced productivity ( p<0.05; either CZP arm vs placebo), and PsA interference with work reduced by 0.8 in placebo patients ( p<0.05; either CZP arm vs placebo) (figure 3). For CZP patients, improvements in participation in family, social and leisure activities were observed at week 24 with an average of 2.3-3.0 days gained per month, compared with just 0.9 days gained in placebo patients ( p<0.05; either CZP arm vs placebo) (figure [fig_ref] Table 3: RAPID-PsA [/fig_ref] for mean difference and CI values).
The number of CZP patients with no days lost of household work was increased by 23.7-26.1% at week 24, compared with an increase of 4.4% in the placebo group. Similar frequency distribution patterns were seen for the number of days with reduced household work productivity, the level of PsA interference with household work productivity, and the number of days missed of family, social and leisure activities.
Cumulative gains in full days of household work over placebo were equivalent to, on average, 10.5 and 12.9 days for CZP 200 mg Q2W and CZP 400 mg Q4W, respectively, over the 24-week period. CZP treated patients also gained over placebo an additional mean of 20.1 and 16.6 days, respectively, of productive household work over the 24-week period and a mean of 5.1 and 7.4 days, respectively, of social, family or leisure activities over the 24 weeks. Clinical and epidemiological research DISCUSSION Baseline data from the RAPID-PsA trial suggest that PsA has a significant impact on employment status and productivity, both within and outside the home. These findings are consistent with previous reports of severe work disability and loss of productivity in PsA patients, [bib_ref] Work disability in psoriatic arthritis: a systematic review, Tillett [/bib_ref] including studies that report a reduced ability to work or volunteer, [bib_ref] Willingness-to-pay stated preferences for 8 health-related quality-of-life domains in psoriatic arthritis: a..., Hu [/bib_ref] or to fulfil work, personal and social lives. [bib_ref] Development of the PsAQoL: a quality of life instrument specific to psoriatic..., Mckenna [/bib_ref] Previous reports also suggest that patients with severe PsA may be forced into early retirement or unemployment due to an inability to perform physical work activities. [bib_ref] Quality-of-life issues in psoriasis and psoriatic arthritis: outcome measures and therapies from..., Mease [/bib_ref] However, compared with a large number of reports on the burden of RA on patient productivity, relatively little is known about the true negative impact of PsA on productivity. There is some evidence to suggest that the employment rate in PsA patients is higher than in the RA population, but significantly lower than that seen in the general population. [bib_ref] Employment across chronic inflammatory rheumatic diseases and comparison with the general population, Mau [/bib_ref] However, age of onset of PsA is in the 30-40s for men and women, an age at which many patients are in the midst of their working careers. Thus, the earlier onset of disease in PsA, compared with RA, could result in greater disability over the course of a PsA patient's working life and may impact significantly on career progression. The financial burden of PsA is likely to be significant and includes both direct costs associated with the provision of healthcare, and indirect costs associated with loss of productivity, due to time off work for medical care, lost wages from time off, decreased productivity at work and unemployment due to disability. [bib_ref] Quality-of-life issues in psoriasis and psoriatic arthritis: outcome measures and therapies from..., Mease [/bib_ref] Previous papers have reported large indirect economic costs associated with PsA in terms of costs associated with patients' short-term absence from work, patients' permanent work disability and caregivers' short-term absence from work. [bib_ref] Cost of illness in rheumatoid arthritis, ankylosing spondylitis, psoriatic arthritis and systemic..., Huscher [/bib_ref] [bib_ref] A cost of-illness study of psoriatic arthritis in Spain, Moreno [/bib_ref] [bib_ref] The psoriatic arthritis cost evaluation study: a cost-of-illness study on tumour necrosis..., Olivieri [/bib_ref] As the indirect costs attributed to lost productivity in the workplace and at home are major contributors to the economic burden of PsA, any improvements in productivity associated with CZP treatment will help reduce the economic burden of this chronic inflammatory disease.
In the RAPID-PsA study, CZP treatment resulted in rapid and substantial improvements in workplace and household productivity. Improved productivity at work and within the home was seen as early as week 4 and continued up to week 24. This included reduced absenteeism and presenteeism for employed patients in both CZP groups, compared with placebo. CZP treatment also resulted in improved productivity within the home, and fewer days of lost participation in family, social and leisure activities per month. These results, alongside clinical improvements seen over the first 24 weeks of CZP treatment in PsA patients, [bib_ref] Effect of certolizumab pegol on signs and symptoms in patients with psoriatic..., Mease [/bib_ref] indicate that CZP may be a valuable treatment option for PsA patients. Improvements in productivity may also contribute to the patients' psychosocial well-being, enhancing these additional aspects of their health-related quality of life.
In line with the current PsA study, CZP plus methotrexate (MTX) has previously been shown to improve productivity in RA patients both outside and within the home. The combination therapy also resulted in greater participation in social activities compared with placebo plus MTX. [bib_ref] Effect of certolizumab pegol with methotrexate on home and work place productivity..., Kavanaugh [/bib_ref] Other TNF inhibitors, namely etanercept, infliximab and golimumab, have also been shown to improve productivity in patients with active PsA. [bib_ref] Effectiveness and safety of etanercept in patients with psoriatic arthritis in a..., Gladman [/bib_ref] [bib_ref] Effect of infliximab therapy on employment, time lost from work, and productivity..., Kavanaugh [/bib_ref] [bib_ref] Patient-reported outcomes and the association with clinical response in patients with active..., Kavanaugh [/bib_ref] However, differences in the measures used to assess the impact of specific treatments on productivity make it difficult to compare between specific TNF inhibitors.
A limitation of the current report is that it only presents results up to week 24 of the RAPID-PsA trial. Consequently, longer-term results from this study or from real-life observational studies will be needed to confirm these initial findings. This is important as currently there is limited evidence from long-term observational studies in terms of the benefits of anti-TNFs on work disability in PsA. Other limitations of the data reported here include the early withdrawal of some patients ( prior to week 24) and use of LOCF methodologies to impute any missing data. Nevertheless, the data reveal a rapid effect of CZP treatment, with improvements seen within 4 weeks of administration, in both CZP dosing regimens (CZP 200 mg Q2W and CZP 400 mg Q4W) across a broad spectrum of productivity-related concepts.
In summary, PsA had a significant burden on patient productivity at study baseline. The rapid improvements seen in CZP-treated patients in paid work productivity, household productivity and participation in family, social and leisure activities, alongside the clinical improvements seen, suggest that CZP is a potentially valuable treatment option for PsA patients. These beneficial effects with respect to PsA patients' workplace and household productivity could ultimately result in gains in the quality of life and in the economic burden of this chronic inflammatory disease, particularly in those of working age.
[fig] Figure 1: RAPID-PsA: employment status at study baseline (randomised set population, observed cases). Other non-employed includes: student, unable to work due to non-arthritis and other non-employed status. CZP, certolizumab pegol; PBO, placebo; Q2W, every 2 weeks; Q4W, every 4 weeks. [/fig]
[fig] *: with household work productivity reduced by ≥50% due to arthritis missed of family, social or leisure activities due to arthritis 3Does not include days counted in the previous question, 'Work days missed due to arthritis'. †0-10 scale, where 0=no interference, 10=complete interference. CZP, certolizumab pegol; PsA, psoriatic arthritis; Q2W, every 2 weeks; Q4W, every 4 weeks; RS, randomised set. [/fig]
[fig] Figure 2: RAPID-PsA: monthly improvements in paid work productivity (employed patients, randomised set population). Left hand panels represent work productivity outcomes up to week 24, right hand panels represent frequency distribution plots at baseline (BL), week 4 and week 24. Non-parametric bootstrap-t test, last observation carried forward. *Certolizumab pegol (CZP) 200 mg Q2W vs placebo (PBO) p-value ≤0.05; ‡CZP 400 mg Q4W vs PBO p-value <0.05; †0-10 point scale (0=no interference, 10=complete interference). [/fig]
[fig] Figure 3: RAPID-PsA: monthly improvements in productivity within the household and in family, social and leisure activities (randomised set population). Left hand panels represent work productivity outcomes up to week 24, right hand panels represent frequency distribution plots at baseline (BL), week 4 and week 24. Non-parametric bootstrap-t test, last observation carried forward. *Certolizumab pegol (CZP) 200 mg Q2W vs placebo (PBO) p-value ≤0.05; ‡CZP 400 mg Q4W vs PBO p-value <0.05; †0-10 point scale (0=no interference, 10=complete interference). [/fig]
[table] Table 1: RAPID-PsA baseline demographics and disease characteristics at study baseline (RS population, observed cases) [/table]
[table] Table 2: RAPID-PsA: burden of PsA on productivity in the workplace and within the home at study baseline (RS population, observed cases) [/table]
[table] Table 3: RAPID-PsA: improvements in productivity in the workplace and within the home at week 24 (RS population, LOCF) [/table]
|
Regulation of mRNA cap methylation
The 7-methylguanosine cap added to the 5 end of mRNA is essential for efficient gene expression and cell viability. Methylation of the guanosine cap is necessary for the translation of most cellular mRNAs in all eukaryotic organisms in which it has been investigated. In some experimental systems, cap methylation has also been demonstrated to promote transcription, splicing, polyadenylation and nuclear export of mRNA. The present review discusses how the 7-methylguanosine cap is synthesized by cellular enzymes, the impact that the 7-methylguanosine cap has on biological processes, and how the mRNA cap methylation reaction is regulated.
# Introduction
Eukaryotic mRNA is modified by the addition of the 7-methylguanosine 'cap' to the first transcribed nucleotide. From yeast to humans, this modification is necessary for efficient gene expression and cell viability. The 7-methylguanosine cap is required for the translation of the majority of mRNAs, and it has also been reported to stabilize mRNA against attack by exonucleases and to promote transcription, splicing, polyadenylation and nuclear export of mRNA. Formation of the 7-methylguanosine cap occurs as the substrate mRNA is being transcribed and is catalysed by enzymes which are recruited to RNA polymerase II. Cellular proteins which regulate mRNA cap methylation are likely to function either by regulating the recruitment of the cap methyltransferase to RNA polymerase II, or by regulating the cap methyltransferase activity. Several cellular factors have now been found which regulate mRNA cap methylation and these are discussed in the present review. Viruses also have mechanisms to ensure efficient mRNA cap methylation. Since this is out of the scope of the present review, readers are directed two other excellent reviews which discuss regulation of cap methylation by viral proteins. Furichi and Shatkinalso provide a comprehensive history of the discovery of the mRNA 7-methylguanosine cap.
## Creation of the 7-methylguanosine cap
The 7-methylguanosine cap is joined to the first transcribed nucleotide via the 5 hydroxyl group, through a triphosphate linkage, to produce m7G(5 )ppp(5 )X, where m7G is 7-methylguanosine, p is a phosphate group and X is the first transcribed nucleotide. This 5 -5 linkage is in contrast with the 3 -5 phosphodiester bond, which links nucleotides in transcribed RNA. The 7-methylguanosine cap is formed by the action of three enzymes. (i) The 5 triphosphate group on nascent RNA, ppp(5 )X, is hydrolysed by an RNA 5 triphosphatase to produce diphosphate-RNA, pp(5 )X. (ii) A guanylyltransferase catalyses the addition of GMP to the diphosphate-RNA to produce the guanosine cap, G(5 )ppp(5 )X via a two-step reversible reaction. Initially, the guanylyltransferase reacts with the α-phosphate of GTP, forming an enzyme-GMP intermediate, Gp-Enz, and releasing pyrophosphate, PPi. The GMP is transferred from Gp-Enz to pp(5 )X to create G(5 )ppp(5 )X, and the enzyme is regenerated. Finally, (iii) an RNMT [RNA (guanine-7-) methyltransferase] catalyses the methylation of the guanosine cap at the N-7 position to produce the 7-methylguanosine cap, m7G(5 )ppp(5 )X. SAM (S-adenosylmethionine) is used as the methyl donor.
Addition of the 7-methylguanosine cap is often referred to as 'capping', and the enzymes that catalyse the reactions are referred to as the 'capping enzymes'. In the present review, the processes of guanosine cap addition and guanosine cap methylation are discussed distinctly and referred to as 'capping' and 'cap methylation' respectively.
Much of the characterization of the biochemical pathway which produces the 7-methylguanosine cap has been carried out in the yeast species Saccharomyces cerevisiae and Schizosaccharomyces pombe, or in vitro using the enzymes encoded by these organisms. The three enzymic activities required to synthesize the 7-methylguanosine cap are present as three separate proteins in S. cerevisiae and S. pombe. Cet1p and Pct1 are the triphosphatases, Ceg1p and Pch1 are the guanylyltransferases, and Abd1 and Pcm1 are the RNMTs in S. cerevisiae and S. pombe respectively. In mammals and other metazoa, the triphosphatase and guanylytransferase activities are found in the same polypeptide, called CE (capping enzyme) or RNGTT (RNA guanylyltransferase and 5 triphosphatase). The mammalian RNMT is a distinct protein. The guanylyltransferases and cap methyltransferases are conserved in structure and mode of action from yeast to metazoa, whereas the RNA triphosphatases are quite different in structure and mode of action in yeast and metazoa.
Higher organisms may have evolved to have the triphosphatase and guanylyltransferase on one polypeptide to permit efficient (b) Reactions that synthesize the 7-methylguanosine cap. The synthesis is described in detail in the text. Enzymes which catalyse reactions are indicated above the reactions, and the names of the enzymes from S. cerevisiae, S. pombe and Homo sapiens are indicated below the reactions.
co-ordination of the reactions required for guanosine cap addition. Some viruses, including vaccinia virus, encode a single polypeptide containing the triphosphatase, guanylyltransferase and methyltransferase, thus permitting efficient capping and cap methylation. The fact that higher organisms have retained a distinct guanosine capping enzyme (CE/RNGTT) and RNMT permits differential regulation of guanosine capping and cap methylation.
## Processes dependent on cap methylation
The 7-methylguanosine cap has been demonstrated to influence several steps of gene expression, including transcription, splicing, polyadenylation, nuclear export of mRNA, translation and mRNA stability, and these are discussed individually below. For some of these processes, the role of the 7-methylguanosine cap and the unmethylated guanosine cap have been compared, whereas for other processes, only the role of the 7-methylguanosine cap has been investigated.
## Transcription elongation
mRNA capping and cap methylation occur 'co-transcriptionally', that is, as the RNA is being transcribed. Furthermore, generation of the 7-methylguanosine cap and the enzymes which catalyse this process promote transcription elongation. This is discussed in more detail below in the discussion of the mechanism of mRNA cap methylation.
## Splicing
RNA splicing is dependent on the 7-methylguanosine cap. In vitro, the splicing reaction has been demonstrated to be dependent on the substrate RNA having a 7-methylguanosine cap and could be inhibited by free 7-methylguanosine. Splicing of the first intron was found to be particularly dependent on the presence of a 7-methylguanosine cap. These findings were consolidated in vivo by work carried out in Xenopus laevis oocytes, where microinjected non-guanosine-capped mRNA, A(5 )ppp(5 )G, was spliced less efficiently than 7-methylguanosine-capped mRNA, m7G(5 )ppp(5 )X. The 7-methylguanosine cap promoted splicing predominantly on the 5 proximal intron. Use of temperaturesensitive mutants of the guanylyltransferase Ceg1 have allowed rapid inhibition of capping in S. cerevisiae. Inhibition of capping led to a defect in splicing, but the severity was dependent on the Ceg1 mutation and was gene-specific. The proportion of genes that are dependent on the 7-methylguanosine cap for splicing is not known for either yeast or mammals.
The dependency of splicing on the 7-methylguanosine cap is mediated by the heterodimeric CBC (cap-binding complex), which consists of CBP80 (cap-binding protein 80) and CBP20. Immunodepletion of this complex from HeLa cell extracts inhibited in vitro splicing reactions. Microinjection of X. laevis oocytes with antibodies raised against CBP20 inhibited CBP20 interaction with the mRNA cap and inhibited splicing. CBC interacts with splicing complex components and is required for efficient recognition of the 5 -splice site on the 5 -proximal intron. These findings support the model that splice sites are defined by interactions which span exons.
## Mrna translation
From yeast to humans, the 7-methylguanosine cap is necessary for the translation of most mRNAs, with the exception of a subset of mRNAs translated by internal ribosome entry. Initially, methylation of the guanosine cap was demonstrated to be necessary for efficient translation in cell-free systems. In vivo, mRNA injected into X. laevis oocytes was translated more efficiently if it contained a 7-methylguanosine cap. In S. cerevisiae, temperature-sensitive mutants of Abd1 were synthesized, which facilitated rapid inhibition of Abd1 in vivo. This resulted in a rapid and almost complete loss of protein synthesis, consistent with mRNA cap methylation being essential for the efficient translation of most mRNAs.
Cap-dependent translation initiates with eIF4F (eukaryotic initiation factor 4F) complex formation on the 7-methylguanosine cap. This is required for the 5 end of mRNA to be recruited to the 43S initiation complex containing the 40S ribosomal subunit, associated initiation factors and the initiator tRNA tRNAi Met. The first mRNA-dependent step of translation is the binding of the eIF4F complex to the 7-methylguanosine cap, which is mediated by the eIF4E subunit. eIF4E has a significantly higher affinity for the 7-methylguanosine cap than the unmethylated guanosine cap, and this has been hypothesized to prevent free cellular GTP from interfering with translation initiation. mRNA polyadenylation mRNA is modified at the 3 end by endonucleolytic cleavage and addition of the poly(A) tail. In vitro, the endonucleolytic cleavage step is dependent on the CBC. A recent study used chromatin immunoprecipitations to demonstrate that the mammalian capping enzyme RNGTT/CE and the mammalian cap methyltransferase RNMT are localized not only at the 5 end of the gene but are also found throughout the gene, including a significant fraction at the 3 end of the gene. Potentially, the capping enzyme and cap methyltransferase could influence polyadenylation at this stage.
In S. cerevisiae, when capping was inhibited rapidly using temperature-sensitive mutants of the capping enzyme Ceg1, polyadenylation of CHY2 mRNA was unaffected. The length of poly(A) tails on total mRNA was investigated and was also not significantly altered following inhibition of capping. Therefore, in S. cerevisiae, polyadenylation of mRNA appears to be largely independent of the 7-methylguanosine cap.
## Nuclear export of mrna
Nuclear export of a subset of mRNAs is dependent on the 7-methylguanosine cap. Nuclear export of microinjected mRNA in X. laevis oocytes could be inhibited by 7-methylguanosinecapped RNA, but not uncapped mRNA. Export of snRNAs (small nuclear RNAs) was also shown to be dependent on the CBC in X. laevis oocytes. However, inactivation of temperaturesensitive mutants of the guanylyltransferase Ceg1 in S. cerevisiae did not alter the distribution of mRNA between the nucleus and cytoplasm, and therefore the 7-methylguanosine cap is not required for the export of the majority of mRNAs in this organism.
## Mrna stability
The presence of a 7-methylguanosine cap can protect mRNA from degradation. When RNA was microinjected into X. laevis oocytes or incubated in mammalian cell extracts, 7-methylguanosine and unmethylated guanosine-capped mRNA was significantly more stable than uncapped mRNA. Subsequently, the 7-methylguanosine cap was found to protect mRNA from degradation by exonuclease activity. Use of temperature-sensitive mutants of the S. cerevisiae guanylyltransferase Ceg1 facilitated rapid inhibition of guanosine capping in vivo. Although some mRNAs were rapidly degraded following loss of Ceg1 activity, others were relatively stable and therefore the necessity for a guanosine cap to stabilize mRNA may be dependent on the RNA sequence, secondary structure, or some other genespecific factor. It is also a possibility that mRNA stability is more dependent on the guanosine cap in higher organsims.
Methylation of the mRNA guanosine cap also has the potential to indirectly regulate mRNA stability. The capping reaction catalysed by the guanylyltransferase is reversible, and the product of the back reaction, uncapped mRNA, is degraded rapidly by exonucleases. Once the guanosine cap is methylated, it is no longer a substrate for the guanylyltransferase, and, since cap methylation appears to be irreversible, it indirectly stabilizes the RNA. Current data, however, suggest that, in vivo, the 5 end of mRNA only encounters a significant concentration of guanylyltransferase during transcription (see below), and therefore mRNA cap methylation may not be required to inhibit the guanosine cap removal catalysed by the guanylyltransferase. Experiments carried out in S. cerevisiae and X. laevis oocytes also indicate that cap methylation plays a relatively minor role in mRNA stability. Unmethylated-capped mRNA and 7-methylguanosine-capped mRNA were equivalently stable when injected into X. laevis oocytes. However, these cells are unusual since they appear to lack decapping activity, and therefore these results may not be able to be extrapolated to all vertebrate cells. Inhibition of the S. cerevisiae cap methyltransferase Abd1 in vivo also resulted in a relatively small fall in the mRNA levels of the genes investigated. In contrast, as described previously, inhibition of the S. cerevisiae guanylyltransferase Ceg1 resulted in almost complete loss of the same mRNAs over the same time course.
## Rnmt catalyses the cap methylation reaction
Methylation of the guanosine cap on the N-7 position is catalysed by the enzyme RNMT. Mammalian RNA guanine-7 cap methyltransferase activities were first isolated from rat liver and HeLa cells based on in vitro cap methylation assays. The first cellular cap methyltransferase to be identified was the RNMT from S. cerevisiae, ABD1, on the basis of sequence homology to the cap methyltransferase domain found in the vaccinia virus D1 subunit. The cap methyltransferases from S. cerevisiae and mammals appear to be isolated as monomers and are able to function effectively as such. The S. pombe cap methyltransferase Pcm1 is isolated in a complex with the guanylyltransferase and P-TEFb (positive transcription elongation factor b), but these factors have not been reported to alter the activity of Pcm1.
Following the identification of ABD1 as a cap methyltransferase, it was demonstrated to be required for cell growth, providing the first genetic evidence that mRNA cap methylation has an essential function in cells. Subsequently the human cap methyltransferase RNMT was also found to be required for mammalian cell survival. Since both RNMT and ABD1 are essential genes, there is not another pleiotropic cap methyltransferase that can compensate for loss of RNMT or ABD1. The cellular cap methyltransferases are highly conserved, including the SAM-binding site/motif I, VLD/ELGCGKG, which is essential for catalysis. In particular, the C-terminus of RNMT and Abd1 are highly homologous, and the C-terminus of RNMT can substitute for ABD1 in S. cerevisiae. Furthermore, the entire S. cerevisiae capping apparatus, that is, the RNA triphosphatase CET1, the guanylyltransferase CEG1, and the cap methyltransferase ABD1, can be replaced by the mammalian bifunctional capping enzyme CE/RNGTT and the cap methyltransferase RNMT. The N-terminal 110 amino acids of Abd1 and the N-terminal 120 amino acids of the human cap methyltransferase RNMT are not essential for in vitro cap methyltransferase activity or for cell growth and survival. At present, the function of the N-terminus of the cellular cap methyltransferases is not known.
When ABD1 was first demonstrated to be an essential gene, it was not clear whether this was due to the cap methyltransferase activity of the enzyme or some other hypothetical function of the protein. Subsequently, a series of point mutations were made in ABD1 (and in the analogous residues in RNMT). All point mutants that inhibited the enzymic activity were essential for cell growth and survival, strengthening the hypothesis that RNMT activity is essential for cell survival. Interestingly, some point mutations in regions that are identical from yeast to humans were not found to be essential for cap methylation or yeast viability. It remains to be seen whether such regions are required under specific growth conditions. Other residues which are essential for RNMT function in vivo are those which ensure correct localization. RNMT is thought to function predominantly, if not entirely, in the nucleus at the initiation of transcription, and was localized to the nucleus by use of an RNMT-GFP (where GFP is green fluorescent protein) fusion protein. The nuclear localization of RNMT is critical for its function. A truncation mutant (human RNMT residues 144-476), which localizes to the cytosol, was enzymically active but unable to rescue cells deficient in RNMT. Three alternate nuclear localization signals were identified in human RNMT, K 80 KRK, K 103 KRKR and K 126 RK.
All discussion of the human cap methyltransferase RNMT refers to the dominant isoform in the present review. However, three splice variant cDNAs transcribed from the RNMT gene, designated hCMT1a, b and c, were cloned from human colon mucosa and transcripts for all three variants were expressed in a variety of tissues. hCMT1c is the predominant isoform of RNMT discussed in the present review and is 476 amino acids long. hCMT1a encodes the same protein as hCMT1c, but has a different 5 -and 3 -UTR (untranslated region). hCMT1b encodes a distinct protein which is 504 amino acids long; it has the same N-terminal 464 amino acids as hCMT1c and differs thereafter. Recombinant hCMT1a/c was demonstrated to be active in an in vitro RNA methyltransferase assay, whereas hCM1b was not. Since the different RNMT isoforms have different UTRs, the potential exists for them to be differentially regulated, and changing the ratio of active hCMT1a/c isoforms to inactive hCMT1b isoform could potentially regulate cellular cap methyltransferase activity. However, this remains speculation at this stage, since hCMT1b has not yet been demonstrated to be expressed at the protein level.
## Mrna cap methylation occurs co-transcriptionally
As discussed above, mRNA cap methylation occurs 'cotranscriptionally', i.e. as the RNA to which it is added is being transcribed. The 5 -end of nascent RNA is guanosine capped, and the guanosine cap is methylated as the RNA emerges from RNA polymerase II, in the vicinity of the RNA polymerase II large subunit CTD (C-terminal domain). The CTD contains approx. 50 repeats (depending on the species) of the consensus sequence YSTPSPS and can be phosphorylated on multiple residues, with Ser 2 and Ser 5 phosphorylation being the most comprehensively characterized. The CTD forms a docking site for the enzymes and factors which mediate RNA modifications, including mRNA cap methylation, splicing and polyadenylation. The CTD residues are phosphorylated and dephosphorylated during distinct stages of gene transcription, and the different patterns of phosphorylation are thought to form a 'CTD code' that dictates which enzymes and factors are recruited, retained or removed from the transcribing polymerase. Thus RNA modifications are correctly co-ordinated in space and time.
RNA polymerase II is recruited to chromatin with the CTD unphosphorylated. During transcription initiation, the CTD is phosphorylated on Ser 5 by CDK7 (cyclin-dependent kinase 7), which is a subunit of the basic transcription factor TFIIH (transcription factor IIH). At this stage, the polymerase pauses under the influence of DSIF [DRB (5,6-dichloro-1-β-D-ribofuranosylbenzimidazole) sensitivity inducing factor] and NELF (negative elongation factor). The enzymes which create the 7-methylguanosine cap in yeast and mammals were found to be recruited preferentially to the RNA polymerase II CTD when it is phosphorylated on Ser 5. The guanylyltransferases also bind to Spt5, a component of DSIF, or the analogous complex in yeast, Spt4-Spt5. Thus the guanosine-capping enzymes and the cap methyltransferase are recruited to their substrate, the 5 end of nascent RNA, as it emerges from the RNA polymerase II complex. Furthermore, the guanylyltransferase is also stimulated by interactions with Ser 5 -phosphorylated CTD and Spt5. Guanosine capping was demonstrated to be enhanced on RNA that is being actively transcribed compared with free RNA. However, stimulation of capping on actively transcribed RNA was only partially dependent on the CTD. This suggests that the capping enzyme may also be recruited by another domain of RNA polymerase II or another component of the transcription apparatus.
## Mrna cap methylation-dependent transcription
Shortly after transcription initiates, RNA polymerase II 'pauses' and transcription stops. Release from this arrest into productive transcription elongation is now recognized to be a major point in the control of transcription. The guanylyltransferases and methyltransferases are recruited to RNA polymerase II during c The Authors Journal compilation c 2010 Biochemical Society pausing, and release from pausing into transcription elongation is partially dependent on these enzymes.
One of the initial findings that suggested a relationship between cap methylation and RNA polymerase II-dependent transcription was that, in vitro, SAH (S-adenosylhomocysteine), an inhibitor of methylation reactions, was found to inhibit the initiation of transcription by RNA polymerase II. SAH did not inhibit transcription elongation, or the activity, of purified RNA polymerase II, and therefore this implied that a methylation reaction which occurred early on during the transcription of a gene was coupled to transcription initiation. Subsequently, the mammalian capping enzyme CE/RNGTT was found to relieve transcriptional repression mediated by NELF. At least in vitro, catalytically inactive CE/RNGTT mutants could also relieve repression mediated by NELF, equivalently to the wild-type enzyme. This implies that it is the CE/RNGTT protein itself, rather than the process of guanosine capping, or the guanosine cap, that relieves repression mediated by NELF. The relationship between guanosine capping and transcription elongation was also observed in vivo when mutations in the yeast capping enzyme Ceg1 caused defects in transcription elongation.
Inactivation and mutation of the S. cerevisiae cap methyltransferase Abd1 also inhibited RNA polymerase II recruitment to promoters and movement through the gene in vivo. Similar to what was observed for capping enzymestimulated transcription, an Abd1 mutant defective for cap methyltransferase activity could promote RNA polymerase II recruitment to most of the genes investigated, suggesting that it is the cap methyltransferase protein itself, rather than the process of cap methylation or the 7-methylguanosine cap, which stimulates transcription.
Transcription elongation is dependent on the P-TEFb. P-TEFb, which contains the kinase CDK9, phosphorylates the RNA polymerase II CTD on Ser 2 and relieves repression mediated by DSIF/NELF, thus promoting transcription elongation. In S. pombe, CDK9 was isolated in a complex with all three enzymes required for cap methylation, i.e. the triphosphatase Pct1, the guanosine-capping enzyme Pch1 and the cap methyltransferase Pcm1. The interaction between the S. pombe cap methyltransferase Pcm1 and CDK9 is such that all detectable Pcm1 can be cleared from cell extracts by an immunoprecipitation with CDK9. Pcm1 was found to mediate recruitment of P-TEFb to chromatin and CTD Ser 2 phosphorylation. In Caenorhabditis elegans, a similiar relationship between the capping enzyme and P-TEFb is implied by the finding that inhibition of expression of the capping enzyme Cel-1 resulted in loss of CTD Ser 2 phosphorylation, without loss of CDK9 expression. It remains to be seen in S. cerevisiae and mammalian systems, where such an avid interaction between CDK9 and the cap methyltransferase has not been documented, whether P-TEFb recruitment is also dependent on the cap methyltransferase.
Although the experiments described above demonstrate that the guanosine-capping enzymes and cap methyltransferases promote transcription elongation, they do not appear to be required, at least for transcription of a subset of genes in S. cerevisiae. Following inhibition of the capping enzyme Ceg1 in S. cerevisiae, transcription of at least some genes occurred independently of capping. Furthermore, following inhibition of the S. cerevisiae methyltransferase Abd1, a subset of genes did not exhibit loss of recruitment of RNA polymerase II. In another study, inhibition of Abd1 using temperature-sensitive mutants lead to a minor loss in transcript level, which paralleled a loss in rRNA expression, which is not predicted to be directly dependent on cap methylation. These studies suggest that the role of the capping enzymes and cap methyltransferases in transcription elongation is gene-specific in S. cerevisiae, and this may be the case in other species too.
## Regulation of mrna cap methylation
The mRNA guanosine cap is methylated predominantly, if not entirely, as RNA polymerase II is paused, shortly after the initiation of transcription. As described above, at this stage the CTD of RNA polymerase II is phosphorylated on Ser 5 , and this forms a docking site for the mRNA cap methyltransferase. Known regulators of mRNA cap methylation either increase the activity of the cap methyltransferase or regulate recruitment of the cap methyltransferase to RNA polymerase II. The regulators of cap methylation described below have typically only been studied in one species or a few closely related ones. However, due to the high degree of homology between all the cap methyltransferases, and the fact that the mammalian cap methyltransferase can functionally substitute for the S. cerevisiae cap methyltransferase in vivo, it is likely that the regulators function with some equivalence in all species in which they are expressed.
## Sam
Certain Abd1 temperature-sensitive mutants cause a severe growth defect at the restrictive temperature, and a screen for suppressors of this phenotype identified SAM synthase, the enzyme which synthesizes the methyl donor SAM as a critical activator of cap methyltransferase activity. Furthermore, the growth defect was also suppressed in a dose-dependent manner by incubation in SAM, which indicates that it is levels of SAM rather than the SAM synthase protein itself that directly stimulates Abd1 activity.
## Cdc34
The same screen for suppressors of temperature-sensitive mutants of Abd1 also identified a ubiquitin-conjugating enzyme, cdc34. Mutations that affect the ubiquitin-conjugating activity of cdc34 or its interaction with the E3-SCF (where SCF is Skp1/cullin/F-box) complex inhibited rescue of the growth defect, correlating the functional interaction of cdc34 with Abd1 with its ubiquitin-conjugating activity. cdc34 did not affect ABD1 transcript levels (or SAM synthase transcript levels, as discussed above). Many possible scenarios exist for how cdc34 may regulate Abd1. cdc34 may positively regulate Abd1 either by directly ubiquitating Abd1 or a co-activator, thus increasing their activity. Alternatively, cdc34 may ubiquitinate a repressor of Adb1 and target it for degradation.
## Importin α
The nuclear transporter importin α was found to interact directly with RNMT in a yeast two-hybrid interaction screen. Importin α enhanced RNMT binding to capped mRNA and enhanced RNMT activity in vitro. Importin α forms a complex with importin β and transports nuclear localization signalcontaining proteins to the nucleus. On encountering a high concentration of RanGTP in the nucleus, this complex dissociates, releasing the transported proteins. Importin β alone had no effect on in vitro RNMT activity, but inhibited importin α binding to RNMT and enhancement of RNMT activity. RanGTP inhibited this action of importin β. Therefore importin α stimulation of RNMT probably predominates in the nucleus where RanGTP levels are high and importin α and β are dissociated.
## Signalling pathways that regulate mrna cap methylation
The potential also exists for mRNA cap methylation to be regulated in response to cellular signalling pathways. Currently, few examples exist and the details are limited.
## Myc
The proto-oncogenes c-Myc and N-Myc (neuroblastoma-derived Myc) are up-regulated in response to growth factors and are essential for the proliferative response. The Myc proteins are weak transcription factors which typically up-regulate mRNA levels by 2-fold. Expression of the Myc proteins was also demonstrated to up-regulate mRNA cap methylation on specific target mRNAs. Using antibodies raised against 7-methylguanosine to purify 7-methylguanosine-capped mRNA, c-Myc was found to up-regulate mRNA cap methylation on its transcriptional targets by up to 8-fold. Significantly, c-Myc was found to up-regulate cap methylation more than transcription for most transcriptional target genes investigated, resulting in greater c-Myc-induced gene expression than had been previously recognized. Myc proteins could also induce mRNA cap methylation on mRNAs that were not transcriptional targets of Myc.
Although Myc is a transcription factor, Myc-induced upregulation of mRNA cap methylation is not simply a result of increased transcription of components of the cap methylation apparatus. Myc mutants lacking transcriptional activity could upregulate mRNA cap methylation of certain transcripts, implying that Myc proteins have a relatively direct effect on mRNA cap methylation. Myc binds close to transcription initiation sites, and was found to bind directly to components of TFIIH, to enhance recruitment of TFIIH to initiation sites, and to increase RNA polymerase II phosphorylation on Ser 5 of the CTD. Since RNMT binds to Ser 5 -phosphorylated CTD, Myc-induced RNA polymerase II phosphorylation is likely to increase the length of time that RNMT is recruited to genes, thus increasing the probability of an mRNA guanosine cap being methylated. It has yet to be proven, however, that Myc promotes increased RNMT recruitment, and the exact mechanism of Mycinduced cap methylation is unresolved.
Although immunoprecipitations carried out with anti-7-methylguanosine antibodies can indicate the relative level of cap methylation of a gene product, they cannot determine the absolute level of cap methylation. However, the fact that cap methylation can be up-regulated more than transcription, or independently of transcription, means that pools of mRNAs are present in the cell which are unmethylated on the guanosine cap. If cap methylation can only occur at the initiation of transcription, then such unmethylated mRNAs will be weakly translated until degraded. If, however, mRNA cap methylation could be regulated post-transcriptionally, then unmethylated mRNAs will represent a dormant pool that could potentially be activated by cap methylation. Although mRNA can be cap methylated in cell extracts in vitro, physiological examples of posttranscriptional cap methylation have not been reported.
## E2f1
E2F1 was also demonstrated to up-regulate mRNA cap methylation for a subset of its transcriptional targets. E2F1 is a transcription factor which promotes cell proliferation and has a number of functional similarities to the Myc proteins, including binding proximal to transcription initiation sites. Presumably E2F1 regulates cap methylation using a similar mechanism to c-Myc. (a) Myc binds to DNA proximal to transcription initation sites. Myc binds to TFIIH subunits, promotes TFIIH recruitment to chromatin and enhances CTD phosphorylation. Then RNGTT and RNMT bind to phosphorylated CTD and catalyse formation of the 7-methylguanosine cap. (b) SAM is the methyl donor for the cap methylation reaction catalysed by the mRNA cap methyltransferase RNMT. SAH is the byproduct of the reaction, which inhibits RNMT. SAHH hydrolyses SAH to adenosine and homocysteine, thus relieving repression of the RNMT.
## Sahh (sah hydrolase)
The enzyme SAHH (also known as AHCY) has the potential to regulate the majority of methylation reactions, including mRNA cap methylation. SAH is a by-product of cellular methylation reactions that inhibits methyltransferases. SAHH hydrolyses SAH, thus relieving repression of methyltransferases. mRNA cap methylation has been proposed to be upregulated by the X. laevis SAHH xSAHH during X. laevis oocyte gastrulation. During gastrulation, xSAHH is gradually translocated from the cytoplasm to the nucleus, coinciding with the increase in zygotic mRNA synthesis. In support of a role for SAHH in co-transcriptional mRNA cap methylation, a fraction of xSAHH was found to co-localize with RNA polymerase II on nascent transcripts, and xSAHH could be immunoprecipitated with the X. laevis cap methyltransferase. It is worth noting that an increase in cap methyltransferase activity was observed during X. laevis maturation, although the mechanism was not explored, and it is not clear whether it was dependent on a change in xSAHH activity.
In mammalian cell systems, SAHH was identified as a Myc target gene that was up-regulated by increased transcription and cap methylation. The up-regulation of SAHH was necessary to permit Myc-induced cap methylation but not transcription. Myc-induced SAHH up-regulation was also necessary for Mycinduced protein synthesis and cell proliferation, thus correlating cap methylation with the core Myc functions. Although SAHH has the potential to regulate all methylation reactions, only Mycinduced cap methyation was SAHH-dependent, whereas Myc-induced total RNA, DNA and protein methylation were not. This suggests that either the RNMT is accutely sensitive to SAH c The Authors Journal compilation c 2010 Biochemical Society levels, or that the local concentration of SAH rises to significant levels around sites of mRNA cap methylation.
## Demethylases and decapping enzymes?
The major mechanism by which the mRNA cap methylation is reversed is 'decapping', which is the removal of the entire 7-methylguanosine cap. There are two major decapping pathways that initiate following repression of translation. Dcp2 and its associated proteins hydrolyse 7-methylguanosine-capped mRNA, producing m7GDP and monophosphate-RNA, which is degraded rapidly in a 5 to 3 direction by exonucleases. DcpS hydrolyses the 7-methylguanosine structure following degradation of the mRNA from the 3 end, releasing m7GMP, and therefore is described as a scavenger enzyme. Decapping can be regulated at several stages, which can affect all, or a specific subset of, mRNAs.
Decapping removes the entire 7-methylguanosine cap. Currently, demethylation of 7-methylguanosine has not been described and, to our knowledge, there is no experimental evidence to suggest that such a demethylase will be identified.
# Conclusion
The 7-methylguanosine cap is essential for mRNA translation and cell viability from yeast to mammals. It also has a role in transcription elongation, mRNA stability and degradation, and mediates other RNA processing events, including splicing, poly(A) tail addition and nuclear export. The effect of the 7-methylguanosine cap on these processes varies in different experimental systems, which raises the possibility that the 7-methylguanosine cap has gene-specific, cell-lineage-specific and species-specific roles. Given the recent observations that mRNA cap methylation can be regulated by cellular transcription factors, the precise role of the 7-methylguanosine cap in has become even more intriguing. |
The associations between psychological distress and healthcare use in patients with non-cardiac chest pain: does a history of cardiac disease matter?
Background: Psychological distress such as somatization, fear of body sensations, cardiac anxiety and depressive symptoms is common among patients with non-cardiac chest pain, and this may lead to increased healthcare use. However, the relationships between the psychological distress variables and healthcare use, and the differences in relation to history of cardiac disease in these patients has not been studied earlier. Therefore, our aim was to explore and model the associations between different variables of psychological distress (i.e. somatization, fear of body sensations, cardiac anxiety, and depressive symptoms) and healthcare use in patients with non-cardiac chest pain in relation to history of cardiac disease. Methods: In total, 552 patients with non-cardiac chest pain (mean age 64 years, 51% women) responded to the Patient Health Questionnaire-15, Body Sensations Questionnaire, Cardiac Anxiety Questionnaire, Patient Health Questionnaire-9 and one question regarding number of healthcare visits. The relationships between the psychological distress variables and healthcare visits were analysed using Structural Equation Modeling in two models representing patients with or without history of cardiac disease.Results: A total of 34% of the patients had previous cardiac disease. These patients were older, more males, and reported more comorbidities, psychological distress and healthcare visits. In both models, no direct association between depressive symptoms and healthcare use was found. However, depressive symptoms had an indirect effect on healthcare use, which was mediated by somatization, fear of body sensations, and cardiac anxiety, and this effect was significantly stronger in patients with history of cardiac disease. Additionally, all the direct and indirect effects between depressive symptoms, somatization, fear of body sensations, cardiac anxiety, and healthcare use were significantly stronger in patients with history of cardiac disease. Conclusions: In patients with non-cardiac chest pain, in particular those with history of cardiac disease, psychological mechanisms play an important role for seeking healthcare. Development of interventions targeting psychological distress in these patients is warranted. Furthermore, there is also a need of more research to clarify as to whether such interventions should be tailored with regard to history of cardiac disease or not.
## Key points
Cardiac anxiety was the main associate to healthcare use Depressive symptoms were only indirectly associated with healthcare use The direct and indirect effects between psychological distress and healthcare use were stronger in patients with a history of cardiac disease Background Non-Cardiac Chest Pain (NCCP) is a common condition [bib_ref] Natural history and predictors of outcome for noncardiac chest pain: a prospective..., Eslick [/bib_ref] with substantial impact on patients' psychological wellbeing, health-related quality of life, and healthcare use [bib_ref] The prevalence and correlates of psychological outcomes in patients with acute non-cardiac..., Webster [/bib_ref] [bib_ref] Health care utilisation in patients with non-cardiac chest pain: a longitudinal analysis..., Hadlandsmyth [/bib_ref] [bib_ref] Quality of life in patients with non-CAD chest pain: associations to fear..., Hadlandsmyth [/bib_ref]. Many patients with NCCP are not convinced by their 'ruled out' cardiac diagnosis, and lack an explanation for their chest pain [bib_ref] Non-cardiac chest pain: a retrospective cohort study of patients who attended a..., Dumville [/bib_ref] [bib_ref] Exploring the information needs of patients with unexplained chest pain, Roysland [/bib_ref] [bib_ref] Everything's fine, so why does it happen?' a qualitative investigation of patients'..., Webster [/bib_ref]. They continue to experience chest pain and avoid activities that they think might be harmful to their heart [bib_ref] Natural history and predictors of outcome for noncardiac chest pain: a prospective..., Eslick [/bib_ref] [bib_ref] Chest pain, Bass [/bib_ref] [bib_ref] The course of nonspecific chest pain in primary care: symptom persistence and..., Glombiewski [/bib_ref] [bib_ref] Short-term cognitive behavioral therapy for non-cardiac chest pain and benign palpitations: a..., Jonsbu [/bib_ref] , leading to substantial use of healthcare and societal resources [bib_ref] Societal costs of non-cardiac chest pain compared with ischemic heart disease-a longitudinal..., Mourad [/bib_ref].
Different psychological problems such as fear of body sensations, cardiac anxiety, and depressive symptoms may be involved in the process of symptom interpretation and healthcare use [bib_ref] Depressive symptoms and healthcare utilization in patients with noncardiac chest pain compared..., Mourad [/bib_ref] [bib_ref] Depressive symptoms, cardiac anxiety, and fear of body sensations in patients with..., Mourad [/bib_ref]. According to the "Fear-Avoidance Model", the reaction to the experience of severe pain is primarily characterized by fear, which is followed by either confrontation or avoidance. While confrontation reduces the fear, avoidance leads to maintenance and exacerbation of the fear [bib_ref] Outline of a fear-avoidance model of exaggerated pain perception-I, Lethem [/bib_ref] [bib_ref] The fear-avoidance model of musculoskeletal pain: current state of scientific evidence, Leeuw [/bib_ref]. Informed by the Fear-Avoidance model, a model was suggested implying that patients who experience recurrent and persistent chest pain that they perceive as threatening may express psychological distress in physical symptoms (i.e. somatization) and experience pain-related fear, which in turn can lead to cardiac anxiety and depressive symptoms, and thus increase healthcare use due to fear of having a cardiac event [fig_ref] Figure 1 a: + b The pre-assumed model of the associations between somatization, fear of... [/fig_ref]. However, this model has not been evaluated in patients with NCCP. Furthermore, patients with history of Cardiac Disease (CD) (i.e. angina pectoris, myocardial infarction, and/or heart failure) can also have NCCP [bib_ref] Chest pain in patients with cardiac and noncardiac disease, Sheps [/bib_ref] [bib_ref] Chest pain of cardiac and noncardiac origin, Lenfant [/bib_ref] , but few studies distinguish between these groups with regard to psychological presentation and healthcare use [bib_ref] Chest pain in patients with cardiac and noncardiac disease, Sheps [/bib_ref] [bib_ref] Chest pain of cardiac and noncardiac origin, Lenfant [/bib_ref] [bib_ref] Noncardiac chest pain in the emergency department: the role of cardiac history,..., Kuijpers [/bib_ref] [bib_ref] Chest pain as a presenting complaint in patients with acute myocardial infarction..., Malik [/bib_ref]. Since patients who have been diagnosed with CD are informed by healthcare providers that they should pay attention to and seek immediate care in case of persistent chest pain [bib_ref] The physician's role in minimizing prehospital delay in patients at high risk..., Dracup [/bib_ref] [bib_ref] Task force on the management of chest pain, Erhardt [/bib_ref] , these patients may consider chest pain as more fearful and threatening than patients with no history of CD. Therefore, our aim was to explore and model the associations between different variables of psychological distress (i.e. somatization, fear of body sensations, cardiac anxiety, and depressive symptoms) and healthcare use in patients with NCCP in relation to history of CD.
# Methods
This is a secondary analysis of a cross sectional study [bib_ref] Depressive symptoms, cardiac anxiety, and fear of body sensations in patients with..., Mourad [/bib_ref]. In the original analysis, the prevalence of fear of body sensations, cardiac anxiety, depressive symptoms and their relations to number of healthcare visits was studied. In this analysis, Structural Equation Modeling (SEM) analyses were used to explore and model the associations between somatization, fear of body sensations, cardiac anxiety, depressive symptoms and healthcare visits in patients with NCCP, divided into two groups based on experience of CD.
## Study participants and procedures
The recruitment process and first results are described in the earlier paper [bib_ref] Depressive symptoms, cardiac anxiety, and fear of body sensations in patients with..., Mourad [/bib_ref]. In brief, 2271 patients older than 18 years who had sought medical care due to chest pain and been diagnosed with NCCP after ruling out acute myocardial infarction between October 2013 and January 2014 were approached within 1 month from the day of discharge from four hospital in southeast Sweden. NCCP was determined using the International Classification of Diseases (ICD) 10-codes: R07.2, R07.3, R07.4, and Z03.4.
The study information, an informed consent form, and the questionnaires were sent by post to all eligible patients together with a pre-stamped envelope for response. Patients had the possibility to contact the research team in case of questions. For study participation, patients were asked to sign and return the informed consent form together with the completed questionnaires. In total, 552 patients were included in the study.
## Data collection and measurements
All data including socio-demographics, health complaints and history of CD was self-reported. Heart failure was included as part of the CD since the diagnosis is cardiac and is often associated with psychological distress. Self-reported data by patients on e.g. chronic diseases is reported to be fairly accurate compared to physician reports [bib_ref] Self-reports and general practitioner information on the presence of chronic diseases in..., Kriegsman [/bib_ref].
## Somatization
The Patient Health Questionnaire-15 (PHQ-15) was used to measure somatization, which is report of somatic symptoms that have no pathophysiological cause [bib_ref] The PHQ-15: validity of a new measure for evaluating the severity of..., Kroenke [/bib_ref] [bib_ref] Measurement and assessment of somatic symptoms, Chaturvedi [/bib_ref]. The PHQ-15 comprises 15 items/physical symptoms with scores between 0 and 30. Scores of 10 or higher indicate at least medium severity. The PHQ-15 has high reliability and validity [bib_ref] The PHQ-15: validity of a new measure for evaluating the severity of..., Kroenke [/bib_ref]. In the present study, the Cronbach's α coefficient was 0.85.
## Fear of body sensations
Fear of body sensations, such as palpitations, dizziness and sweating, was measured with the Body Sensations Questionnaire (BSQ) containing 17 items and scores between 17 and 85. Higher scores indicate more fear of body sensations [bib_ref] Assessment of fear of fear in agoraphobics: the body sensations questionnaire and..., Chambless [/bib_ref]. The BSQ is reliable and valid [bib_ref] Short-term cognitive behavioral therapy for non-cardiac chest pain and benign palpitations: a..., Jonsbu [/bib_ref] [bib_ref] Depressive symptoms, cardiac anxiety, and fear of body sensations in patients with..., Mourad [/bib_ref] [bib_ref] Assessment of fear of fear in agoraphobics: the body sensations questionnaire and..., Chambless [/bib_ref]. In the present study, the Cronbach's α coefficient was 0.93.
## Cardiac anxiety
Cardiac anxiety was measured using the Cardiac Anxiety Questionnaire (CAQ). The questionnaire comprises 18 items and the scores range between 0 and 72, and higher scores indicate greater cardiac anxiety. The CAQ can be divided into three subscales (i.e. fear, avoidance, and heart-focused attention). The CAQ has demonstrated good reliability and validity [bib_ref] The cardiac anxiety questionnaire: development and preliminary validity, Eifert [/bib_ref]. In the present study, the Cronbach's α coefficient was 0.90 for the total CAQ and 0.84, 0.89 and 0.77 for the subscales fear, avoidance, and heart-focused attention.
## Depressive symptoms
The Patient Health Questionnaire-9 (PHQ-9) was used to measure depressive symptoms. The PHQ-9 comprises 9 items and scores between 0 and 27. At a score of 10 or higher, the PHQ-9 has a sensitivity for major depression of 88%, a specificity of 88%, and a positive likelihood ratio of 7.1. The PHQ-9 has shown to have high internal consistency [bib_ref] The PHQ-9: validity of a brief depression severity measure, Kroenke [/bib_ref]. In the present study, the Cronbach's α coefficient was 0.87.
## Healthcare visits
The number of healthcare visits was determined by asking the participants the following self-developed question: "In the last year, how many times did you seek care due to chest pain?" Answers were predetermined to the categories: "1, 2-3, or >3".
## Analytic strategy and statistical analysis
Descriptive statistics, comparative analyses, and SEM were used to analyse the data. To compare background characteristics of the participants with or without history of CD, categorical variables were tested with Chi-Square tests while continuous variables were analysed with Student's t-test or Mann-Whitney U-test, depending on distribution of normality. SEM was used to explore and model the associations between the different variables of psychological distress and healthcare visits in patients with NCCP with or without history of CD. To calculate the differences between the associations in the SEM models of those with or without history of CD, Fisher r to z transformation was used. Analyses of potential nonlinear relationships between somatization, fear of body sensations, cardiac anxiety, and depressive symptoms were performed by means of Distance-Weighted Least Squares regression (DWLS).
SEM allows modeling of the relationships between variables that may have inter-dependent associations. Associations within the model were analysed using maximum likelihood and are described using standardized coefficients. Goodness of fit of the model was analysed with Chi-Square, the Root Mean Square Error of Approximation (RMSEA), and the Comparative Fit Index (CFI). These estimates provide information on how well the model represent our population of patients with NCCP with or without CD. A non-significant Chi-Square value, a RMSEA < 0.06 and a CFI ≥ 0.95 indicate a good model fit.
SPSS version 23.0 was used to perform the descriptive and comparative analyses, and LISREL 8.30 softwarewas used to perform the SEM analyses. Significance levels were set at p < 0.05 for SPSS analyses and t-value of ≥1.96 for LISREL analyses regarding factor loadings and effects. Curve-linearity analyses between the latent variables were performed by SYSTAT.
# Results
## Study participants
Of the 2271 patients who were approached, 680 (30%) agreed to participate, but only 552 (24%) fulfilled the criteria and were included in the study. Study participants had a mean age of 64 (±17) years and 51% were women, 67% were married/cohabiting and 55% were retired [fig_ref] Table 1: Characteristics of patients with non-cardiac chest pain Total a [/fig_ref]. Non-respondents were significantly younger (54 ± 20 years, p < 0.001) and those who declined participation were significantly older (70 ± 17 years, p < 0.001) compared to study participants.
About 34% of the participants had previous diagnosis of CD (i.e. angina pectoris, myocardial infarction, and/or heart failure). These participants were significantly older (71 vs. 60 years), more males (60 vs. 44%), had lower educational levels, reported more conditions (5 vs. 3) and had larger number of healthcare visits compared to patients with no history of CD. Furthermore, patients with history of CD reported significantly higher scores in all psychological distress variables except for fear of body sensations [fig_ref] Table 2: Psychological distress in patients with non-cardiac chest pain, [/fig_ref].
Modeling the associations between somatization, fear of body sensations, cardiac anxiety, depressive symptoms and healthcare use
The SEM analysis of the assumed paths [fig_ref] Figure 1 a: + b The pre-assumed model of the associations between somatization, fear of... [/fig_ref] between somatization, fear of body sensation, cardiac anxiety, depressive symptoms and healthcare visits in patients with NCCP showed a poor fit (Chi-Square statistics 75.94, df = 10, p < 0.001: RMSEA = 0.110: and CFI = 0.95) [fig_ref] Figure 1 a: + b The pre-assumed model of the associations between somatization, fear of... [/fig_ref]. The model showed that there was no direct association between depressive symptoms and healthcare visits. Since a 2015 paper exploring fear avoidance in pain patients and using SEM analysis, reported that depressive symptoms could be seen as a preconditioning variable for fear avoidance and functioning [bib_ref] Is there a role of depressive symptoms in the fearavoidance model? A..., Seekatz [/bib_ref] , our model was revised to explore if depressive symptoms would occur more downstream in the model. Thus, by the possible associations of depressive symptoms with somatization, and/or fear of body sensations, and/or cardiac anxiety, depressive symptoms could have an indirect effect on healthcare visits. In the new model, shown in [fig_ref] Figure 2: The revised model of the associations between somatization, fear of body sensations,... [/fig_ref] , the fit to the data was good (Chi-Square statistics 1.93, df = 5, p = 0.86, RMSEA = 0.000: and CFI = 0.99).
The revised model [fig_ref] Figure 2: The revised model of the associations between somatization, fear of body sensations,... [/fig_ref] showed that depressive symptoms had a direct association to somatization (β = 0.83, p < 0.01), and a direct (β = 0.60, p < 0.01) as well as an indirect (β = 0.18, p < 0.01) association to cardiac anxiety. The total effect of depressive symptoms on cardiac anxiety was 0.78 (p < 0.01). By these paths, an indirect association between depressive symptoms and healthcare visits was found (β = 0.47, p < 0.01). [fig_ref] Table 3: Indirect and total effects in the revised structural equation model for the... [/fig_ref] shows all indirect and total effects in the revised SEM model. Specific analysis of the model based on sex was tested, but no differences were found.
The contribution of cardiac disease in the associations between psychological distress and healthcare use
To explore the contribution of CD, two SEM models of the revised model were analysed; one including patients with history of CD (n = 188) and the other including patients without a history of CD (n = 360). Examination of the fit indices showed good fit of the model for both patients with a history of CD [fig_ref] Figure 3 a: + b The revised model of the associations between somatization, fear of... [/fig_ref] , Chi-Square statistics 9.69, df = 10, p = 0.47, RMSEA = 0.000, and CFI = 0.99) and for patients without [fig_ref] Figure 3 a: + b The revised model of the associations between somatization, fear of... [/fig_ref] , Chi-Square statistics 4. 25, df = 7, p = 0.75, RMSEA = 0.000, and CFI = 0.99). All indirect and total effects in the two models are shown in [fig_ref] Table 4: Indirect and total effects in the revised structural equation model for non-cardiac... [/fig_ref]. The differences in the effects between the groups were analysed using Fisher r to z transformation.
In general, both direct as well as indirect effects were stronger in the model representing patients with history of CD than in the model based on patients without history of CD. In patients with a history of CD, depressive symptoms had a significantly stronger indirect effect on fear of body sensations (β = 0.79 vs. β = 0.47, p < 0.01), cardiac anxiety (β = 0.30 vs. β = 0.11, p < 0.05), and number of healthcare visits (β = 0.62 vs. β = 0.33, p < 0.01) than in patients without a history of CD. Further, in patients with a history of CD, somatization had a significantly stronger direct effect on fear of body sensations (β = 0.95 vs. β = 0.56, p < 0.01), as well as stronger indirect effect on cardiac anxiety (β = 0.36 vs. β = 0.13, p < 0.01), and number of healthcare visits (β = 0.26 vs. β = 0.06, p < 0.01) compared to those without. Additionally, also fear of body sensations had a significantly stronger direct effect on cardiac anxiety (β = 0.38 vs. β = 0.24, p < 0.05), as well as stronger indirect effect on number of healthcare visits (β = 0.28 vs. β = 0.11, p < 0.05) in those with CD than in those without. Finally, the direct effect of cardiac anxiety on healthcare visits was significantly stronger (β = 0.72 vs. β = 0.46, p < 0.01) in patients with CD than in those without. The direct effects of depressive symptoms on cardiac anxiety were about the same in both groups (β = 0.55 vs. β = 0.59).
Depressive symptoms and fear of body sensations seem to be important preconditioning factors for cardiac anxiety in patients with a history of CD. [fig_ref] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear... [/fig_ref] shows that higher scores on the depressive symptoms scale (> 15) for patients with a history of CD (solid line), are associated with cardiac anxiety to a substantially larger extent compared to in patients with no history of CD Sick-leave/disability pension 40 (7) 16 (9) 24 (7) Other 57 (10) 13 (7) 44 (12) Smoking [fig_ref] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear... [/fig_ref] shows that body sensations are associated with cardiac anxiety all over the scale for patients with a history of CD (solid line), but for patients with no history of CD (dashed line) body sensations at scores higher than about 40 cease to induce increases in cardiac anxiety.
# Discussion
Our SEM analyses revealed that depressive symptoms were an important underlying variable, (i.e. lying behind other mechanisms) and influences healthcare use differently in patients with NCCP with or without history of CD. There was no direct association between depressive symptoms and healthcare use, which suggests that depressive symptoms may act more as an underlying variable than an outcome as in the fear avoidance model [bib_ref] The fear-avoidance model of musculoskeletal pain: current state of scientific evidence, Leeuw [/bib_ref] , which is shown in our revised SEM model [fig_ref] Figure 2: The revised model of the associations between somatization, fear of body sensations,... [/fig_ref]. Our findings are supported by a 2015 study [bib_ref] Is there a role of depressive symptoms in the fearavoidance model? A..., Seekatz [/bib_ref] using SEM analysis and showing that depressive symptoms had substantial impact on fear avoidance beliefs and physical functioning. In our revised model, depressive symptoms had strong direct effects on both somatization and cardiac anxiety, and indirect effects on fear of body sensation as well as cardiac anxiety. By these paths, depressive symptoms had a strong indirect effect on healthcare use. In addition, somatization had a strong direct effect on fear of body sensations and an indirect effect on cardiac anxiety. Therefore, in our model depressive symptoms and somatization can be seen as underlying variables. On the other hand, cardiac anxiety which is highly prevalent in patients with NCCP, can be seen as a state variable (i.e. temporarily induced by situations perceived as dangerous), and thus becomes the signal that drives patients to seek care in case of chest pain [bib_ref] Depressive symptoms, cardiac anxiety, and fear of body sensations in patients with..., Mourad [/bib_ref] [bib_ref] Cardiac anxiety in people with and without coronary atherosclerosis, Marker [/bib_ref]. In this case, cardiac anxiety is a result of an arising unfamiliar situation characterized by chest pain that is interpreted as threatening and driving the patients to contact healthcare providers. The revised models based on history of CD acted relatively similar for both groups [fig_ref] Figure 3 a: + b The revised model of the associations between somatization, fear of... [/fig_ref]. Depressive symptoms had in both groups an indirect effect on healthcare visits, although this effect was significantly stronger in patients with history of CD than in those without. Furthermore, this effect was mediated by somatization, fear of body sensations, and cardiac anxiety. Additionally, all the direct and indirect effects between the different variables of psychological distress and healthcare use were significantly stronger in patients with history of CD.
In patients with history of CD, cardiac anxiety was substantially more associated with higher scores of depressive symptoms compared to those without history of CD [fig_ref] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear... [/fig_ref]. This can be explained by stronger associations between somatization and fear of body sensations, and also between fear of body sensations and cardiac anxiety. From [fig_ref] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear... [/fig_ref] relationship between cardiac anxiety and fear of body sensations was found for patients with history of CD, and a curved, leveling of relationship for patients with no history of CD. This means that, for patients without history of CD, there is an increase in cardiac anxiety as a function of fear of body sensations up to a certain level (about a scale value of 40), but above this level, increase in fear of body sensations has no effect on cardiac anxiety. Furthermore, the relationship between fear of body sensations and somatization shows the same difference as that found for the relationship between cardiac anxiety and fear of body sensations in [fig_ref] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear... [/fig_ref]. Accordingly, in contrast to patients without a history of CD, higher levels of somatization were associated with higher levels of fear of body sensations, and higher levels in fear of body sensations were in turn associated with higher levels of cardiac anxiety in patients with history of CD.
The difference between the groups in our study can be explained by the fact that patients with history of CD suffer from fear and anxiety due to earlier experience of physical symptoms [bib_ref] The rust of life": impact of anxiety on cardiac patients, Moser [/bib_ref] [bib_ref] The impact of the intensity of fear on patient's delay regarding health..., Dubayova [/bib_ref] [bib_ref] Fear-avoidance beliefs and cardiac rehabilitation in patients with first-time myocardial infarction, Ahlund [/bib_ref] [bib_ref] Living on a knife edge-the daily struggle of coping with symptomatic cardiac..., Withers [/bib_ref] [bib_ref] Heart-focused anxiety in patients with chronic heart failure before implantation of an..., Bunz [/bib_ref]. In addition, one of the main components in cardiac rehabilitation is to teach patients to pay attention to and seek immediate care in case of persistent chest pain [bib_ref] The physician's role in minimizing prehospital delay in patients at high risk..., Dracup [/bib_ref] [bib_ref] Task force on the management of chest pain, Erhardt [/bib_ref]. This is in line with previous findings revealing that patients perceiving pain as threatening experience pain-related fear and safety seeking behaviour, such as avoidance [bib_ref] The fear-avoidance model of musculoskeletal pain: current state of scientific evidence, Leeuw [/bib_ref] [bib_ref] Chest pain, palpitations and panic, Mayou [/bib_ref] and frequent visits to healthcare providers [bib_ref] Pain catastrophizing: a critical review, Quartana [/bib_ref].
# Limitations
The response rate of about 30% was quite low, but not unusual when approaching a general population and is comparable with data from different surveys [bib_ref] Health care provider surveys in the United States, 2000-2010: a review, Mcleod [/bib_ref]. The study team had no access to medical charts and some of the patients may represent chest pain under evaluation, although their discharge diagnosis was NCCP. The diagnosis of CD was based on diagnosis of angina pectoris, myocardial infarction, and/or heart failure and can be seen as a limitation. Another limitation is that the causal relationship cannot be determined due to use of a cross-sectional design. In addition, the use of self-report data is a limitation.
Although patient report has been found fairly accurate compared to physician report [bib_ref] Self-reports and general practitioner information on the presence of chronic diseases in..., Kriegsman [/bib_ref] , this could be a limitation. Further, patients with history of CD were older, had more comorbidities, larger number of healthcare visits, and reported significantly higher scores in all variables but fear of body sensations. This may mirror that it was not only the CD in itself that may explain the differences between the groups.
# Conclusions
Depressive symptoms only had an indirect association with healthcare use in patients with NCCP regardless of history of CD, and this association was mainly mediated by somatization, fear of body sensations, and cardiac anxiety. Hence, these psychological mechanisms play an important role in healthcare seeking behaviour, especially in those with history of CD, and lead to greater suffering for the individuals and higher healthcare costs. Therefore, there is a need for new innovative interventions targeting this patient group. Internet-delivered cognitive behavior therapy has been found feasible in this patient group. Although differences between the groups may be related to previous experience of a cardiac event, differences regarding group characteristics mentioned in the limitations could potentially confound the results. Therefore, more research is needed to clarify if such interventions should be tailored with regard to history of CD since this may have an impact on the way patients perceive, interpret, and act on their symptoms. The funders were not involved in the study design, in the collection, analysis, and interpretation of the data, in the writing of the report, or in the decision to submit the article for publication.
## Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
[fig] Figure 1 a: + b The pre-assumed model of the associations between somatization, fear of body sensations, cardiac anxiety, depressive symptoms, and healthcare use in patients with non-cardiac chest pain. Chi-Square = 75.94, df = 10, P-value = 0.00000, RMSEA = 0.110, CFI = 0.95. BSQ Body Sensations Questionnaire, CAQ Cardiac Anxiety Questionnaire, PHQ-9 Patient Health Questionnaire-9, PHQ-15 Patient Health Questionnaire-15 [/fig]
[fig] Figure 2: The revised model of the associations between somatization, fear of body sensations, cardiac anxiety, depressive symptoms, and healthcare use in patients with non-cardiac chest pain. Chi-Square = 1.93, df = 5, P-value = 0.85819, RMSEA = 0.000, CFI = 0.99. BSQ Body Sensations Questionnaire, CAQ Cardiac Anxiety Questionnaire, PHQ-9 Patient Health Questionnaire-9, PHQ-15 Patient Health Questionnaire-15 [/fig]
[fig] Figure 3 a: + b The revised model of the associations between somatization, fear of body sensations, cardiac anxiety, depressive symptoms, and healthcare use in patients with non-cardiac chest with (a Chi-Square = 9.69, df = 10, P-value = 0.46772, RMSEA = 0.000, CFI = 0.99) and without history of cardiac disease (b Chi-Square = 4.25, df = 7, P-value = 0.75067, RMSEA = 0.000, CFI = 0.99). BSQ Body Sensations Questionnaire, CAQ Cardiac Anxiety Questionnaire, PHQ-9 Patient Health Questionnaire-9, PHQ-15 Patient Health Questionnaire-15 [/fig]
[fig] Figure 4 a: + b Cardiac anxiety as a function of depressive symptoms and fear of body sensations (BSQ). Solid line indicates non-cardiac chest pain patients with history of, and dashed line chest pain patients with no history of cardiac disease. The lines are smoothed by means of distance-weighted least squares regression (DWLS) [/fig]
[table] Table 1: Characteristics of patients with non-cardiac chest pain Total a (N = 552) History of cardiac disease (n = 188) No history of cardiac disease (n = 360) p-value [/table]
[table] Table 2: Psychological distress in patients with non-cardiac chest pain, (mean ± SD) [/table]
[table] Table 3: Indirect and total effects in the revised structural equation model for the total sample of patients with non-cardiac chest pain, N = 552 [/table]
[table] Table 4: Indirect and total effects in the revised structural equation model for non-cardiac chest pain patients with or without history of cardiac disease [/table]
|
Gut microbiota as the key controllers of “healthy” aging of elderly people
Extrinsic factors, such as lifestyle and diet, are shown to be essential in the control of human healthy aging, and thus, longevity. They do so by targeting at least in part the gut microbiome, a collection of commensal microorganisms (microbiota), which colonize the intestinal tract starting after birth, and is established by the age of three. The composition and abundance of individual microbiota appears to continue to change until adulthood, presumably reflecting lifestyle and geographic, racial, and individual differences. Although most of these changes appear to be harmless, a major shift in their composition in the gut (dysbiosis) can trigger harmful local and systemic inflammation. Recent reports indicate that dysbiosis is increased in aging and that the gut microbiota of elderly people is enriched in pro-inflammatory commensals at the expense of beneficial microbes. The clinical consequence of this change remains confusing due to contradictory reports and a high degree of variability of human microbiota and methodologies used. Here, we present the authors' thoughts that underscore dysbiosis as a primary cause of aging-associated morbidities, and thus, premature death of elderly people. We provide evidence that the dysbiosis triggers a chain of pathological and inflammatory events. Examples include alteration of levels of microbiota-affected metabolites, impaired function and integrity of the gastrointestinal tract, and increased gut leakiness. All of these enhance systemic inflammation, which when associated with aging is termed inflammaging, and result in consequent aging-associated pathologies.
# Introduction
Today, modern humans live markedly longer than their predecessors in the early twentieth century, due to the achievements of modern medicine and lifestyle improvement. The importance of external factors in delaying intrinsic causes of aging-associated pathologies and diseases, i.e., in uncoupling chronological age from physical decline, has been recognized since ancient times. The first prescription to reduce the problems associated with aging, such as dizziness, eye inflammation and ear pain, using a diet of gruel, raw honey, vegetable and fowl is recorded in the book Hygiene, written by the Greek physician Galen in 175 AD. Then over century ago, Elie Metchnikoff postulated that senility is caused by "putrefactive bacterial autotoxins" leaked from the colon, and advocated a diet of fermented milk and a "simple" lifestyle to neutralize these autotoxins, for the first time emphasizing the importance of gut microbiota in human health and aging. The human gut microbiota is a commensal microbial "superorganism" that consists of trillions of Bacteria, Archaea, Eukarya and viruses, where four bacterial phyla (Firmicutes, Bacteroides, Proteobacteria and Actinobacteria) account for about 98% of the microorganisms. This superorganism coevolved with the host to provide numerous essential and mutually beneficial functions. It supports digestion and absorption of food, metabolizes fibers into bioactive short chain fatty acids (SCFAs), generates vitamins and nutrients, maintains the intestinal integrity, regulates host immunity and directly and indirectly protects from pathogens. The gut microbiota is considered to be a master regulator of immune homeostasis, as its absence in germ-free (GF) mice impairs development and maturation of the immune system, while its presence in the gut induces IL-10 and TGFβ-producing Tregs, immunoglobulin A (IgA)-secreting B cells, Th17 cells and type-2 lymphoid innate cells (ILC2). Oral supplementation of mice with 11 commensal strains of bacteria isolated from healthy human feces enhances the host resistance to Listeria monocytogenes infection and improves the therapeutic efficacy of immune checkpoint inhibitors in tumor-bearing mice through upregulation of IFNγexpressing CD8 + T cells in the intestine in a CD103 + dendritic cell and MHC-I molecule-dependent manner. The microbiota also supports maintenance of the intestinal mucus layer and induces production of various factors and secretory IgA from B cells to promote its own growth and to suppress pathogens.
Understanding the role of gut microbiota in human health is hampered by a high degree of variability.
Microbial composition differs depending on the conditions within the gastrointestinal tract, such as the high acidity of the stomach and the small intestine and the slightly acidic to neutral pH of the colon. Recent reports from sequencing human fecal microbiota revealed that composition of the gastro-intestinal microbes is affected by human inter-individual, racial, geographic and lifestyle differences. Furthermore, the abundance of its member composition changes depending on the physical state of the host. In people with morbid obesity, microbial composition in the gut shifts from providing benefit to causing harmful inflammation through at least in part impairing the intestinal epithelial integrity. Similar microbiota change has been proposed to occur also in aging and is thought to be a cause of various pathologies and diseases, such as frailty, neurodegeneration, insulin resistance and type-2 diabetes (T2D), cancer, cardiovascular disease and Alzheimer's disease. Despite an explosion of reports that link gut microbiota to health in aging, the field remains poorly understood and appears to be confusing. Here, we briefly summarize findings from others and also from the authors' The composition of gut microbiota determines inflammation and possibly lifespan of elderly people. The lumen and particularly the mucin layer of the intestine of young adults are colonized by a diverse population of commensal microbes that co-exist with the host in a symbiotic relationship. Members of Verrucomicrobia phylum, particularly Akkermansia muciniphila, support gut barrier integrity and thus prevent leakage and subsequent induction of inflammation. In elderly people, the composition of the gut commensals is changed and microbial diversity is reduced due to accumulation of potentially pro-inflammatory commensals and decrease of beneficial microbes, such as members of Verrucomicrobia. It therefore leads to gut leakiness and consequent systemic inflammation that facilitates aging-associated morbidities and premature death. Although the microbiota of centenarians changes, its diversity and beneficial commensals are retained, thereby controlling overt inflammation and supporting healthy aging group to emphasize the importance of gut microbiota in the healthy aging of humans. Although dysbiosis and consequent inflammation in aging is assumed to be caused by western style diet (see review, we discuss recent results of microbiota sequencing of elderly people from Italy and China, which suggest that the microbiota change could also be intrinsic to aging process. We propose that the change in gut microbiota is a primary cause of agingassociated pathologies and consequent premature death of elderly people. The most compelling evidence was revealed in studies of genetically homogeneous rodents, such as mice aged in the same environment. It showed that natural Commensal microbes are sole producers of beneficial SCFAs. 1, In the gut of healthy young people, beneficial commensal bacterial members of Firmicutes produce SCFAs. 2, SCFAs provide energy to the microbiota and 3, thereby can inhibit the colonization of opportunistic bacteria; and 4, promote the production of protective mucus. 5, SCFAs are also source of energy for enterocytes and 6, immune cells. They can cross the gut epithelium layer to promote immune tolerance by 7, inducing TGFβ producing FoxP3 + regulatory T cells (Tregs) , 8, IL-10 producing T cells; 9, activating anti-inflammatory responses in antigen presenting cells (APCs); and 10, promoting the production of IgA and IgG from B cells. 11, SCFAs also initiate anti-inflammatory responses in neutrophils and 12, affect their recruitment aging decreases the microbiota diversity and aged mice have diminished bacterial biosynthesis of cobalamin (B12) and biotin (B7), and SOS genes associated with DNA repair and enhanced creatine degradation, which is associated with muscle wasting. Cross-sectional analyses of human fecal microbiota suggest that the fecal microbial population is also altered in elderly people, as in aged mice. The biological and, possibly, aging phenotype in humans appears to depend on four subpopulations (modules) of the "core" fecal microbiota. Although it is difficult to link any given microbe to a clinical outcome, as it may require a cooperative action of multiple microbial species, we also discuss recent findings that underscore the importance of Akkermansia muciniphila in healthy aging. A. muciniphila is a bacterium that degrades mucin and provides energy to other beneficial microbes, including SCFA-producing bacteria. It also protects intestinal epithelial integrity via activation of epithelial cells and production of mucus, thus supporting housing of beneficial commensals. Its decrease in the gut of aged mice and, possibly, elderly people leads to gut leakiness and consequent induction of a low-level of systemic inflammation in aging (termed "inflammaging"). This presumably explains why increased levels of proinflammatory cytokines in the circulation are associated with an overall loss of fitness and poor health in the elderly. However, the question remains unresolved whether the loss of Akkermansia causes or is caused by the decrease of microbiota diversity, which also occurs in aging and associates with frailty in elderly people. The gut dysbiosis and the loss of beneficial Gut dysbiosis in the elderly increases risk of aging-associated diseases. The composition of the gut microbiota changes with age, causing a mild inflammation in the elderly. This change can be exacerbated by additional intrinsic and extrinsic factors, such as uptake of antibiotics and diet. Frail elderly people show increased gut dysbiosis, a severe decrease of beneficial commensal bacteria, such as Akkermansia muciniphila and SCFA-producing bacteria, and a marked increase of opportunistic and potentially proinflammatory commensal microbes. It leads to impairment of the intestinal epithelial integrity and increases gut leakiness and translocation of opportunistic bacteria and endotoxin into the circulation, causing a chain of inflammatory events that enhance the risk of developing aging-associated pathologies. For instance, in aged mice, the A. muciniphila loss-caused inflammation recruits and activates CCR2 + monocytes in the omentum, where they upregulate 4-1BB, CD40L and the production of IFNγ and convert B1a B cells into 4BL cells via 4-1BBL/4-1BB axis. The 4BL cells then promote insulin resistance in aged hosts commensals appear to facilitate premature death of elderly people, as fecal microbiota diversity and abundance of A. muciniphila are increased in human centenarians.
Does the gut microbiota change with age?
The gut microbiome is an endogenous ecosystem populated with Bacteria, Archaea, Eukarya and viruses, where four bacterial phyla of Firmicutes, Bacteroides, Proteobacteria and Actinobacteria account for 98% of microorganisms. It coevolved as a symbiotic superorganism with the host to regulate the normal functions of the gut, such as food digestion and absorption of nutrients, and to provide essential vitamins, nutrients, and polyamines. The microbiota degrades undigestible fibers and is thereby exclusively responsible for production of short-chain fatty acids (SCFAs). SCFAs are involved in multiple processes, such as providing an energy source for microbes and colonocytes, controlling microbial functions, combating pathogens, protecting and maintaining intestinal integrity, and regulating immune cells, including differentiation of CD4 + T cells and activation of CD8 + T cells. As such, the absence of microbiota, e.g. SCFA-generating bacteria, in gnotobiotic (germ-free, GF) mice impairs nutrient absorption, dysregulates intestinal morphology, reduces differentiation and maturation of intestinal immune cells such as intraepithelial lymphocytes, Th17 cells, and regulatory T cells (Tregs). This causes shifts in immune responses towards Th2-type cytokines and impairs production of antimicrobial peptides and IgA. The gut microbiota is considered to be a master regulator of immune homeostasis, as it induces IL-10 and TGFβproducing regulatory T cells (Tregs), Th17 cells, ILC2 and IgA-secreting B cells. For instance, members of Bacteroides fragilis and Clostridium strains induce FoxP3 + Treg differentiation and production of IL-10 and TGFβ, resulting in the inhibition of inflammation. The intestinal segmented filamentous bacteria (SFB), Citrobacter rodentium and fungi like Candida albicans induce Th17 cell differentiation and recruitment of neutrophils and other immune cells in the lamina propria (LP), facilitating pathogen clearance.
The microbial colonization of the GI tract starts after the birth of infants by acquiring a few microbial types dominated by the Bifidobacterium genus. The Bifidobacterium genus dominance declines after the first year of infancy as colonization with other microbes and their diversity increase. The infant microbiota is unstable, and its microbial colonization process is influenced by the mode of delivery and nursing, medications, genetic background, age, and geographic/cultural traditions. Breast-fed infants promote Bifidobacterium-dominated commensals with limited diversity primarily due to human milk oligosaccharides (HMOs), which Bifidobacterium species utilize and convert into lactate and short chain fatty acid acetate. In the absence of HMOs, the gut of formula-fed babies is colonized by a diverse population of microbes. The unstable state of the microbiota in infants presumably explains why the gut microbiota of US infants differs from that of non-US ones, as it is dominated with Prevotella genus including 28 Operational Taxonomic Units (OTU). This genus of mucin-degraders is also overrepresented in the fecal microbiota of children from Burkina Faso compared with children living in Italy. The gut microbiota stabilizes when children start eating "solid food" at around 3 years of age and becomes progressively diverse. Fecal microbiome analyses of children and adults of the Amazonas of Venezuela, rural Malawi, and US metropolitan areas indicated that the phylogenic composition and the functional maturation of bacterial communities shift towards adult-like configuration during the first 3 years of childhood. After that age, the gut appears to be protected from microbial colonization, including from supplementation with a cocktail of 11 strains of Bifidobacterium and Lactobacillus, in contrast to having lifelong effects from the diet.
The gut microbiota of adult humans is dominated by the Firmicutes and Bacteroidetes phyla and smaller proportions of Actinobacteria, Proteobacteria, and Verrucomicrobia. Sequencing analyses of human fecal metagenomes from four countries identified well-defined and robust microbial communities (termed enterotypes) represented by different levels of three genera: Bacteroides, Prevotella and Ruminococcus. The three enterotypes do not correlate with gender, nationality, body mass index or age. However, the existence of these robust enterotypes remains debatable and instead the diet is considered to be a major driver of microbial composition and function(also see commentary by Jeffery et al.,. Moreover, the enterotypes have not been identified in a large study of healthy and frail elderly people. Recently, the study of Italian centenarians (99-104 old), and semi-supercentenarian (105-109 old) suggested the presence of a different core microbiota dominated with Ruminococcaceae, Lachnospiraceae and Bacteridaceae families, which decrease with aging. An independent study of Chinese centenarians revealed a negative association between extreme aging and the abundance of Coprococcus, Roseburia and Faecalibacterium genera, belonging to the Lachnospiraceae and Ruminococcaceae families. Despite differences in race, lifestyle and diet of the centenarians, the two studies found 11 shared features among the top 50 microbes, such as comparable changes in members of Blautia, Clostridium cluster XIVa, Faecalibacterium, Escherichia_Shigella, unclassified Lachnospiraceae, Ruminococcaceae, and Erysipelotrichaceae. Other groups reported that the enrichment of Bacteroidetes and Protobacteria abundances and decrease in species of Bifidobacteria and Lactobacilli in aged people. Interestingly, the Italian and Chinese study of centenarians found longevity increases microbial community richness (Chao index and observed operational taxonomic units, OTUs) and the abundance of subdominant but health-related bacterial genera and families, such as Oscillospira, Christensenellaceae, Akkermansia and Bifidobacterium. This presumably implies a healthy state of the elderly cohorts because Oscillospira and Christensenellaceae control leanness and decrease certain inflammatory diseases in humans. A. muciniphila protects the intestinal epithelial integrity, supports beneficial SCFA-producing bacteria and reduces inflammation and metabolic impairments such as insulin resistance. Bifidobacterium generates lactate and SCFA acid acetate and reduces pro-inflammatory microbes. The aging-associated increase of Oscillospira, when compared to children and middle-aged adults, was also noted in recent multivariant unsupervised reanalysis of 16S DNA sequencing data from 371 people ranging from newborn babies to centenarians.
Overall, despite significant interindividual variability and influence of external factors such as diet, medications, type of exercise or mobility, and the geographical locations of the host, the composition of the gut microbiota changes progressively as people age. Although the health status of subjects is often not reported, we can conclude that unlike young and middle-aged adults, the gut of elderly people is reduced in beneficial and enriched in pro-inflammatory commensal microbes. This change is presumably "intrinsic" to the aging process, as it also occurs in genetically homogeneous mice aged and fed in the same condition. Aged mice exhibit a decrease in beneficial gut bacteria, such as A. muciniphila and SCFA-producers in Clostridium members of cluster IV, and an increase pro-inflammatory microbes. The changes in the gut microbiota composition with age are illustrated in .
Does the alteration of the gut microbiota affect the health of elderly people?
Metchnikoff's original concept published over a century ago explains the disabilities of elderly people by accumulation of "putrefactive bacterial autotoxins", which leaked from their colon. He postulated that this leakage results in the conversion of phagocytes into destroyers of healthy tissues. This idea was confirmed centuries later by demonstrating the importance of chronic inflammation on activation as well dysfunction of phagocytic cells. Fecal microbiota markedly differs in people living in community-dwelling and long-term care nursing facilities, consistent with its role in healthy aging. Dysbiosis, increased microbiota instability and loss of its members in community-dwelling adults can diminish numerous benefits of commensal microbes to the host. This probably explains why decrease of the gut microbial diversity escalates as the frailty of elderly humans increases, positively associating the gut dysbiosis of community-dwelling adults and the biological, but not chronological, age. Fecal microbiota of elderly frail people and aged mice is enriched in the Bacteroidetes phylum and the Oscillibacter and Alistipes genera and Eubacteriaceae family and is reduced in Faecalibacterium and Lactobacillus. Similarly, our group reported that elderly macaques and mice acquire increased insulin resistance because of gut dysbiosis, such as marked decrease of beneficial commensals and their metabolite SCFAs in the gut and in the circulation and increase of pro-inflammatory microbes. A primary driver of this process in aged mice was Akkermansia, whose decrease caused intestinal leakage and activation of CCR2 + monocytes to convert innate B1a B cells into pathogenic 4-1BBL + TNF + B cells (termed 4BL cells), which in turn induced insulin resistance via the 4-1BBL/4-1BB axis. Conversely and despite lifestyle and racial differences, both Chinese and Italian centenarians show an increase of the alpha diversity in the top 500 OUTs(although the Simpson reciprocal index of diversity was reduced in Italian centenarians, suggesting that healthy aging may benefit from the retention of richness of commensal microbes in the gut. Overall, the decreased diversity of the intestinal commensal microbes, which often manifests in reduction of beneficial and enrichment of pro-inflammatory members, can have a detrimental consequence in healthy aging and thus longevity.
## The importance of microbial metabolites such as short chain fatty acids in immunity
Although the composition of the gut microbiota changes with age, the core of microbiota does not age per se. Fecal DNA sequencing and PCR analyses mostly reveal change in abundance of its individual members, which as a part of a symbiotic ecosystem then affects the microbial community and their crosstalk with the host. The abundance of Bifidobacterium-dominated commensals in the gut microbiota of breast-fed infants, which generates lactate and a beneficial SCFA (acetate) from milk oligosaccharides, progressively decreases together with the diversification of the microbial community after weaning and transition to adulthood. Their decrease in young adults is complemented by other bacteria that metabolizes fibers to SCFAs, such as by their key producer Firmicutes. This explains why the abundance of Bifidobacterium inversely correlates with inflammaging in elderly people and centenarians, consistent with their benefit in prolonging longevity of mice as inhibitors of proinflammatory cytokines (for example from macrophages) and colonic senescence, and inducers of colonic tight junctions and mucus production.
SCFAs provide energy to commensal microbes, immune cells and the colonic epithelium, and induce production of mucus e.g., they protect the epithelial barrier functions, support the growth of various beneficial commensal microbes, and promote immune tolerance and gut homeostasis. SCFAs also regulate immune responses by directly stimulating immune cells. In airways of patients with cystic fibrosis, the increase of SCFAs, such as acetate, propionate, and butyrate, associates positively with neutrophil infiltration in the sputum and negatively with expansion of Pseudomonas aeruginosa. The SCFAs induce production of IL-8/CXCL8 and release of granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF), which are needed in the recruitment and persistence of neutrophils, while inhibiting the synthesis of nitric oxide synthase involved in airway inflammation. However, the increase of SCFAs, which can be achieved by feeding with a high-fiber diet, inhibits neutrophil recruitment and consequently airway inflammation and thereby enhances survival in influenza-infected mice. In this study, SCFAs also impair infiltration of neutrophils in the lung airways because they recruit bone marrow-derived Ly6c − patrolling monocytes, which upon their differentiation into alternatively activated macrophages reduce expression of neutrophilrecruiting CXCL1 chemokine as well activate effector functions of CD8 + T cells. Extracellular SCFAs are also utilized as substrates for β-oxydation of fatty acids in CD8 + T celland colonocyte metabolismafter they are taken up by several G-protein coupled receptors (GPRs). Butyrate uncouples the tricarboxylic acid cycle from glycolytic input and enhances memory potential and recall responses of antigen-primed CD8 + T cells via GPR41, and GPR43. By targeting GPR109A, GPR41 and GPR43, which are expressed on the surface of macrophages, dendritic cells and neutrophils, butyrate induces anti-inflammatory responses. Similarly, through the interaction with GPR109A, it promotes anti-inflammatory pathways in colonic macrophages and dendritic cells and induces differentiation of Tregs and IL-10-producing T cells. Butyrate upregulates expression of transcription factor Blimp-1 and IL-10 production in Th1 CD4 + T cells (without affecting conversion of Tregs or Th17 cells) via GPR43 and controls colitis. SCFAs from Clostridium species clusters IV and XIV induce and regulate colonic FoxP3 + Tregs, although other SCFA non-producer microbes also support this process as they are required in the restoration of Tregs in GF mice after butyrate supplementation. Butyrate is a potent inhibitor of histone deacetylases (HDACs), and its internalization via the above-noted GPRs or other unknown pathways causes epigenetic modifications of nonhematopoietic and hematopoietic cells, promoting anti-inflammatory immune responses through inhibition of the production of multiple pro-inflammatory cytokines (IFN-γ, IL-6, IL1-β) and induction of the release of antiinflammatory IL-10 and TGF-β. As an inhibitor of HDAC3, butyrate induces differentiation of monocytes to macrophages as well as metabolic and transcriptional changes in macrophages enhancing their bactericidal functions. SCFAs inactivate nuclear factor-κB (NF-κB) and reduce production of TNFα in peripheral blood mononuclear cells and neutrophils. By supporting differentiation of antibody-producing B cells, SCFAs increase both intestinal IgA and systemic IgG responses to combat pathogens. Upon entering the bloodstream and reaching distant organ sites such as brain and lungs, SCFAs control systemic immune responses. By regulating Tregs in the lungs, SCFAs protect against allergic airway diseases. In the brain, SCFAs directly affect microglia maturation and functionand thereby reduce neuroinflammation. However, SCFAs can also exert pro-inflammatory responses under certain conditions. For instance, SCFAs can aggravate colitis-associated inflammation in mice by inducing expression of T-bet and IFNγ in Tregs and in conventional CD4 + T cells. Also, SCFAs can exacerbate inflammation in airway epithelial cells, and the generation of IL-17 and IFNγ producing T cells in the CNS in mice with multiple sclerosis. In the mouse model of Parkinson's disease (PD), SCFAs are linked to acceleration of pathogenic α-synuclein (αSyn) aggregation and motor deficits. The different functions of SCFAs in the gut microbiota are illusrated in .
## Alteration of the scfa levels in elderly people
As discussed above, SCFAs are produced by a number of bacterial populations working in concert with a community of commensals. For example, A. muciniphila not only supports growth of other commensal microbes by liberating oligosaccharides as well maintaining the mucus layer, but also induces butyrate production from Anaerostipes caccae, Eubacterium hallii, and Faecalibacterium prausnitzii, while E. hallii upregulates pseudovitamin B12 to stimulate propionate production from A. muciniphila. Butyrate improves barrier function of the intestinal epithelial cells (IECs) and protects them from C. difficile toxin damage by activating hypoxiainducible factor 1α. This probably explains why the gut of centenarians retains high levels of butyrate and has increased quantities of the butyrate producers Anaerotruncus colihominis et rel. (from Clostridium cluster IV) and Eubacterium limosum et rel. (from Clostridium cluster XV), Bifidobacterium and other the health-associated bacteria, such as Akkermansia and Christensenellaceae. A. muciniphila and its outer membrane protein Amuc_1100* also stimulate IECs to induce mucus production necessary for protection of the intestinal barrier integrity and the support of other beneficial commensals. Therefore, the decrease of this important bacterium causes gut dysbiosis and impairs the intestinal epithelial integrity, which increases gut leakiness and systemic endotoxemia. A resulting chronic proinflammatory state with higher levels of circulating IL-6, IFNs, TNFα, and IL-1 then promotes poor fitness and frailty in elderly humans. Our group recently reported that the decrease of A. muciniphila in the gut of aged mice triggers a chain of inflammatory events that manifests in the increase of insulin resistance. It starts from the reduction of the colonic mucin layer, which presumably leads to the loss of commensal bacteria producing butyrate, such as I. butyriciproducens, F. prausnitzii, R. faecis, and A. butyraticus, and thus a reduction in SCFAs, particularly butyrate, in the gut lumen and in the circulation. The decrease of this important source of energy to colonocytes and suppressor of proinflammatory and pathogenic commensal bacteriafurther enhances dysbiosis and gut leakage in aged mice, thereby sustaining inflammaging. Because butyrate is also a potent inhibitor of TLR4 signaling, its decrease in the circulation enables bacterial stimuli, such as endotoxin, to freely induce inflammation in the omentum and to upregulate expression of 4-1BB, CD40L, and IFN-γ in BM-derived inflammatory CCR2 + monocytes. Upon infiltration into the omentum, these monocytes convert innate B1a B cells into 4BL cellsvia 4-1BBL, CD40, and IFNγR1 signaling. The 4BL cells then promote insulin resistance in aged hosts, at least in part, utilizing the 4-1BBL/4-1BB axis. A similar pathway appears to exist in elderly humans and macaques, as their increased insulin resistance associates with 4BL cells induced by aging-activated monocytes. In elderly people, the level of SCFAs from carbohydrate fermentation is decreased while metabolites from protein fermentation (branched fatty acids, ammonia and phenols) are increased, indicating a shift from saccharolytic fermentation to unfavorable proteolytic activities. This shift occurs progressively as elderly people age, and can be accelerated upon the use of antibiotics or with low-fiber diets. The decrease in the level of SCFAs enhances susceptibility to opportunistic bacterial species and pathogens. For instance, treatment with the antibiotic cefoperazone reduces SCFAs and thereby increases colonization of the fungus Candida albicans in mice. Similarly, antibiotic-induced depletion of butyrate-producing microbes decreases butyrate levels, which leads to epithelial oxygenation and expansion of aerobic bacteria such as Escherichia coli and Salmonella enterica. Mechanistically, the colonocyte oxygenation is primarily linked to decreased activity of the intracellular butyrate sensor peroxisome proliferatoractivated receptor γ (PPAR-γ), which leads to activation of iNOS and increase of nitrate levels. The mechamisms in which gut dysbiosis increases age-related diseases are illustrated in .
## Contribution of other microbial factors in inflammaging
The gut microbiota controls local immune responses, such as the development and regulation of the intestinal immune responses. Germ-free rodents show an alteration in the development of the gut-associated lymphoid tissues (GALT) and maturation of isolated lymphoid follicles (ILFs), leading to significantly lower numbers of T cells, including Tregs, and IgA-producing B cells. On the other hand, dysbiosis induces systemic inflammation, explaining relatively high plasma levels of MCP-1/CCL2, IL-8/CXCL8, TNFα, IL-6, and Creactive protein (CRP) in elderly people living in longterm care facilities. Underscoring the importance of gut microbiota, the inflammatory phenotype in aging is also transferable, as GF mice show signs of inflammaging after transplantation with fecal microbiota of from aged, but not young, mice. Similarly, the fecal transplantation from aged, but not young, mice induces the genration of 4BL cells in GF mice, primarily due to absence of A. muciniphila and SCFA-producer bacteria. However, gut microbiota modulates inflammaging through a plethora of additional mechanisms. For instance, the gut of long lived elderly people is decreased in proinflammatory F. prausnitzii and is increased about 15-fold in E. limosum. Elderly people and centenarians also markedly decrease abundance of Bifidobacterium genus in the gut, presumably explaining increased inflammaging and consequent aging-associated morbidity and mortality. Bifidobacterium is an important producer of lactate, which affects acidity of the environment, and polyamines that scavenge reactive oxygen species (ROS). It also induces stress-response genes, regulates NFκB activation, inhibits pro-inflammatory cytokines from macrophages, suppresses colonic senescence, maintains colonic tight junctions and mucus production, and even prolongs longevity in mice, implying that Bifidobacterium supplementation may also reverse agingassociated impairments in elderly humans. However, recent systematic reviews and meta-analyses of randomized numerous probiotic trials revealed no significant benefit in protection from infection, improving NK cell function, quality of life and mortality of elderly people as compared with placebo controls. Despite importance of probiotics, the key question remains unresolvedwhether they can efficiently colonize the gut of adults and elderly people. Recent study failed to find the colonization evidence in adults supplemented with a cocktail of 11 Bifidobacterium and Lactobacillus strains failed. It is also unclear whether orally given probiotics can survive acidity of the stomach and competition from the commensal microbiota. Lastly, if the bacterial decrease in aging is caused by the immune system, attempts to restore them, for example by supplementing Bifidobacterium, may instead accelerate their loss due to activation of the immune response.
## Concluding remarks
Despite significant interindividual and lifestyle differences, the composition of the gut microbiota of elderly humans markedly differs from that of young and middle-aged adults. The compositional shift coincides with the onset of immune dysregulation and manifestation of aging-associated pathologies, e.g. ≥ 70 years of age. In elderly, particularly frail people, the composition of the gut microbiota shows signs of dysbiosis, such as a marked decrease in diversity of its population due to the accumulation of proinflammatory commensals and reduction of beneficial microbes. The decrease of beneficial microbes, particularly supporters of mucin production and producers of SCFAs, appears to be essential in triggering a chain of inflammatory events, such as the impairment of intestinal barrier integrity and increase of gut leakiness, endotoxemia and subsequent inflammaging and agingassociated morbidities. As in aged fruit flies where intestinal permeability increases mortality, we therefore propose that the gut dysbiosis and leakiness is a major cause of premature death in elderly people. Consistent with ancient philosophers and Elie Metchnikoff's vision and as lifestyle improvements show, extrinsic manipulations can control the ills of the elderly to maintain healthy aging. This concept is validated in modeling studies in rodents, revealing that aging-associated pathologies are reversible. For example, the gut microbiota-induced inflammaging and consequent increased insulin resistance is reversed by restoring A. muciniphila, supplementing with butyrate, or inactivating 4BL cells and monocytes in aged mice and macaques. On the other hand, little is known about the cause(−s) of dysbiosis, the safety and potential health risk of microbiota-based interventions in elderly people. Further research is needed to fully understand the benefit of probiotics and their use in humans. |
Identifications and classifications of human locomotion using Rayleigh-enhanced distributed fiber acoustic sensors with deep neural networks
This paper reports on the use of machine learning to delineate data harnessed by fiber-optic distributed acoustic sensors (DAS) using fiber with enhanced Rayleigh backscattering to recognize vibration events induced by human locomotion. The DAS used in this work is based on homodyne phase-sensitive optical time-domain reflectometry (φ-OTDR). The signal-to-noise ratio (SNR) of the DAS was enhanced using femtosecond laser-induced artificial Rayleigh scattering centers in singlemode fiber cores. Both supervised and unsupervised machine-learning algorithms were explored to identify people and specific events that produce acoustic signals. Using convolutional deep neural networks, the supervised machine learning scheme achieved over 76.25% accuracy in recognizing human identities. Conversely, the unsupervised machine learning scheme achieved over 77.65% accuracy in recognizing events and human identities through acoustic signals. Through integrated efforts on both sensor device innovation and machine learning data analytics, this paper shows that the DAS technique can be an effective security technology to detect and to identify highly similar acoustic events with high spatial resolution and high accuracies.Large-scale facilities and critical infrastructures are currently monitored via the heavy use of security technology such as cameras that operate in the visible or mid-infrared wavelengths. The fact that most of these cameras are equipped with facial recognition software has raised privacy concerns globally. However, these devices only offer the limited information they gather through line of sight, and this information can be distorted by various disguises such as masks or heavy make-up. In addition to imaging technology, vibration-based sensor systems have also been widely used to monitor and detect various intrusion and trespassing events. Sensor networks constructed using electrical seismometers and accelerometers can be used to monitor and to spatially resolve location of vibration caused by human locomotion and other intrusion events. However, extensive deployment of point sensors such as camera or electrical vibration sensors requires significant installation costs, especially in large-scale infrastructures such as national borders, pipelines and railways, etc.Over the last two decades, distributed fiber-optic sensors have been explored and deployed for structural health monitoring and prognoses. A unique trait of fiber-optic sensor technology is its application in distributed sensing. The optical fiber itself can be used as a sensor to perform, at high spatial resolution, both static and dynamic measurements of features such as temperature, strain, and acoustic detections. By exploiting various optical scattering mechanisms (i.e., Rayleigh, Raman, and Brillouin scattering) that naturally occur in optical fibers, distributed measurements can be achieved with a high spatial resolution of 1-10 m using optical fibers spanning up to 100 km in length.Given the large-scale use of fiber-optic sensors in structural health monitoring, the same fiber-optic sensor network can be used in security monitoring. The inherent advantage of this lies in the fact that fiber-optic sensors OPEN
Scientific Reports | (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 www.nature.com/scientificreports/ are less invasive of personal privacy, less easily fooled by disguises and free from line-of-sight detection. Furthermore, installation of fiber-optic sensor networks has the potential to be significantly cheaper, as distributed fiber sensing schemes require only one or two pieces of optical fiber in order to cover a large facility. In the last two decades, several distributed or quasi-distributed fiber sensors were explored for security applications. Fiber sensors such as the Michelson interferometer 1 , Mach-Zehnder interferometer 2 , Sagnac interferometer 3 , fiber Bragg gratings (FBG) network [bib_ref] Unattended ground sensor based on fiber Bragg grating technology, Zhang [/bib_ref] , Rayleigh optical time-domain reflectometry (OTDR) [bib_ref] Blue Rose perimeter defense and security system, Blackmon [/bib_ref] , φ-OTDR [bib_ref] Environmental Monitoring and Ecological Applications; Optical Sensors in Biological, Chemical, and Engineering..., Vdovenko [/bib_ref] , and Brillouin OTDR [bib_ref] Feasibility study of the automated detection and localization of underground tunnel excavation..., Klar [/bib_ref] have been explored for use in intrusion detection through detection of vibration induced by human locomotion. Quasi-distributed FBG sensor networks have been used extensively to spatially locate various intrusion events and maintain perimeter security. Through wavelength-or time-division multiplexing, intrusion events can be detected with high spatial resolution [bib_ref] Real-time activity identification in a smart FBG-based fiber-optic perimeter intrusion detection system, Wu [/bib_ref]. Using various pattern-recognition algorithms such as principal component analysis and K-Nearest Neighbor classifier, movements (e.g. people climbing, knocking, digging, or walking; even cats jumping), can be successfully detected and classified. However, the applications of sensor networks to not only detect but also to identify human locomotion is far more challenging than identifying acoustic event with distinctly different acoustic characteristics demonstrated before. Acoustic signal patterns of human locomotion produced by different individuals are weak and only have subtle difference among each other. Successful identifications of individual using his or her locomotion acoustic pattern requires both high quality sensor data in term of high signal to noise ratio (SNR) and high spatial resolution, as well as robust data analytics algorithms.
Fiber optical Distributed Acoustic Sensor (DAS) based on Rayleigh scattering, is a promising fiber sensor technology used in distributed vibration detection. It uses unperturbed fibers to determine the location and transient waveforms of acoustic events [bib_ref] Advances in fiber-optic distributed acoustic sensors, He [/bib_ref]. Rayleigh scattering-based OTDR can detect acoustic emissions via a wide frequency bandwidth. A 128 km φ-OTDR sensor reportedly detected intrusions with 15 m spatial resolution [bib_ref] 128km fully-distributed high-sensitivity fiber-optic intrusion sensor with 15m spatial resolution, Peng [/bib_ref]. Therefore, DAS is an ideal solution to detect acoustic events with high spatial resolutions. The performance of a φ-OTDR system, such as the sensing dynamic range and sensitivity, is governed by the system SNR, which is limited by weak Rayleigh back-scattering coefficient of an unmodified optical fiber [bib_ref] Polarization discrimination in a phase-sensitive optical time-domain reflectometer intrusion-sensor system, Juarez [/bib_ref]. Most of researchers address low cadence of sensor data through data analytics. Simple data processing techniques, such as moving average and wavelet transformation, have been used to enhance the quality of sensor signals [bib_ref] Distributed vibration sensor based on coherent detection of phase-OTDR, Lu [/bib_ref] [bib_ref] Wavelet denoising method for improving detection performance of distributed vibration sensor, Qin [/bib_ref]. These techniques are time-consuming, and will decrease the maximum detectable frequency of φ-OTDR systems. Multiscale wavelet decomposition [bib_ref] Multi-scale wavelet decomposition and its application in distributed optical fiber fences, Wu [/bib_ref] , Gaussian mixture models [bib_ref] A novel fiber optic based surveillance system for prevention of pipeline integrity..., Tejedor [/bib_ref] [bib_ref] Real field deployment of a smart fiber-optic surveillance system for pipeline integrity..., Tejedor [/bib_ref] , and Morphologic feature extraction [bib_ref] Recognition of a phase-sensitivity OTDR sensing system based on morphologic feature extraction, Sun [/bib_ref] have been used as pattern recognition (PR) tools. However, the fundamental issue of weak backscattering signals in telecom-grade optical fibers still limits the performance of fiber-optic DAS sensors. Poor distributed sensor performance therefore severely limit effectiveness of data analytics to recognize significantly different acoustic patterns, such as human and motor vehicle movement or safe excavator operations and accidental excavator contact with a gas pipeline [bib_ref] Long fiber-optic perimeter sensor: Signature analysis, Madsen [/bib_ref].
To address these challenges, this paper, for the first time to our best knowledge, presents an integrated approach to improve performance of distributed fiber optical acoustic sensors to identify highly similar acoustic events such as human locomotion through innovations at both sensor device levels and through machine learning data analytics. Using a femtosecond laser direct writing scheme, Rayleigh backscattering of optical fibers were enhanced over 35 dB. This fundamentally address poor SNR of DAS using unmodified optical fibers to achieve high-precision dynamic strain measurements. Through deep neural network machine learning data analytics, this paper shows that it is possible to detect and identify human locomotion using distributed fiber sensors with high accuracy.
# Methods
Sensor fabrications. In a typical DAS system using an φ-OTDR scheme, time-resolved Rayleigh backscattering from unmodified telecom single mode fibers is used as a sensing signal to detect phase changes caused by vibrations. However, telecom grade optical fibers have an intrinsically weak Rayleigh scattering profile that severely limits SNR in DAS systems. To mitigate this problem, a deep ultra-violet (DUV) laser was used to produce weak FBG in photosensitized fibers using a phase mask approach [bib_ref] Improved Φ-OTDR sensing system for high-precision dynamic strain measurement based on ultra-weak..., Zhu [/bib_ref] [bib_ref] Distributed OTDR-interferometric sensing network with identical ultra-weak fiber Bragg gratings, Wang [/bib_ref]. Recently, blank DUV laser exposure was used to enhance the Rayleigh scattering profile of telecom fibers [bib_ref] Rayleigh scatter based order of magnitude increase in distributed temperature and strain..., Loranger [/bib_ref]. Enhancement of up to 15 dB in the Rayleigh scattering profile is achievable through DUV radiation. However, both approaches require photosensitization of telecom fibers using a process such as hydrogen loading. Weak FBG and Rayleigh enhancement via DUV laser are not stable at elevated temperatures [bib_ref] High-temperature capabilities and limitations of fiber grating sensors, Morey [/bib_ref].
To address these challenges, an ultrafast laser roll-to-roll direct writing approach was used to induce temperature-stable artificial scattering centers as nano-reflectors for enhancing Rayleigh scattering. The detailed manufacturing process (see [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref] can be found in a previous report [bib_ref] Distributed optical fiber sensors with ultrafast laser enhanced Rayleigh backscattering profiles for..., Yan [/bib_ref]. The femtosecond laser system used for laser fabrication includes a Coherent Mira 900 mode-locked Ti:Sapphire oscillator and a Coherent RegA 9000 regenerative amplifier. The laser system generated 270-fs (FWHM) laser pulses at 800 nm with a 250 kHz repetition rate. By tightly focusing the ultrafast laser on the optic fiber core and using an oil-immersion objective (NA 1.25), nano-reflectors were inscribed at an on-target pulse energy of ~ 160 nJ. The nano-reflectors were inscribed through the protective jacket to preserve the fiber's mechanical integrity. The irradiation time and laser pulse power were carefully adjusted for optimally enhanced level and low insertion loss. The detail laser inscription method and laser processing parameter optimization can be found in ref. [bib_ref] Multiplexable high-temperature stable and low-loss intrinsic Fabry-Perot in-fiber sensors through nanograting engineering, Wang [/bib_ref]. The insertion loss of each Rayleigh-enhanced point went up following an increase in pulse energy when inscribing severe nanodefects. The optimal laser processing condition can produce nano-reflectors in sensing fiber with insertion loss at 0.0012-dB/reflector [bib_ref] Multiplexable high-temperature stable and low-loss intrinsic Fabry-Perot in-fiber sensors through nanograting engineering, Wang [/bib_ref].
Fabrication of weak reflectors in optical fibers does introduce additional attenuation, which could reduce the total sensing length. However, we have demonstrated that through optimization of the femtosecond laser direct writing process, insertion loss of each nano-reflectors can be minimized to 0.0012 dB/point. For example, Scientific Reports | (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 www.nature.com/scientificreports/ a 2.5-km long sensing fiber with 830 nano-reflector points (~ 3-m spacing), will incur 1-dB additional insertion loss. This is a small penalty to pay given a 35 dB backscattering enhancement produced by nano-reflectors. This backscattering enhancement and insertion loss trade-off have been validated by previous works using FBGs as weak reflectors [bib_ref] Multiple-Octave-Spanning vibration sensing based on simultaneous vector demodulation of 499 Fizeau interference..., Li [/bib_ref]. [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref] shows the setup for nano-reflectors fabrication. A stand of telecom single mode fiber (Corning SMF-28e+) was translated via a roll-to-roll setup. The backscattering profile of the optical fiber was monitored by an Optical Backscatter Reflectometer (Luna OBR4600). The femtosecond laser inscribed, at 3-m intervals, six Rayleigh-enhanced points onto a fiber core spanning over 21 m. Each Rayleigh-enhanced point was 2 × 2-μm (cross section) by 7-μm (L) in the middle of the fiber core. Enhancements of over 35 dB were achieved through the laser direct writing process, as shown in [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref]. [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref] shows the φ-OTDR distributed sensor interrogation setup. Unlike weak FBGs, which only reflect light at the FBG wavelength and are sensitive to temperature fluctuation, femtosecond laser-induced nano-reflectors produces broadband Rayleigh backscattering. The increased Rayleigh backscattering signal dramatically improves SNR. This is shown in [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref] , while signals from three Photo detectors (PD1-PD3) from pristine fibers and those from Rayleigh enhanced fibers are compared. The Rayleigh enhanced backscattering points significantly improve SNR and thus precision of measurements. [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref]. A DFB fiber laser with a 6 kHz linewidth and 1550.12 nm output wavelength (NKT Basik C15) was used as the interrogation laser source. The laser was modulated by a semiconductor optical amplifier (SOA, Thorlabs, SOA1013SXS) to generate a 15-ns pulse with a 50 kHz repetition rate. The laser pulses were amplified by an Erbium-doped fiber amplifier to reach 150-nJ/pulse (10 mW output power) and launched into the laser-enhanced fiber (fiber under test). The 3-meter sections of fiber between adjacent Rayleigh-enhanced points formed a fiber acoustic sensor for the DAS system. Therefore, five acoustic sensors were interrogated over a 21-m fiber span. The backscattered optical pulses were redirected to a 3 × 3 coupler through an optical circulator. The backscattered pulses were equally divided into 3 parts, as shown in [fig_ref] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber [/fig_ref]. Two output ports of the 3 × 3 coupler were connected to Faraday rotator mirrors. The optical path difference between the two output ports was tailored to be 3 m, matching the sensing fiber length between adjacent Rayleigh-enhanced points. Vibration-induced physical perturbation along the sensing fiber was then demodulated by three photodetectors (InGaAs Fixed Gain 200 MHz) using the Naval Postgraduate School (NPS) method [bib_ref] Improved demodulation scheme for fiber optic interferometers using an asymmetric 3×3 coupler, Zhao [/bib_ref].
## Das phase demodulation. a schematic of the das system is shown in
The relationship between the mechanical strain and refractive index in sensing fiber was described by the photo-elastic effect that determined the phase shift of transmitting light. The NPS demodulation method for www.nature.com/scientificreports/ phase extraction is briefly described in Eqs. (1) and (2). The three output signals of the 3 × 3 coupler have a 120° constant phase shift, which can be described as:
where φ(t) is the signal phase, C is the DC component of voltage output from photodetectors, B is AC amplitude of photodetector's voltage signals [bib_ref] A symmetric 3x3 coupler based demodulator for fiber optic interferometric sensors, Brown [/bib_ref]. The phase of signal can be decomposed to φ(t) = ϕ(t) + ψ(t) , where ϕ(t) and ψ(t) are phase shifts caused by physical perturbation and environmental noise, respectively. After some differential and integral operations, the output of the NPS algorithm is proportional to the value of φ(t):
where A is the gain of operators in the NPS method. Therefore, phase changes caused by vibrations can be quantitatively detected for further analysis. More technical details on the DAS system using the 3 × 3 coupler demodulation scheme can be found in Ref. [bib_ref] Multiple-Octave-Spanning vibration sensing based on simultaneous vector demodulation of 499 Fizeau interference..., Li [/bib_ref].
Experimental setup. Experiments were carried out in a 15-m-long hallway, shown in [fig_ref] Figure 2: The test site with marks for two tracks and distributed fiber sensors... [/fig_ref]. The distributed fiber sensors were laid in a straight line on the concrete floor. Masking tape was used to introduce mechanic contact between the concrete floor and the sensors. Six Rayleigh backscattering-enhanced points were fabricated at 3-m intervals by an ultrafast laser to form five acoustic sensors for measuring floor vibrations. Two tracks were marked out on the concrete floor using green tape (see [fig_ref] Figure 2: The test site with marks for two tracks and distributed fiber sensors... [/fig_ref]. The two tracks' distance to the optical fiber was 80 cm and 140 cm, respectively.
Data collection and pre-processing. Pedestrian identification was reported previously based on electronic vibration point sensors. Data analytic approach such as support vector machine (SVM) [bib_ref] Indoor person identification through footstep induced structural vibration, Pan [/bib_ref] and iterative transductive learning algorithm [bib_ref] Footprintid: Indoor pedestrian identification through ambient structural vibration sensing, Pan [/bib_ref] have been used to delineate acoustic signal produced by solo walking pedestrian. Using distributed fiber sensors, a great diversity of locomotion data was harnessed in our experiments. These include acoustic signals produced by both walking and running persons (solo and pairs), and signals differentiated in human locomotion at different tracks, shoes and pushing cart. Multiple acoustic events induced by eight individuals were listed in [fig_ref] Table 1: Description of dataset from events involving human movements [/fig_ref]. Both shallow and deep neural network machine learning algorithm were used to analyze and to classify sensor data. For acoustic events produced by a solo person, each participant repeated a given event (e.g., running or walking) 26 times on both tracks. Additional variations, such as acoustic signals generated by the same individual but wearing different shoes, were also recorded. Both human locomotion and mechanical system like pushing a cart was introduced as extra noise in order to test the robustness of the algorithm. For events involving two participants, random combinations of the individuals running and/or walking were selected.
(1) www.nature.com/scientificreports/ The multiple sensors and high bandwidth data acquisition led to an enormous amount of data. In total, acoustic signals generated by eight different events were recorded, producing 1196 sets of acoustic signals for machine learning application. The bandwidth for acoustic signals captured by the data acquisition system was capped at 2-kHz for each sensor (4-kHz sampling rate at 14 bits). Each sensor continually recorded a 1-s acoustic signal for ten sequential frames. The idle time between segmented frames was due to limited operational capability in the computer-based data acquisition and processing. Therefore, for each acoustic event, 200 k data points were recorded (4 k sampling/s × 5 sensors × 10 s).
[formula] � V out1 V out2 V out3 � = C + Bcosφ(t) C + Bcos � φ(t) − 2 3 π � C + Bcos � φ(t) + 2 3 π � (2) Vout = √ 3Aφ(t) = √ 3A[ϕ(t) + ψ(t)] [/formula]
Before applying machine learning methods for classification, the data were preprocessed to reduce computation times for the various machine learning algorithms. First, all acoustic data was transformed to frequency domains by Fast Fourier Transform (FFT) in order to offer neural networks direct access to the global properties of the signals.shows an example of raw acoustic signal harnessed by fiber sensors and its first 256 frequency components after FFT. Both our results shown inand past studies have shown that human locomotion produce low-frequency vibration signals [bib_ref] Vibration and sound signatures of human footsteps in buildings, Ekimov [/bib_ref]. Therefore, a non-causal low-pass filter was applied to the signal in order to reduce the computational burden of the subsequent algorithm without losing relevant information. The order of a non-causal filter is infinite and the cutoff frequency is 32 Hz, which corresponds to the first www.nature.com/scientificreports/ 32 frequency components retained in each spectrum. A sinc filter was then used to smoothly interpolate these 32 amplitudes to 256 points, facilitating the deep learning algorithm shown below. As a result, the experimental data for an acoustic event (e.g., walking or running) consists of 1280 data points, which is 15 times less than the original dataset. The preprocessed data was arranged into a 3D format of 256 (frequency dimension) × 10 (s) × 5 (sensors).show the filtered and smoothed acoustic signal in both temporal and frequency domains.shows both overlay curves plot of acoustic data and stacked intensity maps for visualization. It is evident that the data preprocessing using the low-pass filter keeps the predominant features of signals. Time-domain data presented inshows simultaneous measurements of five sensors in response to the running motion of a single individual. It is easy to deduce that 3 foot-strikes occurred during this 1-s segment measurement. Given that Sensor 4 and 5 measured weak vibration signals, the foot-strikes of the runner occurred between the 0 and 9-m section of the hallway from the location of the first enhanced point. This is consistent with the actual event.
Disclaimer. This work was funded by the Department of Energy, National Energy Technology Laboratory, an agency of the United States Government, through a support contract with Leidos Research Support Team (LRST). Neither the United States Government nor any agency thereof, nor any of their employees, nor LRST, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
# Results
Sensor data. Variations in vibration intensity of the foot strikes also imply such things as running direction and/or imbalance in a person's stride. These are cues for identifying the runner. However, identification through data analysis using the qualitative "eye-ball" approach in the time domain, or quantitative analysis in the frequency domain, become extremely difficult if a large number of participants or a wide range of events are involved. This challenge is shown in [fig_ref] Figure 4: Plots of 1-s time-domain acoustic signals from five sensors are illustrated as... [/fig_ref] , which presents sensor data from various events, including walking or running with different footwear, walking while pushing a cart, and locomotion involving multiple participants. To highlight the frequency features of various events, only first 256 frequency components from vibration data www.nature.com/scientificreports/ in frequency domain is presented in [fig_ref] Figure 4: Plots of 1-s time-domain acoustic signals from five sensors are illustrated as... [/fig_ref]. This is consistent with the vibration feature generated by human locomotive motions.
## Supervised machine learnings. vibration-related information collected by the distributed fiber sensors
includes high-frequency signals simultaneously harnessed from many locations. The enormous volume of data demands powerful analysis tools for extracting the covert features of specific individuals or acoustic events. Identification of a person or event through acoustic data relies on the extraction of subtle, complex patterns, such as the frequency/rhythm of human locomotion, imbalances in a person's stride, and the temporal characteristics of foot impact (see [fig_ref] Figure 4: Plots of 1-s time-domain acoustic signals from five sensors are illustrated as... [/fig_ref]. Though it is difficult to delineate these features using predictive data analyses, a machine learning approach could potentially handle the large amount of data from DAS and unveil patterns associated with specific individuals or acoustic events. The indoor pedestrian identification was reported previously based on electrical vibration sensors to classify slow-walking person by using SVM [bib_ref] Indoor person identification through footstep induced structural vibration, Pan [/bib_ref] and iterative transductive learning algorithm [bib_ref] Footprintid: Indoor pedestrian identification through ambient structural vibration sensing, Pan [/bib_ref]. For this paper, a great diversity of locomotion data including walking and running at different tracks, shoes and pushing cart were harnessed by distributed optical fiber sensor. Both of supervised and unsupervised machine learnings were employed to analyze data in both the time and frequency domains. They included machine learning algorithms like SVM-based error-correcting output codes, multiple-hiddenlayer convolutional neural network and unsupervised adaptive clustering.
## Convolutional neural network (cnn)
. CNN is a well-known machine learning algorithm that uses PR to identify images and various one-dimensional data. Its multiple layers of convolution filters extract local features of the input data and map them onto various categories of classification. During the non-linear activation and pooling processes, covert characteristics of acoustic events can be manifested and interpreted at the end of a fully connected layer. The acoustic signal harnessed from each sensor at a 1-s time frame was organized into a onedimensional data sequence. Data acquired sequentially over ten 1-s time frames were arranged into a 2D data image. These 2D data images were then compiled into 3D data structures by stacking data from multiple sensors. By transforming into a 3D tensor within the frequency domain, the global properties of various intrusions can be efficiently accessed by CNN. [fig_ref] Figure 5: Architecture of the CNN [/fig_ref] shows the CNN structure used to identify movement. The temporal acoustic signals were filtered and smoothed to become 256 frequency components. Ten 1-s frames and five sensors carried out in our experiment increased this vector to a 3D array as 256 (frequency components) × 10 (s) × 5 (sensors), which was fed into CNN.
Three layers of convolutional stages were applied in the CNN structure including 2 × 3 or 3 × 3 convolutional filters and 2 × 2 max pooling. The nonlinear activation function was Rectified Linear Unit, and max pooling was used between each layer to reduce data in each channel. Each layer of a CNN has two goals: increase the size of extracted features and reduce the size of raw data, so that the entire network transforms large size raw data into only useful features. The first goal is achieved through having increasing number of channels from one layer to the next. Each channel represents a feature and each unit in the channel decide whether the feature is present at that layer. The second goal is achieved through pooling. Pooling is similar to the process when an image being zoomed out. Nearby units compare their outputs and keep only one of them. In maxpooling, the one output they keep is the greatest output among them, so that the presence of the feature at that location can be passed onto the next layer. After each layer, the increase of channel numbers, which implied more features, were arranged and extracted. Based on those features, the output stage was set to be a fully connected softmax predicting the probability of fitting into different categories. Softmax is a function to map the non-normalized output of a network to a probability distribution over predicted output classes. It is the standard layer to transform the latent features extracted by the previous convolutional layers into the final classification result. Performance of the CNN was gauged for supervised machine learning, while various acoustic data generated by eight different events [fig_ref] Table 1: Description of dataset from events involving human movements [/fig_ref] were used to train the CNN.
The recognition goals were set to be individuals, tracks, and shoes when using corresponding dataset for training. Acoustic signal dataset generated by various locomotion events summarized in [fig_ref] Table 1: Description of dataset from events involving human movements [/fig_ref] were partitioned into training dataset and testing dataset. For example, among 208 datasets of eight participants' solo walking events, one hundred and twenty-eight datasets were randomly selected for training, while the rest were used for testing. details number of datasets used for training and for testing, respectively. The acoustic signals generated www.nature.com/scientificreports/ by solo participants (running or walking) were also used to train neural networks to classify acoustic signals generated by two participants. Each algorithm was run ten times in order to obtain the range of uncertainty. To describe the prediction outcome as interval might be more informative when the lower limit of classification accuracy indicates the worse-case performance usually referring to the false alarms rate of system. For datasets generated by a single person walking, CNN achieved greater than 78.75% accuracy in specifically identifying the individual (out of eight possibilities) who generated the acoustic signature. CNN achieved even greater accuracy (> 85%) in identifying the specific track(s) used by an individual. When the acoustic signal was generated by simultaneous walking and cart-pushing, CNN achieved better than 93% accuracy in identifying the event, thanks to the presence of the cart. CNN failed to identify the specific individual when different shoes were worn. However, when this dataset was singled out to train CNN, the neural network achieved over 80% accuracy in classifying the signals generated from different pairs of shoes.
Each time, the classifier achieved an accuracy of around 76.25% or higher in identifying the person making the movements, and over 80% accuracy in recognizing carts, tracks, and shoes, with no major difference in regard to walking or running. However, when the classifier was used to identify people in a two-person (grouped) activity, its accuracy decreased to 60.26% in identifying either person, and 28.46% in identifying both. In addition to a person's identity, the classifier was used to identify movements, generating > 54.62% (either) or 46.92% (both) accuracy. A plausible cause for the classifier's failure in this respect is that one subject's signal may be drowned out by the other's. An advanced mathematical model needs to be developed to address this problem.
Long short-term memory (LSTM). LSTM is Recurrent Neural Network's popular architecture for dynamical systems and real-time temporal signal processing. Its robustness is proven in speech recognition and intrusion detection applications. By letting the output serve as its own input, the recurrent layer creates, in effect, a feedback loop or "memory" to track intrinsic dependencies during dataflow. Furthermore, the memory is controlled in the location of being received (input gate), retained (forget gate), or used in the output (output gate) by relying on nonlinear functions performing as regulators. In this work, inputs of the LSTM are two-dimensional array, which consist of raw data harnessed by five fiber sensors. The LSTM layer has 100 hidden units. The output of the LSTM layer is the final output of each unit. A fully connected layer following the LSTM layer transform the output from these 100 hidden units into 8 features. A softmax layer transforms these 8 features into the classification result (among 8 categories). In regard to data generated by eight people running, a comparison is made between CNN and LSTM (see [fig_ref] Table 3: Classification results using different machine learning algorithms or neural networks [/fig_ref]. LSTM's low accuracy in distinguishing among the eight people reveals its weakness in segmented signal recognition. With limited operational capability in terms of computer-based . Classification results using CNN for different scenarios. The number of datasets for training and testing, respectively, is included in parentheses. www.nature.com/scientificreports/ data acquisition/processing, LSTM is neither competent nor adequate for completing the longstanding process of acoustic signal extraction. Furthermore, its analysis of segmented temporal data is inferior to feature recognition in the frequency domain in this application.
## Error-correcting output codes (ecoc), k-nearest neighbor (knn), and naïve bayes (nb) classifiers.
Three more methods were introduced to examine the robustness in the frequency domain: the ECOC, KNN, and NB classifiers. These methods provide strong outcomes in a variety of applications. There are well-developed and fully integrated libraries and toolboxes for these methods in most scientific programming languages, including MATLAB. ECOC is a multi-class extension of the SVM algorithm which combines lots of binary classifiers to solve multi-class problems. One-vs-one coding was used in this experiment, which brought 28 binary classifiers to distinguish eight participants as eight classes. Each binary classifier was a linear SVM that used a box constraint of 1 (a parameter which prevents overfitting from outlier). The KNN classifier is an instance-based supervised classification method. It classifies the test sample to the class to which a plurality of its k nearest neighbors belong. The distance function k was optimized to be set as 1. It simply assigned the test sample to the class of its nearest neighbor. NB classifier is a probabilistic classification method based on Bayesian inference. It assumes the statistic independence of each component of the input variable. Given the training data, a Gaussian probability model with empirical prior was learned for each class. The test samples were classified into multiple classes corresponding to the maximum posteriori estimate of the model parameters. Detailed descriptions on these three frequency-domain machine learning algorithms can be found in Ref.. [fig_ref] Table 3: Classification results using different machine learning algorithms or neural networks [/fig_ref] presents the accuracies to classify one person walking and running from eight participants by using various machine learnings. Among all machine learning algorithms tested in this work, CNN produces best outcomes. This is consistent with other works where CNN yielded better results for spectrum analysis [bib_ref] Spectrum analysis of EEG signals using CNN to model patient's consciousness level..., Liu [/bib_ref] [bib_ref] Spectrum analysis and convolutional neural network for automatic modulation recognition, Zeng [/bib_ref]. Machine learning algorithms often have strong assumptions about the statistic natures of the data. For instance, NB has a strong assumption about the distribution of data, and KNN assumes that the data clusters with respect to a given metric. The assumptions behind CNN are relatively generic, and it has in addition benefits of its depth, which hierarchically extracts features for the data. This might be the reason for higher classification accuracy than other approaches.
Unsupervised machine learning. Unlike supervised learning, unsupervised learning methods extract features from the training data without the use of any labels. This eliminates the tedious work of labeling a large amount of training data needed for supervised learning schemes. Clustering is an example of unsupervised learning by partitioning the data into different clusters based on certain criteria. K-means clustering with Euclidean distance is a basic clustering problem commonly used for feature extraction. Its objective is to minimize the sum of total variances within the clusters. While the exact solution of K-means clustering is NP-hard, Lloyd's algorithm provides an effective method of relaxation by iteratively updating the centroids of the clusters.
Let (x i ) 1≤i≤N be the datapoints. Denote the centroids of cluster k at iteration j by c kj . Lloyd's algorithm has two steps in each iteration: (1) classify each datapoint x i into the cluster whose centroid is closest to x i (i.e., argmin k �x i − c kj � ); (2) calculate the centroids for the next iteration based on the current clustering where C kj is set of the indices of all datapoints in cluster k at iteration j . Despite its efficiency for well-behaved datasets, Lloyd's algorithm may show a slow convergency and does not guarantee the optimality of the solution. The parameter K also greatly impacts the clustering results.
One of the difficulties of unsupervised learning is the interpretation of results. This is because the K-means algorithm itself does not label the clusters. For that reason, we proceeded with a hierarchical two-class clustering, and at each step we compared the result with potential features. Instead of directly predict the precise category a sample belongs to, a hierarchical two-class clustering makes several different binary classifications between two larger categories in sequence. For instance, the algorithm decides first whether the sample is a running sample or a walking sample. If it's a running sample, then it decides which direction the subject is running. This approach is more suitable for unsupervised learning, as it may be unknown to the algorithm how many precise categories there are. It should be mentioned that the accuracy resulting from the comparison is not an objective evaluation of the unsupervised learning method, but simply an indication of the type of features the algorithm prioritizes and finds. When the algorithm makes unsupervised binary classifications, human operators don't know which criteria these classifications are based on, so classifications made by the algorithm must be combined with the known characteristics of the data in order to understand whether a classification according to the type, or direction, or location of the motion. These comparisons and the associated accuracies are shown in [fig_ref] Table 4: Classification results using K-means clustering for various scenarios [/fig_ref]. [fig_ref] Table 4: Classification results using K-means clustering for various scenarios [/fig_ref] summarizes the results from implementing K-means clustering. The range of uncertainty was not investigated, since unsupervised machine learning finds relations inside the data itself and organizes them into several clusters. Therefore, they were not divided into training and validation groups. The top cluster depends on the subject and how the data are fed to the algorithm. As in direction recognition, if data for each subject are fed individually, the accuracy reaches 92.31%. If the data of all eight subjects are fed together, the accuracy drops as low as 53.85% for certain subjects. This can be explained by the fact that more subjects in the input data equal more secondary features in the dataset-a potential source of confusion for the algorithm. Precise and successful identification must result from sifting through a multitude of possibilities and determining a specific combination of features, such as the strength, duration, and location of steps, as well as the speed and rhythm of movement. However, unsupervised learning only extracts those features it deems most important. Without www.nature.com/scientificreports/ deterministic guidance, some covert-and possibly useful-characteristics may be lost. As a result, the accuracy of the algorithm is highly variable in both human identification and direction of movement. Unsupervised learning normally has poor performance than supervised learning, which is consistent with results summarized in [fig_ref] Table 4: Classification results using K-means clustering for various scenarios [/fig_ref]. However, in some applications, supervised learning is inaccessible due to the lack of labeled samples for training. Therefore, studies of unsupervised learning for human locomotion identification can be very meaningful as it could reveals level of performance of machine learning algorithms without much prior knowledge of dataset. These results are not simply replaceable by supervised learning studies.
# Conclusion
This paper describes a bottom-up approach to improve efficacy of machine learning algorithm. Through SNR improvement of distributed fiber sensors, high quality and high spatial resolution data can be harnessed to provide better training of machine learning algorithm to identify subtle difference existed in datasets. Artificial Rayleigh scattering features were induced to enhance weak intrinsic Rayleigh scattering in single-mode fibers by more than 35 dB through femtosecond laser direct writing schemes. This led to improved SNR and sensitivity for distributed fiber acoustic sensors, thus generating high-quality acoustic data carrying characteristics of human locomotion. Such large amount acoustic data was used to train various machine learning algorithms, bringing higher accuracy in identifying individuals and various events. Machine learning is a powerful tool in PR, proven to manifest covert features used in classification. It meets the requirements for both human identification and locomotion recognition, with an accuracy of over 76.25% when using supervised deep neural networks and 77.65% when using unsupervised machine learning algorithms. Despite these strides, the potency of artificial intelligence algorithms relies on both the quantity and quality of training data. This paper gives a detailed example of how high spatial resolution and bandwidth data from distributed fiber sensors can be very valuable and new stream of data sources for artificial intelligence.
[fig] Figure 1: Enhanced backscattering by embedding nano-reflectors into sensing fiber. (a) Roll-to-roll setup of ultrafast laser direct writing. (b) Six nano-reflectors with Rayleigh enhancement of over 35 dB along 21 m of fiber. (c) Schematic diagram of the homodyne φ-OTDR sensing system inscribed with nano-reflectors. (d) Temporal profiles of intrinsic and enhanced Rayleigh backscattering signal detected by three photodetectors (PD1-3). (SOA: semiconductor optical amplifier; EDFA: erbium-doped fiber amplifier; DAQ: data acquisition; FRMs: Faraday rotation mirrors). Scientific Reports | (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 [/fig]
[fig] Figure 2: The test site with marks for two tracks and distributed fiber sensors on the ground. The example shows a simultaneous event featuring one person running and another walking. [/fig]
[fig] Figure 3: (a) Raw acoustic signals in temporal and frequency domain; and (b) acoustic signals after data preprocessing. Acoustic signals harnessed from all five sensors are overlaid together. Temporal data shown in (a) and (b) are also visualized using intensity maps. Scientific Reports | (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 [/fig]
[fig] Figure 4: Plots of 1-s time-domain acoustic signals from five sensors are illustrated as overlaid curve plots and stacked intensity maps for visualization. The spectrum of first 256 frequency components after preprocessing are presented as well.Scientific Reports| (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 [/fig]
[fig] Figure 5: Architecture of the CNN. Scientific Reports | (2020) 10:21014 | https://doi.org/10.1038/s41598-020-77147-2 [/fig]
[table] Table 1: Description of dataset from events involving human movements. [/table]
[table] Table 3: Classification results using different machine learning algorithms or neural networks. [/table]
[table] Table 4: Classification results using K-means clustering for various scenarios. [/table]
|
Pleomorphic spermatic cord liposarcoma: A case report and review of management
a b s t r a c tLiposarcoma of the spermatic cord is very rare, representing about 7% of para testicular sarcomas. It is considered to be one of the highest malignancy grades.We present a case of a liposarcoma of the spermatic cord in a 45-year-old male complaining of a progressive painless swelling in the right inguinoscrotal region. Ultrasonography and computed tomography findings were compatible with liposarcoma of the spermatic cord. We performed a right radical orchiectomy with a wide resection of the mass. Histological examination confirmed the diagnosis and showed a pleomorphic subtype.The mainstay of management of spermatic cord liposarcoma is wide excision with radical orchiectomy. The most important factors for prognosis are the histologic subtype and surgical margin status. Adjuvant radiotherapy should be considered in cases at high risk for local recurrence. Long-term surveillance is mandatory.Liposarcoma of the spermatic cord is an uncommon para testicular tumor which should be part of the differential diagnosis of inguinoscrotal mass. A radical inguinal orchiectomy with wide resection of the soft tissue mass and the spermatic cord are the key to longest local and systemic disease-free survival.
# Introduction
Liposarcomas are the most common of the soft tissue sarcomas, accounting for approximately 20% of all mesenchymal malignancies. Liposarcoma of the spermatic cord (LSC) is very rare, representing about 7% of para testicular sarcomas. They are frequently misdiagnosed after surgery for inguinal hernia, inguinal lymphadenectomy or testicular malignancy. Until now, the published literature on LSC has been limited to case reports with limited clinical information. To our knowledge only about 350 cases have been described in the literature. Pleomorphic liposarcoma is a rare subtype of liposarcoma and accounts for less than 5% of all Liposarcomas; it is considered to be one of the highest malignancy grades with high invasion, metastasis and recurrence. We share our experience on the presentation of pleomorphic LSC in a 45 years-old male, and we review the management of this rare entity. This procedure was performed by a junior resident with 5 years of specialized training and it has been reported in line with the SCARE 2020 criteria.
## Presentation of case
A 45-year-old male with a medical history of right inguinal hernia surgery, presented to our department with a progressive painless swelling in the right inguinoscrotal region that was present for over 6 months, with rapid progression in last 2 months. There was no recent history of trauma, lower urinary tract symptoms, drug intake nor family history of cancer. On clinical examination, there was a firm and nodular scrotal mass. The right testis was not identified during palpation; the trans-illumination test was negative and there was no associated inguinal lymphadenopathy. Ultrasonography (US) revealed a heterogeneous hyperechogenic mass in the right hemi-scrotum with mixed echogenicity. Computed tomography (CT) scan showed a 14 × 8 cm mass lesion in the right hemi-scrotum which contained fat and soft tissue. The US and CT patterns were compatible with the fat-containing tumor, especially liposarcoma. Pre-operative imaging didn't reveal any lymph node or distant metastasis. Testicular tumor mark- ers, such as Alpha-fetoprotein, ß-human chorionic gonadotropin (ß-hCG), and lactate dehydrogenase (LDH) isoenzymes, were all normal.
Surgery was performed using a right inguinal approach. A yellowish soft tissue mass with abunch of grapesäppearance was closely attached to the spermatic cord; the testis was atrophic. Radical orchiectomy with wide resection of the mass was performed.
Gross examination showed a 14 × 12 × 8 cm wellcircumscribed, encapsulated and nodular mass (Figs. 2 and 3). Microscopic examination revealed multinuclear giant lipoblasts with hyperchromatic nuclei and multivacuolated cytoplasm consistent with a pleomorphic subtype of LSC. The testis was atrophic, and it showed no pathological changes. The resected margins were negative and free of neoplastic cells.
The patient had a good postoperative clinical course with no complications and was discharged on the third postoperative day. Adjuvant radiotherapy was proposed; however, the patient refused any adjuvant treatment. Follow up with CT scans after 6 months were recommended. There have been no signs of any local recurrence during 1-year follow-up period.
# Discussion
Liposarcomas of the spermatic cord are rare. They occur mostly in adults between 50 and 60 years. They generally remain asymptomatic for years. The duration of symptoms, reported in the literature, ranged from 1 week to 5 years.
The typical clinical characteristics of LSC are painless, nontender, nodular scrotal mass which progresses slowly and then increases rapidly in size. In our case, total duration of symptoms was 6 months with rapid progression in last 2 months. Pain and tenderness have been reported in 10-15% of cases. Rapid growth, large size, and symptomatic presentation are features suggestive of malignancy.
Sometimes the tumors are challenging to diagnose and are often mistaken as inguinal hernia, lipoma, hydrocele, epididymal cyst, or testicular tumors. Initially, inguinal and scrotal masses are evaluated by ultrasound in order to distinguish between intratesticular and extra testicular, solid and cystic masses. Once the lesion is defined as solid, it is considered malignant until proven otherwise. CT scan is more useful for the diagnosis of liposarcoma, distinguishing tissue characteristics, morphologic features and tumor location and it determines the extent of the mass into the neighboring tissue. However, there are no pathognomonic features for differentiating between benign and malignant masses. Magnetic resonance imaging (MRI) provides good information for the precise localization of the tumor and delineates its extension.
Tumor staging is based on histological examination and grading, and the presence of metastases. According to the World Health Organization's International Classification of Disease, there are seven recognized classifications of liposarcoma: well, differentiated, myxoid, round cell, pleomorphic, mixed, dedifferentiated, and not otherwise specified. Low-grade subtypes are histologically well-differentiated and have no metastatic potential, but lesions may recur locally. Local recurrence rate after conservative surgery alone reported in literature is up to 30%. High-grade subtypes (round-cell and pleomorphic) are rarer, but they are associated with a higher rate of recurrence and hematogenous metastasis to the lungs and bone. The majority of cases of LSC reported in the literature were well-differentiated. In the present case, pathological examination of the specimen revealed pleomorphic subtype.
Due to its rarity, there is no gold standard treatment for LSC, including pleomorphic subtype. Radical orchiectomy, with wide excision of the soft tissue mass and excision of all potentially contaminated tissues from as close to the deep ring as possible, is the cornerstone of treatment of this neoplasm. A positive surgical margin seems to be the main risk factor for early local recurrence and distant metastasis. Wide excision has demonstrated microscopic residual disease in 27% of apparently complete excision. For this reason, the majority of reports acknowledge the importance of complete resection with clear microscopic margins and re-resection with wide resection in cases with positive margins in order to achieve a negative margin. Routine lymph node dissection is currently not justified as the locoregional lymph nodes are rarely involved.
The use of adjuvant treatment (chemotherapy/radiotherapy) remains controversial, due to the paucity of data in the literature. The role of adjuvant radiotherapy is not clear. Two prospective randomized trials about the role of the addition of radiation to surgery in soft tissue sarcoma, showed that adjuvant radiotherapy can improve locoregional control and disease-free survival; however, this advantage did not translate into an overall survival benefit. In another study, it has been demonstrated that postoperative radiotherapy can significantly decreased the 10-year local recurrence rate in patients with high-grade lesions (p = 0.0028). Therefore, adjuvant radiotherapy is recommended for highgrade tumors, lymphatic invasion, inadequate margins, or relapses. Based on this limited literature, there is no definitive role for chemotherapy in the management of localized liposarcoma of the spermatic cord.
For patient follow-up, the consensus is to perform close and regular follow-up with imaging at 3, 6, 12 and 24 months. These cancers have a known risk of local recurrence, hence a long-term follow-up up to 10 years is mandatory.
# Conclusion
SCL is an uncommon paratesticular tumor which should be part of the differential diagnosis of inguino-scrotal mass. Up to now radical inguinal orchiectomy with wide resection of the soft tissue mass and the spermatic cord are the key to longest local and systemic disease-free survival. Lymph node dissection is not justified. The role of adjuvant radiotherapy remains controversial; however, it should be considered in cases at high risk for local recurrence, such as pleomorphic SCL. Chemotherapy has not proved its benefice until now.
## Declaration of competing interest
The authors report no declarations of interest.
# Funding
We have no sources of funding for our research.
# Ethical approval
This study is exempt from ethnical approval in our institution.
## Consent
Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.
# Author contribution
Ebey Babe Nejib: writing the initial draft. Sahbi Naouar: critical revision of the manuscript. Bilel Faidi: critical revision of the manuscript. Rayen Lahouar: writing of the final draft. Badreddine Ben Khalifa: case report design. Rafik El Kamel: final review.
## Registration of research studies
Not applicable.
## Guarantor
Professor El Kamel Rafik.
## Provenance and peer review
Not commissioned, externally peer-reviewed. |
Spatial arrangement of several flagellins within bacterial flagella improves motility in different environments
Bacterial flagella are helical proteinaceous fibers, composed of the protein flagellin, that confer motility to many bacterial species. The genomes of about half of all flagellated species include more than one flagellin gene, for reasons mostly unknown. Here we show that two flagellins (FlaA and FlaB) are spatially arranged in the polar flagellum of Shewanella putrefaciens, with FlaA being more abundant close to the motor and FlaB in the remainder of the flagellar filament. Observations of swimming trajectories and numerical simulations demonstrate that this segmentation improves motility in a range of environmental conditions, compared to mutants with single-flagellin filaments. In particular, it facilitates screw-like motility, which enhances cellular spreading through obstructed environments. Similar mechanisms may apply to other bacterial species and may explain the maintenance of multiple flagellins to form the flagellar filament.
A ctive motility enables bacteria to establish themselves in their favorable environmental niche. For propulsion many bacterial species employ flagella, long helical proteinaceous filaments extending from the cell surface into the surrounding environment. They are rotated at the filaments' base by membrane-embedded motors, which commonly allow rotational switches to enable navigation [bib_ref] The bacterial flagellar motor and its structural diversity, Minamino [/bib_ref]. Each flagellar filament is composed of thousands of copies of flagellin subunits 2,3 in a process that is tightly regulated at the expression level. Typically, expression of the flagellin genes requires alternative sigma factors, and production and export of the flagellins is initiated once assembly of the basal body and the hook joint structure is completed [bib_ref] Coordinating assembly of a bacterial macromolecular machine, Chevance [/bib_ref]. The flagellins are then transported through the filament and assembled at the tip of the growing flagellum [bib_ref] Structure analysis of the flagellar cap-filament complex by electron cryomicroscopy and single-particle..., Yonekura [/bib_ref] [bib_ref] Complete atomic model of the bacterial flagellar filament by electron cryomicroscopy, Yonekura [/bib_ref].
Many bacterial species, such as Escherichia coli, harbor a single flagellin gene so that the resulting flagellar filament consists of only one type of subunit. However, about 45% of the flagellated bacterial species possess more than a single flagellin gene [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] , with numbers ranging up to fifteen copies as reported for Magnetococcus sp. MC-1 [bib_ref] KEGG for linking genomes to life and the environment, Kanehisa [/bib_ref]. Salmonella Typhimurium with two distinct flagellins undergoes phase variation and only utilizes one of the two flagellins at a time to build the filament [bib_ref] Phase variation in Salmonella, Lederberg [/bib_ref]. In contrast, most other bacterial species with multiple flagellins, such as Bacillus thuringiensis, Bdellovibrio bacteriovorus, Campylobacter jejuni, Caulobacter crescentus, Helicobacter pylori, Sinorhizobium meliloti, and Vibrio spp., have been shown to assemble the flagellar filament from all or at least most flagellins encoded in their genome [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] [bib_ref] Identification of two expressed flagellin genes in the insect pathogen Bacillus thuringiensis..., Lovgren [/bib_ref] [bib_ref] Characterizing the flagellar filament and the role of motility in bacterial prey-penetration..., Lambert [/bib_ref] [bib_ref] Structural and functional analysis of two Campylobacter jejuni flagellin genes, Nuijten [/bib_ref] [bib_ref] The organization of the Caulobacter crescentus flagellar filament, Driks [/bib_ref] [bib_ref] A family of six flagellin genes contributes to the Caulobacter crescentus flagellar..., Ely [/bib_ref] [bib_ref] Identification, characterization, and spatial localization of two flagellin species in Helicobacter pylori..., Kostrzynska [/bib_ref] [bib_ref] Mutational analysis of the Rhizobium lupini H13-3 and Sinorhizobium meliloti flagellin genes:..., Scharf [/bib_ref] [bib_ref] Polar flagellar motility of the Vibrionaceae. Microbiol, Mccarter [/bib_ref] [bib_ref] Vibrio fischeri flagellin A is essential for normal motility and for symbiotic..., Millikan [/bib_ref] [bib_ref] Contribution of six flagellin genes to the flagellum biogenesis of Vibrio vulnificus..., Kim [/bib_ref]. For B. bacteriovorus, C. crescentus, and H. pylori, it was demonstrated that the different flagellins exhibit a spatial arrangement within the filament [bib_ref] Characterizing the flagellar filament and the role of motility in bacterial prey-penetration..., Lambert [/bib_ref] [bib_ref] The organization of the Caulobacter crescentus flagellar filament, Driks [/bib_ref] [bib_ref] Identification, characterization, and spatial localization of two flagellin species in Helicobacter pylori..., Kostrzynska [/bib_ref].
The role of different flagellins within the flagellar filament is still mostly unclear. While in some systems one flagellin exhibits the major structural or functional role, a large degree of functional redundancy occurs in other multi-flagellin systems. However, depending on the species, loss of certain flagellins in such systems may result in changes in the filament's morphology and function, e.g., with respect to swimming speed, adhesion, or cytotoxicity [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] [bib_ref] Characterizing the flagellar filament and the role of motility in bacterial prey-penetration..., Lambert [/bib_ref] [bib_ref] Contribution of six flagellin genes to the flagellum biogenesis of Vibrio vulnificus..., Kim [/bib_ref] [bib_ref] Flagellar phase variation of Salmonella enterica serovar Typhimurium contributes to virulence in..., Ikeda [/bib_ref]. It has thus been suggested that the composition of the filament may be influenced by the environmental conditions [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] [bib_ref] Polar flagellar motility of the Vibrionaceae. Microbiol, Mccarter [/bib_ref].
Species of the genus Shewanella possess a single polar flagellum in the form of a left-handed helix, which pushes the cell during counterclockwise (CCW) rotation and pulls the cell upon switching to clockwise (CW) rotation. With few exceptions, the gene cluster encoding the components of the Shewanella polar flagellar system harbors two distinct flagellins [bib_ref] Specificity of motor components in the dual flagellar system of Shewanella putrefaciens..., Bubendorfer [/bib_ref] [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] [bib_ref] Genetic and molecular characterization of flagellar assembly in Shewanella oneidensis, Wu [/bib_ref]. In this study, we explored the role of the two flagellins, FlaA and FlaB, of the polar flagellar system in S. putrefaciens CN-32. We determined how FlaA and FlaB production may result in spatial arrangement within the flagellar filament, which enabled us to create mutants with defined filament compositions. Subsequent microscopy and three-dimensional tracking of swimming trajectories demonstrated that the composition of the filament significantly affects the flagellar morphology and the motion during free-swimming and spreading through soft agar. In agreement with corresponding simulations on flagellar stability under rotation, the results strongly implied that the basal flagellar part formed by the flagellin FlaA benefits swimming under a range of different conditions. With the switch to FlaB in the remainder of the flagellin, the cells maintain the ability to wrap the filament around the cell body and execute a screw-like movement [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] , which is highly beneficial for spreading through complex environments.
# Results
Characterization of S. putrefaciens FlaA and FlaB. S. putrefaciens possesses a monopolar and a lateral flagellar system. We previously showed that the polar system predominantly mediates propulsion, chemotactic behavior and backward screwing motion [bib_ref] Specificity of motor components in the dual flagellar system of Shewanella putrefaciens..., Bubendorfer [/bib_ref] [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] [bib_ref] Secondary bacterial flagellar system improves bacterial spreading by increasing the directional persistence..., Bubendorfer [/bib_ref]. In the following, only this polar filament will be addressed, and all experiments described here were conducted using a strain unable to produce the lateral flagellar filaments due to the lack of the corresponding flagellin-encoding genes (ΔflaA 2 B 2 ). For simplification, this strain will henceforth be referred to as wild type (FlaAB).
The two genes encoding the polar flagellins, flaA (Sputcn32_2586) and flaB (Sputcn32_2585), are arranged next to each other [bib_ref] Specificity of motor components in the dual flagellar system of Shewanella putrefaciens..., Bubendorfer [/bib_ref] and encode proteins of 273 (FlaA) and 271 (FlaB) amino acids with estimated molecular masses of 28.6 (FlaA) and kDa. They exhibit a high degree of conservation (86% identity; 91% similarity); most differences occur within the flagellin variable region which presumably faces the environment (Supplementary [bib_ref] Structure of the bacterial flagellar protofilament and implications for a switch for..., Samatey [/bib_ref] [bib_ref] Complete atomic model of the bacterial flagellar filament by electron cryomicroscopy, Yonekura [/bib_ref].
We previously showed that both FlaA and FlaB are produced at estimated molecular masses of~37 and~35 kDa, respectively. This is higher than the masses deduced from the amino acid sequence (~28.5 kDa) [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] , indicating modification of both flagellins, e.g., by glycosylation. To determine if the flagellins of S. putrefaciens are glycosylated, we introduced in-frame deletions into the genes encoding PseG (Sputcn32_2626) and Maf-1 (Sputcn32_2630), as the deletion of the orthologs in S. oneidensis results in completely non-glycosylated flagellins [bib_ref] Posttranslational modification of flagellin FlaB in Shewanella oneidensis, Sun [/bib_ref] [bib_ref] Analyzing the modification of the Shewanella oneidensis MR-1 flagellar filament, Bubendorfer [/bib_ref]. Immunoblotting analysis of PAGE-separated crude extracts (Supplementary [fig_ref] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB [/fig_ref] revealed that in both ΔpseG and Δmaf-1 mutants the flagellins occur as a single band at a position according to their estimated molecular mass. Thus, within the cell population, both FlaA and FlaB are produced and both are likely modified by glycosylation. The difference in the molecular mass, presumably caused by a more extensive modification of FlaA, enabled us to determine the presence of each flagellin by immunoblotting, which helped to unravel the regulation of flagellin production later in this study.
Spatial arrangement of FlaA and FlaB in the filament. To explore the distribution of FlaA and FlaB in filament assembly at the single cell level, we performed fluorescence microscopy on cells in which FlaA, FlaB, or both, can be selectively labeled by coupling with a maleimide-ligated fluorophore (FlaA-Cys and FlaB-Cys). The corresponding serine/threonine to cysteine substitutions in the flagellin proteins do not affect swimming motility or flagellin decoration by glycosylation [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] and ref. [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. We found that FlaA and FlaB exhibit a spatial arrangement within the flagellar filament , e; [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref]. FlaA almost exclusively forms the base of the filament close to the cell body, which accounts for about 8-28% (17% average) of the fully assembled flagellum. The residual filament is predominantly composed of FlaB with some interspersed FlaA. Mutants completely lacking FlaB (ΔflaB, FlaA stub) produced very short filaments, which barely supported swimming (Supplementary , e, i, m) while ΔflaA mutants (FlaB-only) formed flagellar filaments that were not noticeably different from the wild-type filament in morphology or length f; Supplementary . This spatial flagellin arrangement strongly suggested sequential production and/or export of FlaA and FlaB.
Previous studies on the closely related species S. oneidensis indicated that flaA and flaB are not organized in an operon with a single promoter but are regulated individually [bib_ref] Genetic and molecular characterization of flagellar assembly in Shewanella oneidensis, Wu [/bib_ref] [bib_ref] Effects of FlrBC on flagellar biosynthesis of Shewanella oneidensis, Shi [/bib_ref]. Accordingly, q-RT-PCR revealed that in S. putrefaciens the flaB transcript is about twice as abundant as that of flaA , also indicating individual regulation. We therefore monitored the production of FlaA and FlaB in mutants deleted in genes rpoN, flrA, and fliA, encoding main regulators of flagellation. Generally, in the group of polarly flagellated gammaproteobacteria, RpoN (σ 54 ) acts as a major general regulator of flagellation, FlrA is required for production of numerous structural flagellar components, and the sigma factor FliA (σ 28 ) is required for the production of late flagellar components such as flagellin. FliA is released from its inactivating anti-sigma factor FlgM upon completion of the basal body and the hook to avoid premature production of large amounts of flagellin 4 . Western immunoblotting revealed that, in S. putrefaciens, production of both FlaA and FlaB strictly requires RpoN and FlrA. In contrast, in a ΔfliA mutant FlaA as well as the flagellin chaperone FliS, which is required for export, were produced in normal amounts while FlaB was absent [fig_ref] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB [/fig_ref]. Hence, production of FlaA can already start before the hook-basal body complex is finished, but FlaB production requires hook completion and the subsequent export of the anti-sigma factor FlgM. This regulatory arrangement indicated that a pool of FlaA is already present and ready to be exported and assembled into the filament as soon as Overview of the different types of polar flagellar filaments constructed in this study. In all strains the genes for the lateral flagellins were deleted. a-d Genetic organization of the polar flagellins and genetic modifications to obtain different filament types. In the wild-type context flaA lies upstream of flaB and transcription is controlled by a FliA-independent promoter (cp. [fig_ref] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB [/fig_ref]. In contrast, transcription of flaB is controlled by a stronger, FliA-dependent promoter and mediates production of the major part of the flagellar filament. Expression of both genes relies on RpoN and FlrA. A functional filament is only produced when at least one of the flagellar genes is expressed from the flaB promoter. If no flagellin is expressed from the flaB promoter only short filament stubs are formed (see . Gene deletions are marked with a cross, swapping of the flagellin gene sequences is marked with an asterisk. PflaA/B = flaA/B promoter. e-h Micrographs of cells with fluorescently labeled flagellar filaments displaying the outcome of the genetic editing of the flagellin genes. Panel e also shows the spatial distribution of the flagellins FlaA and FlaB in the wild-type filament (see also [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref]. The FlaA portion is marked with an arrow. The increased fluorescence of FlaA compared to that of FlaB is likely due to a better accessibility of the cysteine residue during the fluorescent labeling process. Scale bars represent 2 µm. i-l Radial expansion of cells with different types of flagellar filaments in 0.25% soft agar. Wild-type and FlaB-only cells spread well in this structured environment while the spreading ability of FlaA-only and FlaBA cells is strongly impaired. The numbers in the center of the colonies indicate the relative spreading compared to the wild type in percent (% ± s.d.) of three individual experiments NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-018-07802-w ARTICLE the hook is completed, while FliA-dependent production of FlaB is initiated [fig_ref] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB [/fig_ref]. Accordingly, a strain with only flaB under control of the flaA promoter produced very short FlaB filaments , f, m). In contrast, placing only flaA under control of the flaB promoter resulted in flagellar filaments exclusively consisting of FlaA , which were of almost normal length (~6.5 µm) and also exhibited a left-handed helix. However, pitch, diameter and therefore number of turns of FlaA-only helical filaments differed significantly from FlaB-only or wild-type filaments . Finally, when placing flaB under the control of the flaA promoter and vice versa, a flagellar filament with FlaB forming a short basal segment was formed, while the residual filament consisted of FlaA Supplementary [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref]. Accordingly, the overall filament of these FlaBA mutants had a highly similar geometry as the FlaA-only filament .
Based on these results we propose that the sequential production pattern depends on two individual promoters driving the expression of flaA and flaB, respectively. The length of the proximal filament segments formed under control of the weaker FliA-independent flaA promoter showed a wide range, indicating flagellin production from this promoter varies substantially at the single cell level. Accordingly, overproduction of FlaA from a plasmid resulted in aberrantly long flagellar filaments , indicating that the amount of available flagellin monomers is an important length determining factor. However, the data did not explain why flagellin production Filament composition affects flagella-mediated swimming. The different morphologies of the FlaA-and FlaB-only filaments suggested that the filament composition affects the swimming behavior of the cells [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] [bib_ref] Effect of viscosity on bacterial motility, Schneider [/bib_ref]. We therefore compared the spreading ability of the wild-type cells (FlaAB), FlaA-only (with flaA being expressed from the flaB promoter, resulting in filaments of normal length), FlaB-only and FlaBA mutants through soft agar.
Here, the cells have to navigate through an intricate network of polysaccharide strands . Under these conditions, FlaB-only mutants spread as well as wild-type cells, while cells driven by FlaA-only or FlaBA filaments exhibited a pronounced decrease in spreading. As the inability of these filaments to promote normal spreading in soft agar could be due to general differences in free swimming, we compared the free-swimming patterns of wild-type (FlaAB) cells to those driven by FlaAonly, FlaB-only and FlaBA filaments by three-dimensional holographic tracking [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] ; example tracks given in panels i and j). This was carried out at normal and, to simulate flagellar performance at high load, at elevated viscosity (10% wt · vol −1 Ficoll 400). From the trajectories obtained, three major swimming parameters were deduced: run duration, swimming velocity, and turning angle. The run duration represents the time interval of swimming between directional changes. A shift toward shorter run durations may negatively affect effective spreading in soft agar while very long runs may hamper navigation. We found that under conditions of normal viscosity the distribution of run durations was very similar for all four strains with a descending slope . P(t) indicates the probability of observing a run with a duration in the range t to t+dt. c, d Velocity distribution for the different filament types in regular medium and medium with increased viscosity, respectively. P(v) indicates the probability of observing a cell swimming with a speed in the range v to v+dv. e, f Turning angle histograms for the different filament types in regular medium and medium with increased viscosity, respectively. Low and high turning angles correspond to weak deviations from straight swimming and nearreversals, respectively. P(θ) indicates the probability that a particular re-orientation event results in a change of direction in the range θ to θ+dθ. from short runs (1 s or less) toward longer runs up to about 20 s [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref]. Under conditions of elevated viscosity [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] , the distributions of run durations remained similar except for the FlaBA mutant that exhibited a pronounced shift towards shorter runs, indicating a lower spreading ability of this strain. The second major parameter, the swimming velocity, directly affects the speed of spreading. At normal viscosity, all four strains displayed a similar top speed and velocity distributions with two major subpopulations, a slow and a fast one [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref]. The slow population, which did not include completely non-motile cells, was comparable in size for the wild type and the FlaA-only and FlaB-only strains, but much broader for the FlaBA mutant. The mean speed increased from 5 µm s −1 for the first three strains to 15 µm s −1 for the FlaBA mutant. The velocity of the fast-moving subpopulation was different for all four strains: The wild-type cells had the highest speed (~63 µm s −1 ), followed by FlaA-only cells (~57 µm s −1 ), FlaBA cells (~53 µm s −1 ) and FlaB-only cells (~49 µm s −1 ). At elevated viscosity [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] , the velocity distribution with a slow and a fast swimming major population persisted for all four strains, albeit at a lower overall speed. Under these conditions, FlaA-only cells were the fastest (~38 µm s −1 ), confirming earlier reports stating that a smaller helix diameter of a bacterial flagellum promotes swimming at elevated viscosity [bib_ref] Effect of viscosity on bacterial motility, Schneider [/bib_ref]. The FlaA-only mutant was followed by wild-type and FlaB-only cells (~28-30 µm s −1 ) and FlaBA (~22 µm s −1 ). Notably, the size of the slow subpopulation of wild-type and FlaB-only cells increased compared to normal viscosity, but not the one of the other two strains.
The third major parameter, the turning angle, indicates how efficiently the cells can switch direction. Very high angles between 150°and 180°would indicate forward-backward (run-reverse) swimming without efficient directional changes and thus a decrease in spreading efficiency. All four strains showed a wide variation of turning angles from 0°to 180°. Under conditions of normal viscosity [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] , the turning-angle profiles of wild-type, FlaA-only and FlaB-only cells were rather similar. However, at elevated viscosity [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] , the wild-type cells maintained the wide distribution of turning angles, while for FlaA-only and FlaB-only mutants a large subpopulation exhibited directional changes at very high angles in a run-reverse mode, which would decrease efficient spreading. The FlaBA-mutant cells showed an aberrant behavior under both conditions, as a major subpopulation of the cells displayed rather low turning angles (0-60°), resulting in a higher directional persistence.
The trajectory data on free-swimming cells demonstrated that the composition of the flagellar filament significantly affects the free-swimming behavior of the cells. However, the free-swimming capability of the different strains did not correlate well to the spreading capability in soft agar. This was particularly eminent in the case of cells driven by FlaA-only filaments, which did not spread well in soft agar but outperformed FlaB-only mutants, the best spreaders in soft agar, in almost every aspect of free swimming. Discrepancies with the agar-based experiments are to be expected, as the tracking experiments measure the individual characteristics of cells in the absence of external perturbations, such as mechanical interactions with their environment. In contrast, the agar assays capture a holistic picture of motility, integrating speed, reversal rate, response to mechanical environment, etc. across a whole population [bib_ref] Diffusive transport without detailed balance in motile bacteria: does microbiology need statistical..., Cates [/bib_ref] , which cannot be adequately simulated by simply increasing the viscosity as in our free-swimming assays. We concluded that the poor spreading capability of the FlaA-only and FlaBA mutants in structured environments is not due to obvious deficits in free-swimming capabilities. The source of the differences must therefore be connected with other properties of the flagellum.
The FlaA segment stabilizes the flagellar filament. The analysis of swimming speed at high viscosity showed a pronounced increase in size of the slow subpopulation of wild-type and FlaBonly cells, but not of FlaBA and FlaA-only strains [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref].
Corresponding single-cell trajectories indicated that a drop in speed frequently followed a directional switch after which the cells continued at lower speed, but accelerated again after another switch of direction [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref]. We previously showed that under conditions of increased viscosity and while swimming backward with the filament pulling the cell, S. putrefaciens can exhibit a filament instability, which results in wrapping of the flagellar filament around the cell and a slow movement in a screw-like fashion [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. Therefore, we anticipated that the drop in swimming speed of wild-type and FlaB-only cells was due to screw formation. An analysis of the four strains' flagellar behavior at normal and high viscosity [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref] confirmed that at normal viscosity only few cells of the wild-type population (<5%) exhibited screw formation during backward swimming [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. Notably, under these conditions, already about half of the cells of FlaB-only mutants displayed screw formation, while this flagellar behavior was completely absent for FlaA-only or FlaBA cells. The screw-like swimming phenotype became significantly more pronounced at high viscosity, where about half of all backward-swimming wildtype and more than 85% of FlaB-only cells displayed the screwing behavior. Notably, also at increased viscosity, not a single FlaAonly or FlaBA mutant cell exhibited screwing motility, strongly indicating that a filament mainly or exclusively formed by FlaA is more stable when rotating CW and pulling the cells. This finding strongly suggested that the filaments' base formed by FlaA stabilizes the flagellum to prevent premature screw formation during free swimming. To further support this hypothesis, we introduced a second synthetic flaA gene into the chromosome directly downstream of the native flaA (FlaAAB; , c, e). As anticipated, the higher level of FlaA production increased the length of the basal FlaA segment on average from 17% to~24% of the overall filament. In this strain, the number of cells forming the screw at high viscosity dropped from about 50% to about 25%, and at normal viscosity almost no screws (2%) were observed . Similarly, we constructed a FlaBBA mutant strain to probe if a longer FlaB segment at the filament's base affects the stability of the flagellum (Supplementary , d, f). The length of the FlaB basal section increased from~6.8% to~8.1%, but we did not observe any screw formation. Based on these findings, we conclude that the FlaA segment stabilizes the flagellar filament against screw formation to an extent which depends on the length of this segment.
The findings of the screw formation analysis correlate well with the observed differences in free-swimming speed distributions and spreading through obstructed environments such as soft agar. Wild-type and FlaB-only filaments readily form flagellar screws under conditions of high viscosity, which increases the population of slow free-swimming cells (cp. [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] as free swimming in screwing mode does not allow effective propulsion [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. Accordingly, this increase in the slow population of free-swimming cells is absent in FlaBA and FlaA-only mutants, which lack the ability of screw formation. However, the latter two strains are hampered in spreading through structured and obstructed environments, the soft agar plates , strongly indicating that screw formation is significantly contributing to movement when the cells can make contact with a solid environment. In contrast, the differences in efficient turning angles observed under freeswimming conditions at high viscosity cannot be attributed to screw formation: compared to the wild-type FlaB-only cells, that are prone to screw formation, as well as FlaA-only cells, which are incapable of screw formation, tend to turn at high angles in less efficient forward-backward movement. Notably, in contrast to the mutants, the wild-type cells driven by the native FlaAB filament always performed very well under all conditions tested.
Numerical simulations on behavior of the flagellar filament. The mechanical properties of filaments and their effects on the dynamics can also be studied in numerical simulations. In an established model for filaments, the flagellum is replaced by a string of beads that are connected to each other by elastic bonds with preferred bonding angles, so that in equilibrium a helix with prescribed radius and pitch forms. The interaction with the surrounding fluid is given by resistive force theory (see ref. [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. Such a model reproduced the screw formation of the helix in dependence on the motor torque and thus the force acting on the flagellar helix. Here, we extend our previous model to the case of two flagellins by adjusting the parameters locally to those of the corresponding flagellins, while keeping all other parameters fixed. Specifically, FlaA forms narrow elongated helices (radius: 0.175 µm; pitch 1.18 µm) whereas FlaB forms wider and shorter helices (radius: 0.315 µm; pitch 1.91 µm; Supplementary .
In the simulations we can switch between the two flagellins at continuous positions along the flagellum. This allows us to study the properties of the flagellum as a function of the partitioning into FlaA and FlaB and of the motor torque. As the screw forms after about 60 ms 23 , the simulation was conducted for this period of time. For the wild-type flagellum composed of FlaA close to the cell and FlaB further away, we find that it is stable to screw formation for motor torques of less than about 5pNμm. For larger torque, screws form, but the required torque varies with FlaA fraction: It reaches a local maximum in stability for a fraction of about 24%, where a torque of 8.5 pNμm is needed to form the screw [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref]. This simulated region of enhanced stability corresponds well to our experimental findings that an increase to an average of 24% proximal FlaA, as observed in the FlaAAB mutant, increases the filament's stability towards screw formation compared to that of the wild type . The model shows a second region with increased stability for FlaA fractions above about 40%, but that could not be reliably probed with our mutants. Turning the order of FlaA and FlaB around, no screws are formed until the FlaB part exceeds 40% , and then the required torque falls with increasing fraction. This is in line with our experiments in which FlaB forms up to 10-12% of the flagellar filament's base and no screw formation occurs. Thus, the simulation not only reproduces our experimental findings but also suggests that introducing a proximal FlaA filament segment with a smaller diameter and pitch is already sufficient to result in the observed stabilization of the flagellar filament against screw formation, indicating the geometry of the helix as a key factor in the determination of the mechanical properties.
The simulations also allow to extract the propulsive force which the flagellum exerts on the cells. Up to the formation of the screw, the propulsive force F is proportional to the motor torque M, i.e., F = βeff · M. A larger value of the slope βeff indicates a more efficient transfer of torque into driving force. The data in [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref] show the linear relation between F and M for the cases of a pure FlaB flagellum and admixtures of 20 and 80% FlaA. The inset shows that the slope βeff increases strongly with the FlaA content until about 20% and varies little thereafter. The data indicate a variation of the onset of screw formation depending on the FlaA fraction and show that in a range between 16 and 20% the cells achieve efficient forward propulsion and maintain the ability to form screws at high torque. This is well within the range of proximal segments we determined for wild-type cells (~8% tõ 28%; [fig_ref] Figure 4: Experimentally observed and simulated screw formation [/fig_ref]. With respect to this pronounced variability in length of the proximal FlaA segment within the flagellar filament, the experimental data in concert with the simulations strongly suggest that S. putrefaciens forms a highly heterogenous cell population with respect to propulsion and the ability of screw formation. This may not be optimal at the single cell level but distributes the trade-off between torque generation and propulsive efficiency among the population, which allows efficient spreading of at least a fraction of the population under a wide range of conditions.
# Discussion
About half of all flagellated bacteria possess more than a single flagellin-encoding gene, and it was demonstrated for several species that their flagellar filament is, in fact, assembled from more than one building block. However, it is still unclear why so many bacteria have maintained this arrangement. In this study, we showed that the S. putrefaciens flagellum is assembled from two different flagellins, FlaA and FlaB, in a spatially organized fashion and explored how this affects flagellar geometry and function. We provided evidence that the particular spatial assembly by two different flagellins benefits cellular motility when considering a wide range of environmental conditions, free swimming under normal and elevated viscosity and spreading through structured environments. We show that a proximal FlaA segment stabilizes the flagellar filaments to provide a compromise between propulsion and the ability to wrap the flagellum around the cell, which benefits motility in structured environments. In S. putrefaciens, the length of this proximal stabilizing segment shows a high variability, leading to a heterogenous population well equipped for different environmental conditions. Taken together, our results provide quantitative support for the hypothesis that structural aspects of filament assembly drive the maintenance of multiple flagellin-encoding genes [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref].
How do the bacteria form a spatially organized filament? Our study strongly suggests that the arrangement is achieved by sequential production and export of different amounts of FlaA and FlaB, based on individually regulated promoters rather than differences in assembly efficiency as previously proposed for S. oneidensis 31 . A very similar spatial pattern of FlaA/FlaB occurs in the flagellum of the latter species , strongly indicating a similar timing of flagellin production and export, which therefore appears to be conserved among Shewanella sp [bib_ref] Genetic and molecular characterization of flagellar assembly in Shewanella oneidensis, Wu [/bib_ref] [bib_ref] Effects of FlrBC on flagellar biosynthesis of Shewanella oneidensis, Shi [/bib_ref] [bib_ref] Two residues predominantly dictate functional difference in motility between Shewanella oneidensis flagellins..., Sun [/bib_ref]. A similar sequential flagellin production may underlie the spatial flagellin distributions in other flagellar filaments composed of two flagellins, e.g. in Campylobacter coli or Helicobacter pylori, while the more complex flagellar filaments of Caulobacter crescentus or Bdellovibrio bacteriovorus, both produced from six individual flagellin subunits, likely require a more complex regulation [bib_ref] Flagellin redundancy in Caulobacter crescentus and its implications for flagellar filament assembly, Faulds-Pain [/bib_ref] [bib_ref] Characterizing the flagellar filament and the role of motility in bacterial prey-penetration..., Lambert [/bib_ref] [bib_ref] The organization of the Caulobacter crescentus flagellar filament, Driks [/bib_ref] [bib_ref] Identification, characterization, and spatial localization of two flagellin species in Helicobacter pylori..., Kostrzynska [/bib_ref] [bib_ref] Role of two flagellin genes in Campylobacter motility, Guerry [/bib_ref]. However, several important aspects of regulation remain to be elucidated in S. putrefaciens: Flagellin expression and production from the flaA promoter is not constitutive, and flagellar filaments produced from this promoter never reach normal length . Furthermore, the length of the proximal flagellar segment also depends on whether flaA or flaB is expressed from the flaA promoter . We therefore hypothesize that, in the wild type, at the onset of FlaB production flaA expression is restricted or even shut down by a yet unknown mechanism at the transcriptional and posttranscriptional level. Flagellin expression from differentially controlled promoters suggest that Shewanella sp. are capable of regulating the composition and properties of the flagellar filament by tuning the amount of incorporated FlaA in response to yet unknown environmental signals. In S. oneidensis, production of FlaA has been shown to be under control of the flagellar regulators FlrB and FlrC 28 , which may therefore be involved in this regulation. It is also yet unclear which differences between the flagellins determine the different filament morphologies, as FlaA and FlaB are highly conserved. For S. oneidensis it has been proposed that the functional difference between FlaA and FlaB is due to four amino-acid residues within the flagellins 31 . Three of these residues are conserved in S. putrefaciens FlaA and FlaB, but did not affect swimming upon substitution . Thus, other residues or factors, such as the differences in flagellin glycosylation, account for the difference in filament morphology and function.
Our study shows that an assembly of the flagellar filament by one or two different flagellins affects its mechanistic properties, which has significant consequences for several aspects of flagellar function in different environments. Directional changes of many monopolarly flagellated bacteria are mediated by combined bending of flagellar filament and hook, the structure which joins filament and motor, during the typical run-reverseflick movement [bib_ref] Secondary bacterial flagellar system improves bacterial spreading by increasing the directional persistence..., Bubendorfer [/bib_ref] [bib_ref] Bacterial flagellum as a propeller and as a rudder for efficient chemotaxis, Xie [/bib_ref] [bib_ref] Bacteria can exploit a flagellar buckling instability to change direction, Son [/bib_ref] [bib_ref] Dynamic instability in the hook-flagellum system that triggers bacterial flicks, Jabbarzadeh [/bib_ref]. Thus, the proximal FlaA segment in S. putrefaciens likely has a positive effect on robust directional changes at high and fast swimming at normal viscosity. In contrast, a proximal FlaB segment as in the FlaBA mutant filament rather suppresses directional switches at high angles. The trajectories of this mutant with an enrichment in small turning angles also point towards rather short motor breaks or very short intervals of rotational switching. This may occur more frequently at elevated viscosity, as indicated by the shorter run periods under these conditions, which may be explained by differences in load associated with the composition of the flagellar filament [bib_ref] Switching of the bacterial flagellar motor near zero load, Yuan [/bib_ref] [bib_ref] Coupling between switching regulation and torque generation in bacterial flagellar motor, Bai [/bib_ref]. We have previously demonstrated that S. putrefaciens may wrap its polar flagellar filament around the cell body in a spiral-like fashion upon CW rotation, and this process starts with an instability of the proximal filament segment at high load [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. Our experimental and simulation data strongly suggest that the spatial localization of FlaA at the filament's base stabilizes the flagellum, which may be already conferred by the smaller diameter and pitch of the flagellar helix in this region. Spatial distributions with one flagellin primarily localized at the base of the flagellar filament were reported previously for C. crescentus [bib_ref] The organization of the Caulobacter crescentus flagellar filament, Driks [/bib_ref] and H. pylori [bib_ref] Identification, characterization, and spatial localization of two flagellin species in Helicobacter pylori..., Kostrzynska [/bib_ref] and suggest that these segments fulfill a similar role in optimizing flagella-mediated swimming in these species.
Beyond monopolarly flagellated S. putrefaciens, the screwing mode of movement with the flagellar filament(s) wrapped around the cell body was shown for the lophotrichously flagellated species Burkholderia sp. RPE64, Aliivibrio fischerii, and Pseudomonas putida [bib_ref] A polar bundle of flagella can drive bacterial swimming by pushing, pulling,..., Hintsche [/bib_ref] and bipolarly flagellated Helicobacter suis [bib_ref] Bipolar lophotrichous Helicobacter suis combine extended and wrapped flagella bundles to exhibit..., Constantino [/bib_ref]. In all cases, full flagellar screw formation significantly slowed down the speed of free-swimming cells. More efficient propulsion by screwing motility was observed when the cells had surface contact [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. In many of their natural environments flagellated bacteria have to move through structured environments, such as sediments, mucus layers or biofilm matrices, where the bacteria constantly are at risk of getting stuck in passages too narrow for cells to pass. In these environments, polarly flagellated bacteria may employ the flagella-mediated screwing motility to escape from traps or to enable the passage through mucus-filled ducts of higher organisms [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] , however, clear evidence for such a role was missing so far. In this study, we provided a mutant (FlaA-only), which is completely incapable of screwing motility. This strain retains vigorous free swimming, but spreading through soft agar is drastically diminished, demonstrating that this mode of motility very likely gives a significant advantage for moving through structured environments. We expect screwing motility to be similarly important for numerous other polarly flagellated bacterial species in structured environments.
# Methods
Bacterial strains, growth conditions, and media. Shewanella and Escherichia coli strains that were used in this study are listed in . Shewanella strains were grown in LB medium at room temperature or 30°C, E. coli in LB medium at 37°C. Selective media were supplemented with 50 mg ml −1 kanamycin and/or 10% (wt vol −1 ) sucrose when appropriate. For the 2,6-diaminoheptanedioic acid (DAP)-auxotroph E. coli WM 3064 media were supplemented with DAP at a final concentration of 300 µM. To solidify media, LB agar was prepared using 1.5% (wt vol −1 ) agar for regular plates and 0.15-0.25% (wt vol −1 ) select agar (Invitrogen) for soft-agar plates.
Strains and vector constructions. Gene knock-out and substitution strains of Shewanella were constructed by sequential double homologous recombination. Briefly, in-frame deletions were generated by combining approximately 500-bp fragments of the up-and downstream regions of the designated gene. Only a few codons (typically around six) were left in the genome to prevent disruption of regulatory elements for other genes. In-frame insertions were constructed and re-integrated into those deletion strains in basically the same fashion. Substitution of single or multiple nucleotides was done by inserting a modified fragment into the corresponding deletion strain. Regular genetic manipulations of S. putrefaciens CN-32 and S. oneidensis MR-1 were introduced into the chromosome by double homologous recombination via the suicide plasmid pNPTS-R6K 41 . Plasmids were constructed using standard Gibson assembly protocols [bib_ref] Enzymatic assembly of DNA molecules up to several hundred kilobases, Gibson [/bib_ref] and introduced into Shewanella cells by conjugative mating with E. coli WM3064 as donor. Plasmids and corresponding oligonucleotides are listed in , respectively.
The S. putrefaciens strains featuring two flagellin genes under the control of the flaA promoter were constructed as described above using synthetic flagellin genes (Genescript) that were codon optimized for E. coli K-12 to prevent gene loss by homologous recombination with the native flagellin gene (sequences see Supplementary . The synthetic gene was introduced downstream of the native gene with a copy of the native Shine-Dalgarno sequence of flaA and a spacer of 12 random base pairs. The FlaA-overproduction plasmid was constructed using standard Gibson assembly protocols [bib_ref] Enzymatic assembly of DNA molecules up to several hundred kilobases, Gibson [/bib_ref] and introduced into Shewanella cells by conjugative mating with E. coli WM3064 as donor. Both flaA and fliS were introduced into pBTOK downstream of the anhydrotetracycline-inducible promoter each with their own AGGAGG Shine-Dalgarno sequence and with a spacer of 12 random base pairs. Gene expression was induced after 1 h of growth in LB with 10 ng ml −1 anhydrotetracycline.
Flagellar filament staining and microscopy. Flagellar filaments were visualized by exchanging surface-exposed threonine residues of the flagellin monomers to cysteine residues and selectively coupling maleimide-ligated fluorescent dyes to the SH-group of cysteine [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. To rule out that cysteine substitutions within the flagellin negatively affect the swimming behavior we performed Western Blot analysis, soft agar assays and single cell speed analysis.
For microscopy, cells of an exponentially growing LB culture (OD600 = 0.6) were harvested by centrifugation (1200 × g, 5 min, room temperature) and resuspended in 50 μl PBS. To prevent flagella being sheared off, the pipette tip was always cut off when pipetting cells. 0.5-1 μl Alexa Fluor 488 C5 maleimide (Thermo Fisher Scientific) or CF™ 488 maleimide (Sigma-Aldrich) was added and the cell suspension was incubated in the dark for 15 min. Cells were sedimented again and carefully washed with 1 ml of PBS to remove residual unbound dye. After final resuspension in LM100 medium (10 mM HEPES, pH 7.3; 100 mM NaCl; 100 mM KCl; 0.02% (wt vol −1 ) yeast extract; 0.01% (wt vol −1 ) peptone; 15 mM lactate) the cells were kept shaking in the dark until microscopy, but never longer than 30 min. Image sequences of typically 150-300 frames and 20-50 ms exposure time were taken at room temperature using a custom microscope setup (Visitron Systems) based on a Leica DMI 6000 B inverse microscope (Leica) equipped with a pco.edge sCMOS camera (PCO), a SPECTRA light engine (lumencor), and an HCPL APO ×63/1.4-0.6 objective (Leica) using a custom filter set (T495lpxr, ET525/50m; Chroma Technology).
Determining flagellar screw formation. To determine the screw formation frequency of the four types of flagellar filament mutants, an aliquot of stained cells was loaded on a swim slide and monitored about 100 µm away from the glass surfaces. For each filament type movies of three biological replicates were recorded on subsequent days in regular LM medium and LM supplemented with 15 % (wt vol −1 ) Ficoll® 400 (diluted from a 50 % (wt vol −1 ) stock solution). This small, highly branched polymer has been observed to act as a purely viscous contribution to the solvent rheology in experiments with E. coli [bib_ref] Flagellated bacterial motility in polymer solutions, Martinez [/bib_ref]. Several image sequences of 200 frames at 30 ms exposure time were taken for each condition. Screw formation was determined manually for backward swimming cells or cells switching to backward swimming. For each condition about 300 backward swimming events were counted.
Determining flagellar helix parameters. To measure the pitch, diameter and axis length of the flagellum the cells were prepared as described above and supplemented with 50 µM phenamil prior to loading on the swimslide. Phenamil stopped or at least slowed down the rotation of the Na + -driven motor so that images of the helical waveform of the flagella could be obtained. The proportion of the proximal flagellin fragment could only be measured for FlaA, as this flagellin shows brighter fluorescence and could be clearly distinguished from the cell body and the remaining part of the flagellum. Parameters were measured using the ImageJ distribution Fiji [bib_ref] Fiji: an open-source platform for biological-image analysis, Schindelin [/bib_ref].
To determine the handedness of the different types of flagellar filaments z-stack image sequences of fluorescently labeled flagella were recorded as described above. Shifting the focal plane through the cell body shows characteristic patterns for lefthanded and right-handed helices. The flagellar helices of all flagella types were found to be left-handed.
Protein separation and western immunoblotting. For western immunoblotting, cell lysates of exponentially growing LB cultures were obtained by centrifuging cells corresponding to an OD600 = 10 and resuspending the cell pellet in Laemmli buffer [bib_ref] Cleavage of structural proteins during the assembly of the head of bacteriophage..., Laemmli [/bib_ref]. Prior to protein separation by SDS/PAGE using 12% (wt · vol −1 ) polyacrylamide gels the samples were heated to 95°C for 5 min Subsequently, the proteins were transferred to a nitrocellulose Roti-PVDF membrane (Roth) by semidry transfer. The polar flagellins FlaA and FlaB (of S. putrefaciens and S. oneidensis alike) were detected with polyclonal antibodies which were raised against the N-terminal conserved region of S. MR-1 FlaB (Eurogentec Deutschland) in the dilution of 1:500. As secondary antibody anti-rabbit IgG-horseradish peroxidase (Thermo Fisher Scientific, prod. # 31460) was used at a dilution of 1:20,000. FLAG-tagged FliS was detected with a monoclonal, horseradishperoxidase-conjugated antibody raised against the FLAG-tag (Sigma Aldrich, prod. # A8592) in the dilution of 1:1000. The horseradish peroxidase signal was detected with the CCD System LAS 4000 (Fujifilm) after incubating the membranes with SuperSignalH West Pico Chemiluminescent Substrate (Thermo Scientific) for one minute. If portions of gels or blots are shown the full gels or blots can be found in [fig_ref] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB [/fig_ref].
Soft-agar spreading assay. Spreading ability in complex environments was carried out in soft-agar medium. LB medium supplemented with 0.15-0.25% (wt vol −1 ) select agar (Invitrogen) was heated carefully and cooled to 30-40°C before pouring the plates. After solidification 2 µl of exponentially grown cultures were spotted on the plates and incubated in a moist environment over night at room temperature. Plates were scanned before the swim colonies merged using an Epson V700 Photo Scanner. Cells to be directly compared were always incubated on the same plate.
Statistics of filament parameters and screw formation. Significance of the relative transcription levels (qRT-PCR) and the filament parameter measurements were tested with a two-sided two-sample t-test or Welch's t-test (P < 0.05, Bonferroni corrected) in R version 3.3.2. Mean, Median and SD were calculated in Microsoft Excel 2013. For each flagellar filament type pitch, diameter, axis length and if possible the proportion of the proximal filament fragment of the flagellum of 50 cells were measured. The actual length of the filament (arc length) was calculated in Excel 2013. Significance of screw formation (only if screws were observed at all) was tested pairwise with two-sided Fisher's exact test (P < 0.05, Bonferroni NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-018-07802-w ARTICLE corrected) in R version 3.3.2. 95% confidence intervals were calculated with the exact binomial test. For each flagellar filament type and condition about 100 backward swimming events were counted on 3 subsequent days, resulting in a total of about 300 counts for each filament type and condition. All measurements and the corresponding statistics are summarized in .
Isolation of total RNA and qRT-PCR. Total RNA of exponentially growing S. putrefaciens CN-32 cells (OD600 = 0.5, three biological replicates) was extracted using the hot-phenol method [bib_ref] Evidence for two functional gal promoters in intact Escherichia coli cells, Aiba [/bib_ref]. Briefly, the cells were incubated in SDS-acetate buffer (20 mM sodium acetate, 1 mM EDTA, 1% SDS, pH 5.5), with one and a half volumes of hot phenol (pH 4.3) at 65°C for 10 min, inverting the tube every 2 min. After centrifugation, the aqueous phase was extracted twice using phenolchloroform-isoamyl alcohol (25:24:1) and RNA was precipitated with 1/8 volume of 2 M sodium acetate solution (pH 5.2) and 2.5 volumes of ice cold ethanol (>96% v v −1 ). After incubation at −20°C overnight, the pellet was washed with ice cold ethanol (70% v v −1 ), dried and resuspended in RNase-free water. The quality of the RNA was determined by agarose gel electrophoresis and concentration was measured at 260 nm. Residual DNA was removed with the Turbo DNA-free kit (Applied Biosystems) according to the manufacturer's instructions. The RNA extract was then applied for random-primed first-strand cDNA synthesis using BioScript reverse transcriptase (Bioline) according to the manufacturer's instructions. RNA and cDNA samples were stored at −80°C. The cDNA samples were used as a template for qRT-PCR (C1000 Thermal Cycler with the CFX96 Real-Time System, Bio-Rad) using the SYBR green detection system, MicroAmp Optical 96-well reaction plates and Optica adhesive covers (Applied Biosystems). As flaA and flaB gene sequences are very similar the specificity of the used oligonucleotides (Supplementary [bib_ref] A new mathematical model for relative quantification in real-time RT-PCR, Pfaffl [/bib_ref] and used to estimate the differences in transcript amounts of the two flagellins.
Holographic cell tracking. Cells were grown according to the microscopy protocol above, but without the flagellar staining step. The cells were loaded into glass sample chambers measuring approximately 5 mm × 20 mm × 0.5 mm constructed from glass slides and UV-curing glue. The sample was illuminated using a fibercoupled laser diode (wavelength λ = 642 nm) mounted in place of the condenser lens assembly on a Nikon Ti U inverted microscope. The sample was imaged using a standard bright field objective lens (×10, NA 0.3) onto a CMOS camera (Mikrotron MC-1362). Movies were acquired at 50 Hz for around 60 s, at a resolution of 1024 × 1024 pixels. This arrangement gave a sensitive volume of around 1.4 mm × 1.4 mm × 0.5 mm in which to track cells. This resulted in 4887 tracks for the wild type, 4611 tracks for FlaB-only, 1883 tracks for FlaA-only and 4850 tracks for FlaBA in regular medium and 1340 tracks for the wild type, 2558 tracks for FlaB-only, 3019 tracks for FlaA-only and 1623 tracks for FlaBA in high viscosity medium. Note that even the smallest data set adds up to over 5.5 h of recorded and analyzed swim time. The movies were analyzed offline to extract cell coordinates in each frame, and to assemble these coordinates into cell tracks using custom-written software routines [bib_ref] Holographic microscopy of holographically trapped three-dimensional structures, Lee [/bib_ref] [bib_ref] 3D Localization of weak scatterers in digital holographic microscopy using Rayleigh-Sommerfeld back-propagation, Wilson [/bib_ref] [bib_ref] Sperm navigation along helical paths in 3D chemoattractant landscapes, Jikeli [/bib_ref] [bib_ref] High-speed holographic microscopy of malaria parasites reveals ambidextrous flagellar waveforms, Wilson [/bib_ref]. Cell tracks were smoothed with piecewise cubic splines to remove noise 52 and analyzed to extract the instantaneous velocity, tumble angle and run time distribution.
The raw data was processed stepwise to obtain cell coordinates. We obtained a background image from each movie sequence by finding the median pixel value at each position in the image. We divided each image in a sequence by this static background, and applied the Rayleigh-Sommerfeld back-propagation approach [bib_ref] Holographic microscopy of holographically trapped three-dimensional structures, Lee [/bib_ref] , to produce a stack of numerically refocused images at various distances from the original optical plane. A spatial bandpass filter was applied during the reconstruction to reduce pixel noise; the filter preserved objects between 2 and 30 pixels in size. At our relatively low magnification (×10 total magnification), cells appear point-like in the resulting images. We applied a Sobel-type filter 49,52 to obtain the position of each cell in three dimensions.
To reconstruct cell tracks an extract instantaneous velocities a custom software routine was written to connect the positions of cells between frames in order to assemble three-dimensional tracks. The position resolution carried an uncertainty of approximately ±0.5 µm in the directions perpendicular to the optical axis, and ±1 µm parallel to the optical axis in each frame, at this magnification. All cells were also subject to translational Brownian motion as they swam; the combination of high-frequency pixel noise and Brownian motion complicated efforts to calculate the time-derivatives of position, and therefore the swimming speed. To smooth short-time fluctuations associated with these noise sources, the cell tracks were fitted using piecewise cubic splines. This allowed us to reliably extract instantaneous cell velocities. This technique also allows us to reject the cells that are moving by Brownian alone. The spline-smoothed trajectory of a non-motile particle has a mean-squared displacement that scales with time, whereas a swimming cell's mean squared displacement scales with time squared. By setting a threshold for the slope of the mean-squared displacement at short times, and a threshold for the total mean squared displacement after 2 s, we can effectively eliminate non-motile particles.
Extraction of tumble angle and run duration distribution was done by calculating the cell's instantaneous direction at each time point, along with the change in direction, θ(t), between subsequent time points. These data were scanned using a standard peak search algorithm to detect the position of maxima in dθ/dt above a threshold of 5°s −1 . These maxima were associated with reorientation events; residual effects of Brownian motion after the spline smoothing are sufficient to allow pauses followed by a resumption of swimming (equivalent to a reorientation angle near zero degrees) to be detected as reorientation events. Example data are shown in [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] ; 13a shows a single cell track (wild type, the same track shown in [fig_ref] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types [/fig_ref] , and 13b displays a graph of dθ/dt against time in which the events noted as tumbles are highlighted with blue circles. The total change in swimming direction at each re-orientation event was calculated by comparing a cell's swimming direction 0.25 s before a maximum in θ(t) with the swimming direction 0.25 s after the maximum. The time between subsequent re-orientation events was used to build a distribution of run durations. Only tracks with two or more re-orientation events contributed to the run duration distributions.
Numerical model for the flagellar filament. The numerical model for the flagellar filament is based on on that proposed by Vogel and Stark 53 , and the same as the one used in Kühn et al. [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref]. It is based on a representation of the flagellum as a set of beads with mutual interactions such that they form a helix with prescribed radius R and pitch P at rest. The differences between FlaA and FlaB are reflected in the local radius R and pitch P of the helix: FlaA forms narrow helices with R = 0.175 µm and P = 1.18 µm, whereas FlaB forms wider helices with R = 0.315 µm and P = 1.91 µm. The torque-force relation is obtained by direct numerical integration of the equations of motion at fixed torque, and then by computing the mean force over 10 rotation periods to average out the oscillations that are connected with the positioning of the motor relative to the axis of the filament. As in the previous simulations [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] , we observe that polymorphic transitions within the filament reduce the required torque, and thus have a quantitative but not a qualitative effect on the changes in mechanical properties [bib_ref] Motor-driven bacterial flagella and buckling instabilities, Vogel [/bib_ref] [bib_ref] On torque and tumbling in swimming Escherichia coli, Darnton [/bib_ref].
With s the arclength along the rod, the torsional state of the rod is described by the rotational strain vector Ω(s). Deviations from the rest state give quadratic contributions to a free energy
[formula] F K ¼ Z L 0 A 2 Ω 1 À Ω 0;1 2 þ Ω 2 À Ω 0;2 2 þ C 2 Ω 3 À Ω 0;3 2 ds;ð1Þ [/formula]
where A is the bending and C the torsional rigidity. Polymorphism is introduced by calculating the free energy relative to the different equilibrium states and taking the minimum: if the flagellum is stretched too far from one local minimum it can fall into the basin of attraction of another one and relax to a different configuration. For FlaB we find two relevant equilibrium configurations, but for FlaA there is no evidence for a second one. We take the same elastic coefficients A and C for all minima, since we do not have independent numbers for the different equilibrium states (neither for the polymorphic states nor for FlaA vs FlaB).
In addition to the torsional contribution the free energy also contains a global harmonic spring potential
[formula] F S ¼ ðK=2Þ R L 0 ð∂r=∂sÞ 2 dsð2Þ [/formula]
with an elastic constant K that keeps the variations in length within 0.1%. The interaction between flagellum and fluid is calculated within resistive force theory. There are three local friction coefficients that depend on the geometry of the flagellum:
[formula] γ ? ¼ 4πη lnð0:09l=r f Þ þ 1=2ð3Þ [/formula]
is the friction coefficient for motion perpendicular to the centerline, with η the viscosity, r f the radius of the filament and l ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi 4π 2 R 2 þ P 2 p the contour length of one helical turn. The tangential friction coefficient is
[formula] γ jj ¼ 2πη lnð0:09l=r f Þð4Þ [/formula]
and the rotational coefficient is
[formula] γ r ¼ 4πηr 2 fð5Þ [/formula]
As in Vogel and Stark and Kühn et al. [bib_ref] Bacteria exploit a polymorphic instability of the flagellar filament to escape from..., Kühn [/bib_ref] [bib_ref] Motor-driven bacterial flagella and buckling instabilities, Vogel [/bib_ref] , the equations of motion are integrated with an embedded Cash-Karp method [bib_ref] A variable order Runge-Kutta method for initial value problems with rapidly varying..., Cash [/bib_ref] , and determine the positions and velocities of the attached tripoid following the procedure proposed by Chirico and Langowski 57 . The runs are started with a left-handed helix with P S = 1.91 µm and R S = 0.315 µm for segments containing FlaB, and with P S = 1.18 µm and R S = 0.175 µm for segments containing FlaA. The contour length of the flagellum was set to Lc = 6 µm, corresponding to 2 to 4 helical turns, depending on the FlaA to FlaB ratio. The various parameters in the model for the flagellum and their values are summarized in .
Code availability. The full code for the three-dimensional tracking has been published in the corresponding references [bib_ref] 3D Localization of weak scatterers in digital holographic microscopy using Rayleigh-Sommerfeld back-propagation, Wilson [/bib_ref] [bib_ref] Simultaneous two-color imaging in digital holographic microscopy, Farthing [/bib_ref]. The full code for simulations of flagellar screw formation is available upon request from the corresponding authors.
## Data availability
The tracking data are available at the York Research Database (https://doi.org/ 10.15124/675b6083-2a6e-43d1-8511-b7c18bb1be9a). Other data supporting the findings of the study are available in this article and its Supplementary Information files, or from the corresponding authors upon request. A reporting summary for this Article is available as a Supplementary Information file.
[fig] Figure 2: Schematic model of the sequential production and export of FlaA and FlaB. The figure summarizes our current working model of how spatial arrangement of the two flagellins is achieved in Shewanella putrefaciens, depicted in four subsequent stages of the flagellar assembly. a The current model of flagellin expression suggests that transcription is inhibited as long as the alternative sigma factor 28 (FliA) is blocked by the anti-sigma factor FlgM. In S. putrefaciens, this only applies to FlaB while FlaA production does not depend on FliA (seeSupplementary b). Therefore, a pool of FlaA is already present while the basal body is still being assembled. b Once the hook-basal-body complex is completed the export apparatus changes its specificity from hook proteins (FlgE) to the flagellins and other late assembly proteins. At this stage also the anti-sigma factor FlgM gets exported and FliA-dependent transcription of FlaB is initiated. c The already produced FlaA monomers get exported first and assemble at the base of the flagellar filament. d The increasing production of FlaB monomers passes that of FlaA and potentially FlaA transcription and/or translation may be reduced or even completely terminated (indicated by the question marks). Thus, FlaB constitutes the majority residual part of the flagellar filament. (Pol, RNA polymerase; Rib, ribosome; σ, flaA promoter only produces filament stubs, which indicates cessation of the expression from this promoter that may occur at the onset of flaB expression. [/fig]
[fig] Figure 3: Free-swimming behavior mediated by the four different flagellar filament types. Several thousand cell tracks were obtained for each strain by holographic microscopy. The diagrams show the relative counts to account for different sample sizes. a, b Run duration histograms for wild-type (FlaAB, dark grey solid lines), FlaA-only (orange mixed dashed lines), FlaB-only (blue long dashed lines) and FlaBA (light grey short dashed lines) cells in regular medium (upper panel) and medium with increased viscosity (lower panel) [/fig]
[fig] Figure 4: Experimentally observed and simulated screw formation. a The probability of screw formation of wild-type (FlaAB) and FlaB-only cells increases in high viscosity, while for FlaA-only and FlaBA screw formation was never observed. Significance was tested for all filament combinations under both conditions and for each filament between the two conditions. If no screw was observed at all, significance was not tested. All tested combinations were significant (P < 0.05, Bonferroni corrected). Error bars indicate 95% confidence intervals. About 300 cells were counted for each strain.b Observation of screw formation for varying flagellin compositions at different motor torques after a simulation time of t = 60 ms. The simulation was carried out for flagella with an increasing number of FlaA segments, starting with a flagellum completely composed of FlaB (bottom of the diagram) and successively exchanging the segments to a FlaA configuration starting from the filament's base. Yellow squares indicate regular backward rotation, blue dots indicate screw formation. The color coding represents the z-position of the flagellum's free end, with negative values indicating a position below the motor segment (position 0). For comparison with the actual flagellin composition in the wild type experimentally measured FlaA proportions are given on the right. Data points are displayed as individual values (gray dots) measured for 50 filaments. The box spans the central half of the data, the black bar indicates the median. A movie of the simulation with FlaA-only and FlaB-only filaments is provided as Supplementary Movie 2. c Force-torque relation extracted from the numerical simulation. The mean force F on the flagellum varies linearly with motor torque M for both forward and backward swimming. Shown are the relations for a FlaB-only flagellum (blue diamonds), a 20% admixture of FlaA (gray stars) and an 80% admixture (orange triangles). The continuous lines are linear fits to the data. The slope βeff in F = βeff M versus the FlaA content is shown in the inset: it increases rapidly up to about 20% FlaA and more slowly for larger fractions. A higher value of βeff indicates a more efficient transfer of torque into driving https://doi.org/10.1038/s41467-018-07802-w | www.nature.com/naturecommunications [/fig]
|
Primary uterine diffuse large B-cell lymphoma involving the urinary bladder with urinary cytology mimicking carcinomas: A case report
We report a rare case of a 69-year-old woman in whom diffuse large B-cell lymphoma (DLBCL) originated from the uterus and involved the urinary bladder. The cervical smears of the case mostly consisted of discohesive atypical round cells, which were highly suggestive of lymphoma; however, in voided urine smears, a majority of the cells formed large aggregates of degenerated cells, mimicking those of urothelial carcinoma (UC). The smears also represented some small loose clusters, in which tumor cells formed short chains with nuclear molding, mimicking those of small cell carcinoma. The cytodiagnosis got defi nitive when we identifi ed the atypical cells that showed CD20+/CD3-/cytokeratin-/NSE-immunophenotype. These are of particular concern as they may have misleading similarities to other epithelial neoplasms when examining lymphoma involving the urinary bladder. Accordingly, this case highlights the importance of immunocytochemistry to rule out malignant lymphoma when encountering large and/or small loose clusters of atypical round cells on urinary cytology.
# Introduction
Extranodal non-Hodgkin's lymphomas (NHLs) make up approximately 25% of all cases of lymphomas, with the most common sites being gastrointestinal tract and skin. [bib_ref] Primary cervical lymphoma: Report of two cases and review of the literature, Dursun [/bib_ref] Despite the increasing incidence of NHLs during the last few decades, only 1-1.5% arises from female genital organs. [bib_ref] Primary cervical lymphoma: Report of two cases and review of the literature, Dursun [/bib_ref] Among them, the ovaries, uterus, and fallopian tubes are the most common sites, and only 0.12% of all NHLs originate from the uterine cervix. [bib_ref] Primary malignant reticuloendothelial disease involving the vagina, cervix, and corpus uteri, Chorlton [/bib_ref] The most common subtype of cervical NHLs is diffuse large B-cell lymphoma (DLBCL); therefore, the lymphoma cells in smears typically appear as discohesive cells with scant cytoplasm, indistinct cellular membranes, irregular nuclear contours, one or more large prominent nucleoli, and irregular clearing of the nuclear chromatin. [bib_ref] Malignant lymphoma of the genitourinary tract, Ferry [/bib_ref] We report a rare case of DLBCL in the uterine cervix directly invading the urinary bladder, in which tumor cells were identified in the urine smears. Interestingly, the voided urine smears consisted predominantly of cohesive clusters mimicking those of epithelial neoplasms arising in the urinary bladder.
## Case report
A 69-year-old menopausal woman presented to the urology department of our institution with a 1-month history of lower abdominal discomfort, back pain, and sudden hematuria.
## Case report
This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as the author is credited and the new creations are licensed under the identical terms. Journal of Cytology / July 2015 / Volume 32 / Issue 3 patient underwent a cystoscopy procedure which revealed a smooth, large protruded lesion of the posterior bladder wall. Cytological smears of voided urine were interpreted as a lymphoma with a confirmation by immucytochemistry. Subsequent computed tomography (CT) scan and magnetic resonance imaging (MRI) of the pelvis revealed a solid tumor (5.0 × 5.0 cm in dimension) occupying the uterine cervix and lower part of the body [ ], thickening of the posterior bladder wall, and bilateral hydronephrosis. A few pelvic lymph nodes were slightly enlarged. Speculum examination showed a large, irregular surfaced mass with no pedicle arising from the anterior lip of the cervix. Specimen from the cervix and endometrium were taken and sent to cytopathology; cytologic interpretation was of malignant lymphoma [ and pathological diagnosis was that of DLBCL, nongerminal center subtype.
The patient was hospitalized and systemic CT scan revealed slightly enlarged systemic lymph nodes. A staging bone marrow aspirate showed microscopic neoplastic infiltrate. These findings were compatible with stage IV of DLBCL arising from the uterus where the cervix appears to be the primary site of disease. The patient was successfully treated with one course of cyclophosphamide, hydroxydaunorubicin, oncovin, and prednisone or prednisolone (CHOP) and five courses of CHOP with the drug rituximab (R-CHOP) chemotherapy. She has maintained complete remission for 3 years since the final course of the chemotherapy.
Cytopathological fi ndings A voided urine specimen was collected and processed by the centrifuge method. The sample was centrifuged at 700 g for 3 min, fixed in 95% alcohol, and stained with May-Grünwald-Giemsa and Papanicolaou methods. The smears consisted predominantly of large aggregates with marked cellular crowding, nuclear overlapping, and short chains composed of degenerated medium-sized round cells [ ]. These cells contained relatively abundant cytoplasm and reduced round nuclei, resulting in a relatively low nuclear to cytoplasmic ratio. Some of the cells were arranged as short chains with nuclear molding but chromatin smearing was rare [ ]. However, discohesive fairly uniform cells were scattered around the clusters [ ]. Differentiating between epithelial neoplasms and malignant lymphoma was difficult from these findings; therefore, we performed immunocytochemistry. We employed the polymerperoxidase method without antigen retrieval, using Histofine Simple Stain MAX-PO (Nichirei Biosciences, Tokyo, Japan) and monoclonal antibodies against CD20 [L26, diluted 1/500; DAKO Cytomation (DC, Glostrup, Denmark)], CD3 (2GV6, prediluted; Roche, Basel, Switzerland), cytokeratin (AE1/ AE3, 1/100; DC), and neuron-specific enulase (NSE) (BBS/NC/ VI-H14, 1/100; DC). The immunophenotype of CD20+/CD3-/ cytokeratin-/NSE-was consistent with a B-cell lymphoma [ [fig_ref] Figure 3: Immunocytochemistry of voided urine specimen [/fig_ref].
Meanwhile, the cervical smears, collected and processed by the centrifuge method, demonstrated a dispersive cellular population with single cells of intermediate to large size in a clear background. Majority of these cells were discohesive and contained a moderate amount of cytoplasm and round nuclei with finely granular chromatin and a few small nucleoli. Cell borders were conspicuous and nuclear contours were relatively smooth. Histology of the cervical biopsy revealed a highly cellular tumor of fairly monotonous lymphoid proliferation with a diffuse infiltrating pattern. Immunohistochemistry on formalin-fixed paraffin-embedded sections showed that the tumor cells were positive for CD45, CD20, CD79a, and multiple myeloma oncogene 1 (MUM1) and negative for CD3, CD5, CD10, cyclin D1, Bcl-2, Bcl-6,
# Discussion
Cytologic evaluation of voided urine specimen is the most common method for the detection and diagnosis of neoplasms in the urinary bladder. Typical cytologic features of bladder lymphoma are discohesive population of atypical large lymphocytes, whereas those of urothelial carcinoma (UC) are cohesive clusters with marked cellular crowding and nuclear overlapping. [bib_ref] Hematologic lesions in urine cytology, Murro [/bib_ref] Thus, neoplastic cells with lack of cohesion are generally considered as a useful finding for distinguishing lymphoma cells from epithelial cells. However, we described here an unusual cytomorphological pattern of DLBCL presenting predominantly as cohesive clusters of cells on urine smear, mimicking epithelial neoplasms.
It is known that some specific types of lymphoma are prone to form aggregates in cytological specimens. For example, anaplastic large cell lymphoma and adult T-cell leukemia/ lymphoma may show clusters of highly atypical cells. [bib_ref] Cytological features of diffuse large B-cell lymphoma can mimic metastatic carcinoma on..., Mohiuddin [/bib_ref] Apart from these specific types, other types of lymphoma may form cohesive groups of degenerated neoplastic cells in the urine. Indeed, some previous reports represented cases of high-grade DLBCL and low-grade lymphoma of the mucosa-associated lymphoid tissue (MALT) type where neoplastic cells appeared as clusters with cellular crowding and cohesion, mimicking those of UC. [bib_ref] Cytologic features of recurrent lymphoma involving the urinary bladder, Quinn [/bib_ref] [bib_ref] Malignant lymphoma case with urinary cytology mimicking that of urothelial carcinoma, Takahashi [/bib_ref] In addition, short chains with nuclear molding of lymphoma cells may have misleading similarities to small cell carcinoma. [bib_ref] Primary small cell neuroendocrine carcinoma of the urinary bladder with coexisting high-grade..., Bui [/bib_ref] In the urine specimen, neoplastic cells of small cell carcinoma often show loose clusters and isolated single cells with scant cytoplasm, mimicking those of lymphoma cells. [bib_ref] Primary small cell neuroendocrine carcinoma of the urinary bladder with coexisting high-grade..., Bui [/bib_ref] These are of particular concern and may have misleading similarities to epithelial neoplasms when examining lymphoma involving the urinary bladder.
In addition, the type of preparation may affect interpretation. Compared with the centrifuge method, thin layer preparations are more cellular with better cell preservation and cleaner backgrounds. [bib_ref] Hematologic lesions in urine cytology, Murro [/bib_ref] Such characteristics facilitate the identification of isolated malignant cells. However, cells in centrifuge method preparations are prone to form relatively tight clusters. [bib_ref] Evaluation of a centrifuge method and thin-layer preparation in urine cytology, Zardawi [/bib_ref] Therefore, cytodiagnosis of lymphoma involving the urinary bladder should be based on the awareness that these patterns contain potential pitfalls and misleading mimics, and should be supported by adequate immunocytochemical analysis.
## Financial support and sponsorship
The authors declare that no funding was received for this study.
## Confl icts of interest
There are no conflicts of interest.
[fig] Figure 1, Figure 2: (a) Pelvic magne c resonance imaging (MRI) shows a solid tumor occupying the uterine cervix. Extensive thickening of the posterior wall of the urinary bladder is also noted (b) Histology of biopsy from the uterine cervix shows diff use large B-cell lymphoma (H and E, ×400) (a) Cervical smear shows discohesive, large atypical lymphocytes with high nuclear-cytoplasmic ratio, finely granular chromatin, and conspicuous nucleoli (b) Voided urine sediment shows a large aggregated cluster of degenerated lymphocytes of intermediate size (c) Short chains of degenerated lymphocytes with nuclear molding in voided urine (d) Discohesive large lymphocytes in voided urine, which are morphologically similar to the atypical lymphocytes in cervical smear (Pap, ×600) Journal of Cytology / July 2015 / Volume 32 / Issue 3and TdT that were consistent with DLBCL, nongerminalcenter type. [/fig]
[fig] Figure 3: Immunocytochemistry of voided urine specimen. (a) The intermediate-sized atypical lymphocytes are nega ve for cytokera n (b) nega ve for CD3 (c) posi ve for CD20, and (d) nega ve for NSE. (IHC, x600) [/fig]
|
Delineating the mechanisms and design principles of Caenorhabditis elegans embryogenesis using in toto high-resolution imaging data and computational modeling
a b s t r a c tThe nematode (roundworm) Caenorhabditis elegans is one of the most popular animal models for the study of developmental biology, as its invariant development and transparent body enable in toto cellular-resolution fluorescence microscopy imaging of developmental processes at 1-min intervals. This has led to the development of various computational tools for the systematic and automated analysis of imaging data to delineate the molecular and cellular processes throughout the embryogenesis of C. elegans, such as those associated with cell lineage, cell migration, cell morphology, and gene activity. In this review, we first introduce C. elegans embryogenesis and the development of techniques for tracking cell lineage and reconstructing cell morphology during this process. We then contrast the developmental modes of C. elegans and the customized technologies used for studying them with the ones of other animal models, highlighting its advantage for studying embryogenesis with exceptional spatial and temporal resolution. This is followed by an examination of the physical models that have been devised-based on accurate determinations of developmental processes afforded by analyses of imaging data-to interpret the early embryonic development of C. elegans from subcellular to intercellular levels of multiple cells, which focus on two key processes: cell polarization and morphogenesis. We subsequently discuss how quantitative data-based theoretical modeling has improved our understanding of the mechanisms of C. elegans embryogenesis. We conclude by summarizing the challenges associated with the acquisition of C. elegans embryogenesis data, the construction of algorithms to analyze them, and the theoretical interpretation.
# Introduction
Caenorhabditis elegans is a transparent free-living nematode that dwells in decaying organic samples in soil and feeds on microorganisms such as bacteria. Its embryogenesis occurs inside an axisymmetric oval egg with a length and width of 50 lm and 30 lm, respectively, over a period of 10-12 h at 20°C [bib_ref] The life cycle of the nematode Caenorhabditis elegans. I. Wild-type growth and..., Byerly [/bib_ref]. After hatching, C. elegans grows from 250 to 850 lm in length via four larval stages and over a period of 44-46 h at 20°C [bib_ref] The life cycle of the nematode Caenorhabditis elegans. I. Wild-type growth and..., Byerly [/bib_ref] [bib_ref] Regulation of Caenorhabditis elegans body size and male tail development by the..., Soete [/bib_ref]. The tiny size of C. elegans means that a microscope is usually needed to observe its anatomy and behavior. Its life cycle from zygote to egg-laying adult is only 3 days at room temperature, and as each individual lays 300 eggs in 60 h, it is an efficient and low-cost organism for biological research, particularly genetics research [bib_ref] The life cycle of the nematode Caenorhabditis elegans. I. Wild-type growth and..., Byerly [/bib_ref] [bib_ref] Caenorhabditis elegans as a powerful tool in natural product bioactivity research, Ha [/bib_ref]. There are two sexes of C. elegans: hermaphrodite and male. Hermaphrodites' progeny generated by selffertilization are nearly all hermaphrodites with only 0.1 % males, while those produced by mating are 50 % males. The hermaphrodite progeny generated by self-fertilization have the same genome as their hermaphrodite parent, which enables molecular biologists to perform experiments with high genotypic and phenotypic stability [fig_ref] Figure 1: C [/fig_ref].
Since the early 20 th century, it has been known that adult C. elegans have a constant number of somatic cells; this is an example of the deterministic phenomenon known as eutely [bib_ref] Die Zellkonstanz und ihre Beziehungen zu anderen zoologischen Vorwürfen, Martini [/bib_ref] [bib_ref] Eutely or cell constancy in its relation to body size, Van Cleave [/bib_ref]. Thus, although C. elegans has numerous types of organs and tissues, such as the skin, pharynx, neuron, muscle, and gonad, hermaphrodite embryos comprise only 558 living cells and male embryos comprise only 560 living cells, with the corresponding adults comprising 959 and 1,031 cells, respectively [bib_ref] Cell lineage and the control of Caenorhabditis elegans development, Kenyon [/bib_ref]. This has enabled every cell's behavior and function during its lifespan to be fully recorded, revealing the highly invariant somatic-cell lineages of the embryonic and postembryonic stages [fig_ref] Figure 1: C [/fig_ref] [bib_ref] Post-embryonic development in the ventral cord of Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Post-embryonic cell lineages of the nematode, Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Cell lineages of the embryo of the nematode Caenorhabditis elegans, Deppe [/bib_ref] [bib_ref] The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis..., Kimble [/bib_ref] [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref]. However, C. elegans cells do show variations in the timing and orientation of their cell division [bib_ref] Assessing normal embryogenesis in Caenorhabditis elegans using a 4D microscope: Variability of..., Schnabel [/bib_ref] [bib_ref] A quantitative model of normal Caenorhabditis elegans embryogenesis and its disruption after..., Richards [/bib_ref] [bib_ref] Systems-level quantification of division timing reveals a common genetic architecture controlling asynchrony..., Ho [/bib_ref]. The variety of cell fates is a consequence of progressive cell-fate specification that occurs after fertilization. First, the zygote (P0) undergoes four successive rounds of asymmetric division to give rise to the soma (i.e., four somatic founder cells: AB, EMS, C, and D) [fig_ref] Figure 1: C [/fig_ref]. This leaves one remaining germline precursor cell (P4), which undergoes one round of symmetric division to generate two primordial germ cells (Z2 and Z3) [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] MEX-5 and MEX-6 function to establish soma/germline asymmetry in early C. elegans..., Schubert [/bib_ref] [bib_ref] Establishment and maintenance of heritable chromatin in C. elegans, Gangishetti [/bib_ref]. Second, somatic cells are further differentiated upon induction by cellular signaling, such as by Wnt signaling originating from the embryo posterior, which also induces the asymmetric localization of cellular contents in daughter cells and leads to fate specification [bib_ref] Wnt signaling polarizes an early C. elegans blastomere to distinguish endoderm from..., Thorpe [/bib_ref] [bib_ref] Wnt signaling and an APC-related gene specify endoderm in early C. elegans..., Rocheleau [/bib_ref] [bib_ref] Quantitative differences in nuclear b-catenin and TCF pattern embryonic cells in C...., Zacharias [/bib_ref] , or by Notch signaling transducted through intercellular physical contact, which drives the differentiation of AB descendants at the 4-and 12-cell stages [bib_ref] The maternal genes apx-1 and glp-1 and establishment of dorsal-ventral polarity in..., Mello [/bib_ref] [bib_ref] glp-1 and inductions establishing embryonic axes in C. elegans, Hutter [/bib_ref] [bib_ref] Two maternal genes, apx-1 and pie-1, are required to distinguish the fates..., Mango [/bib_ref]. These stereotypical developmental patterns exhibit the key characteristics of metazoan development, which allows for the study of various biological processes. For instance, in C. elegans, there are only two intestinal precursor cells (i.e., Ea and Ep, specified by Wnt signaling) undergoing gastrulation and there's only one excretory/kidney cell (i.e., ABplpappaap, specified by Notch signaling) [bib_ref] Wnt/ Frizzled signaling controls C. elegans gastrulation by activating actomyosin contractility, Lee [/bib_ref] [bib_ref] lin-12 and glp-1 are required zygotically for early embryonic cellular interactions and..., Moskowitz [/bib_ref]. In addition, as C. elegans has many genes controlling various physiologies and behaviors, it is an excellent animal model for the mechanistic study of many types of processes [bib_ref] The genetics of behaviour, Brenner [/bib_ref] [bib_ref] The genetics of Caenorhabditis elegans, Brenner [/bib_ref] [bib_ref] The DNA of Caenorhabditis elegans, Sulston [/bib_ref]. For example, as a pilot project for the Human Genome Project, the C. elegans genome was completely sequenced in 1998. This revealed that its genome consists of 19,000 proteincoding genes, while the number of genes in the human genome was estimated as 30,000-40,000 [bib_ref] Genome sequence of the nematode C. elegans: A platform for investigating biology, Elegans Sequencing [/bib_ref]. These studies have required the development of genetic and genomic technologies that, together with sequencing data, have founded a new era of data-driven systems biology studies of worms and other species [bib_ref] The genome sequence of Caenorhabditis briggsae: A platform for comparative genomics, Stein [/bib_ref] [bib_ref] Genomics in C. elegans: So many genes, such a little worm, Hillier [/bib_ref] [bib_ref] WormBase in 2022 -Data, processes, and tools for analyzing Caenorhabditis elegans, Davis [/bib_ref]. This has resulted in substantial improvements in our understanding of how genes and proteins function and interact during development and respond to environmental stimuli.
Studies using C. elegans have generated several fundamental discoveries that have had significant scientific, social, and economic effects. For example, the 2002 Nobel Prize in Physiology or Medicine was awarded to Sydney Brenner, John E. Sulston, and H. Robert Horvitz, ''for their discoveries concerning genetic regulation of organ development and programmed cell death" [bib_ref] Working with worms: Caenorhabditis elegans as a model organism, Meneely [/bib_ref]. C. elegans was established as a model animal nearly half a century ago by Brenner, who described the induction of C. elegans mutants and uncovered its huge pool of genetic units [bib_ref] The genetics of behaviour, Brenner [/bib_ref] [bib_ref] The genetics of Caenorhabditis elegans, Brenner [/bib_ref]. Sulston and coworkers performed laborious naked-eye studies of C. elegans to trace its complete cell lineage, from zygote to adult, which serves as a repeatable reference for the migration, division, and differentiation of every C. elegans cell [bib_ref] Post-embryonic development in the ventral cord of Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Post-embryonic cell lineages of the nematode, Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref]. Horvitz and co-workers exploited the invariant cell apoptosis pattern in C. elegans development to identify the ''death genes" and ''survival genes" that control apoptosis and the genes that regulate the removal of dead cells, which has facilitated much research in developmental and cell biology, especially in cancer biology [bib_ref] Genetic control of programmed cell death in the nematode C. elegans, Ellis [/bib_ref] [bib_ref] Two C. elegans genes control the programmed deaths of specific cells in..., Ellis [/bib_ref] [bib_ref] Caenorhabditis elegans gene ced-9 protects cells from programmed cell death, Hengartner [/bib_ref] [bib_ref] Genes required for the engulfment of cell corpses during programmed cell death..., Ellis [/bib_ref]. Subsequently, Andrew Z. Fire and Craig C. Mello were awarded the 2006 Nobel Prize in Physiology or Medicine ''for their discovery of RNA interference -gene silencing by double-stranded RNA," whereby the expression of selected genes is suppressed by double-stranded RNA homologous to these genes' RNA [bib_ref] Working with worms: Caenorhabditis elegans as a model organism, Meneely [/bib_ref] [bib_ref] Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans, Fire [/bib_ref] [bib_ref] RNA interference in Caenorhabditis elegans, Conte [/bib_ref]. This geneknockdown method and cellular-resolution data on the development of C. elegans have facilitated high-throughput identification and analysis of cell-and lineage-specific gene functions [bib_ref] Systems-level quantification of division timing reveals a common genetic architecture controlling asynchrony..., Ho [/bib_ref] [bib_ref] The regulatory landscape of lineage differentiation in a metazoan embryo, Du [/bib_ref] [bib_ref] Digital development: A database of cell lineage differentiation in C. elegans with..., Santella [/bib_ref] [bib_ref] Timing of tissuespecific cell division requires a differential onset of zygotic transcription..., Wong [/bib_ref]. Furthermore, the 2008 Nobel Prize in Chemistry was awarded to Osamu Shimomura, Martin Chalfie, and Roger Y. Tsien ''for the discovery and development of the green fluorescent protein, GFP" [bib_ref] Working with worms: Caenorhabditis elegans as a model organism, Meneely [/bib_ref]. Chalfie, a neurobiologist, pioneered the use of GFP to visualize gene expression in the mechanosensory neurons of living C. elegans larvae, which paved the way for direct monitoring of gene or cellular/tissue activities in vivo at both spatial and temporal scales [bib_ref] Green fluorescent protein as a marker for gene expression, Chalfie [/bib_ref]. Thus, the above-mentioned discoveries and technological developments have established a solid foundation for quantitative, systematic, and single-cell-based research on C. elegans embryogenesis.
## Cellular-resolution studies of embryogenesis
In this section, we review methods used for cell tracing and lineage tracking (hereinafter referred to as lineaging) and cell morphology reconstruction in C. elegans. The developmental characteristics of and technologies that have been used to study the ascidian, fruit fly, and zebrafish are briefly introduced and compared with the developmental characteristics of and technologies that have been used to study C. elegans, to demonstrate its strength as a model system. [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. The branch length (cell cycle length) of the tree is an average of 46 wild-type embryos, and the blue bar at the bottom terminal of each branch denotes the standard deviation of the corresponding cell cycle length. (C) Images of fluorescently labeled cell nuclei and cell membranes (middle row) and the corresponding nucleus tracing (upper row) and membrane segmentation (lower row); a scale bar (white line) denotes 10 lm (the original data are from [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. (D) Images of cells with fluorescently labeled cell nuclei (left) and hlh-1 promoter reporters (middle) and the corresponding cell tracing and expression profiling with expression intensity scaling (right); a scale bar as in C (all of these subfigures are reproduced from [bib_ref] Multilevel regulation of muscle-specific transcription factor hlh-1 during Caenorhabditis elegans embryogenesis, Guan [/bib_ref]. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
## Cell tracing and lineaging
Given the nematode species' stereotypical developmental pattern in terms of their cell position, cycle, and fate, it has been recognized that customized methods are needed for tracing nematode cell lineage [bib_ref] Die Zellkonstanz und ihre Beziehungen zu anderen zoologischen Vorwürfen, Martini [/bib_ref] [bib_ref] Über die Entstehung des Gegensatzes zwischen den Geschlechtszellen und den somatischen Zellen..., Boveri [/bib_ref] [bib_ref] Das Nervensystem von Ascaris lumbricoides und megalocephala, I, Goldschmidt [/bib_ref] [bib_ref] Das Nervensystem von Ascaris lumbricoides und megalocephala, II, Goldschmidt [/bib_ref]. Although it is relatively easy to identify cells in a juvenile or adult nematode, it is extremely difficult to identify cells in a developing embryo, in particular during later embryogenesis when (compared with early embryogenesis) there are more cells and those from the same lineage and with the same fate look similar. This means that researchers must be able to recognize patterns while the blastomere is in rapid cleavage and the cell volume is continuously decreasing. This problem was taken seriously when Sydney Brenner showed his vision in the utility of C. elegans in biological research [bib_ref] The genetics of behaviour, Brenner [/bib_ref] [bib_ref] The genetics of Caenorhabditis elegans, Brenner [/bib_ref] [bib_ref] The DNA of Caenorhabditis elegans, Sulston [/bib_ref]. Several of his postdoctoral researchers and others devoted themselves to establishing the cell lineage pattern of C. elegans and in around 1980, after years of laborious observation and recording, achieved the historic milestone of a full description of its cell lineage, from zygote to adult. Remarkably, this was achieved using manual techniques, aided by Nomarski microscopy, and it provided a reliable reference for further research using this organism [bib_ref] Post-embryonic development in the ventral cord of Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Post-embryonic cell lineages of the nematode, Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Cell lineages of the embryo of the nematode Caenorhabditis elegans, Deppe [/bib_ref] [bib_ref] The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis..., Kimble [/bib_ref] [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref]. The nomenclature for cell identity was established based on an arbitrary lineal origin and the relative location of offspring cells with respect to the anterior-posterior, left-right, and dorsal-ventral body axes [bib_ref] Post-embryonic cell lineages of the nematode, Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref]. As mentioned, the zygote is named P0, and its successively derived germline precursor cells inherit the prefix ''P" and are labeled with a suffix [bib_ref] Regulation of Caenorhabditis elegans body size and male tail development by the..., Soete [/bib_ref] [bib_ref] Caenorhabditis elegans as a powerful tool in natural product bioactivity research, Ha [/bib_ref] to denote their generation. P4 divides symmetrically to produce two primordial germ cells, Z2 and Z3, which then arrest until the larval stage. The somatic founder cells generated by P0, P1, P2, and P3 are named AB, EMS, C, and D, respectively. The daughter cells of a somatic cell take their name from their mother cell's name and its initial location relative to its sister cell (''a" for anterior, ''p" for posterior, ''l" for left, ''r" for right, ''d" for dorsal, and ''v" for ventral). For example, the daughter cell of AB that is located in the anterior relative to its sister cell is named ''ABa," and the daughters of ABa that are located on either side along the left-right axis are respectively named ''ABal" and ''ABar." Hence, all of the cells proliferated throughout C. elegans embryogenesis are unambiguously identified and named [fig_ref] Figure 1: C [/fig_ref].
Given that manual determination of the entire embryonic cell lineage of C. elegans was extremely laborious, it was not feasible to use this approach for determining the entire embryonic cell lineage of a mutant or perturbed embryo [bib_ref] Cell lineages of the embryo of the nematode Caenorhabditis elegans, Deppe [/bib_ref] [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref]. In addition, manually collected cell-lineage data are only qualitative or semiquantitative. However, 15 years after the publication of the pioneering study of Sulston and co-workers, a method for the quantitative tracking of C. elegans embryonic cell lineage was developed based on the use of a multifocal-plane time-lapse video recording system and customized tracing/visualizing software, Biocell [bib_ref] Assessing normal embryogenesis in Caenorhabditis elegans using a 4D microscope: Variability of..., Schnabel [/bib_ref] [bib_ref] Cortical and cytoplasmic flow polarity in early embryonic cells of Caenorhabditis elegans, Hird [/bib_ref]. This method enables systems-level studies, such as cell sorting and developmental variability in each individual cell, with cell nuclei in Nomarski images up to the 400-cell stage used for manual cell identification [bib_ref] Global cell sorting in the C. elegans embryo defines a new mechanism..., Schnabel [/bib_ref] [bib_ref] Global cell sorting is mediated by local cell-cell interactions in the C...., Bischoff [/bib_ref]. In 2006, Bao and coworkers developed a platform for automated cell tracing and lineaging called StarryNite, in which cell nuclei are ubiquitously labeled with a histone::GFP fusion protein that increases the contrast between the nuclei and cytoplasms, and thus facilitates segmentation of images acquired by time-lapse three-dimensional (3D) confocal microscopy [fig_ref] Figure 1: C [/fig_ref] [bib_ref] Automated cell lineage tracing in Caenorhabditis elegans, Bao [/bib_ref] [bib_ref] The lineaging of fluorescentlylabeled Caenorhabditis elegans embryos with StarryNite and AceTree, Murray [/bib_ref]. The labeled nuclei are used for cell identification, which is followed by manual editing to remove errors produced by StarryNite [bib_ref] The lineaging of fluorescentlylabeled Caenorhabditis elegans embryos with StarryNite and AceTree, Murray [/bib_ref] [bib_ref] AceTree: A tool for visual analysis of Caenorhabditis elegans embryogenesis, Boyle [/bib_ref]. Image acquisition is performed up to the 550-cell stage at 1-min intervals, and these intervals can be reduced to 10 s to observe a transient process [bib_ref] A quantitative model of normal Caenorhabditis elegans embryogenesis and its disruption after..., Richards [/bib_ref] [bib_ref] Systems-level quantification of division timing reveals a common genetic architecture controlling asynchrony..., Ho [/bib_ref] [bib_ref] Automated cell lineage tracing in Caenorhabditis elegans, Bao [/bib_ref] [bib_ref] The lineaging of fluorescentlylabeled Caenorhabditis elegans embryos with StarryNite and AceTree, Murray [/bib_ref] [bib_ref] Computable early Caenorhabditis elegans embryo with a phase field model, Kuang [/bib_ref]. The software StarryNite enables the extraction of information on cell position, division timing and axis, and lin-eage/division history from 250 3D time-lapse image stacks consisting of 20,000 two-dimensional (2D) images, while a companion program, AceTree, provides a graphical user interface for manual checking and editing, and visualization and analysis of tracking results (e.g., by plotting a lineage tree) [bib_ref] AceTree: A tool for visual analysis of Caenorhabditis elegans embryogenesis, Boyle [/bib_ref]. Manual editing of a 350-cell embryo and a 550-cell embryo can be performed by an expert in 0.5-2 h and 8-16 h, respectively depending on image quality [bib_ref] A quantitative model of normal Caenorhabditis elegans embryogenesis and its disruption after..., Richards [/bib_ref] [bib_ref] Systems-level quantification of division timing reveals a common genetic architecture controlling asynchrony..., Ho [/bib_ref]. Such cell tracing and lineaging approaches are feasible until the muscle starts twitching [bib_ref] WormGUIDES: An interactive single cell developmental atlas and tool for collaborative multidimensional..., Santella [/bib_ref] , as this prevents the acquisition of meaningful images.
The above-described analytical tools have been continually improved to achieve better performance [bib_ref] Using machine learning to speed up manual image annotation: Application to a..., Aydin [/bib_ref] [bib_ref] A hybrid blob-slice model for accurate and efficient detection of fluorescence labeled..., Santella [/bib_ref] [bib_ref] A semi-local neighborhood-based framework for probabilistic cell lineage tracing, Santella [/bib_ref] [bib_ref] AceTree: A major update and case study in the long term maintenance..., Katzman [/bib_ref]. When introducing another fluorescent marker to label the gene of interest by promoter-or protein-fusion, the lineage tracking system is able to generate the lineal of the gene in each cell of a developing embryo [bib_ref] Highthroughput in vivo analysis of gene expression in Caenorhabditis elegans, Hunt-Newbury [/bib_ref] [bib_ref] Automated analysis of embryonic gene expression with cellular resolution in C. elegans, Murray [/bib_ref] [bib_ref] Automated lineage and expression profiling in live Caenorhabditis elegans embryos, Murray [/bib_ref]. As such, the software tool AceBatch was developed to automate this task and adapted to the existing systems constituted by StarryNite and AceTree [bib_ref] Automated analysis of embryonic gene expression with cellular resolution in C. elegans, Murray [/bib_ref] [bib_ref] Automated lineage and expression profiling in live Caenorhabditis elegans embryos, Murray [/bib_ref]. This is exemplified by the lineal expression profiling of the muscle-specific transcription factor hlh-1 [fig_ref] Figure 1: C [/fig_ref] [bib_ref] Multilevel regulation of muscle-specific transcription factor hlh-1 during Caenorhabditis elegans embryogenesis, Guan [/bib_ref]. Thus far, expression patterns during C. elegans development have been collected for thousands of genes, thereby allowing inference of regulatory control in particular pathways [bib_ref] The regulatory landscape of lineage differentiation in a metazoan embryo, Du [/bib_ref] [bib_ref] Digital development: A database of cell lineage differentiation in C. elegans with..., Santella [/bib_ref] [bib_ref] Highthroughput in vivo analysis of gene expression in Caenorhabditis elegans, Hunt-Newbury [/bib_ref] [bib_ref] Automated analysis of embryonic gene expression with cellular resolution in C. elegans, Murray [/bib_ref] [bib_ref] Automated lineage and expression profiling in live Caenorhabditis elegans embryos, Murray [/bib_ref] [bib_ref] Multilevel regulation of muscle-specific transcription factor hlh-1 during Caenorhabditis elegans embryogenesis, Guan [/bib_ref] [bib_ref] De novo inference of systems-level mechanistic models of development from live-imaging-based phenotype..., Du [/bib_ref] [bib_ref] Inference of cellular level signaling networks using single-cell gene expression data in..., Huang [/bib_ref] [bib_ref] Gene regulatory effects inference for cell fate determination based on single-cell resolution..., Huang [/bib_ref]. This systems-level approach to the collection of spatial and genomic information is readily adaptable to other nematode species, which has facilitated comparative and evolutionary biology studies [bib_ref] Comparative analysis of embryonic cell lineage between Caenorhabditis briggsae and Caenorhabditis elegans, Zhao [/bib_ref] [bib_ref] Twenty million years of evolution: The embryogenesis of four Caenorhabditis species are..., Memar [/bib_ref] [bib_ref] Investigating spatio-temporal cellular interactions in embryonic morphogenesis by 4D nucleus tracking and..., Guan [/bib_ref]. Furthermore, cell lineage data obtained by the above-described method have been used to establish statistical references for C. elegans embryogenesis to enable cellular-resolution systematic characterization of developmental properties and to build visualization software, such as Dev-scape, WormGUIDES, and STAR [bib_ref] WormGUIDES: An interactive single cell developmental atlas and tool for collaborative multidimensional..., Santella [/bib_ref] [bib_ref] Systematic quantification of developmental phenotypes at single-cell resolution during embryogenesis, Moore [/bib_ref]. The above-mentioned analytical tools are summarized in [fig_ref] Table 1: Analytical software developed for studies of C [/fig_ref].
## Cell morphology reconstruction
Cell morphology is the 3D space that is surrounded by a cell membrane, and is defined by multiple parameters, including cell shape, volume, surface area and relationship to and area of contact with other cells. Cell morphology is a result of both intracellular and intercellular forces, and experimental and theoretical studies have shown that it reveals the mechanical state of a (multi)cellular system and plays a pivotal role in many critical biological processes during metazoan development [bib_ref] Contact area-dependent cell communication and the morphological invariance of ascidian embryogenesis, Guignard [/bib_ref] [bib_ref] Spatiotemporal dynamics of single cell stiffness in the early developing ascidian chordate..., Fujii [/bib_ref] [bib_ref] Periodic oscillations of Myosin-II mechanically proofread cell-cell connections to ensure robust formation..., Zhang [/bib_ref] [bib_ref] Self-organized patterning of cell morphology via mechanosensitive feedback, Dye [/bib_ref] [bib_ref] A fluid-to-solid jamming transition underlies vertebrate body axis elongation, Mongera [/bib_ref] [bib_ref] Rigidity percolation uncovers a structural basis for embryonic tissue phase transitions, Petridou [/bib_ref] [bib_ref] On mechanics and morphology, Klopper [/bib_ref] [bib_ref] Mechanics of development, Goodwin [/bib_ref]. During C. elegans embryogenesis, cell morphology is involved in many functions. First, cell shape is associated with cell fate and cell lineage. For example, AB and P1 are the first two cells derived from P0, and the descendants of AB are more spherical than those of P1. Forinstance, the gut cells, a sublineage derived from P1, are roughly cuboid and are packed in a regular array after gastrulation [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref] [bib_ref] Asymmetric organelle positioning during epithelial polarization of C. elegans intestinal cells, Brandt [/bib_ref]. Moreover, the severely deformed cell shape that comprises filopodia and lamellipodia structures can mediate the movement of cells, such as ABpl during the left-right asymmetric chiral morphogenesis from the 7-to 8-cell stages [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref] [bib_ref] Chiral forces organize left-right patterning in C. elegans by uncoupling midline and..., Pohl [/bib_ref] [bib_ref] Left-right patterning in the C. elegans embryo: Unique mechanisms and common principles, Pohl [/bib_ref]. Second, cell volume, which is determined by the symmetric or asymmetric division of a mother cell, influences the nucleus-to-cytoplasm ratio and the molecular contents allocated to the daughter cell, which regulates this cell's fate, cycle length and the robustness of its arrangement [bib_ref] Power law relationship between cell cycle duration and cell volume in the..., Arata [/bib_ref] [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref] [bib_ref] Anti-correlation of cell volumes and cellcycle times during the embryogenesis of a..., Fickentscher [/bib_ref] [bib_ref] Volume segregation programming in a nematode's early embryogenesis, Guan [/bib_ref]. Third, cell-cell contact serves as a physical basis for signal transduction between specific cell pairs that regulates the signal-receiving cell's orientation during division (e.g., Wnt signaling from C to ABar at the 8-cell stage) and fate specification (e.g., Notch signaling from MS to ABalp and ABara instead of to ABala and ABarp at the 12-cell stage) [bib_ref] The maternal genes apx-1 and glp-1 and establishment of dorsal-ventral polarity in..., Mello [/bib_ref] [bib_ref] glp-1 and inductions establishing embryonic axes in C. elegans, Hutter [/bib_ref] [bib_ref] Two maternal genes, apx-1 and pie-1, are required to distinguish the fates..., Mango [/bib_ref] [bib_ref] Multiple Wnt signaling pathways converge to orient the mitotic spindle in early..., Walston [/bib_ref]. Moreover, cell-cell contact geometry coordinates the orientation of cells during division by directing cortical myosin flows via intercellular mechanical force [bib_ref] Cell contacts orient some cell division axes in the Caenorhabditis elegans embryo, Goldstein [/bib_ref] [bib_ref] Combinatorial contact cues specify cell division orientation by directing cortical myosin flows, Sugioka [/bib_ref] [bib_ref] In vitro reconstitution of spatial cell contact patterns with isolated Caenorhabditis elegans..., Hsu [/bib_ref]. These findings highlight the emerging demand for a reliable cell-level-accurate 3D time-lapse morphological atlas of C. elegans embryos. Given that all cell nuclei can be fluorescently labeled and automatically traced in a C. elegans embryo, this approach has served as a convenient way to obtain data from which to infer the morphological properties of cells (e.g., their relationships to neighboring cells) in wild-type and mutant C. elegans embryos and those of other nematodes. The morphometric inference was first used 25 years ago when the whole-embryo cell lineage, cell division, and cell position during C. elegans embryogenesis were documented quantitatively, where the cell-cell contacts could be estimated from nucleus-based cell positions and raw images [bib_ref] Assessing normal embryogenesis in Caenorhabditis elegans using a 4D microscope: Variability of..., Schnabel [/bib_ref]. [bib_ref] Structural complexity of early embryos: A study on the nematode Caenorhabditis elegans, Bignone [/bib_ref] used these data and distance geometry methods (which had been established for studying protein-folding problems) to systematically predict a cell-cell contact map of C. elegans embryo based on the distance between cells defined by nucleus positions [bib_ref] Structural complexity of early embryos: A study on the nematode Caenorhabditis elegans, Bignone [/bib_ref]. More commonly, 3D cell centers represented by cell nuclei are transformed into cell morphologies via the Delaunay triangulation and Voronoi tessellation approaches. First, each point in a continuous space is assigned to its closest cell (center); second, every cell is allocated a convex polyhedral region, i.e., forming a Voronoi diagram [bib_ref] Nouvelles applications des paramètres continus à la théorie des formes quadratiques. Premier..., Voronoi [/bib_ref] [bib_ref] Voronoi diagrams and Delaunay triangulations: Ubiquitous siamese twins, Liebling [/bib_ref]. Voronoi diagrams have been widely used in combination with 3D time-lapse cell-nucleus position data of C. elegans embryos for characterizing and visualizing embryonic morphology, which has greatly facilitated studies in many areas of biology [bib_ref] Topological coding and visualization grammar of the development of C. elegans, Krämer [/bib_ref] [bib_ref] Development and validation of computational models of cellular interaction, Smallwood [/bib_ref] [bib_ref] Spatio-temporal reference model of Caenorhabditis elegans embryogenesis with cell contact maps, Hench [/bib_ref] [bib_ref] Quantitative semi-automated analysis of morphogenesis with single-cell resolution in complex embryos, Giurumescu [/bib_ref] [bib_ref] The Homeobox genes of Caenorhabditis elegans and insights into their spatiotemporal expression..., Hench [/bib_ref] [bib_ref] Compensatory cell movements confer robustness to mechanical deformation during embryonic development, Jelier [/bib_ref] [bib_ref] Cell neighbor determination in the metazoan embryo system, Wang [/bib_ref] [bib_ref] Establishment of signaling interactions with cellular resolution for every cell cycle of..., Chen [/bib_ref] [bib_ref] Systems properties and spatiotemporal regulation of cell position variability during embryogenesis, Li [/bib_ref] [bib_ref] Hierarchical deep reinforcement learning reveals a modular mechanism of cell movement, Wang [/bib_ref]. Moreover, this implementation has been used to delineate several key mechanisms in C. elegans embryogenesis that are supported by experimental evidence, such as Notch signaling and the positional variability mediated by physical contact, cell adhesion, and gap junctions [bib_ref] Establishment of signaling interactions with cellular resolution for every cell cycle of..., Chen [/bib_ref] [bib_ref] Systems properties and spatiotemporal regulation of cell position variability during embryogenesis, Li [/bib_ref].
However, although the generation of a Voronoi diagram is an efficient way to process and depict temporal series of 3D images that contain hundreds of cell nuclei, it is an indirect inference of cell surface and thus has inherent disadvantages [bib_ref] Limits of applicability of the Voronoi tessellation determined by centers of cell..., Kaliman [/bib_ref]. First, it identifies two cell nuclei dividing during cytokinesis as separate cells, even though they remain within the membrane of their mother cell. Second, the nucleus inside a cell can exhibit drastic movement without the body or shape of the cell being obviously affected, especially during the mitotic process [bib_ref] Spindle oscillations during asymmetric cell division require a threshold number of active..., Pecreaux [/bib_ref] [bib_ref] Cell biology of the Caenorhabditis elegans nucleus, Cohen-Fix [/bib_ref]. Third, a Voronoi diagram partitions space into convex polyhedra with planar surfaces, whereas a deformed cell may have severely curved and irregular surfaces [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref] [bib_ref] Chiral forces organize left-right patterning in C. elegans by uncoupling midline and..., Pohl [/bib_ref] [bib_ref] Left-right patterning in the C. elegans embryo: Unique mechanisms and common principles, Pohl [/bib_ref]. Fourth, the partitioned polyhedra are packed tightly against those representing neighboring cells, which means that an authentic anuclear cavity, such as a blastocoel, cannot be established [bib_ref] Cell polarity and gastrulation in C. elegans, Nance [/bib_ref]. Thus, a Voronoi analysis usually requires a large sample size of cells or embryos to reach sufficiently high statistical reliability and, when applied to a cell-specific prob-lem, must be checked by comparison with direct observations of membrane morphology to avoid the generation of false-positive or false-negative results [bib_ref] Establishment of signaling interactions with cellular resolution for every cell cycle of..., Chen [/bib_ref].
A few studies have extracted cell morphology based on empirical data on cell membranes acquired by general image segmentation software or manual annotation [bib_ref] Power law relationship between cell cycle duration and cell volume in the..., Arata [/bib_ref] [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref] [bib_ref] Anti-correlation of cell volumes and cellcycle times during the embryogenesis of a..., Fickentscher [/bib_ref] [bib_ref] Compensatory cell movements confer robustness to mechanical deformation during embryonic development, Jelier [/bib_ref] [bib_ref] Internalization of multiple cells during C. elegans gastrulation depends on common cytoskeletal..., Harrell [/bib_ref]. However, although C. elegans cell membranes can be tagged with a fluorescent marker to enable cellular boundaries to be determined, existing image-segmentation methods are difficult to apply to the labeled C. elegans embryos due to their relatively small size. Therefore, there is an urgent need for an automatic, reliable, and efficient segmentation tool for studies of C. elegans cellular morphology. Accordingly, in the past 5 years, many researchers have developed experimental methods to label cell membranes with fluorescent proteins and designed new segmentation algorithms customized for processing 3D time-lapse series of images of C. elegans embryogenesis [fig_ref] Table 2: Comparison of algorithms used for reconstructing cell morphology during C [/fig_ref] Optimization-based cell Membrane Segmentation (BCOMS), in which biological knowledge is included as a constraint in the segmentation process, i.e., it is assumed that the embryo size is nearly invariant and the nucleus is surrounded by a cell membrane [bib_ref] Biologically constrained optimization based cell membrane segmentation in C. elegans embryos, Azuma [/bib_ref]. In the evaluation of 25 shape features, this method exhibited an average deviation of 5.6 % from the results generated by manually annotating a 24-cell-stage embryo. In 2019, Cao et al. integrated several existing algorithms to preprocess raw images before watershed segmentation and devised a computational pipeline named 3D Membrane Morphological Segmentation (3DMMS) [bib_ref] 3DMMS: Robust 3D membrane morphological segmentation of C. elegans embryo, Cao [/bib_ref]. For C. elegans embryogenesis up to the 86-cell stage, the dice ratio (i.e., the overlap between segmentation output and ground truth) of 3DMMS (0.97) is greater than that of BCOMS. In another vein, Thiels et al. designed the pipeline spheresDT/Mpacts-PiCS and showed that a mechanical model considering intracellular and intercellular forces can substantially refine the segmented cell morphologies [bib_ref] Cell tracking and shape retrieval in membrane-labeled embryos, Thiels [/bib_ref] [bib_ref] Analysis of initial cell spreading using mechanistic contact formulations for a deformable..., Odenthal [/bib_ref] [bib_ref] The effect of cortical elasticity and active tension on cell adhesion mechanics, Smeets [/bib_ref]. It's a remarkable finding that cell mechanics-which are difficult to directly observe or measure in vivo-can be inferred from a combination of an appropriate mechanical model and 3D cell-membrane images [bib_ref] Cell tracking and shape retrieval in membrane-labeled embryos, Thiels [/bib_ref] [bib_ref] Cell-to-cell heterogeneity in cortical tension specifies curvature of contact surfaces in Caenorhabditis..., Fujita [/bib_ref] [bib_ref] A scheme for 3-dimensional morphological reconstruction and force inference in the early..., Xu [/bib_ref].
CShaper is a novel segmentation algorithm based on deep learning that takes advantage of a new transgenic strain of C. elegans that ubiquitously expresses GFP in nuclei and the red fluorophore mCherry in cell membranes during embryogenesis [fig_ref] Figure 1: C [/fig_ref] [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. CShaper performs automated segmentation of cell membranes to reconstruct cell morphology during C. elegans embryogenesis. Thus, in combination with an automated lineaging algorithm, CShaper enables the automated generation of cellular morphologies and the resolution of cell identities from the 4-to 350-cell stages at an interval of 1.5 min [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. Evaluations of dice ratios, Hausdorff distances (i.e., the largest distance from a voxel in the segmented region to its closest voxel in the ground truth region), and F 1 values (i.e., the proportion of correctly segmented cells) have demonstrated that CShaper outperforms five existing segmentation algorithms (3DUNet, CellProfiler, FusionNet, RACE, and SingleCellDetector) when applied to the C. elegans embryo [bib_ref] CNN-based preprocessing to optimize watershed-based cell segmentation in 3D confocal microscopy images, Eschweiler [/bib_ref] [bib_ref] CellProfiler 3.0: Next-generation image processing for biology, Mcquin [/bib_ref] [bib_ref] A deep fully residual convolutional neural network for image segmentation in connectomics, Quan [/bib_ref] [bib_ref] Real-time three-dimensional cell segmentation in large-scale microscopy data of developing embryos, Stegmaier [/bib_ref] [bib_ref] Learn to segment single cells with deep distance estimator and deep cell..., Wang [/bib_ref]. This increased performance of CShaper is due to its deep learning architecture, the large sample size it uses for training, and its customization to C. elegans embryogenesis. For example, for a specific time point, over 95 % of cells are segmented before the 200-cell stage, although the loss ratio increases to 18 % during the 200-to 350cell stages. In addition, CShaper can reconstruct the cavities inside an embryo, i.e., the non-nucleated empty space, from tracked cell-nucleus positions. Thus, 17 wild-type embryos have been imaged, segmented, and linearly normalized to form a statistical reference consisting of quantitative morphological information, such as cell volumes, cell surface areas, cell-cell contact relationships, and areas, which has been publicly documented as a resource. Each embryo was imaged and segmented until the complete divisions of AB128 (which reach the ninth generation of the AB lineage and yield 256 progeny), MS16 (which reach the sixth generation of the MS lineage and yield 32 progeny), E8 (which reach the fifth generation of the E lineage and yield 16 progeny), C8 (which reach the fifth generation of the C lineage and yield 16 progeny), D4 (which reach the fourth generation of the D lineage and yield 8 progeny), and germline precursor cell P4 (which yields 2 progeny) [fig_ref] Figure 1: C [/fig_ref]. As a result, a list of cells that each have 17 morphological trajectories were generated, with the smallest cell having an average volume of 8.5 lm 3 . The reference validates 10 cell-cell contacts that have been proposed to be involved in signaling, quantifies the variability of cell size (volume and surface area) and cell-cell contacts (relationship, area, and duration), and shows that cell irregularity is lineage-dependent and likely coupled with cell motility.
Currently, membrane-based cell morphology remains unclear beyond the 350-cell stage in the C. elegans embryo, where many morphogenetic events (e.g., ventral cleft closure, dorsal intercalation, epidermal enclosure, elongation, and most apoptosis) occur. The methodology of cell tracing and lineaging for late developmental stages has been established in nematode species because the fluorescently labeled nucleus signal is localized and thus there is a high signal-noise ratio [bib_ref] A novel cell nuclei segmentation method for 3D C. elegans embryonic time-lapse..., Chen [/bib_ref] [bib_ref] A high-fidelity cell lineage tracing method for obtaining systematic spatiotemporal gene expression..., Mace [/bib_ref] [bib_ref] Comparative proteome analysis between C . briggsae embryos and larvae reveals a..., An [/bib_ref]. However, beyond the 350-cell stage, the performance of membrane-based image segmentation decreases drastically over time, irrespective of the algorithm used. Consequently, there are four major challenges in the reconstruction of cell morphology: (1) the cell volume becomes exponentially smaller but the image resolution of microscopy remains constant over embryo development; (2) scattering and loss of laser intensity through the embryo body decrease image quality, especially in images of the last focal plane to be collected;
(3) resolution of the z-axis (in the axial direction) is much lower than that of the x-and y-axes (in the lateral directions), resulting in blurred images for the cell boundary that is parallel to a focal plane; and (4) the embryo boundary is formed by only a single layer of the fluorescently labeled cell membrane, whereas the cell-cell interfaces inside an embryo consist of two layers of membrane, so the membranes of cells located on an embryo surface are more difficult to be correctly segmented (due to reduced fluorescence intensity) than those of cells inside an embryo. In principle, higher-quality images could be obtained by increasing laser intensity; however, this increases phototoxicity, which may perturb normal development, and increases the photobleaching of fluorescent proteins. Finally, compared with other organisms' embryonic cells, the relatively smaller size of C. elegans embryonic cells presents a significant challenge to both fluorescence imaging and image segmentation. Technology improvements in all of the above-mentioned aspects are necessary for enabling the reliable reconstruction of 3D time-lapse cell morphology in the late embryonic development of C. elegans.
In terms of algorithmic improvements, one possible strategy could be to consolidate the advantages of various segmentation algorithms, such as those involving deep learning, biological constraints, and mechanical modeling [fig_ref] Table 2: Comparison of algorithms used for reconstructing cell morphology during C [/fig_ref]. In terms of imaging improvements, the performance of cell segmentation would be enhanced in a C. elegans strain that expressed a more uniform and brighter membrane tag across developmental stages and cell types than current strains. In addition, compared with confocal microscopy, Light Sheet Fluorescence Microscopy (LSFM) has a faster imaging speed and causes less phototoxicity and photobleaching, while Structured Illumination Microscopy (SIM) has a resolution approaching the diffraction limit [bib_ref] Optical sectioning deep inside live embryos by selective plane illumination microscopy, Huisken [/bib_ref] [bib_ref] Light sheet fluorescence microscopy: A review, Santi [/bib_ref] [bib_ref] Faster, sharper, and deeper: Structured illumination microscopy for biological imaging, Wu [/bib_ref] [bib_ref] Recent advances in structured illumination microscopy, Ma [/bib_ref]. Thus, to further improve the performance of segmentation, image acquisition should be explored using LSFM-based and SIM-based imaging systems that have been developed and adapted for C. elegans embryogenesis, such as inverted selective plane illumination microscopy, dual-view inverted selective plane illumination microscopy, and instant structured illumination microscopy [bib_ref] Inverted selective plane illumination microscopy (iSPIM) enables coupled cell identity lineaging and..., Wu [/bib_ref] [bib_ref] Untwisting the Caenorhabditis elegans embryo, Christensen [/bib_ref] [bib_ref] Spatially isotropic four-dimensional imaging with dual-view plane illumination microscopy, Wu [/bib_ref] [bib_ref] Dual-view plane illumination microscopy for rapid and spatially isotropic imaging, Kumar [/bib_ref] [bib_ref] Using stage-and slit-scanning to improve contrast and optical sectioning in dualview inverted..., Kumar [/bib_ref] [bib_ref] Isotropic light-sheet microscopy and automated cell lineage analyses to catalogue Caenorhabditis elegans..., Duncan [/bib_ref] [bib_ref] An in toto approach to dissecting cellular interactions in complex tissues, Shah [/bib_ref] [bib_ref] Small RNAs couple embryonic developmental programs to gut microbes, Ohno [/bib_ref].
## Comparison with other animal models
Cell tracing and lineaging and cell morphology reconstruction have also been performed in organisms other than C. elegans, such as two other invertebrates-the ascidian Phallusia mammillata and the fruit fly Drosophila melanogaster-and a vertebrate, the zebrafish (Danio rerio) [fig_ref] Table 3: Developmental characteristics and technical limits of cell tracing and lineaging and cell... [/fig_ref]. Like that of C. elegans, the embryonic development of P. mammillata involves an invariant cell lineage and cell-level stereotypic spatial patterns, i.e., each cell can be uniquely identified in different embryos based on its reproducible developmental behaviors, such as fate and position [bib_ref] Contact area-dependent cell communication and the morphological invariance of ascidian embryogenesis, Guignard [/bib_ref] [bib_ref] Eutely, cell lineage, and fate within the ascidian larval nervous system: determinacy..., Meinertzhagen [/bib_ref]. The P. mammillata cell lineage was traced up to its tailbud stage (the 850-cell stage) based on nucleus recognition using the software Mov-IT [bib_ref] A workflow to process 3D+time microscopy images of developing organisms and reconstruct..., Faure [/bib_ref]. In addition, P. mammillata cell-membrane segmentation was achieved up to the neurula stage (the 750-cell stage) using images obtained with multiview light-sheet microscopy [bib_ref] Contact area-dependent cell communication and the morphological invariance of ascidian embryogenesis, Guignard [/bib_ref] , which is around twice the number of cells for which this has been achieved in C. elegans. However, (1) the volume of the P. mammillata embryo is 60-70 times that of C. elegans, meaning that the smallest P. mammillata cell that can be segmented is much larger than that of C. elegans; and (2) rigorous manual curation must be incorporated into the P. mammillata cell membrane-segmentation procedure. Cell morphology reconstruction was also attempted in D. melanogaster. Compared with C. ele-gans and P. mammillata, D. melanogaster has a much larger embryo and its cell lineage and spatial patterns are not invariant at the cellular level. After fertilization, D. melanogaster undergoes 13 rounds of pseudo-synchronous divisions to form a syncytium, followed by blastoderm cellularization involving 6,000 cells (Stage 5), and subsequently gastrulation onset (Stage 6). Wholeembryo cell segmentation has been achieved up to D. melanogaster gastrulation (Stage 6), and nucleus-based cell tracing has been attempted up to Stage 11, when the germ band is subdivided into metameric units by parasegmental furrows [bib_ref] Real-time three-dimensional cell segmentation in large-scale microscopy data of developing embryos, Stegmaier [/bib_ref] [bib_ref] Fast, accurate reconstruction of cell lineages from large-scale fluorescence microscopy data, Amat [/bib_ref] [bib_ref] Quantitative 4D analyses of epithelial folding during Drosophila gastrulation, Khan [/bib_ref]. However, cell identity is only partially determinable in a D. melanogaster embryo as cell tracing and lineaging are not achievable in the first several rounds of cell cycles post-fertilization, as during this period the cells are sparsely located inside the embryo instead of on the periphery of the embryo and are thus difficult to distinguish. Automated reconstruction of cell morphology has also been attempted in D. rerio. The software RACE has been developed to extract cell morphology from D. rerio embryos in the gastrula period (the 5,500-cell stage, 6 h after fertilization), and Mov-IT has been developed to perform nucleus tracing up to the segmentation period (the 9,500-cell stage, 11 h after fertilization) [bib_ref] Real-time three-dimensional cell segmentation in large-scale microscopy data of developing embryos, Stegmaier [/bib_ref] [bib_ref] A workflow to process 3D+time microscopy images of developing organisms and reconstruct..., Faure [/bib_ref] [bib_ref] Stages of embryonic development of the zebrafish, Kimmel [/bib_ref].
Although automated cell tracing and lineaging and cell morphology reconstruction are feasible in all four animal models, C. elegans embryogenesis has the most condensed developmental trajectories out of all organisms. For instance, C. elegans gastrulation begins at the 26-cell stage, while P. mammillata gastrulation begins at the 112-cell stage, and both D. melanogaster and D. rerio gastrulations begin after thousands of cells are formed [bib_ref] Cell polarity and gastrulation in C. elegans, Nance [/bib_ref] [bib_ref] A workflow to process 3D+time microscopy images of developing organisms and reconstruct..., Faure [/bib_ref] [bib_ref] Cell-cycle control in oocytes and during early embryonic cleavage cycles in ascidians, Mcdougall [/bib_ref]. Moreover, C. elegans embryogenesis ends with 550 living cells, the identity of each of these can be resolved by nucleus-based cell tracing and lineaging, and the fate of each of these has been documented [bib_ref] The embryonic cell lineage of the nematode Caenorhabditis elegans, Sulston [/bib_ref] [bib_ref] Automated cell lineage tracing in Caenorhabditis elegans, Bao [/bib_ref] [bib_ref] The lineaging of fluorescentlylabeled Caenorhabditis elegans embryos with StarryNite and AceTree, Murray [/bib_ref] [bib_ref] AceTree: A tool for visual analysis of Caenorhabditis elegans embryogenesis, Boyle [/bib_ref] ; this can be advantageous for systematically studying developmental events at the cellular level. Thus, researchers studying C. elegans to answer various ques-tions in metazoan embryogenesis may benefit by focusing on only a single cell or all cells in an embryo.
## Physical models
In this section, we review the application of physical models for elucidating two biological phenomena, i.e., cell polarization and morphogenesis, and understanding the underlying genetic programs selected by evolution.
## Reaction-diffusion modeling of cell polarization
Cell polarization is responsible for many significant biological processes in the zygote, in germline precursor cells, and in their derived somatic cells during C. elegans embryogenesis [bib_ref] Physically asymmetric division of the C. elegans zygote ensures invariably successful embryogenesis, Jankele [/bib_ref]. Cell polarization in P0-P3 occurs sequentially, and many of its associated regulatory pathways have been revealed by experiments. Thus, the cell polarization in those cells has served as the basis for theoretical studies in recent years.
## Known molecular mechanisms
Cell polarization plays a decisive role in the segregation of cell contents, the orientation of cell division, and the specification of cell fate and has therefore been a focus in developmental biology for many years (reviewed in [bib_ref] The PAR proteins: Fundamental players in animal cell polarization, Goldstein [/bib_ref]. Four consecutive rounds of asymmetric divisions governed by cell polarization occur in P0, P1, P2, and P3 during C. elegans embryogenesis, thereby providing a framework for research on cell polarization [fig_ref] Figure 2: Polarization of zygote and germline precursor cells and specification of somatic founder... [/fig_ref] [bib_ref] Cytoplasmic localization and asymmetric division in the early embryo of Caenorhabditis elegans, Griffin [/bib_ref]. In each asymmetric division, a new somatic founder cell is generated, while the other daughter cell-the next P cell-remains multipotent. After that, the last P cell, P4, undergoes symmetric division and yields two primordial germ cells, Z2 and Z3, due to its too small size and loss of cell polarity [bib_ref] A cell-size threshold limits cell polarity and asymmetric division potential, Hubatsch [/bib_ref]. Cell polarity is first estab- [bib_ref] Real-time three-dimensional cell segmentation in large-scale microscopy data of developing embryos, Stegmaier [/bib_ref] Note: The first and second columns of this table are adapted from [bib_ref] The maternal-to-zygotic transition: A play in two acts, Tadros [/bib_ref] [bib_ref] The maternal-to-zygotic transition revisited, Vastenhouw [/bib_ref] , except those for P. mammillata [bib_ref] Cell-cycle control in oocytes and during early embryonic cleavage cycles in ascidians, Mcdougall [/bib_ref] , and the second column depicts the embryonic stages when zygotic transcription is essential. The smallest cell volume for cell morphology reconstruction represents the smallest cell reported to have been segmented. The value listed for D. melanogaster is obtained from the segmentation outputs of RACE [bib_ref] Real-time three-dimensional cell segmentation in large-scale microscopy data of developing embryos, Stegmaier [/bib_ref] , and this value is assumed to apply to D. rerio as well, considering that the same microscopy method and segmentation algorithm was used for both species.
lished in P0 and is triggered by the entry of sperm, with the entry site determining the posterior of the embryo [fig_ref] Figure 2: Polarization of zygote and germline precursor cells and specification of somatic founder... [/fig_ref] [bib_ref] Specification of the anteroposterior axis in Caenorhabditis elegans, Goldstein [/bib_ref]. Two groups of membrane-bound partitioning-defective (PAR) proteins generate cell polarity by mutually inhibiting their membrane asso-ciation and thus localize on opposite sides of a cell [fig_ref] Figure 2: Polarization of zygote and germline precursor cells and specification of somatic founder... [/fig_ref]. At the 1-cell stage, the embryo is embedded in an ellipsoidal eggshell whose major axis becomes the anterior-posterior (a-p) axis of the embryo by polarization, and the PAR proteins in the a and p regions are therefore named aPAR (PAR-3, PAR-6, and PKC-3) and pPAR (PAR-1 and PAR-2) [bib_ref] Cell polarity: Getting the PARty started, Boxem [/bib_ref]. The localization of PAR proteins conveys positional information to their downstream molecules, such as MEX-5/MEX-6, PIE-1, and P granule, which are thus polarized and spatially separated. Additionally, a few molecules, such as actomyosin, CDC-42, CHIN-1, and possibly some downstream molecules (e.g., MEX-5/MEX-6), are involved in the interaction network and cortical activity that facilitate cell polarization [bib_ref] MEX-5 and MEX-6 function to establish soma/germline asymmetry in early C. elegans..., Schubert [/bib_ref] [bib_ref] Going with the flow: Insights from Caenorhabditis elegans zygote polarization, Gubieda [/bib_ref] [bib_ref] Polarization of the C. elegans zygote proceeds via distinct establishment and maintenance..., Cuenca [/bib_ref].
A key aim of systems biology is to integrate mathematical and physical approaches with biological knowledge to interpret how genetic circuits determine the dynamics and function of a network system, as this enables the delineation of biological mechanisms and establishes a sound foundation for synthetic biological studies [bib_ref] Construction of a genetic toggle switch in Escherichia coli, Gardner [/bib_ref] [bib_ref] A synthetic oscillatory network of transcriptional regulators, Elowitz [/bib_ref] [bib_ref] Defining network topologies that can achieve biochemical adaptation, Ma [/bib_ref] [bib_ref] Designing synthetic regulatory networks capable of self-organizing cell polarization, Chau [/bib_ref]. The determination that aPAR and pPAR shuttle between the membrane and the cytosol, diffuse, and interact with each other provides a good basis for such studies [bib_ref] PAR proteins diffuse freely across the anterior-posterior boundary in polarized C. elegans..., Goehring [/bib_ref]. Many auxiliary molecular interactions (motifs) and coupling mechanisms (such as advective transport in the cortex) can be exploited for characterizing the delicate control of cell-polarization patterning and its underlying design principles [bib_ref] Asymmetric cell division from a cell to cells: Shape, length, and location..., Seirin-Lee [/bib_ref].
## Critical role of the mutual inhibition of apar and ppar
The knowledge obtained from empirical studies of P0-P4 has allowed mathematical and physical modeling-based investigations of the theoretical principle of cell polarization in C. elegans. Tostevin et al. (2008) introduced one-dimensional reaction-diffusion equations to model the distributions of aPAR and pPAR in the cortex and cytosol of C. elegans zygote, based on which they suggested that these proteins' simple mutual inhibition is sufficient to generate their anterior-posterior domain separation [bib_ref] Modeling the establishment of PAR protein polarity in the one-cell C. elegans..., Tostevin [/bib_ref]. In a comprehensive computational study of protein network circuits capable of cell polarization, [bib_ref] Designing synthetic regulatory networks capable of self-organizing cell polarization, Chau [/bib_ref] showed that mutual inhibition is a core mechanism in such circuits [bib_ref] Designing synthetic regulatory networks capable of self-organizing cell polarization, Chau [/bib_ref]. Using a combination of computational and experimental techniques, demonstrated that the advective flow of actomyosin in the cortex of C. elegans zygote is a critical trigger and facilitator of the robust segregation of PAR proteins and established a mechanochemical scheme for cell polarization [bib_ref] Polarization of PAR proteins by advective triggering of a pattern-forming system, Goehring [/bib_ref] [bib_ref] Mechanochemical pattern formation in the polarization of the one-cell C. elegans embryo, Bois [/bib_ref]. In other research, [bib_ref] Actomyosin regulation and symmetry breaking in a model of polarization in the..., Kravtsova [/bib_ref] provided theoretical evidence that actomyosin contractility is required to stabilize the polarity pattern and that this mechanical cue can only break cell symmetry in the presence of certain chemical interactions [bib_ref] Actomyosin regulation and symmetry breaking in a model of polarization in the..., Kravtsova [/bib_ref]. Seirin- [bib_ref] Self-organization and advective transport in the cell polarity formation for asymmetric cell..., Seirin-Lee [/bib_ref] examined the morphometric features of the polarity distribution and found that polarity proteins' advective transport in the membrane and cytosol and their mass concentrations in these regions regulate the width of the posterior domain [bib_ref] Self-organization and advective transport in the cell polarity formation for asymmetric cell..., Seirin-Lee [/bib_ref]. Seirin-Lee et al. (2020) also mathematically and numerically explored the polarity distribution from upstream to downstream molecules (i.e., from PARs to MEX-5/MEX-6 to PIE-1) [bib_ref] Transitions to slow or fast diffusions provide a general property for in-phase..., Seirin-Lee [/bib_ref]. [bib_ref] Guiding self-organized pattern formation in cell polarity establishment, Gross [/bib_ref] probed how guiding cues and biochemical feedback affect C. elegans cell polarization and manifest a transition point where the system changes from a guide-dominated to a feedback-dominated state [bib_ref] Guiding self-organized pattern formation in cell polarity establishment, Gross [/bib_ref]. In addition to the explorations of the C. elegans 1-cell embryo, Lim et al. (2021) used computational modeling, fluorescent imaging, and genetic perturbation to scrutinize the differentiated C. elegans 2-cell embryo, which revealed that the balance of the most upstream PAR proteins independently determines whether the symmetric or asymmetric mode of division occurs [bib_ref] A balance between antagonizing PAR proteins specifies the pattern of asymmetric and..., Lim [/bib_ref].
The effect of geometric cues on cell polarization is also a popular research theme. [bib_ref] Cortical geometry may influence placement of interface between Par protein domains in..., Dawes [/bib_ref] used a reaction-diffusion model and an Allen-Cahn equation in two-dimensional space to elucidate that the cortical thickness distribution can determine the position of the interface between the exclusive domains of aPAR and pPAR [bib_ref] Cortical geometry may influence placement of interface between Par protein domains in..., Dawes [/bib_ref]. Aras et al. [fig_ref] Figure 1: C [/fig_ref] used a phase field model to simulate the actomyosin contraction dynamics during the first cell polarization and division in the C. elegans embryo; this revealed that the rigid eggshell of the embryo generates asymmetric cortical tension that restricts changes in cell shape and consequently affects cell morphology and the timing of polarity establishment [bib_ref] The importance of mechanical constraints for proper polarization and psuedo-cleavage furrow generation..., Aras [/bib_ref]. Another study, Gebele et al. (2020), found that the local ratio of membrane-surface-to-cytosolic-volume, which is high at the two terminals of an embryo and low in the middle of an embryo, plays an important role in the formation of polarity by mediating protein shuffling between the membrane and the cytosol [bib_ref] Geometric cues stabilise long-axis polarisation of PAR protein patterns in C. elegans, Geßele [/bib_ref]. The cell size was also determined to be crucial for the establishment of polarity, due to the reaction-diffusion system describing the interplay between aPAR and pPAR not being linearly scalable; this explains why the asymmetric division doesn't take place in the last germline precursor cell, P4 [bib_ref] A cell-size threshold limits cell polarity and asymmetric division potential, Hubatsch [/bib_ref]. In theory, the location of the transition region between anterior and posterior domains is sensitive to changes in cell shape when a cell is relatively small, which implies that additional programmed regulation might be employed to ensure the precision and robustness of cell polarization [bib_ref] Long time behavior and stable patterns in highdimensional polarity models of asymmetric..., Morita [/bib_ref].
## Investigation of additional motifs and regulations
In addition to the mutual inhibition exhibited by PAR proteins, other proteins and their interactions and feedback have been extensively investigated using theoretical methods. [bib_ref] PAR-3 oligomerization may provide an actinindependent mechanism to maintain distinct par protein..., Dawes [/bib_ref] studied the self-association of aPAR, which is realized by PAR-3 oligomerization, and proved that it functions in establishing and stabilizing polarity regardless of cortical activity [bib_ref] PAR-3 oligomerization may provide an actinindependent mechanism to maintain distinct par protein..., Dawes [/bib_ref]. In another work, Seirin-Lee et al. (2020), minimal network analysis and eFAST sensitivity analysis revealed that the active regulation of aPAR by CDC-42 helps maintain the anterior and posterior polarity domains and polarization itself [bib_ref] CDC-42 interactions with Par proteins are critical for proper patterning in polarization, Seirin-Lee [/bib_ref]. Given the fact that the depletion of the downstream protein MEX-5/MEX-6 is harmful to the polarity of upstream PAR proteins in later germline precursor cells, Seirin-Lee (2021) hypothesized their inhibitive regulation on aPAR's association on membrane and active regulation on aPAR's dissociation from membrane; computation showed that both two assumed loops can reinforce PAR polarity [bib_ref] MEX-5 and MEX-6 function to establish soma/germline asymmetry in early C. elegans..., Schubert [/bib_ref] [bib_ref] Polarization of the C. elegans zygote proceeds via distinct establishment and maintenance..., Cuenca [/bib_ref] [bib_ref] The role of cytoplasmic MEX-5/6 polarity in asymmetric cell division, Seirin-Lee [/bib_ref]. Further research on cell polarization in P2, which has the opposite polarity to its ancestors P1 and P0 in vivo, showed that the abovementioned effect was due to the contact-based MES-1/SRC-1 signaling from P2 0 s sister cell, EMS, that enhances the on-rate of PAR-2 [bib_ref] Positioning of polarity formation by extracellular signaling during asymmetric cell division, Seirin-Lee [/bib_ref].
Overall, these above-described theoretical explorations have delineated crucial aspects of the diverse and precise control of cell polarization in C. elegans. These programs give rise to the stereotypical development of C. elegans embryos and are vital for optimal cell proliferation, morphogenesis, and differentiation [bib_ref] Volume segregation programming in a nematode's early embryogenesis, Guan [/bib_ref] [bib_ref] Physically asymmetric division of the C. elegans zygote ensures invariably successful embryogenesis, Jankele [/bib_ref] [bib_ref] Principles of PAR polarity in Caenorhabditis elegans embryos, Hoege [/bib_ref]. Other experimental observations, such as that the LGL-1 is localized in the posterior of C. elegans, can be included to form a complete interaction network for future studies [bib_ref] Principles of PAR polarity in Caenorhabditis elegans embryos, Hoege [/bib_ref].
## Mechanical modeling of cell positioning and cell morphology
Many mechanical models have been used to investigate embryonic morphogenesis in various species. As the early development of C. elegans is highly precise and repeatable in individual embryos, for the past two decades researchers have attempted to reproduce the progression of its cellular arrangement in silico.
## Multi-particle model
In 2003, Kajita et al. reported their physical model that describes the cell shape and the intercellular and intracellular mechanical interactions in the early embryo of C. elegans [fig_ref] Figure 3: Mechanical modeling of cell positioning and cell morphology in C [/fig_ref] [bib_ref] Computer simulation of the cellular arrangement using physical model in early cleavage..., Kajita [/bib_ref]. In this model, multiple particles with a certain mass are used to represent various subcellular structures; for example, (1) the cell membrane consists of particles that interact with each other via a spring and damper; (2) these membrane particles are assigned an inner pressure and bending force to control the cell size and cell shape, respectively; and (3) the membrane particles are supported by a centrosome, which serves as the cell center and can model cell division by separating into two distant and mutually repulsive centrosomes. This model was shown to reconstruct cell shape dynamics for 1-to 4-cell morphogenesis restricted by an ellipsoidal eggshell that were similar to real dynamics and to classify four different shapes of embryonic structures (i.e., diamond-, parallelogram-, square-, and T-shaped structures) as its mechanical parameters are varied. Further analysis revealed that the forces between centrosomes and between the membrane and the centrosome during cytokinesis are important for the generation of the diamond-shaped structure seen in vivo, while the maximum distance between centrosomes directly controls cell elongation and skewing and is therefore the primary factor determining cell patterning.
Kajita et al. noted key differences between the cell shapes generated by their model and those observed in experiments and thus developed a better model that simulates the rounding and stiffening during cell division [bib_ref] Physical modeling of the cellular arrangement in C. elegans early embryo: Effect..., Kajita [/bib_ref]. This is realized by reinforcing the repulsion between membrane particles and between the membrane and the centrosome and the bending force of the membrane, which reproduces the curved interface between AB and P1 at the end of the 2-cell stage (i.e., where AB is protruding toward P1). The researchers optimized the mechanical parameters by examining how frequently the model generates the diamond-shaped structure, which revealed that both cell rounding and cell stiffening affect the in silico pattern formation. However, tests demonstrated that cell stiffening is more important than cell rounding in the model. The advantage of the multi-particle model is its ability to describe subcellular components (e.g., the membrane, centrosome, and contractile ring) in sufficient detail to reproduce cell morphology. However, the approximately 20 parameters of the model that have biological significance are difficult to measure or fit experimentally, which limits further application of the model.
## Coarse-grained model
The coarse-grained model simplifies a cell by representing it as a single point, with the interaction (e.g., repulsion and attraction) between cells being pairwise and distance-dependent, so it incurs low computational costs [fig_ref] Figure 3: Mechanical modeling of cell positioning and cell morphology in C [/fig_ref]. As this model has minimal parameters it can reveal some fundamental rules of a mechanical system. In 2013, Fickentscher et al. adopted the coarse-grained model to determine whether cell arrangement patterns are primarily an outcome of mechanical interaction and relaxation [bib_ref] Mechanical cues in the early embryogenesis of Caenorhabditis elegans, Fickentscher [/bib_ref]. In their form of the model, the cells are treated as repellent elastic balls enveloped by an ellipsoidal eggshell and governed by Langevin equations with random noise, and cell division and cell adhesion are not considered. Surprisingly, the simulated close-packing configurations at the 4-, 8-, and 12-cell stages are within the natural range measured by imaging cell nucleus positions via lightsheet microscopy, indicating the role of mechanical cues in determining cell arrangement. Furthermore, they found that the formation of the planar diamond-shaped pattern at the 4-cell stage, in which the dorsal-ventral axis is established, is primarily controlled by repulsive force exerted by the eggshell. Fickentscher et al.
(2016) then incorporated cell division into their model and also asked ''how do individual cells know that dividing right now is a good choice for a meaningful and robust mechanically driven arrangement of all cells in the embryo?" [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref]. They concluded in this and another study that the experimentally observed anticorrelation between cell cycle length (T) and cell volume (V) is significant and lineage-specific [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref] [bib_ref] Anti-correlation of cell volumes and cellcycle times during the embryogenesis of a..., Fickentscher [/bib_ref]. After experimentally fitting the formulized T-V relationship, they found that a simulation of the cell arrangement at the 24-cell stage (i.e., prior to the onset of gastrulation) resembles the in vivo situation. Their simulations with a range of values of parameters suggest that programmed celldivision asymmetry provides sufficient time for cells to relax into their required positions after cell division. Moreover, as the number of cells increases and their size decreases as embryogenesis advances, the system becomes more sensitive to local perturbations and thus more prone to producing irrecoverable defects.
Tian et al. (2020) concentrated on the cell division sequence that is highly conserved between C. elegans embryos and other Caenorhabditis species and showed that both the interval of relaxation and the order of cell division events must be controlled for the correct cell arrangement pattern to be generated [bib_ref] Twenty million years of evolution: The embryogenesis of four Caenorhabditis species are..., Memar [/bib_ref] [bib_ref] Investigating spatio-temporal cellular interactions in embryonic morphogenesis by 4D nucleus tracking and..., Guan [/bib_ref] [bib_ref] Why and how the nematode's early embryogenesis can be precise and robust:..., Tian [/bib_ref]. In addition, they found that the synchrony of cell divisions is programmed in a particular manner to ensure that cells move collectively, so that cell arrangement patterns do not bifurcate. Moreover, they determined that the cell division orientation known to be regulated by cell-cell contact geometry and cell-cell signaling is important for generating the stereotypical cell arrangement pattern in vivo.
In summary, the above-described theoretical studies using coarse-grained models have highlighted that cell division is carefully programmed to ensure that there is a precise and robust progression of cell arrangement during C. elegans embryogenesis.
The coarse-grained model has also been applied to study specific biological phenomena. In 2017, Yamamoto et al. reported that the RNAi knockdown of the genes dpy-1 and lon-1 generated C. elegans eggshells with variable aspect ratios, in which at the 4-cell stage they identified a variety of cell arrangement patterns (i.e., diamond, pyramid, T-shaped, and linear patterns) [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref]. The distribution of cell arrangement patterns with respect to aspect ratio cannot be reproduced by a coarse-grained model that considers only cell-cell repulsion but can be reproduced by the improved coarse-grained model that considers cell-cell attraction in addition to cell-cell repulsion [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref] [bib_ref] Mechanical cues in the early embryogenesis of Caenorhabditis elegans, Fickentscher [/bib_ref] [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref]. Fluorescence imaging of the cell adhesive protein HMR-1 showed that it accumulates in significantly lower amounts at the interface between EMS and P2 than at the interface between other cells, and the improved model-which incorporates this in the form of an asymmetric adhesion map-captures the empirical distribution of cell arrangement patterns [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref]. Moreover, the RNAi-mediated elimination of cell adhesion generates a different distribution of cell arrangement patterns, confirming the role of a cell-cell adhesion map in determining the cell arrangement pattern. Another study revealed that microinjecting an actin inhibitor into the AB cell to interfere with its adhesion function effectively increased the level of HMR-1 accumulation at the EMS-P2 interface, suggesting that a cell-non-autonomous mechanism programs the distribution of cell-cell adhesion. Therefore, it appears that mechanical interaction (e.g., cell adhesion) and spatial confinement serve as key regulatory factors that specify the cell arrangement pattern in the C. elegans 4-cell embryo [bib_ref] Physical constraints on early blastomere packings, Giammona [/bib_ref].
Recently, Miao et al. (2022) applied a coarse-grained model that considers both intercellular repulsion and attraction and used experimentally measured cell-division order and volume segregation ratio as inputs to simulate cell division and motion up to the 330-cell stage of C. elegans [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. These studies have shown that even though the orientation of each cell division is set to be nearly tangential to the eggshell, cell internalization occurs spontaneously in all conditions and mimics the behavior of gastrulation, during which a cell originally in contact with the eggshell moves inward the embryo such that the cell becomes surrounded by other cells. The researchers have also found that the lateral compression and pressure between cells becomes increasingly stronger because of the reduction in cell size during proliferation, and that the internalization of a cell effectively decreases such stress and lowers the potential energy of the entire system. Moreover, they have used force analysis and numerical simulation to predict that bistable states exist during the 15-to 44-cell stages and that a larger cell has a stronger tendency to internalize than a smaller cell. In the real C. elegans embryo, the gut precursor cells E2 (Ea and Ep) become the largest of all cells in the embryo due to the initiation of tissue-specific zygotic transcription and consequent lengthened cell cycle and gastrulate during the 26-to 44-cell stages, which is very close to the theoretically predicted one [bib_ref] Timing of tissuespecific cell division requires a differential onset of zygotic transcription..., Wong [/bib_ref] [bib_ref] Cell polarity and gastrulation in C. elegans, Nance [/bib_ref]. Furthermore, systematic virtual experiments using the coarse-grained model showed how boundary geometry, cellular mechanical properties such as stiffness and adhesion, and cellspecific active regulation affect cell internalization.
Overall, studies of embryonic morphogenesis at the 4-cell and gastrulation stages have suggested that a simple coarse-grained model can recapitulate the fundamental features of the embryonic system of C. elegans, thereby enabling the design and optimization strategy of the genetic programs to be rationally interpreted.
In 2020, Guan et al. revised the coarse-grained model by optimizing the formulation and parameters and inputting experimentally measured cell volumes and eggshell shape to determine the latest developmental stage to which it can be applied with sufficient precision [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref] [bib_ref] Reconstructing the multicellular structure of a developing metazoan embryo with repulsion-attraction model..., Guan [/bib_ref]. They found that the model can reproduce cell positions up to the 26-cell stage (at which time they are assumed to be at mechanical equilibrium) with positional variation that is less than a cell's radius for all cells. Moreover, they determined that the model performs well even at the 51-cell stage, except for three cells that exhibit a deviation larger than their radii, suggesting that the system at this stage remains largely governed by mechanical cues but that additional modifications (e.g., inhomogeneous cell stiffness and cell adhesion) might have a nonnegligible effect on some cells. In addition, these researchers found that the cell-cell contact relationship conserved among embryos is poorly reproduced by the coarse-grained model, although the inclusion of a heterogeneous cell-cell adhesion map derived from cell-cell contact data somewhat improves the reconstruction performance of the model [bib_ref] Reconstructing the multicellular structure of a developing metazoan embryo with repulsion-attraction model..., Guan [/bib_ref].
The limitation of the coarse-grained model was also discussed by [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref] in light of its failure to reproduce the T-reverse-type pattern in an elongated eggshell with depleted cell adhesion [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref]. The discrepancy between coarse-grained simulations and experiments is not surprising, given that the model oversimplifies each cell (i.e., as a single point) and therefore does not include information on cell morphology. Models with higher complexity (e.g., the multi-particle model and the phase field model) are needed to answer questions on cell morphology and more complex cellular properties.
## Phase field model
The phase field model, which was originally developed in materials science to describe the concentration, interaction, and separation of multiple components, uses a diffusible scalar (phase) field to represent an object and track its interfacial interactions with other objects [fig_ref] Figure 3: Mechanical modeling of cell positioning and cell morphology in C [/fig_ref] [bib_ref] Thermodynamische theorie der kapillarität unter voraussetzung stetiger dichteänderung, Van Der Waals [/bib_ref] [bib_ref] Free energy of a nonuniform system. I. Interfacial free energy, Cahn [/bib_ref] [bib_ref] Phase-field models for microstructure evolution, Chen [/bib_ref] [bib_ref] Phase-field simulation of solidification, Boettinger [/bib_ref]. In the past two decades, this approach has been used in cell modeling, wherein a cell is treated as a 2D or 3D field and thus has interfacial interactions with other cells and with an external boundary, and the shape of a cell is programmable to determine its response to mechanical and biochemical stimuli [bib_ref] Phase-field model of cellular migration: Threedimensional simulations in fibrous networks, Moure [/bib_ref] [bib_ref] Plasticity of cell migration resulting from mechanochemical coupling, Cao [/bib_ref]. Recently, this approach has been applied to examine C. elegans embryonic morphogenesis.
In 2019, Jiang et al. reported a phase field model for simulating cell growth, cell division, cell-cell contact, and the mechanical moduli of a cell and modeled cell compaction and packing under soft constraints [bib_ref] A diffuse interface framework for modeling the evolution of multi-cell aggregates as..., Jiang [/bib_ref]. They tested the model in two dimensions and found that it generates cell morphologies and positions that are similar to those experimentally observed in the C. elegans embryo. Kuang et al. (2022) utilized the 3D time-lapse cell morphology data of the CShaper system to construct a more sophisticated 3D phase field model that takes into account cell surface tension, cell-eggshell and cell-cell repulsion, cell-cell attraction, and cell volume constriction [bib_ref] Computable early Caenorhabditis elegans embryo with a phase field model, Kuang [/bib_ref] [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. They showed that by inputting in vivo data of cell-division order, orientation, and volume segregation ratio into the model and performing step-by-step model construction and verification, the model accurately reconstructs cell morphology and the conserved cell-cell contact map [fig_ref] Figure 3: Mechanical modeling of cell positioning and cell morphology in C [/fig_ref]. Furthermore, they found that the cell adhesion program can be inferred without a priori knowledge by comparison of the simulated morphologies with the experimentally observed morphologies [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref] [bib_ref] Planar asymmetries in the C. elegans embryo emerge by differential retention of..., Dutta [/bib_ref]. The model is also able to characterize the in vivo system in space and time, such that it can account for the timing and orientation of cell division and the cell-cell attraction matrix programmed from the 6-to 8-cell stages. However, despite the outstanding accuracy of the phase field model, high computational time and memory costs are incurred during its computation of the partial differential equation and the dense grid, which hamper its further application. Accordingly, the phase-field framework was recently upgraded by resolution refinement, formula optimization, and numerical scheme improvement, which enables it to compute over 100 cells with considerable efficiency and accuracy.
The phase-field method has also been applied to solve the problem of T-reversed-type cell arrangement at the 4-cell stage, which was observed in a genetically perturbed egg with a high aspect ratio and weak/no cell adhesion, and which cannot be reproduced by a coarse-grained model [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref] [bib_ref] The extra-embryonic space and the local contour are crucial geometric constraints regulating..., Seirin-Lee [/bib_ref]. Seirin-Lee et al. (2022) used a 2D phase field model and real eggshell morphologies extracted from experimental images to reproduce the T-reversed-type pattern under a condition with experimental parameters [bib_ref] The extra-embryonic space and the local contour are crucial geometric constraints regulating..., Seirin-Lee [/bib_ref]. They assigned an ellipsoid and a capsule eggshell with the same aspect ratio in the simulation and found that the capsule eggshell, which is more similar than the ellipsoid eggshell to the real eggshell in a T-reversed-type embryo, leaves more extra-embryonic space for cell movement. This was verified by experimental data, establishing that this extra-embryonic space is another geometric factor that modulates the cell arrangement in concert with mechanical parameters, such as cell stiffness and cell adhesion.
In summary, the quantitative morphology data obtained from the above-described studies have effectively guided the development of comprehensive mechanical models that serve as ground truth. This has deepened our understanding of the in vivo developmental system of C. elegans and enhanced our ability to predict its behavior.
## Summary and outlook
Rapid technological advancements in live-cell imaging allow rapid acquisition of high-resolution images of molecular and cellular events throughout development. Thus, what scientific questions can or should we ask and what roles can computational modeling play in this regard? Developmental biologists have delineated the developmental patterns of C. elegans and uncovered details of its developmental processes and regulatory mechanisms by direct observation and/or genetic perturbation experiments. These explorations have mainly been conducted using genetic approaches, such as the identification of stepwise Wnt and Notch signaling events between specific interacting cell pairs [bib_ref] Wnt signaling polarizes an early C. elegans blastomere to distinguish endoderm from..., Thorpe [/bib_ref] [bib_ref] Wnt signaling and an APC-related gene specify endoderm in early C. elegans..., Rocheleau [/bib_ref] [bib_ref] Quantitative differences in nuclear b-catenin and TCF pattern embryonic cells in C...., Zacharias [/bib_ref] [bib_ref] The maternal genes apx-1 and glp-1 and establishment of dorsal-ventral polarity in..., Mello [/bib_ref] [bib_ref] glp-1 and inductions establishing embryonic axes in C. elegans, Hutter [/bib_ref] [bib_ref] Two maternal genes, apx-1 and pie-1, are required to distinguish the fates..., Mango [/bib_ref] [bib_ref] lin-12 and glp-1 are required zygotically for early embryonic cellular interactions and..., Moskowitz [/bib_ref] [bib_ref] Multiple Wnt signaling pathways converge to orient the mitotic spindle in early..., Walston [/bib_ref]. These have been followed by more systems-level studies, aided by 4D live-cell imaging data, which have taken into account all of the cells derived from different lineages with different fates and have accurately measured the positioning and division timing and orientation of each cell throughout embryogenesis [bib_ref] Assessing normal embryogenesis in Caenorhabditis elegans using a 4D microscope: Variability of..., Schnabel [/bib_ref] [bib_ref] Automated cell lineage tracing in Caenorhabditis elegans, Bao [/bib_ref] [bib_ref] The lineaging of fluorescentlylabeled Caenorhabditis elegans embryos with StarryNite and AceTree, Murray [/bib_ref] [bib_ref] AceTree: A tool for visual analysis of Caenorhabditis elegans embryogenesis, Boyle [/bib_ref]. This quantitative single-cellresolution 4D data and the stereotypical procedures of C. elegans embryogenesis have drawn the attention of mathematicians and physicists, who have attempted to provide theoretical explanations for various observations, such as why the developmental program has the design. For instance, in modeling cell positioning, physical simulations show why the timing and orientation of cell division and cell-cell adhesion are programmed as they are in vivo [bib_ref] Computable early Caenorhabditis elegans embryo with a phase field model, Kuang [/bib_ref] [bib_ref] Setting the clock for fail-safe early embryogenesis, Fickentscher [/bib_ref] [bib_ref] Why and how the nematode's early embryogenesis can be precise and robust:..., Tian [/bib_ref] [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref]. In mechanical terms, this programming ensures the precise and robust realization of the developmental path of embryonic morphology [bib_ref] Wnt signaling polarizes an early C. elegans blastomere to distinguish endoderm from..., Thorpe [/bib_ref] [bib_ref] Wnt signaling and an APC-related gene specify endoderm in early C. elegans..., Rocheleau [/bib_ref] [bib_ref] Power law relationship between cell cycle duration and cell volume in the..., Arata [/bib_ref] [bib_ref] Anti-correlation of cell volumes and cellcycle times during the embryogenesis of a..., Fickentscher [/bib_ref] [bib_ref] Volume segregation programming in a nematode's early embryogenesis, Guan [/bib_ref] [bib_ref] Multiple Wnt signaling pathways converge to orient the mitotic spindle in early..., Walston [/bib_ref] [bib_ref] Cell contacts orient some cell division axes in the Caenorhabditis elegans embryo, Goldstein [/bib_ref] [bib_ref] Combinatorial contact cues specify cell division orientation by directing cortical myosin flows, Sugioka [/bib_ref] [bib_ref] Planar asymmetries in the C. elegans embryo emerge by differential retention of..., Dutta [/bib_ref] [bib_ref] Differential activation of the DNA replication checkpoint contributes to asynchrony of cell..., Brauchle [/bib_ref]. Physical modeling also provides answers to the perennial question of why and how the C. elegans developmental procedure is so reproducible. However, there remains a large gap between data and theory, as there are much more data than theory and modeling can absorb, interpret, and exploit. Accordingly, there is a need for more theorists to be involved in studies to assist in the interpretation of high-quality quantitative data. Seamless collaboration between experimentalists and theorists will be crucial for developing a deep understanding of the embryogenesis program from both genetic and physical perspectives by making full use of data and models.
Reliable quantitative data serve as ground truth to verify or falsify theoretical and computational models. The physicist Richard Feynman wrote on his blackboard ''what I cannot create, I do not understand"; similarly, reconstituting a biological process in silico is a way to test how well the process is understood [bib_ref] What I cannot create, I do not understand, Way [/bib_ref]. The increasing amount of data on C. elegans embryogenesis will greatly aid the construction of accurate theories, algorithms, and mathematical language to describe this organism's development. Conversely, theoretical endeavors will stimulate new experiments, new applications of technologies, and the generation of new data. For example, the in situ measurement of cell mechanical properties (e.g., cell-cell adhesion mediated by HMR-1) is extremely difficult to perform when the signal-to-noise ratio of fluorescence drastically decays during blastomere cleavage [bib_ref] An asymmetric attraction model for the diversity and robustness of cell arrangement..., Yamamoto [/bib_ref] [bib_ref] Planar asymmetries in the C. elegans embryo emerge by differential retention of..., Dutta [/bib_ref]. This impedes accurate modeling, especially of developmental stages with dozens to hundreds of cells [bib_ref] Computable early Caenorhabditis elegans embryo with a phase field model, Kuang [/bib_ref] [bib_ref] Reconstructing the multicellular structure of a developing metazoan embryo with repulsion-attraction model..., Guan [/bib_ref]. This problem also occurs in the reconstruction of cell morphology from imaging data beyond the 350-cell stage [bib_ref] Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deeplearning-based..., Cao [/bib_ref]. For these and other reasons, we believe that in many cases the advancement of data generation and theory should be complementary and mutually stimulating.
The interaction and integration of quantitative data and modeling are essential to answer two key questions that have been posed by the developmental biologist Lewis Wolpert: ''Will the egg be computable?" and ''Do we understand development?" [bib_ref] Do we understand development?, Wolpert [/bib_ref] [bib_ref] The ''computable egg": Myth or useful concept?, Biasuz [/bib_ref]. These questions are related but not the same. On the one hand, even if egg development could be perfectly simulated in silico, it is not guaranteed that this would inevitably result in a comprehensive understanding of the process. This is analogous to the fact that simulating the exact trajectory of all of the air molecules in a room would not necessarily result in an understanding of thermodynamics. On the other hand, it is possible that a certain level of understanding can be achieved without detailed and ''realistic" modeling (again, the example of thermodynamics). Thus, the answer to one question would help to answer the other question, and vice versa. Overall, models aid understanding, and models of different scales and complexity are needed to answer different questions.
## Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Our gratitude is also extended to Kai Kang for his assistance in data visualization.
[fig] Figure 1: C. elegans embryogenesis and cell lineage analysis. (A) Simplified diagram of adult hermaphrodite C. elegans. (B) Embryonic cell lineage trees composed of the cells with a complete lifespan recorded from the 4-to 350-cell stages [/fig]
[fig] Figure 2: Polarization of zygote and germline precursor cells and specification of somatic founder cells in early embryogenesis of C. elegans. (A) Consecutive asymmetric divisions from P0 to P3 that sequentially generate somatic founder cells (AB, EMS, C, and D) and symmetric division in P4 that ultimately generates primordial germ cells (Z2 and Z3). (B) Polarization of upstream aPAR and pPAR (shown with distributions on the membrane) and downstream MEX-5/MEX-6 and PIE-1 (shown with distributions in the cytosol) in P0 that is induced by sperm entry. The actomyosin meshwork that flows from the posterior to the anterior in P0 and drives the initial regionalization of aPAR is depicted as thin lines. The color of each component is listed on the right. The stable final boundary between the anterior and the posterior in P0 is represented by a dashed line that splits the zygote into two future daughter cells (the larger cell = the somatic founder cell AB; the smaller cell = the germline precursor cell P1). This schematic is adapted from[174]. (C) Fluorescent labeling of aPAR and pPAR, exemplified by PAR-6 and PAR-2, respectively; both images are reproduced from[163] with permission. A bar representing an actual length of 10 lm is depicted in the bottom-right corner of each experimental image. (D) Schematic showing the migration of aPAR (blue) and pPAR (cyan) between the membrane and the cytosol (straight arrows). Their diffusion and mutual inhibition are indicated by curved and gray blunt-ended arrows, respectively, while the red rods represent the cell cortex. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) [/fig]
[fig] Figure 3: Mechanical modeling of cell positioning and cell morphology in C. elegans embryogenesis. (A) Schematic diagrams of the multi-particle model (left), the coarse-grained model (middle), and the phase field model (right) at the 4-cell stage. (B) Embryonic morphologies acquired from in silico simulation by the phase field model (first row) and in vivo experiment (second row) from the 2-to 8-cell stages. The subfigures of in silico embryonic morphologies are reproduced from[57], and the in vivo embryonic morphologies with their segmentation outputs are rendered by CShaper, with a scale bar representing an actual length of 10 lm depicted beneath each[76]. [/fig]
[fig] Funding: This work was supported by funding from the National Natural Science Foundation of China (Grant No 12090053 and 32088101) to G. Guan and C. Tang and from the Hong Kong 18, and FNRA-IG/21-22/SCI_02) to Z. Zhao. [/fig]
[table] Table 1: Analytical software developed for studies of C. elegans embryonic cell lineage. [/table]
[table] Table 2: Comparison of algorithms used for reconstructing cell morphology during C. elegans embryogenesis. [/table]
[table] Table 3: Developmental characteristics and technical limits of cell tracing and lineaging and cell morphology reconstruction in four widely used animal models. [/table]
|
Non-coding RNAs as potential biomarkers in osteosarcoma
Osteosarcoma (OS) is a primary solid malignant tumor that occurs most frequently in the metaphysis of long bones. More likely to happen to children and adolescents. OS has high mortality and disability rate. However, the etiology and pathogenesis of OS have not been fully understood till now. Due to the lack of effective biomarkers, OS cannot be precisely detected in the early stage. With the application of next-generation and high-throughput sequencing, more and more abnormally expressed non-coding RNAs(ncRNAs) have been identified in OS. Growing evidences have suggested the ncRNAs, such as microRNAs (miRNAs), long non-coding RNAs (lncRNAs), circular RNAs (circRNAs), have played an important role in the tumorigenesis and progression of OS. Thus, they can be served as novel biomarkers for diagnosis, treatment and prognosis. This review summarized the application of ncRNA as biomarkers in OS in detail, and discussed the limitation and future improvement of the potential biomarkers.
# Introduction
Osteosarcoma (OS), as a common kind of primary malignant tumor of bone, has a high mortality [bib_ref] Tumor-associated macrophages in osteosarcoma: From mechanisms to therapy, Cersosimo [/bib_ref] , accounting for about 2.4% of children with malignant tumors and 20% of all primary bone cancers. Depending on statistics, there are about 800 new cases in the United States each year, among which about 50% are children and adolescents [bib_ref] Childhood and adolescent cancer statistics. Ca, Ward [/bib_ref]. Studies have shown that the incidence of OS in men is higher than that in women, which is 1.27 times that of women.. OS occurs in the epiphysis of long bones more frequently, such as the proximal tibia, distal femur, proximal humerus, and other parts of the fastest growing bone . OS has a high disability and mortality, and is prone to metastasis . The prognosis of patients with early metastasis are very poor, the 5-year survival rate of patients with metastasis is lower than 20%. However, the early symptoms of OS are not typical, it was usually in the late stage when found even with metastasis, resulting in the high mortality rate. Early diagnosis and treatment are very important for the prognosis of OS. With the continuous improvement of the diagnosis and treatment methods, a variety of therapies have emerged, including targeted therapy and immunotherapy, which have effectively reduced the overall mortality of the patients. Nevertheless, the prognosis is still far from ideal. Therefore, it is urgent to carry out in-depth research on the related molecular mechanisms of the occurrence and development. By looking for new and effective biomarkers for detection, diagnosis, therapy and prognosis, the disability rate and mortality rate can be reduced in some extent, so as to prolong the survival time of the patients.
There are a wide kinds of tumor markers. At present, the discovered species include nucleic acid, protein, glucose, small molecular metabolites, cytokines, circulating tumor cells (CTCs), and so on, which play an important role in the diagnosis, treatment and prognosis of tumors. In the past few decades, bioinformatics has provided new ideas for the functions of early diagnosis and treatment with its advantages of simplicity, noninvasive and high efficiency, further play an important role in early diagnosis and prognosis evaluation. Compared with protein biomarkers, ncRNAs tumor markers have the advantage of replicability, stable expression, less affected by in vivo and in vitro stimulation. In clinic, the commonly used ncRNAs tumor markers of OS include microRNA (miRNA), long noncoding RNA (lncRNA), and circular RNA (circRNA) [bib_ref] Long noncoding RNADUXAP8 regulates TOP2A in the growth and metastasis of osteosarcoma..., Yang [/bib_ref]. Biomarkers have become a new potential due to their advantages of easy sample acquisition, less trauma to the human body, higher sensitivity and specificity. These biomarkers have played significant roles in non-invasive diagnosis and detection during the occurrence, development and prognosis of tumor. Non-coding RNA (ncRNAs), because of its own advantages, has become a major participant in tumor markers, such as in hepatocellular carcinoma, lung cancer, breast cancer, et al. [bib_ref] Non-coding RNAs in hepatocellular carcinoma: Molecular functions and pathological implications, Wong [/bib_ref] [bib_ref] Identification and validation of immunerelated lncRNA prognostic signature for breast cancer, Shen [/bib_ref].
At present, the researches focused on ncRNAs tumor markers and relevant action pathways have revealed the important roles in the occurrence, metastasis and prognosis. These biomarkers can help to determine the accuracy of the diagnosis, decide on the appropriate treatment time, make personalized treatment plans, and explore new therapeutic targets. Many ncRNAs biomarkers and corresponding protein products have been reported to be overexpressed or underexpressed in OS, such as miR-223 [bib_ref] MiR-223/Ect2/ p21 signaling regulates osteosarcoma cell cycle progression and proliferation, Xu [/bib_ref] , COP9 signalosome subunit 3 (COPS3) , lncRNA small nucleolar RNA host gene 4 (lncRNA SNHG4) [bib_ref] LncRNA SNHG4 promotes tumour growth by sponging miR-224-3p and predicts poor survival..., Xu [/bib_ref] , miR-142 [bib_ref] Role of miR-142 in the pathogenesis of osteosarcoma and its potential as..., Shabani [/bib_ref] , etc. Although the reports on types and functions of ncRNAs in OS are still limited, they might express some diagnostic, stage, treatment and prognosis targets. ncRNAs such as miRNAs, lncRNAs, circRNAs studied before were summarized in this review. miRNA is a small ncRNA molecule, which mainly regulates gene expression by degrading mRNA or inhibiting the translation process after transcription. The discovery of lncRNA mainly comes from microarray technology, the second generation highthroughput transcriptome sequencing technology, single cell sequencing technology, etc., and plays an important role in maintaining the activities of living cells. circRNA is a wide range of ncRNA that is stable in nature, highly conservative and specific. In this paper, by searching the relevant databases, we searched the ncRNA related to the occurrence and development of OS, and expounded its role in the occurrence and development of OS. The specific literature screening process is shown in [fig_ref] FIGURE 1: Flow chart of literature selection and collection [/fig_ref]. The specific roles of some of the relevant ncRNAs are presented in [fig_ref] TABLE 1: Identification of NcRNA Tumor Biomarkers analyzed in multiple studies [/fig_ref].
In this review, the roles of ncRNAs who had the potential as biomarkers in the diagnosis, stage, treatment, and prognosis of OS were discussed in depth, which has great potential to provide reference for relevant studies.
## Diagnostic biomarkers
Diagnosing OS at the early stage can cure diseases faster and earlier, significantly improve the prognosis of patients. Abnormal expression of ncRNA biomarkers can be used as potential biomarkers for diagnosis of OS. For miRNAs, there were significant differences in serum and plasma levels of miR-21 between patients with OS and healthy people, which found can be used as biomarkers of OS. Studies have shown that miR-21 was over-expressed in OS tissues and cell lines, and it had great potential as a biomarker for OS diagnosis. [bib_ref] Identification of serum miR-124 as a biomarker for diagnosis and prognosis in..., Cong [/bib_ref] found that the serum miR-124 level in patients with OS was significantly lower than that in the periostitis group and the healthy control group. The level of miR-124 normalized after tumor resection, and the area under the area under curve (AUC) for serum miR-124 was 0.846, with sensitivity of 79.8% and specificity of 86.0%, which proved that miR-124 had the potential to be the biomarker for diagnosis of OS. [bib_ref] MiR-143 inhibits EGFR-signaling-dependent osteosarcoma invasion, Wang [/bib_ref] found that compared with the corresponding para-carcinoma tissue, the miR-143 level was significantly decreased in OS, which proved that miR-143 can be used as diagnostic biomarker of OS.
Fraxetin (FXT), had been reported to be associated with the development of various tumors [bib_ref] Fraxetin exerts anticancer effect in glioma by suppressing MiR-21-3p, Yao [/bib_ref]. Through examining the expression of FTX in 25 OS and the adjacent tissues,found that FTX were significantly upregulated in OS tissues, and knocking out FTX gene could inhibit the survival rate, invasion, and migration force of OS cells, as well as promote the apoptosis of OS cells. LncRNA can not only directly regulate OS cells, but also affect the growth, proliferation, apoptosis and invasion of OS cells through the regulation of mRNA or miRNA, playing a key regulatory role in the occurrence and development of OS. The research studied by [bib_ref] Application of long noncoding RNAs in osteosarcoma: Biomarkers and therapeutic targets, Li [/bib_ref] found that lncRNA can be regulated in the occurrence and development of OS by competing endogenous RNAs (ceRNA), Wnt/β-Catenin and other pathways. In short, lncRNA can be served as biomarkers for early screening of OS, but their specificity and sensitivity need to be further verified clinically.
When it comes to circRNAs, recently more and more studies have shown that circRNAs can be used as a kind of biomarkers in the early diagnosis of OS, with high accuracy and specificity.reported that circ_0008717 owned high expression in OS tissues, compared with the corresponding paracancerous tissues . Molecular sponge could bind to miR-203 to exert its carcinogenic effect. Its AUC was 0.782 (95% CI: 0.682-0.862), sensitivity was 0.80, and specificity was 0.73, which could be used as a potential biomarker for the diagnosis of OS. The commonly used biomarkers for clinical OS diagnosis are serum alkaline phosphatase (AKP) and lactate dehydrogenase (LDH), but the comparison of diagnostic efficacy between these biomarkers and OS biomarkers has not been reported.found that hsa_circ_0003074 was highly expressed in plasma of OS patients and had a high efficacy in distinguishing OS patients from healthy volunteers, with the AUC of 0.93, higher than LDH (AUC = 0.83) and alkaline phosphatase (ALP) (AUC = 0.88). Thus, circRNAs also had the potential to be a diagnostic biomarker, which need further studies.
lncR-SOX21-AS1 Overexpression miR-7-5p/IRS2 --High expression leads to the proliferation, migration and invasion of OS cells Chen and Chen, It can be observed that many articles have proved that ncRNAs also has the potential as a diagnostic marker. In addition to the function in diagnosis, the levels of OS ncRNAs biomarkers also play an important role in determining the disease stage of OS at diagnosis. Abnormal expression of OS related genes and biomarkers can induce the activity, proliferation and differentiation of osteoblasts. The abnormal degree can directly reflect the pathological staging of patients and affect the prognosis. Enneking staging has become an important basis of OS surgical staging which can better remind the prognosis, and help to select an appropriate treatment. However, research on the relationship between ncRNAs biomarkers and tumor staging is still very lacking. A variety of ncRNAs biomarkers are closely related to the clinical stage of Enneking and have been widely used to determine the clinical stage of OS. The more malignant the lesion is, the higher Enneking stage and higher possibility of distant metastasis the tumor have.
It was found that the expression of lncRNA GNAS antisense (lncRNA GNAS-AS1) could significantly increase in OS cells and tissues, which were positively correlated with Enneking stage and distant metastasis. Its high expression often predicted the short overall survival time. Through regulating miR-490-3p, lncRNA GNAS-AS1 could play a major role in cell proliferation, migration and invasion, which could be an independent prognostic predictor of OS, providing a new therapeutic strategy for OS. The expression of circRNA-mitochondrial tRNA translation optimization 1 (circ_MTO1) in OS is low, its expression level is significantly correlated with Enneking staging, and its high expression in the tumor meant that the Enneking staging was lower in OS patients who had better neoadjuvant chemotherapy response and longer disease-free survival (DFS) . Except for the above, the increased expression of circ_001569 in OS was positively related to tumor size, Enneking stage or Tumor Node Metastasis (TNM) stage and lung metastasis [bib_ref] Circ_001569 regulates FLOT2 expression to promote the proliferation, migration, invasion and EMT..., Xiao [/bib_ref]. The results reported by [bib_ref] hsa_circ0021347 as a potential target regulated by B7-H3 in modulating the malignant..., Wang [/bib_ref] showed that circ_0021347 was significantly downregulated in OS tissues and cell lines, negatively correlated with TNM staging and positively correlated with patient survival. circ_0021347 can target B7 Homolog3 (B7-H3), which showed a strong negative correlation with the expression of B7-H3 in OS and exerted anti-cancer effect by negatively regulating the expression of B7-H3. In a word, the emergence of ncRNAs biomarkers in patients provides a reliable non-invasive method for the diagnosis and clinical staging of OS. However, when exploring new serum biomarkers in the future, the sensitivity and specificity of existing biomarkers should not be ignored, so as to provide accurate guidance information for clinical practice.
## Therapeutic biomarkers
Neoadjuvant chemotherapy is the standard treatment at present, and its survival rate has been greatly improved. However, the survival rate of patients with lung metastasis and chemotherapy resistance is still very low. Although the introduction of neoadjuvant chemotherapy has greatly improved the 5-year survival rate of OS, a large number of patients still have poor response to chemotherapy. Even after surgical resection and chemotherapy, there still exist a high risk of local recurrence or distant metastasis, leading to poor prognosis. It is necessary to develop new therapeutic methods for OS clinically, which require us to clarify the molecular mechanisms of the pathogenesis and development of OS. The screening of new molecular markers is important for the prognosis and treatment of OS.
The ncRNAs have good stability, which can be used as potential cancer biomarkers and treatment targets [bib_ref] An emerging function of circRNA-miRNAs-mRNA axis in human diseases, Rong [/bib_ref]. Previous studies have explained the mechanism of OS from multiple drug resistance-related genes, miRNAs, circRNAs and other aspects. Significant changes of miRNAs expression profile in drug-resistant OS cells indicate that miRNAs can participate in the development of drug resistance by regulating various targets and signaling pathways. The treatment based on miRNAs mainly included blocking the expression of oncogenic miRNAs and restoring the expression of tumor-inhibiting miRNAs genes. For instance, miR-21, which was highly expressed in many cancer types, had higher
## Figure 2
The role of partial ncRNAs in OS.
## Frontiers in genetics
frontiersin.org expression level in the sera of OS than that in the healthy, which has been used clinically as a biomarker of chemotherapy sensitivity and prognosis [bib_ref] Identification of serum microRNA-21 as a biomarker for chemosensitivity and prognosis in..., Yuan [/bib_ref]. Related studies have proved that transforming growth factor-β1 (TGF-β1) inhibitor treatment reduced the inhibitory effects of miR-21 knockdown on OS cell proliferation. miR-21 inhibition may inhibit OS cell proliferation by targeting PTEN and regulating the TGF-β1 signaling pathway [bib_ref] miR-143-3p inhibits the proliferation, migration and invasion in osteosarcoma by targeting FOSL2, Sun [/bib_ref]. In addition, the miRNAs profile is associated with tumor response which can be a preferred tool for predicting tumor susceptibility to treatment.
With the deepening understanding of genetic biomarkers, the therapeutic effect of lncRNAs has become a hot research topic in recent years, which is superior to known protein encoded gene biomarkers in predicting drug responses. At present, the abnormal expression of lncRNAs has been found in many patients. These abnormal expressions can affect the processes of drug outflow, apoptosis, DNA repair, cell cycle, proliferation, autophagy, etc. At the same time, these abnormal expressions participate in the chemotherapy resistance of OS by regulating the expression of different target genes and related signaling pathways. Distal-less homeobox 6 antisense 1 (DLX 6-AS1), which was highly expressed in OS patients, exerted its capability of inhibiting the proliferation, invasion and metastasis of OS cells by targeting the miR-641/homeobox protein Hox-A9 (HOXA9) signaling pathway . Upregulation of SOX2 overlapping transcript lncRNA transcript variant 7 (Sox2OT-V7) in OS can directly target miR-142/miR-22 to inhibit its expression, especially in OS tissues and cell lines that were resistant to chemotherapy. When knocking out Sox2OT-V7 in OS cells, the drug-resistant U2OS/Dox cells can be re-sensitive to chemotherapy drugs [bib_ref] LncRNA Sox2OT-V7 promotes doxorubicin-induced autophagy and chemoresistance in osteosarcoma via tumor-suppressive miR-142/miR-22, Zhu [/bib_ref]. Small nucleolar RNA host genes 4 (SNHG 4) can play a role in the occurrence and development of OS through the miR-224-3p/ DOCK7 pathway, which provide a new method in the treatment of OS [bib_ref] CircRNA CDR1as/miR-7 signals promote tumor growth of osteosarcoma with a potential therapeutic..., Xu [/bib_ref]. identified a positive correlation between the expression of lncRNA and antisense non-coding RNA in the INK4 locus (ANRIL) and the resistance of two therapeutic drugs for OS, cisplatin and doxorubicin. It was found that the drug resistance of MG-63/ DXR cells transfected with lncRNA NR_036444 decreased significantly, the proportion of cells in the G (1) phase increased. The proportion of cells in the later stage of apoptosis also increased, which might play an important role in regulating the drug resistance to doxorubicin, and might become a useful biomarker to evaluate chemotherapy sensitivity and predict prognosis of OS in the future [bib_ref] Sensitivity of doxorubicin-resistant osteosarcoma cells to doxorubicin regulated by long non-coding RNA..., Zhu [/bib_ref].
CircRNAs can play positive or negative regulatory roles in the occurrence and development of OS. In the fact, circRNAs can further participate in the regulation of chemotherapy resistance and metastasis of OS through "sponging" miRNAs as a tumor activating or inhibiting factor. The treatment based on miRNAs mainly included blocking the expression of oncogenic miRNAs and restoring the expression of tumor-inhibiting miRNAs genes. The carcinogenic effects of circ_0009910 were partially dependent on the JAK2/STAT3 pathway, which was involved in the apoptosis and proliferation of cells, thus affecting the progression of OS [bib_ref] Lenalidomide in multiple myeloma: An evidencebased review of its role in therapy, Richardson [/bib_ref]. And circ_001569 may promote resistance through the Wnt/βcatenin pathway [bib_ref] TWIST interacts with betacatenin signaling on osteosarcoma cell survival against cisplatin, Wu [/bib_ref].
Drug resistance is the main limiting factor for the effectiveness of cancer treatment. Some drugs can quickly relieve the tumor, but on the one hand, they can also produce resistance, thus increasing the difficulty of treatment [bib_ref] A view on drug resistance in cancer, Vasan [/bib_ref]. At present, more and more ncRNAs have been found to be associated with drug resistance of OS [bib_ref] Crosstalk between lncRNAs and miRNAs in gastrointestinal cancer drug resistance, Raei [/bib_ref]. The reversal of targeted ncRNAs is likely to be a potential method for reversing drug resistance of OS, which requires us to consider how to select the key ncRNAs from a large number of candidates ncRNAs. In the future, we should actively carry out clinical trials or transformation research based on targeted ncRNAs therapy, clarify the detailed mechanism of ncRNAs in OS drug resistance, and further apply them to clinical treatment of OS.
## Prognostic biomarkers
OS has a poor prognosis due to chemo-resistance and/or metastases. ncRNAs biomarkers of OS can affect the prognosis in many ways. Clarifying the relationship between ncRNAs biomarkers and the prognosis of OS, can help to optimize the current treatment plan and choose the right time of surgery and chemotherapy.
A variety of miRNAs also have great significance for the prognosis and progression of OS. The expression of miR-140 was related to the chemotherapy sensitivity of OS xenografts, which was involved in a wide range of chemotherapy resistance mechanisms by inhibiting HDAC4-mediated decrease in cell proliferation in G (1) and G (2) stages and participating in chemotherapy resistance. miR-140 could inhibit the proliferation of U2OS cells, but its inhibitory effect on MG-63 cell was weak . miR-223 has proved to be a potential prognostic marker for a variety of cancers in the past. Recently, some scholars have observed its function in OS. They found that miR-223 may be related to OS metastasis, and it has great potential as a potential biomarker for diagnosis and prognosis of OS (AUC:0.926) [bib_ref] miRNA-223 is a potential diagnostic and prognostic marker for osteosarcoma, Dong [/bib_ref].
LncRNAs can play important roles in drug sensitivity and cancer metastasis. lncR-BC050642 was significantly increased in OS tissues and cell lines, playing an important role in promoting the proliferation of OS cells, who can also be an independent biomarker of OS prognosis [bib_ref] The potential value of lncRNA-BC050642 in osteosarcoma origination and outcomes, Yang [/bib_ref]. In OS tissues, the significantly upregulated lncR-01614 indicated a worse prognosis for patients. After knocking out lncR-01614, the proliferation, invasion and Frontiers in Genetics frontiersin.org metastasis of OS cells could be inhibited. This effect was achieved through the miR-520a-3p/SNX3 regulatory axis, which had the potential to be a new clinical prognostic marker for OS. Titin-antisense RNA1 (TTN-AS1) could act as a carcinogen, playing a role in cancer through the miR-134-5p/MBTD1. Its high expression was also significantly related to the poor prognosis, and it had a potential to be a biomarker . CircRNA LRP6 (circ_LRP6) was highly expressed in OS, which can promote the progress of OS by inhibiting the expression of APC and KLF2. Its high expression usually indicated the short-term disease-free survival time and overall survival time . Studies have established an experimental system to study the in vivo and in vitro effects of the interaction between circ_001621/miR-578/VEGF, and the results show that circ_001621 played an important role in OS, and can increase the malignant degree, directly inhibit the expression of miR-578, significantly upregulate the expressions of VEGF, cyclin dependent kinase 4 (CDK4) and matrix metalloprotein 9 (MMP-9) in MG-63 and U2OS cells. Further it can promote the proliferation and migration of OS cells and provide a new therapeutic target for advanced OS [bib_ref] Circular RNA circ_001621 promotes osteosarcoma cells proliferation and migration by sponging miR-578..., Ji [/bib_ref] [fig_ref] FIGURE 3: Crosstab of the roles of selected ncRNAs in osteosarcoma [/fig_ref]. It was found that circ_0008717 and lung metastasis are independent prognostic factors through Cox multivariate regression analysis. After knocking down the expression of circ_0008717, the proliferation, migration and invasion of tumor cells can be reduced, and apoptosis can be promoted. circ_0081001 was another potential biomarker of OS prognosis, which was selected from chemotherapy and chemotherapy-sensitive OS cell lines, showing high levels in advanced OS, chemotherapy-resistant and lung metastasis [bib_ref] Circular RNA circ_0081001 knockdown enhances methotrexate sensitivity in osteosarcoma cells by regulating..., Wei [/bib_ref] [fig_ref] FIGURE 3: Crosstab of the roles of selected ncRNAs in osteosarcoma [/fig_ref].
Although some ncRNAs as prognostic biomarkers have been identified in OS, the outcome was not fully validated. As there have been some studies on the prognostic markers of OS, there is no practical application. The function of biomarkers for the diagnosis and prognosis of OS, especially patients with metastasis, may be an urgently needed tool for early diagnosis and identifying potential therapeutic targets.
## Perspectives and future opportunities
The early symptoms of OS are not typical. It is easy to metastasize, and the treatment often lagging behind with poor effect. Although the diagnosis and treatment of OS have made Frontiers in Genetics frontiersin.org continuous progress in recent years, the 10-year survival rate of patients with metastatic OS is still less than 20% [bib_ref] Bone microenvironment and osteosarcoma metastasis, Yang [/bib_ref]. As a result, new early diagnosis and treatment methods of OS are urgently needed. Biomarkers refer to biochemical indicators that can mark changes or possible changes in the structure or function of systems, organs, tissues, cells, and subcellular. The development of biomarkers provides new insights for the early diagnosis and treatment of diseases, which can assist clinicians in initial diagnosis especially for patients with metastatic OS, guide the treatment method, judge disease stages, evaluate the safety and effectiveness of new drugs or therapies in target population, providing a series of key genes and approaches for elucidating the molecular mechanism of OS. Therefore, it is urgent to search for useful biomarkers and therapeutic targets for clinical application. ncRNAs can be produced in the early stage of disease, which is an important part of epigenetics research. It can be used as a key regulatory factor to regulate the expression of related genes and participate in the processes of cell development, differentiation, proliferation, transcription, post-transcriptional modification, apoptosis, and cell metabolism [bib_ref] ncRNAs and thermoregulation: a view in prokaryotes and eukaryotes, De La Fuente [/bib_ref]. Moreover, it can naturally connect related genetic networks to affect various basic protein effect factors that drive specific cellular biological responses and determine cell fate [bib_ref] Non-coding RNAs as regulators in epigenetics (Review), Wei [/bib_ref]. As a carcinogen or anti-cancer factor, it plays an important role in the occurrence and development of a variety of cancers, with a certain stability and richness. ncRNAs can enter the circulatory system, which is expected to become a potential biomarker of OS [bib_ref] The role of non-coding RNAs in oncology, Slack [/bib_ref]. The application of these ncRNAs in the research and development of anti-OS drugs, related therapeutic targets and biomarkers in the field of OS was discussed in detail in this paper. At present, the newly discovered tumor biomarkers are one of the hot spots in oncology research. The research on ncRNAsencoded peptides or proteins has opened up a new research field for the diagnosis and treatment of tumors. As for OS, ncRNAs can be produced in the early stage of the disease, with certain stability and richness. When them entered the circulatory system, they can play an important role in cell function and affect its clinical manifestations to a certain extent. It is expected to become an effective diagnosis and treatment method. However, current research on ncRNAs also has the following problems. Firstly, the related research on ncRNAs and OS is limited and not in-depth. The current research mainly focuses on basic in vitro experiments. Further animal experiments and clinical trials are needed to confirm it in the future. Secondly, the targeted anticancer drugs of ncRNAs can weaken or even eliminate drug resistance of the patients, and further enhance the therapeutic effect. However, these drugs still have the disadvantages of insufficient specificity and utilization rate. The absorption and biodistribution of some drugs still need further solution. There is still a long way to go before the approval and commercialization of ncRNAs-targeted anticancer drugs.
Thirdly, ncRNAs have many targets and complex regulatory networks. At present, most research on ncRNAs are independent and scattered. The lack of specificity in the diagnosis and treatment of OS targeting ncRNAs still needs to be resolved. Finally, there still exist problems such as small sample size in the current research on the relationship between ncRNAs and OS, and lacking of epidemiological evaluation and functional exploration of candidate biomarkers. Only a few ncRNAs have been fully studied, and the possibility of its clinical application is still uncertain. In addition, most ncRNAs cannot be detected by standard methods such as quantitative PCR. It is an urgent to find new detection optimization methods and advanced laboratory platform conditions.
In summary, the expression of the biomarker is significantly related to the tumor size, distant metastasis, TNM stage, and Enneking surgical stage of the patients, which has the ability to distinguish OS from the healthy. It is important to look for a reliable clinical biomarker in the field of the early diagnosis and treatment of OS. These biomarkers are helpful to clarify the molecular mechanisms related to OS, and may become new targets for OS diagnosis and treatment. The higher the level of biological evolution, the higher the proportion of ncRNAs in the genome. Constituting almost 60% of the transcriptional output in human cells, ncRNAs have been shown to regulate cellular processes and pathways in developmental and pathological contexts [bib_ref] Non-coding RNA networks in cancer, Anastasiadou [/bib_ref]. Therefore, searching for more types of ncRNAs and its more extensive and diverse biological functions and action mechanisms are our current main tasks. The role of these ncRNAs in the proliferation, occurrence and development of OS cells should be deeply studied, so as to better predict and diagnose OS, provide meaningful theoretical guidance for individualized treatment and drug research. In the future, ncRNAs may continue to better help researchers and clinicians find molecular features, help in differential diagnosis, personalized treatment and determine prognosis et al. Although the applications of ncRNAs are still at an early stage, it deserves expectation can be put into clinical practice especially those candidates that can be readily detected. In the follow-up study, we would further verify the efficacy of diagnosis, treatment and prognosis of OS in large samples, and combine multiple ncRNAs to build a model to assist the diagnosis of clinical OS. Benign biomarkers should affect biological function of the disease, and the impact of promising ncRNA on the function of OS will be further explored. In addition, the biomarkers found in urine or blood by non-invasive procedures is ideal better than collecting tissues. Furthermore, there are great challenges associated with RNAbased therapeutics. On the one hand, RNA drugs can easily approach the targets based on nucleotide hybridization. On the other hand, the delivery method is a challenging process. As a promising new biomarker, ncRNA research is undergoing a rapid change and entering a new area of development. To keep up the pace, better techniques for the detect of new useful ncRNAs need Frontiers in Genetics frontiersin.org to be implemented, and the evaluation of those potential biomarker should be validated in a larger sample size. In the near future, our team will still be committed to the above studies of OS. Discovering new targets by high throughput sequencing methods, and follow-up studies on pathways are our research goals which can promise the revolution of OS. It is believed that with the further popularization and deepening of research, there will be new thoughts on the early prevention, accurate diagnosis and safer and more effective treatment of OS, and further application of ncRNAs to the diagnosis and treatment of other types of cancer.
# Author contributions
JL conceived the study. LF and ZZ conducted the study and drafted the application sections, they contributed equally to this work. YL contributed to the writing and review of the manuscript. All authors read, revised and approved the final manuscript.
[fig] FIGURE 1: Flow chart of literature selection and collection.Frontiers in Genetics frontiersin.org 03 [/fig]
[fig] FIGURE 3: Crosstab of the roles of selected ncRNAs in osteosarcoma. [/fig]
|
A case of bilateral persistent sciatic artery with unilateral aneurysm: An 18-year period of graft patency after excision of aneurysm
Background:A persistent sciatic artery (PSA) is a rare vascular anomaly with an estimated incidence of 0.02-0.04% and with a high rate of complications such as aneurysm formation, thromboembolism, and ischemia, that may lead to amputation.Case Report:We present a case of a female patient with complete symptomatic ambilateral PSA and with unilateral aneurysm. The aneurysm was excised and the PTFE graft was interposed at the aneurismal sac (femoro-popliteal bypass could not be performed because of the high degree hypoplasia of the superficial femoral artery). The graft endured continuous compression and stretching during regular daily life of the patient. At check-up 18 years after the operation, the Doppler ultrasound showed a patent graft and no new aneurismal dilatation of the sciatic artery.Conclusions:To our knowledge the follow-up of the presented case is the longest reported so far in the literature. The uneventful course of the patient confirms that classical aneurysmectomy still constitutes one of the treatment options of PSA aneurysm. key words: persistent sciatic artery - PSA aneurysm - vascular anomaly Full-text PDF:
# Background
Persistent sciatic artery (PSA) is a rare, but potentially serious vascular anomaly [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref]. In mammals, during the early fetal period, it connects the internal iliac artery with the popliteal artery and is the major blood supply to the lower limb in embryonic development. At that time the trunk runs down along with the sciatic nerve on the dorsal surface of the hip joint and the extensor muscles of the thigh. During further fetal development (after 6 weeks), due to position changes of the pelvis, which deteriorates the blood supply, the sciatic artery undergoes gradual involution. In mammals the role takes over femoral artery [bib_ref] A persistent sciatic artery aneurysm with an associated internal iliac artery aneurysm, Batchelor [/bib_ref] [bib_ref] Persistent sciatic artery: embryology, pathology, and treatment, Brantley [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Hassan [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Persistent sciatic vessels, varicose veins, and lower limb hypertrophy: an unusual case..., Parry [/bib_ref]. Remains of the sciatic artery take part in the formation of the inferior gluteal artery, deep femoral artery, fibular artery and the arteries of the foot. In the literature this vascular anomaly is called "persistent sciatic artery", "persistent axial artery", or "persistent ischiadico-femoral trunk" [bib_ref] Persistent sciatic artery: embryology, pathology, and treatment, Brantley [/bib_ref] [bib_ref] Persistent sciatic artery: an uncommon cause of intermittent claudication, Kritsch [/bib_ref] [bib_ref] Persistent sciatic artery: clinical, embryologic, and angiographic features, Mandell [/bib_ref] [bib_ref] Bilateral persistent sciatic arteries with unilateral complicating aneurysm, Aziz [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref]. In healthy adults, remnants of the sciatic artery are represented by a descending branch of the inferior gluteal artery, which at the level between the sciatic tuberositas and major trochanter forms a long and very thin branch called the committing artery of the sciatic nerve, which accompanies the sciatic nerve down to the popliteal fossa [bib_ref] Persistent sciatic artery: embryology, pathology, and treatment, Brantley [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Hassan [/bib_ref].
In the most common ("complete") form of this vascular anomaly, the internal iliac artery and its continuity PSA leave the minor pelvis and enter the thigh through the lower portion of the greater sciatic foramen, below the pyriform muscle, running down to the posterior femoral compartment and accompanying sciatic nerve. PSA has a variable relationship to the sciatic nerve, occasionally lying within the nerve sheath [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] A persistent sciatic artery aneurysm with an associated internal iliac artery aneurysm, Batchelor [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Persistent sciatic artery: an uncommon cause of intermittent claudication, Kritsch [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref]. At the level of the popliteal fossa it joins the popliteal artery. The profound femoral artery is usually preserved, whereas the superficial femoral artery in most cases (over 70%) is hypoplastic or even aplastic and ends in the form of several small branches at the distal part of the thigh [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] A persistent sciatic artery aneurysm with an associated internal iliac artery aneurysm, Batchelor [/bib_ref] [bib_ref] Persistent sciatic artery: embryology, pathology, and treatment, Brantley [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Persistent sciatic artery: an uncommon cause of intermittent claudication, Kritsch [/bib_ref] [bib_ref] Persistent sciatic artery: clinical, embryologic, and angiographic features, Mandell [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref].
The persistent sciatic artery usually runs a torturous course, is fragile and prone to vasculopathy, atherosclerosis, and aneurysm formation mostly due to hypertension and chronic trauma [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref].
The potentially serious complications of PSA include ischemia due to occlusive thrombosis or thromboembolization, formation and rupture of an aneurysm. The amputation rate in the course of PSA is estimated at between 10-15% [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] Bilateral persistent sciatic arteries with unilateral complicating aneurysm, Aziz [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref].
We present a rare case of the complete symptomatic bilateral PSA with hypoplastic femoral arteries and unilateral aneurysm, treated successfully, with an 18-year period of uneventful follow-up.
## Case report
In 1991 a 63-year-old non-smoking female was referred to the outpatient vascular service of our clinic, with a 2-year history of pain of varying intensity, localized in the left buttock, radiating down along the left thigh with concomitant paresthesias of the left foot. She was initially treated at a neurological unit. During the second year she experienced 2 incidents of acute pain of the left toe accompanied by the symptoms of the "blue toe" syndrome. Six months before, she began to feel discomfort, compression, and pulsatile mass in her left buttock. On admission, ankle/brachial index (AB/I) on the left was 0.67 and on the right 0.86. On physical examination, Cowie's sign was present -absent femoral pulse with palpable popliteal pulse [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Brancaccio [/bib_ref]. In the left buttock there was a pulsatile mass of 5×10 cm with slight bruit heard over it. The initial angiograms [fig_ref] Figure 1: Angiogram showing bilateral persistent sciatic arteries Type IIa [/fig_ref] revealed ambilaterally enlarged internal iliac arteries which ran down the thighs in a posterior location and joined the popliteal arteries. On the left side there was a saccular aneurysm at the level of the greater trochanter. The superficial femoral arteries were hypoplastic, and tapered gradually, ending as small branches at the distal thighs. It was diagnosed as PSA (with aneurysm), type IIa, according to Pillet [bib_ref] The ischiopopliteal arterial trunk: a report of two cases, Pillet [/bib_ref].
The patient was operated on. At surgery the aneurysm was directly exposed and excised [fig_ref] Figure 3: Excised aneurismal sac [/fig_ref]. To restore the blood flow, a PTFE graft #12 (12 cm long) was implanted in the place of the aneurismal sac in an end-to-end way [fig_ref] Figure 4: Intraoperative view of the interposed graft in the place of the removed... [/fig_ref]. The postoperative course was uneventful. The AB/I of the left extremity rose from preoperative 0.67 to 0.91.The patient was discharged on the 12 th postoperative day, free of symptoms and without any neurological deficit from the sciatic nerve.
The patient had been lost from follow-up since then. However, we managed to call her for check-up on June 2009, when she was 82. The general condition of the patient was fairly good. Her daily activity was adequate to her age. She did not report any complaints that could be directly attributed to the blood supply to the lower limbs. AB/I of her left foot was 0.76 and on the right AB/I was 0.81.The Doppler ultrasound showed a patent graft and no new aneurismal dilatation of the sciatic arteries.
# Discussion
It has been estimated that in approximately 0.02-0.04% of adults the sciatic artery persists as the main artery supplying the lower limb [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Hassan [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Persistent sciatic artery: clinical, embryologic, and angiographic features, Mandell [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref]. One needs to be especially aware of the PSA presence in the case of hip surgery, since there may be a certain overlap of symptoms, or during renal transplantation. Division of the internal iliac artery for the graft renal artery anastomosis would result in limb ischemia if there was a PSA [bib_ref] Persistent sciatic artery: clinical, embryologic, and angiographic features, Mandell [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref]. Histological study of resected sciatic artery usually demonstrates severe atherosclerotic changes, with a deficit in the collagen components of the vessel wall. PSA is a rare condition, but this artery is especially prone to develop atherosclerotic changes, embolism and aneurismal formation [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref].
Whereas major arteries normally follow a course on the flexor aspects of articulations, the PSA courses posteriorly in the buttock and thigh and is subjected to repeated mechanical trauma (eg, compression against the edge of a chair while sitting or overstretching when walking). This may result in aneurysmal formation (unilateral aneurysm is found in over 40% of cases of PSA, while bilateral aneurysm in more than 12% of PSA cases), typically under the gluteus maximus muscle at the level of the greater trochanter [bib_ref] Limb ischemia associated with persistent sciatic artery aneurysms-a report of 2 cases, Wilson [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Hassan [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Bilateral persistent sciatic arteries with unilateral complicating aneurysm, Aziz [/bib_ref] [bib_ref] An aneurysm at the back of a thigh: a rare presentation of..., De Boer [/bib_ref] [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref].
Aneurysmal complications such as rupture, thrombosis or embolism constitute a high risk of acute ischemia and are still connected with a substantial rate of limb amputations [bib_ref] Bilateral persistent sciatic arteries with unilateral complicating aneurysm, Aziz [/bib_ref] [bib_ref] Role of computed tomographic angiography in the detection and comprehensive evaluation of..., Jung [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref] , which is why early surgical intervention is increasingly regarded as preferable in case of the symptomatic PSA with aneurysm [bib_ref] Persistent sciatic artery: embryology, pathology, and treatment, Brantley [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref]. Treatment of PSA depends on the type of sciatic artery involved. In a case where the sciatic artery exists as the primary arterial supply to the extremity in the so-called "complete" form of PSA (approximately two-thirds of reported cases) revascularization of the distal extremity is necessary [bib_ref] A persistent sciatic artery aneurysm with an associated internal iliac artery aneurysm, Batchelor [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Hassan [/bib_ref] [bib_ref] Persistent sciatic artery: an uncommon cause of intermittent claudication, Kritsch [/bib_ref] [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref] [bib_ref] Symptomatic persistent sciatic artery, Brancaccio [/bib_ref]. Multiple surgical techniques are available, including interposition grafting or femoro-popliteal bypass grafting in combination with ligation or resection of the aneurysm [bib_ref] A persistent sciatic artery aneurysm with an associated internal iliac artery aneurysm, Batchelor [/bib_ref] [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] A ruptured aneurysm in persistent sciatic artery: a case report, Ishida [/bib_ref].
Recently, endovascular techniques such as percutaneous embolization with coils have been tried, providing symptomatic improvement, especially in cases of incomplete PSA where revascularization is usually unnecessary [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref] [bib_ref] Endovascular interventions on persistent sciatic arteries, Gabelmann [/bib_ref] [bib_ref] Angioplasty via retrograde popliteal approach in a stenosed persistent sciatic artery, Littler [/bib_ref] [bib_ref] Persistent sciatic artery: clinical, surgical, and angiographic aspects, Mclellan [/bib_ref]. In cases requiring continuity of the sciatic artery, percutaneous endovascular management of the lesion with stent-graft represents a further treatment option [bib_ref] Combined percutaneous endovascular and open surgical approach in the treatment of a..., Maldini [/bib_ref] [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref] [bib_ref] Endovascular interventions on persistent sciatic arteries, Gabelmann [/bib_ref] [bib_ref] Angioplasty via retrograde popliteal approach in a stenosed persistent sciatic artery, Littler [/bib_ref] [bib_ref] Persistent sciatic artery: clinical, surgical, and angiographic aspects, Mclellan [/bib_ref]. However, whether or not endovascular procedures represent a reasonable alternative to classic surgical techniques is yet to be determined, because a stent graft in this location bears the potential risk for graft dislodgment or compression over the hip joint, and its durability remains an open question [bib_ref] Endovascular stent graft repair of a persistent sciatic artery aneurysm, Fearing [/bib_ref].
In our case, because of poor perfusion, and the fact that the left distal superficial femoral artery was hypoplastic and did not supply adequate flow to the lower extremity, we could not limit our procedure to the exclusion of the aneurismal sac, or perform a femoro-popliteal bypass. Additionally, endovascular procedures were not available to us at that time.
That is why we decided to interpose the graft at the place of the excised aneurismal sac. We were aware that, theoretically, placing the graft in the same position as the aneurysm was not the optimal solution, because such a graft could be exposed to incidents of repeated trauma or stretching, as it had been the sciatic artery. It could increase the risk of leaks at the anastomotic sites or cause graft thrombosis. However, in this case the graft endured continuous compression during regular daily life of the patient over a period of years.
In an extensive English language review of this subject, van Hooft et al noticed that presented follow-up was short (6-12 months) and poorly described in most articles, thus therapeutic strategies could not be deduced from the available literature [bib_ref] The persistent sciatic artery, Van Hooft [/bib_ref].
conclusions To our knowledge the follow-up of the presented case is the longest reported so far in the literature.
references:
[fig] Figure 1: Angiogram showing bilateral persistent sciatic arteries Type IIa (according to Pillet) with saccular aneurysm indicated by arrows -on the left. [/fig]
[fig] Figure 2: Schematic diagram presenting vasculature to the lower extremities: A -aorta, IIA -internal iliac artery, A.sac -aneurysmal sac, SA -sciatic artery, SFA -superficial femoral artery, PA -popliteal artery. [/fig]
[fig] Figure 3: Excised aneurismal sac (7.5 cm in diameter). [/fig]
[fig] Figure 4: Intraoperative view of the interposed graft in the place of the removed aneurysmal sac. [/fig]
|
Rehabilitation for the management of knee osteoarthritis using comprehensive traditional Chinese medicine in community health centers: study protocol for a randomized controlled trial
Background: It is becoming increasingly necessary for community health centers to make rehabilitation services available to patients with osteoarthritis of the knee. However, for a number of reasons, including a lack of expertise, the small size of community health centers and the availability of only simple medical equipment, conventional rehabilitation therapy has not been widely used in China. Consequently, most patients with knee osteoarthritis seek treatment in high-grade hospitals. However, many patients cannot manage the techniques that they were taught in the hospital. Methods such as acupuncture, tuina, Chinese medical herb fumigation-washing and t'ai chi are easy to do and have been reported to have curative effects in those with knee osteoarthritis. To date, there have been no randomized controlled trials validating comprehensive traditional Chinese medicine for the rehabilitation of knee osteoarthritis in a community health center. Furthermore, there is no standard rehabilitation protocol using traditional Chinese medicine for knee osteoarthritis. The aim of the current study is to develop a comprehensive rehabilitation protocol using traditional Chinese medicine for the management of knee osteoarthritis in a community health center.Method/design: This will be a randomized controlled clinical trial with blinded assessment. There will be a 4-week intervention utilizing rehabilitation protocols from traditional Chinese medicine and conventional therapy. Follow-up will be conducted for a period of 12 weeks. A total of 722 participants with knee osteoarthritis will be recruited. Participants will be randomly divided into two groups: experimental and control. Primary outcomes will include range of motion, girth measurement, the visual analogue scale, and results from the manual muscle, six-minute walking and stair-climbing tests. Secondary outcomes will include average daily consumption of pain medication, ability to perform daily tasks and health-related quality-of-life assessments. Other outcomes will include rate of adverse events and economic effects. Relative cost-effectiveness will be determined from health service usage and outcome data.Discussion:The primary aim of this trial is to develop a standard protocol for traditional Chinese medicine, which can be adopted by community health centers in China and worldwide, for the rehabilitation of patients with knee osteoarthritis.
# Background
Knee osteoarthritis (KOA) increases in prevalence with age and causes pain, loss of functional independence and psychological impairment, and reduces the overall quality of life of affected individuals [bib_ref] Health-related quality of life among older adults with arthritis, Dominick [/bib_ref] [bib_ref] OARSI recommendations for the management of hip and knee osteoarthritis. Part III:..., Zhang [/bib_ref] [bib_ref] Risk factors for incident radiographic knee osteoarthritis in the elderly: the Framingham..., Felson [/bib_ref]. Based on epidemiological results, the incidence rate of osteoarthritis in the Chinese population is approximately 10% and the number of patients with KOA is greater than 50 million. KOA has been identified as one of the most serious diseases that can affect an individual during his or her lifetime. It is a degenerative disorder caused by the interplay of genetic, metabolic, biochemical and biomechanical factors and results in pain, swelling and malformation of the knee joint [bib_ref] The Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors, Fryback [/bib_ref]. KOA is prevalent and chronic and is associated with physical and psychological dysfunction. It reduces quality of life by inhibiting or preventing sufferers from carrying out activities such as walking, climbing stairs, doing housework, socializing and engaging in outdoor pursuits [bib_ref] OARSI recommendations for the management of hip and knee osteoarthritis. Part I:..., Zhang [/bib_ref]. Furthermore, dysfunction due to KOA strongly influences employment status, productivity [bib_ref] Osteoarthritis: quality of life, comorbidities, medication and health service utilization assessed in..., Rosemann [/bib_ref] [bib_ref] Psychosocial factors and the pain experience of osteoarthritis patients: new findings and..., Somers [/bib_ref] and increases obesity rates [bib_ref] Pathogenesis and management of pain in osteoarthritis, Dieppe [/bib_ref] , especially among older people. Pain, swelling, malformation and muscle weakness are often interrelated, leading to a downward cascade in physical as well as mental function [bib_ref] Physical functioning over three years in knee osteoarthritis: role of psychosocial, local..., Sharma [/bib_ref] [bib_ref] Preoperative quadriceps strength is asignificant predictor of knee function two years after..., Eitzen [/bib_ref].
During rehabilitation, patients with KOA require good pain control with few adverse events (AEs). Clinical guidelines advocate conservative non-drug strategies for the treatment of KOA [bib_ref] Tugwell P: OARSI recommendations for the management of hip and kneeosteoarthritis, Part..., Zhang [/bib_ref] , such as muscle-strengthening exercises, electromagnetic therapy and acupuncture. These conventional rehabilitation therapies are focused on controlling pain as well as reducing physical disability and impairment, while minimizing the potentially harmful side effects of medication [bib_ref] OARSI recommendations for the management of hip and knee osteoarthritis, part I:..., Zhang [/bib_ref]. These therapies can improve KOA patient outcomes [bib_ref] The relationship between disease symptoms, life events, coping and treatment, and depression..., Sale [/bib_ref] , but are often only offered in high-grade hospitals and not in community health centers (CHCs). Consequently, the direct and indirect health-care costs associated with KOA frequently become a major burden to patients. After their discharge, patients will often abandon the rehabilitation therapy techniques they have learned and the KOA will become progressively worse. Therefore, as the patients aging and comorbidities increasing, a more convenient, effective, simple and inexpensive approach to rehabilitation is necessary for those with KOA.
The use of traditional Chinese medicine (TCM) for the treatment of KOA has increased substantially. In Chinese medicine, KOA is considered to originate in the tendons and bones, but manifests as viscera insufficiency and biao or branch excess. The invasion of evil pathogens, due to wind-cold-dampness, is linked to blood stasis, which then leads to wei zheng and bi zheng [bib_ref] Experiment study on the suppression of sodium nitroprussiate-induced chondrocyte apoptosis by Touguxiaotong..., Li [/bib_ref]. Wei zheng is treated by viscera reinforcement and bi zheng is treated by wind-cold-dampness purging and blood activation. TCM has been used for health care in China for thousands of years. Some TCM rehabilitation methods, such as Chinese medical herbal patches, traditional exercises (developed from simplified t'ai chi and yi jinjing) and acupuncture have been reported to improve KOA in patients clinically [bib_ref] Efficacy of acupuncture for chronic knee pain: protocol for a randomised controlled..., Hinman [/bib_ref] [bib_ref] Traditional chinese herbal patch for short-term management of knee osteoarthritis: a randomized,..., Wang [/bib_ref] [bib_ref] Individually integrated traditional Chinese medicine approach in the management of knee osteoarthritis:..., Cao [/bib_ref] [bib_ref] Enrolling older adults with cognitive impairment in research: lessons from a study..., Tsai [/bib_ref]. The comprehensive TCM rehabilitation protocol (CTCMRP) comprises Chinese medical herb fumigation-washing (done to dispel wind-cold-dampness and to activate blood circulation in the knee), traditional exercise training, acupuncture and bloodletting via puncture. CTCMRP can enhance muscle strength, improve balance, reduce knee pain, stimulate the governor vessel and spine, and can regulate the function of the viscera, qi and blood in the zang-fu organs.
However, results from randomized controlled trials examining TCM treatments for KOA are contradictory and there is no standard rehabilitation protocol for KOA treatment using TCM (SRPTCM) [bib_ref] Ten years therapy survey of knee osteoarthritis: lack of acceptable evaluation standard..., Li [/bib_ref]. Therefore, in the current study a protocol was designed in which KOA patients will be treated using CTCMRP in CHCs. The data from this study will shed light on the use of TCM for KOA rehabilitation.
# Methods/design
## Study design
This will be a cluster randomized controlled trial with blinded assessment [fig_ref] Figure 1: Study flow chart. [/fig_ref]. The trial will comprise a 4-week intervention, using the CTCMRP and conventional therapy (CT), and a 12-week follow-up period. The protocol will be conducted at CHCs run by the Affiliated People's Hospital of Fujian University of Traditional Chinese Medicine (Fujian province, China), the Comprehensive Traditional and Western Medicine Hospital of Guangdong province (Guangdong province, China), the Chengdu Military Region General Hospital of the Chinese People's Liberation Army (Sichuan province, China) and the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine (Guangdong province, China). To include a broad range of patients, this protocol includes CHCs in different areas. Fujian University of Traditional Chinese Medicine (FJTCM) is responsible for training and supervising rehabilitation therapists at all clinical sites. The evidence-based medicine center, which will be used for documentation and data analysis, is an independent site that performs data monitoring, randomization and statistical evaluation and will be blind to the intervention. Patients will be recruited through CHC outpatient clinics. The Human Research Ethics Committee at Fujian University of Traditional Chinese Medicine (Fujian province, China) approved the study.
## Participants
Patients suffering from pain, swelling, muscle weakness and stiffness of the knee will tested for KOA. All participants will be required to provide signed, informed consent prior to beginning baseline assessment and rehabilitation.
## Inclusion criteria
Participants will have been diagnosed with KOA according to the criteria in the Guide Diagnosis of Bone Arthritis (2007) published by the Chinese Medical Association. Participants may be male or female and must be between 40 and 75 years of age. Participants will be considered to have early-to middle-stage KOA. Participants will understand the research and will have signed the informed consent.
Symptoms will include pain, swelling, stiffness or malfunction of the knee according to the assessment scale.
## Exclusion criteria
Patients are excluded if they have:
severe knee swelling a muscle strength of grade 2 or below according to a manual muscle test a history of trauma to the knees serious heart, kidney, liver or nervous system disease, or lower limb gangrene due to diabetes. lower limb dysfunction caused by spinal, hip or ankle joint disease, or prior knee joint surgery a tumor of the knee joint a cognitive or psychological disorder received traditional exercise training and/or tuina for the knee within the last 4 weeks
## Recruitment
Outpatients with KOA will be recruited from participating CHCs in China. All participants will be selected according to the above inclusion and exclusion criteria. Baseline measurements will include the general and demographic data for the subjects, assessment of the knee prior to testing, medication use, previous treatments, KOA complications and duration of KOA symptoms. Due to ethical considerations, non-steroidal, anti-inflammatory drugs will be permitted (and registered) as required for patients with visual analogue scale (VAS) scores over 5. All patients will be asked to refrain from seeking other forms of treatment during the trial. During the follow-up period, patients will be asked to report whether they have received other forms of treatment.
## Randomization and allocation
Following an epidemiological survey, the CHC randomization list for this study will be computer generated using SAS software (Version 8.2, SAS Institute, Cary, NC, USA) and concealed at an independent evidence-based medical center. Employees at the medical center will also be blinded to the intervention. Treatment allocation will start when it has been determined that a study participant meets the inclusion criteria and has signed the informed consent.
The CHC randomization program will assign subjects to each group and each participant's group number will be displayed (CTCMRP versus CT rehabilitation) on the back of the form. Baseline measurements will not be viewed until the end of the study to ensure blinded assessment. The allocation list will be protected and held by FJTCM. In the event of an emergency, the individual's randomization code and group allocation can be identified. Therapists at each participating CHC will provide treatment, as usual, to patients who are not eligible or willing to participate.
## Sample size
In cluster randomized trials, the inspection efficiency comes from the random unit number (cluster number) rather than the size. There are two important factors when estimating the sample number: (1) The number within each cluster group and (2) the intraclass correlation coefficient estimate. The total number of subjects (n) from all of the CHCs was calculated according to the estimation formula [bib_ref] Sample estimation: Cluster randomization test design, Guo [/bib_ref] :
[formula] n ¼ μ α þ μ β 2 π 1 1−π 1 ð Þþπ 2 1−π 2 ð Þ ½ 1 þ m−1 ð Þρ ½ π 1 −π 2 ð Þ 2 [/formula]
In a trial test of this protocol, the curative effect rate after rehabilitation of KOA was 70% in the control group (π 1 ) and 82.3% in the experimental group (π 2 ); this will be used as the main index. The number of qualified participants recruited from each CHC (m) will not be less than 17. Prior data indicate that the intra-CHC correlation coefficient (ρ) is likely to be, at most, 0.020, andμ α = 0.05 and μ β = 0.10.
There are 38 CHCs. Based on the formula, each group (control and experimental) will comprise 323 participants. A total number of 722 subjects will be included to allow for a dropout rate of 10%.
# Ethical issues
The Ethics Committee of the FJTCM approved this protocol (Approval No. 2012-031). Each participating center will obtain local institutional review board approval. All study participants will sign an informed consent sheet prior to participation.
## Baseline measurements
Baseline descriptive data will be obtained via a questionnaire [fig_ref] Table 1: Baseline descriptive data General data Agent of the person who conducted test,... [/fig_ref].
## Interventions
The participants in the experimental group will undergo CTCMRP rehabilitation treatment. This consists of Chinese medical herb fumigation-washing, traditional exercises (developed from simplified t'ai chi and yi jinjing), acupuncture and tuina. Patients in the control group will undergo CT rehabilitation including physical factor therapy and kinesitherapy. The duration of rehabilitation will be 4 weeks, which is a common duration in clinical practice. During the rehabilitation period the therapist will monitor proper performance, exercise intensity and will progress the exercises, if necessary.
## Health education
Two strategies will be used for health education: (1) Educational brochures on KOA, published by FJTCM, will be sent to participants. Information in these brochures will include the concept, epidemiology, etiology, common symptoms, diagnosis, treatments and precautions that should be taken. (2) The rehabilitation therapist will introduce and explain the contents of the educational brochures to all participants in person.
The aim is to help participants learn about and understand KOA and to implement strategies to help alleviate the disease. Moderate activity, aerobic training (swimming), avoidance of long periods of running or jumping, decreasing or avoiding climbing stairs, and reducing weight if necessary, can all help to improve KOA.
## Experimental group
The experimental group will be treated according to the CTCMRP [fig_ref] Table 2: TCM rehabilitation based on syndrome using appropriate technology [/fig_ref].
In all cases, Chinese medical herb fumigation-washing with traditional exercises (to a basic prescription) will be used. If there is significant swelling in the joint, basic rehabilitation plus cupping and bloodletting will be used. When there is a significant decrease in joint strength, basic rehabilitation plus tuina will be used. When there is significant joint pain, basic rehabilitation plus acupuncture will be used.
## Chinese medical herb fumigation-washing
The right or left knee joint will be washed twice daily, for 6 days, with a decoction of hai tong pi [bib_ref] Beijing: China Chinese medicine press, Wang [/bib_ref]. The recipe for this is: Erythrina indica Lam (15 g), moxa (15 g), Phryma leptostachya (15 g), Lycopodium clavatum (15 g), radix aconiti (10 g), radix aconiti agrestis (10 g), Pericarpium zanthoxyli (10 g), radix angelicae (10 g), radix clematidis (10 g), cassia twig (10 g), safflower carthamus (5 g) and Ligusticum wallichii (5 g). Patients will undergo a 6-day course of treatment on four separate occasions .
## Traditional exercises
Simplified t'ai chi and yi jinjing exercises will be performed twice daily with each patient completing one set of 15 repetitions for both knees .
## Cupping and bloodletting intervention
Patients will undergo cupping as well as being punctured for bloodletting. This will be done twice weekly, at points ST 34 and SP 10, using a percusso-punctator. This will not be done in patients with skin ulcers, dermatitis or hemophilia.
## Tuina
This treatment is based on the theory of meridians and will be done once a day (15 repetitions of 3 to 5 minutes each) as in clinical practice [fig_ref] Table 4: Tuina intervention [/fig_ref].
## Acupuncture
Patients with a pain score over 3, according to the VAS scale, will undergo acupuncture. While the patient is lying down, needles will be inserted into the following acupuncture points: Ex-LE 5, Ex-LE05, Ex-LE 2 and Ashi. The treatment will last 30 minutes and will be carried once per day. Single-use Huatuo needles (0.35 × 40 mm) will be used and will be disposed of into a bedside sharps container upon completion of treatment.
## Control group
The control group will be treated with physical factor rehabilitation and kinesitherapy [bib_ref] Beijing: People's Medical Publishing House, Nan [/bib_ref].
## Thermal design power
Participants will be asked to undergo Te Ding Dian Ci Pu (TDP, a machine that can release specific frequency of electromagnetic wave to treat disease) for at least 30 minutes per day. The distance between the TDP and the knee must be from 30 cm to 50 cm.
## Kinesitherapy
Kinesitherapy is designed to strengthen lower limb muscles and incorporates exercises commonly used in clinical practice. It is based on clinical trials showing that such exercise programs can reduce pain and improve function [bib_ref] Exercise for osteoarthritis of the hip or knee, Fransen [/bib_ref]. The aim of kinesitherapy is to strengthen the quadriceps, hamstrings and other muscles around the knee. Participants will do this twice daily under the direction of the therapist .
## Timelines
Project funding was approved in July 2012 and commenced in October 2012. Ethics approval was obtained from the Medical Ethics Committee of FJTCM in September 2012. Recruitment commenced in October 2012 and will be completed in October 2013. The trial is due for completion in October 2015 at which time all participants will have completed the 3-month follow-up. Outcomes will be assessed at baseline, and at 4-, 10-and 12-weeks post-baseline. Telephone contact and follow-up clinic visits will be used to assess participant adherence to the study protocol.
## Follow-up
During the 12-week follow-up period, trained personnel will contact the participants each week via telephone.
Participants will be asked about the KOA and whether they have received other treatments. Participants will be asked to attend a follow-up clinic at 6 and 12 weeks to assess pain, swelling, range of motion, muscle strength, average walking distance, stair climbing ability, activities of daily living (ADL), quality of life (QOL) and medication usage.
## Outcome assessment
Outcome measures have been selected based on those recommended for clinical trials of KOA [bib_ref] Osteoarthritis clinical trials: candidate variables and clinimetric properties, Bellamy [/bib_ref] [fig_ref] Table 6: Summary of measures to be collected [/fig_ref].
## Primary outcomes
Average knee pain will be self-assessed using a 100 mm visual analogue scale (VAS) with terminal descriptors of 'no pain' and 'worst pain possible'. The VAS pain measurement has demonstrated reliability in KOA [bib_ref] Osteoarthritis clinical trials: candidate variables and clinimetric properties, Bellamy [/bib_ref]. Knee girth will be determined using a flexible plastic measuring tape to measure the transverse plane circumference of the knee at the mid-patellar height, with the patient in a supine position. Girth measurement has shown acceptable reliability for determination of gross changes in knee swelling after knee surgery [bib_ref] Reliability of lower extremity girth measurements after anterior cruciate ligament reconstruction, Soderberg [/bib_ref].
The knee's range of motion (ROM) will be measured using a standard long-arm goniometer. To determine knee flexion ROM, patients will be asked to slide the heel actively towards the buttocks and maximal active knee flexion will be measured. Knee ROM in patients with KOA has been shown to be adequately reliable with a coefficient of 0.96 for flexion and 0.81 for extension [bib_ref] Goniometric reliability in a clinical setting. Elbow and knee measurements, Rothstein [/bib_ref].
The manual muscle test (MMT) will be performed according to Medical Research Council (MRC) guidelines Traditional exercise intervention (developed from simplified t'ai chi and yi jinjing)
## Phase
Description and progression Around the knee Push the muscle and warm using the palm of the hand and the base of the thumb (the thenar eminence) repeatedly Kinesitherapy
## Intervention description
Passive movement (i) The patient sits on a bed with the limb hanging over the edge of a bed. A rolled towel or pad is placed under the shallow depression at the back of the knee joint (popliteal fossa).
(ii) The therapist holds the bottom (distal) of the lower leg and pulls toward the foot for 10 seconds. This is repeated five times consecutively.
(iii) The therapist uses his or her hands to flex and extend the knee. Patients will be informed that this will be a passive motion and may uncomfortable but should not be painful.
Active movement (i) The participant lies on his or her back, lifts and straightens one leg with a weight attached to the ankle. The heel is lifted so that it is approximately 18 cm above the bed and held in that position for 30 seconds. This is repeated three times consecutively.
(ii) The participant lies on his or her back, stretches out the knee and performs a set of 15 quadriceps contractions keeping the leg in the same position. This is repeat twice a day for 4 weeks. [bib_ref] Quantitation of muscle function in children: a prospectivestudy in Duchenne muscular dystrophy, Scott [/bib_ref]. The British MRC has developed a muscle scale that grades muscle strength between 0 and 5. According to the muscle scale, 0 = no palpable contraction, total paralysis; 1 = flicker or trace of muscle contractions; 2 = active movement with gravity eliminated; 3 = full range with gravity eliminated; 3-= partial range against gravity; 3+ = full range against gravity with no resistance possible; 4 = slight resistance; 4-= mild resistance; 4+ = moderate resistance; 5-= barely detectable weakness and 5+ = normal motor power. Walking performance will be assessed using the six-minute walking test . This measures the distance that participants are able to walk in 6 minutes while wearing their usual footwear. Patients will be allowed to use a cane, if necessary. Standardized feedback will be provided to patients during each test as described previously [bib_ref] Altered loading during walking and sit-to-stand is affected by quadriceps weakness after..., Mizner [/bib_ref]. The 6 MW test will be done twice and the test results averaged. A longer walking distance than baseline will indicate better walking performance. The 6 MW test has excellent test-retest reliability from 0.95 to 0.97 and a low coefficient of variation (10.4%) [bib_ref] Age-and gender-related test performance in community dwelling elderly people: Six-Minute Walk Test, Steffen [/bib_ref].
The stair-climbing test (SCT) is a physical performance test that measures the time it takes to ascend and descend a flight of stairs. Patients will be asked to complete the test as quickly as they can while still feeling safe and comfortable. Patients will be allowed to use a device if necessary, though participants will be encouraged to perform the test without the aid of a device if possible. Rejeski [bib_ref] Assessing performance-related disability in patients with knee osteoarthritis, Rejeski [/bib_ref] found that a similar stair-climbing task had an excellent test-retest reliability coefficient of 0.93.
Based on the Guidelines for the Clinical Research of Chinese Medicine New Drugs (2002) [bib_ref] Beijing: Chinese medicine and technology publishing house, Zheng [/bib_ref] , the curative effect for KOA is calculated as:
Curative effect ¼ Total score before treatment-Total score after treatment ð Þ Â 100=Total score before treatment where the total score before and after treatment is based on the measures above. The standard comprehensive effect:
clinical cure: curative effect ≥ 90% obvious effects: 75% ≤ curative effect < 90% effective: 35% ≤ curative effect < 75% invalid: curative effect < 35%
The efficiency of a treatment is calculated as:
Efficiency % ð Þ ¼ ðNumber of clinical cures þ Number of cases with obvious effects þ Number of effective curesÞ Â 100=Total number of cases
## Secondary outcomes
The average daily consumption of non-steroidal, antiinflammatory drugs.
ADL performance will be assessed using the modified Barthel index (MBI) [bib_ref] Development of a Chinese version of the Modified Barthel Index -validity and..., Leung [/bib_ref] , which is based on 10 items and is scored on a scale from 0 to 100. A higher score indicates more independence [bib_ref] Rehabilitation of stroke patients using traditional Thai massage, herbal treatments and physical..., Sibbritt [/bib_ref].
QOL will be evaluated using the medical outcomes short form 36 health questionnaire (SF-36). The SF-36 has been found to be reliable and is easy to administer [bib_ref] Comparison of methods for the scoring and statistical analysis of SF-36 health..., Ware [/bib_ref] [bib_ref] Generic and condition-specific outcome measures for people with osteoarthritis of the knee, Brazier [/bib_ref]. The questionnaire has eight scales for different domains of health as follows: physical functioning, bodily pain, physical state, general health, vitality, emotional state, social functioning and mental health. Each scale is scored between 0 and 100 with 100 representing the best score possible. The PCS (physical component scales, summary scales of physical functioning, bodily pain, role-physical and general health) in SF-36 will be the main focus of this investigation as it is a composite of all scales representing the physical aspects of health.
## Safety assessments
Trial coordinators and therapists at each CHC will record adverse events and the details of these events, throughout the study and during the follow-up period. Participants will be able to report adverse events at any time during the trial and will receive advice accordingly. Serious adverse events will be reported to the Human Research Ethics Committee (HREC) of FJTCM.
## Economic evaluation
The cost of interventions will be assessed during the rehabilitation period, including the length of any hospital stay, emergency department visits, general medical treatment, visits to a specialized physician and nursing care. The intervention costs of high-grade hospitals will be compared to those of CHCs. Medicare national fee structures will be used to evaluate the costs of these services since out-of-pocket expenses for these services often do not accurately reflect the service value. Finally, the number of workdays lost by participants during rehabilitation will be used to assess indirect costs.
# Statistical analysis
Statistical analysis will be conducted using SAS software (version 8.2, SAS Institute, Cary, NC, USA). All tests will be two-tailed with P values of less than 0.05 considered statistically significant. Descriptive statistics will be presented for each group as the mean ± standard deviation and all enumeration data will be presented as constituent ratio and rate in the outcomes from baseline to each time point. Changes in differences between groups in the primary and secondary outcomes, before and after treatment and during the followup period (at 4 weeks and 12 weeks), will be analyzed using repeated measures ANOVA or a rank-sum test.
Changes in effect sizes between pre-and post-intervention will be performed using the Wilcoxon test and chi square test. The effect size will be evaluated with safety set (SS), full analysis set (FAS), per-protocol set (PPS) and costeffectiveness analysis. Demographic and CHC characteristics, as well as baseline data, will be analyzed using the FAS to assess the baseline and CHC comparability between groups. The primary analysis of the data will be undertaken using FAS with the principle of intention-totreat and the PPS. The secondary analysis of the data will be undertaken using PPS. To account for missing data, multiple imputation of missing follow-up measures will be performed per the methods used for missing data. The PPS will be used to assess adherence, the SS to assess adverse events and the cost-effectiveness analysis to assess health economics.
# Discussion
KOA is a major public health problem and this project, funded by the State Administration of Traditional Chinese Medicine, has two aims. First, the curative effect of CTCMRP for KOA rehabilitation will be assessed and CTCMRP will be examined to develop a SRPTCM for KOA rehabilitation in China and worldwide. Secondly, self-management for KOA patients will be improved and their financial burden reduced.
In this study, the rehabilitation therapists will undergo rigorous training and accreditation to administer the CTCMRP intervention. Regular meetings will be held throughout the study to exchange views. The outcome measures will include self-reported measures of pain, the SF-36, QOL and activities of daily living. Knee ROM, the 6 MW test, the stair-climbing test and the comprehensive effect will assess functional performance, strength and physical activity levels. In addition, a health economic assessment will be used to justify the cost-effectiveness of treatments in the current economic climate and a safety assessment will be done to assess adverse events.
The biggest limitation of this protocol is that there is no placebo control and it is not double-blinded. However, the evidence-based medicine center will be blind to the intervention to decrease possible bias. The study also lacks long-term follow-up and evaluation. The 4-week intervention period reflects what is done in actual clinical practice and is long enough to test whether a rehabilitation modality is effective in the short term and whether the efficacy is sustained during the follow-up period. Due to ethical considerations, non-steroidal anti-inflammatory drugs will be permitted as required.
If successful, this project offers a low-cost solution for efficiently linking millions of hospitalized KOA patients with effective outpatient treatment. CTCMRP rehabilitation therapy for KOA can be done in a CHC or at home instead of in a high-grade hospital, thus reducing the financial burden for the patient.
[fig] Figure 1: Study flow chart. [/fig]
[table] Table 1: Baseline descriptive data General data Agent of the person who conducted test, test start date, the patient's address, name of accompanying family members and the patient's telephone number Demographic data of subject Gender, age, nationality, level of education, profession, height and weight Assessment of knee before test Visual analogue scale (VAS), knee girth, knee's range of motion (ROM), manual muscle test (MMT), stair-climbing ability, walking distance, comprehensive effect, average daily amount of non-steroidal anti-inflammatory drugs taken, activities of daily living (ADL), quality of life (QOL), safety assessments and economic evaluation Other Complications, medication use, previous treatment and duration of KOA symptoms [/table]
[table] Table 2: TCM rehabilitation based on syndrome using appropriate technology [/table]
[table] Table 4: Tuina intervention [/table]
[table] Table 6: Summary of measures to be collected [/table]
|
IL-14α as a Putative Biomarker for Stratification of Dry Eye in Primary Sjögren’s Syndrome
## Figure 1. a fabricated western blot image opened in imagej. the information along the top of the image indicates that the image is currently in 8-bit mode, using an inverting lut (look-up table). the inverting lut ensures that dark bands will be recorded as higher density values.
3. Choose the Rectangular Selections tool from the ImageJ toolbar. Draw a rectangle around the first lane. ImageJ assumes that your lanes run vertically (so individual bands are horizontal), so your rectangle should be tall and narrow to enclose a single lane. If you draw a rectangle that is short and wide, ImageJ will switch to assuming the lanes run horizontally (individual bands are vertical), leading to much confusion. it over to the next lane. You can also use the arrow keys to move the rectangle, though this is slower. Center the rectangle over the lane left-to-right, but don't worry about lining it up perfectly on the same vertical axis. ImageJ will automatically align the rectangle on the same vertical axis as the 1st rectangle in the next step. 6. Press 2 or go to Analyze>Gels>Select Next Lane to set the rectangle in place over the 2nd lane. A 2 will appear in the lane when the rectangle is placed. 7. Repeat Steps 5 + 6 for each subsequent lane on the gel, pressing 2 each time to set the rectangle in place [fig_ref] Figure 3: Place the rectangle over each lane, pressing 2 each time to set... [/fig_ref]. 8. After you have set the rectangle in place on the last lane (by pressing 2), press 3, or go to Analyze>Gels>Plot Lanes to draw a profile plot of each lane. 9. The profile plot represents the relative density of the contents of the rectangle over each lane. The rectangles are arranged top to bottom on the profile plot. In the example western blot image, the peaks in the profile plot [fig_ref] Figure 4: The profile plot from the example western blot [/fig_ref] correspond to the dark bands in the original image [fig_ref] Figure 3: Place the rectangle over each lane, pressing 2 each time to set... [/fig_ref]. Because there were four lanes selected, there are four sections in the profile plot. Higher peaks represent darker bands. Wider peaks represent bands that cover a wider size range on the original gel. 10. Images of real gels or western blots will always have some background signal, so the peaks don't reach down to the baseline of the profile plot. [fig_ref] Figure 5: Real western blots often have some background noise, so the peak doesn't... [/fig_ref] shows a peak from a real blot where there was some background noise, so the peak appears to float above the baseline of the profile plot. It will be necessary to close off the peak so that we can measure its size. 11. Choose the Straight Line selection tool from the ImageJ toolbar [fig_ref] Figure 6: Use the straight line tool to close off the base of each... [/fig_ref]. For each peak you want to analyze in the profile plot, draw a line across the base of the peak to enclose the peak [fig_ref] Figure 5: Real western blots often have some background noise, so the peak doesn't... [/fig_ref]. This step requires some subjective judgment on your part to decide where the peak ends and the background noise begins. 12. Note that if you have many lanes highlighted, the later lanes will be hidden at the bottom of the profile plot window. To see these lanes, press and hold the space bar, and use the mouse to click and drag the profile plot upwards. 13. When each peak has been closed off at the base with the Straight Line selection tool, select the Wand tool from the ImageJ toolbar [fig_ref] Figure 8: Choose the Wand tool to highlight each peak of interest on the... [/fig_ref]. 14. Using the spacebar and mouse, drag the profile plot back down until you are back at the first lane. With the Wand tool, click inside the peak [fig_ref] Figure 9: Click inside the 1st peak of interest with the Wand tool [/fig_ref]. Repeat this for each peak as you go down the profile plot. For each peak that you highlight, measurements should pop up in the Results window that appears. 15. When all of the peaks have been highlighted, go to Analyze>Gels>Label Peaks. This labels each peak with its size, expressed as a percentage of the total size of all of the highlighted peaks. 16. The values from the Results window [fig_ref] Figure 10: The output from selecting peaks in the profile plot and labeling the... [/fig_ref] can be moved to a spreadsheet program by selecting Edit>Copy All in the Results window. Paste the values into a spreadsheet. Note: If you accidentally click in the wrong place with the Wand, the program still records that clicked area as a peak, and it will factor into the total area used to calculate the percentage values. Obviously this will skew your results if you click in areas that aren't peaks. If you do happen to click in the wrong place, simple go to Analyze>Gel>Label Peaks to plot the current results, which displays the incorrect values, but more importantly resets the counter for the Results window. Go back to the profile plot and begin clicking inside the peaks again, starting with the 1st peak of interest. The Results window should clear and begin showing your new values. When you're sure you've click in all of the correct peaks without accidentally clicking in any wrong areas, you can go back to Analyze>Gels>Label Peaks and get the correct results.
# Data analysis
With your data pasted into a spreadsheet, you can now calculate the relative density of the peaks. As a reminder, the values calculated by ImageJ are essentially arbitrary numbers, they only have meaning within the context of the set of peaks that you selected on the single gel image you've been working on. They do not have units of µg of protein or any other real-world units that you can think of. The normal procedure is to express the density of the selected bands relative to some standard band that you also selected during this process.
1. Place your data in a spreadsheet. One of the peaks should be your standard. In this example we'll use the 1st peak as the standard. 2. In a new column next to the Percent column, divide the Percent value for each sample by the Percent value for the standard (the 1st peak in this case, 26.666). 3. The resulting column of values is a measure of the relative density of each peak, compared to the standard, which will obviously have a relative density of 1. 4. In this example, the 2nd lane has a higher Relative Density (1.86), which corresponds well with the size and darkness of that band in the original image . Recall that these data are for the upper row of bands on the original western blot image. 5. If you want to compare the density of samples on multiple gels or blots, you will need to use the same standard sample on every gel to provide a common reference when you calculate Relative Density values. See the sections below for more detailed discussion of these requirements. 6. In order to test for significant differences between treatments in an experiment, all of your gels or blots will need to be scanned and quantified using this method, and the values will be expressed in terms of Relative Density, or you can treat Relative Density as a fold-change value (i.e. a Relative Density difference of 2 between a control and treatment would indicate a 2-fold change in expression). If you will be using analysis of variance techniques to test your data, you may need to ensure that your Relative Density values are normally distributed and that there is homogeneity of variance among the different treatments. 7. It should be noted here that some researchers make the extra effort to include a set of serial dilutions of a known standard on each blot. Using the serial dilution curve and the quantification techniques outlined above, it should be possible to express your sample bands in terms of picograms or nanograms of protein A more involved example using loading-controls.
We'll use [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref] as a representative western blot. On this blot, we will pretend that we loaded four replicate samples of protein (four pipette loads out of the same vial of homogenate), so we expect the densities in each lane to be equivalent. The upper row of bars will represent our protein of interest. The lower set of bars will represent our loading-control protein, which is meant to ensure that an equal amount of total protein was loaded in each lane. This loadingcontrol protein is a protein that is presumably expressed at a constant level regardless of the treatment applied to the original organisms, such as actin (though many people will question the assertion that actin will be expressed equivalently across treatments). Looking at [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref] , we had hoped to load equivalent amounts of total protein in each lane, but after running the western blot, the size and intensity of the lower bars in each lane varies quite a lot. The two left lanes appear equivalent, but the 3rd lane has half the density (gray value) compared to lanes 1+2, while lane 4 has half the density and half the size compared to lanes 1+2. Because our loading controls are so different, the density values of the upper set of bands may not be directly comparable.
We'll use ImageJ's gel analysis routine to quantify the density and size of the blots, and use the results from our loading-controls (lower bands) to scale the values for our protein of interest (upper bands).
1. Open the western blot image in ImageJ.
2. Make sure that the image is in 8-bit mode: go to Image>Type>8-bit.
3. Use the rectangle tool to draw a box around the entire 1st lane (both upper and lower bars included. 4. Press "1" to set the rectangle. A "1" should appear in the middle of the rectangle. 5. Click and hold in the middle of the rectangle and drag it over the 2nd lane. 6. Press "2" to set the rectangle for lane 2. A "2" should appear in the middle of the rectangle. 7. Repeat steps 5 + 6 for each subsequent lane, pressing "2" to set the rectangle over each subsequent lane (see [fig_ref] Figure 3: Place the rectangle over each lane, pressing 2 each time to set... [/fig_ref]. 8. When you have placed the last lane (and pressed "2" to set it in place), you can press "3" to produce a plot of the selected lanes (see [fig_ref] Figure 4: The profile plot from the example western blot [/fig_ref]. 9. The profile plot essentially represents the average density value across a set of horizontal slices of each lane. Darker blots will have higher peaks, and blots that cover a larger size range (kD) will have wider peaks. In our example western blot, the bands are perfect rectangles, but you will notice some slope in the profile plot peaks, as ImageJ is applying a bit of averaging of density values as it moves from top to bottom of each lane. As a result, the sharp transition from perfect white to perfect black on the bands of lane 1 is translated into a slight slope on the profile plot due to the averaging. 10. On our idealized western blot used here, there is no background noise, so the peak reaches all the way down to the baseline of the profile plot. In real western blots, there will be some background noise (the background will not be perfectly white), so the peaks won't reach the baseline of the profile plot [fig_ref] Figure 5: Real western blots often have some background noise, so the peak doesn't... [/fig_ref]. As a result, each plot will need to have a line drawn across the base of the peak to close it off. 11. Choose the Straight Line selection tool from the ImageJ toolbar. For each peak you want to analyze in the profile plot, draw a line across the base of the peak to enclose the peak [fig_ref] Figure 6: Use the straight line tool to close off the base of each... [/fig_ref]. This step requires some subjective judgment on your part to decide where the peak ends and the background noise begins. 12. When each peak of interest is closed off with the straight line tool, switch to the Wand tool. We will use the wand tool to highlight each peak of interest so that ImageJ can calculate its relative area+density. 13. We will start by highlighting the loading-control bands (lower row) on our example western blot. Beginning at the top of the profile plot, use the wand to click inside the 1st peak [fig_ref] Figure 7: The peak from Figure 5 is shown with a line drawn across... [/fig_ref]. The peak should be highlighted after you click on it. Continue clicking on the loading-control peaks for the other lanes. If a lane is not visible at the bottom of the profile plot, hold down the space bar and click-and-drag the profile plot upwards to reveal the remaining lanes. 14. When the loading control peak for each lane has been highlighted with the wand, go to Analyze>Gel>Label Peaks. Each highlighted peak will be labeled with its relative size expressed as a percentage of the total area of all the highlighted peaks. You can go to the Results window and choose Edit>Copy All to copy the results for placement in a spreadsheet. 15. Repeat steps 13 + 14 for the real sample peaks now. We are selecting these peaks separately from the loading-control peaks so that those areas are not factored into the calculation of the density of our proteins-of-interest. As before, use the Wand tool to click inside the area of the peak in the 1st lane, then continue clicking inside the peaks of the remaining lanes. When finished, go to Analyze>Gel>Label Peaks to show the results.
Copy the results to a spreadsheet alongside the data for the loading-control bands [fig_ref] Figure 9: Click inside the 1st peak of interest with the Wand tool [/fig_ref].
## Figure 19. results from the loading-control bands (labeled "stds" here) and the sample protein peaks (the upper row of bands on the original western blot).
# Data analysis
1. With all of the relative density values now in the spreadsheet, we can calculate the relative amounts of protein on the western blot. Remember that the "Area" and "Percent" values returned by ImageJ are expressed as relative values, based only on the peaks that you highlighted on the gel. Start the analysis by calculating Relative Density values for each of the loading-standard bands. In this case, we'll pretend that Lane 1 is our control that we want to compare the other 3 lanes to. Divide the Percent value for each lane by the Percent value in the control (Lane 1 here) to get a set of density values that is relative to the amount of protein in Lane 1's loading-control band [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. 2. Next we'll calculate the Relative Density values for our sample protein bands (upper row on the example western blot). We carry out a similar calculation as step 1, dividing the Percent value in each row by the Percent value of our control's protein band (Lane 1 here). Note: Recall that because some of our loading-control bands were wildly different on the original western blot, we can't simply use the Relative Density values from our Samples calculated in Step 2 as the final results. Now it is necessary to scale the Relative Density values for the Samples by the Relative Density of the corresponding loading-control bands for each lane. We do this based on the assumption that the proportional differences in the Relative Densities of the loading-control bands represent the proportional differences in amounts of total protein we loaded on the gel. In our example western blot, we have evidence of massively different amounts of total protein in each sample (poor pipetting practice, probably). 4. When we created our imaginary example western blot, we expected to get equivalent amounts of the protein of interest in each lane, because we (hypothetically) loaded equivalent sample volumes of protein from the same vial of homogenate in each lane. During the loading process, something went awry, and we got very different densities on the western blot. But after carrying out the loading-control corrections above, we find that our Adjusted Densities for each of the 4 lanes are roughly equivalent (all are approximately = 1), indicating that there are equivalent ratios of the protein of interest contained within the sample of total protein in each lane, as we would expect.
The example above represents a special case. In reality, you will typically be skilled enough to pipette equivalent amounts of total protein into your gel lanes, presumably because you have used a process like the BCA Protein Assay to quantify the total amount of protein present in each of your homogenate samples. This can obviate the need for the loading-control bands and the associated controversy about whether the protein you're using to assess equal loading is truly expressed consistently across all treatment levels.
## Extending the example to multiple western blots
If you can apply equal amounts of total protein to each lane of a gel, then you only need to concern yourself with having an appropriate standard sample placed on each western blot you run for the project. For example, you might purchase an aliquot of purified Human Hsp70 and place 1µl in a lane on each gel you run when probing for Hsp70. Or you might save some money by creating your own standard sample by mixing aliquots of several of your homogenates to make a large quantity of mixed homogenate that you can use to load a standard volume onto a single lane of each gel for your entire project (running out of this mixture part way through your project would be a Bad Thing). In this case you may not know the true concentration of the protein of interest in your mixed homogenate, but you will at least be able to get an equivalent amount of total protein (including the protein of interest) on each gel.
The goal of the standard sample is to account for variations in antibody binding efficiency (for example if you're re-using a mixture of antibody on multiple western blots to save money), for variations in blocking and washing efficiency, for variation in the decay rate of a chemiluminescent reagent used to visualize the antibodies, or for variation in the exposure time of your x-ray film. In any of these cases, the intensity of the signal you get out of your western blot can vary from blot to blot [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. You can use a standard sample on each gel to normalize every other band on that same blot, and then compare across multiple blots because every band in your dataset is normalized to the same standard (be it 10ng of Human Hsp70 or a mixture of several homogenates). Finally, if you are using a loading-control protein band (lower row on these blots) on each gel, and a standard protein sample on multiple gels, you can correct both for differences in total protein loaded in a lane (using the loading-control band) and for differences in exposure between gels (using the standard sample).
For example, consider the two gels in [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. The 1st (white) gel has minimal background noise, and there is some variation the amount of protein loaded in each lane, as revealed by the size and density of the loading control bands of the samples. The 1st lane on the left is our standard sample used on each gel. The 2nd (gray) gel has more background noise due to a different exposure time, poor washing, or any number of possible problems. Again, there are different protein quantities loaded in the 4 lanes, based on the size and density of the loadingcontrol bands on the lower half. The 1st lane on the left is the same standard protein sample as that used on the 1st gel.
We would start by analyzing the two gels separately, using the protocol at the beginning of this document.
1. Calculate the relative density of the loading-control bands (lower row on this gel). Use the loading-control band for the standard in lane 1 to do this calculation. 2. Calculate the relative density of the bands for the protein band of interest (upper row on this gel). Use the protein band of interest for the standard in lane 1 to do this calculation. 3. Adjust the relative densities calculated in Step 2 using the relative densities for the loading-control bands calculated in Step 1. Simply divide the relative density from
Step 2 for each lane by the corresponding relative density from Step 1.
After you've scaled the density of the protein of interest for each sample to the standard sample, and adjusted the relative density based on the loading-control band, you are left with an estimate of the relative density of the protein of interest in each lane on your gel (relative to the standard in lane 1). Repeat this for each separate gel. When finished, you'll note that because every sample lane is normalized to the standard sample in Lane 1 on that particular gel, and every gel contains the same standard sample, your cross-gel comparison is already accomplished, since every sample is now normalized to the same standard sample [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. After all of this work, you're left with (adjusted) Relative Densities for each of your samples (the yellow cells in [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. At this point you can ignore the values of the standard samples, since you only needed those to ensure that your cross-gel comparisons were scaled properly. Now you can finally begin calculating average relative density values for your experimental control samples (maybe these are samples 3+6) and your experimental treatment samples (perhaps these are samples 1+4 and 2+5) to carry out the real comparison of protein expression under your treatment conditions. Because your relative density values are all normalized to the standard sample, you may wish to eventually re-scale your relative density values once again using your experimental controls' average value as a baseline.
## A few extra details:
With regard to the ImageJ gel analysis routine, there has been some question of what the values reported by ImageJ correspond to. The images below may help illustrate what ImageJ is measuring. [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. A comparison of "area" values returned by ImageJ.
In [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref] above, I have drawn out a set of fake "bands" in Adobe Illustrator. The gray value and area of each band are listed above the band. Additionally, I have included the "area" value returned by ImageJ after plotting the bands and clicking in each peak with the Wand tool. Note that these "area" values are a RELATIVE measure of the size and density of each peak you clicked with the wand tool. When you halve the area of a band, but maintain the same gray value (compare lanes 1+2), the value reported by ImageJ is half as large. By the same token, if you halve the gray value but maintain the same area (compare lanes 1+5), the value reported by ImageJ is halved. [fig_ref] Figure 2: The first lane has been selected with the Rectangular Selections tool 4 [/fig_ref]. ImageJ will return similar area values when bands have equivalent numbers of pixels of equal darkness, but vary in shape.
[fig] Figure 2: The first lane has been selected with the Rectangular Selections tool 4. After drawing the rectangle over your first lane, press the 1 key or go to Analyze>Gels>Select First Lane to set the rectangle in place. The 1st lane will now be highlighted and have a 1 in the middle of it. 5. Use your mouse to click and hold in the middle of the rectangle on the 1st lane and drag [/fig]
[fig] Figure 3: Place the rectangle over each lane, pressing 2 each time to set the rectangle in place. [/fig]
[fig] Figure 4: The profile plot from the example western blot. [/fig]
[fig] Figure 5: Real western blots often have some background noise, so the peak doesn't reach down to the baseline ( = perfect white). [/fig]
[fig] Figure 6: Use the straight line tool to close off the base of each peak of interest. [/fig]
[fig] Figure 7: The peak from Figure 5 is shown with a line drawn across the base of the peak to enclose the area of the peak. [/fig]
[fig] Figure 8: Choose the Wand tool to highlight each peak of interest on the profile plot. [/fig]
[fig] Figure 9: Click inside the 1st peak of interest with the Wand tool. The peak should be highlighted and a measurement should pop up in the Results window. [/fig]
[fig] Figure 10: The output from selecting peaks in the profile plot and labeling the peaks. In this case, the results are from selecting the upper band in each of the 4 lanes in the example western blot from Figure 1. [/fig]
[fig] Figure 11: Calculating relative density for each of the 4 highlighted peaks. We use Sample 1 as the standard for this gel. The Relative Density is calculated by dividing the Percent value for each sample by the Percent value for the standard. [/fig]
[fig] Figure 12: An example western blot with a lower row of loading-control bands. Often these bands will represent a protein that is thought to be expressed equivalently in all treatments, such as actin. [/fig]
[fig] Figure 13: Each lane has been selected by moving the rectangle over it and pressing "2" to set the rectangle in place. [/fig]
[fig] Figure 14: Profile plot of each lane (arranged with lane 1 at the top, lane 4 at the bottom). You may need to scroll through the window to see all of the lanes. [/fig]
[fig] Figure 15: Real western blots often have some background noise, so the peak doesn't reach down to the baseline ( = perfect white). [/fig]
[fig] Figure 16: The peak from Figure 15 is shown with a line drawn across the base of the peak to enclose the area of the peak. [/fig]
[fig] Figure 17: Click inside the loading-control peak with the wand tool. Click on each of the loading-control peaks for the remaining lanes (ignore the protein lanes to the left for now). [/fig]
[fig] Figure 18: Area and Percent values for our loading-control bands from the example western blot. Lanes 1 and 2 are equal, as we expect by looking at the original bands on the western blot(Figure 12). [/fig]
|
National Mass Drug Administration Costs for Lymphatic Filariasis Elimination
Background: Because lymphatic filariasis (LF) elimination efforts are hampered by a dearth of economic information about the cost of mass drug administration (MDA) programs (using either albendazole with diethylcarbamazine [DEC] or albendazole with ivermectin), a multicenter study was undertaken to determine the costs of MDA programs to interrupt transmission of infection with LF. Such results are particularly important because LF programs have the necessary diagnostic and treatment tools to eliminate the disease as a public health problem globally, and already by 2006, the Global Programme to Eliminate LF had initiated treatment programs covering over 400 million of the 1.3 billion people at risk.Methodology/Principal Findings: To obtain annual costs to carry out the MDA strategy, researchers from seven countries developed and followed a common cost analysis protocol designed to estimate 1) the total annual cost of the LF program, 2) the average cost per person treated, and 3) the relative contributions of the endemic countries and the external partners. Costs per person treated ranged from $0.06 to $2.23. Principal reasons for the variation were 1) the age (newness) of the MDA program, 2) the use of volunteers, and 3) the size of the population treated. Substantial contributions by governments were documented -generally 60%-90% of program operation costs, excluding costs of donated medications.Conclusions/Significance: MDA for LF elimination is comparatively inexpensive in relation to most other public health programs. Governments and communities make the predominant financial contributions to actual MDA implementation, not counting the cost of the drugs themselves. The results highlight the impact of the use of volunteers on program costs and provide specific cost data for 7 different countries that can be used as a basis both for modifying current programs and for developing new ones.
# Introduction
Lymphatic filariasis (LF), commonly known as elephantiasis, is a profoundly disfiguring parasitic disease caused by thread-like nematode worms. The World Health Organization (WHO) places the number of people at risk in 83 countries at 1.307 billion.Globally, the reduced productivity as a consequence of LF disability has been well recognized. The chronic and debilitating burden of LF maintains the cycle of poverty not only in infected individuals but also in entire endemic communities. [bib_ref] The economic burden of lymphatic filariasis in India, Ramaiah [/bib_ref] [bib_ref] The economic burden of lymphatic filariasis in northern Ghana, Gyapong [/bib_ref] Indeed, as a disease of poverty, LF is endemic in 43 of the 50 countries classified as least developed nations. [bib_ref] The elimination of lymphatic filariasis: A strategy for poverty alleviation and sustainable..., Galvez Tan [/bib_ref] [bib_ref] Editorial: Lymphatic filariasis endemicity-an indicator of poverty?, Durrheim [/bib_ref] The 1997 World Health Assembly resolved to eliminate LF as a public health problem after LF had been identified as one of only a small number of diseases classified as potentially eradicable.The principal strategy for elimination relies on once-yearly concurrent administration of two drugs, albendazole with diethylcarbamazine (DEC) or albendazole with ivermectin -both regimens shown to be highly effective in removing microfilariae from the blood for a full year after treatment. [bib_ref] Efficacy of single dose combinations of albendazole, ivermectin and diethylcarbamazine for the..., Ismail [/bib_ref] Administration of these once-yearly, single-dose regimens to people in at-risk communities for 4-6 years makes feasible the prospect of interrupting transmission and thereby eliminating LF, [bib_ref] The global programme to eliminate lymphatic filariasis, Ottesen [/bib_ref] largely because the reproductive life span of the adult worm is estimated to be 4-6 years. While the delivery of DEC fortified salt is another strategy that has been applied in some regions of the world to eliminate LF, this strategy was not in the purview of the current studies.
Demonstrating that LF elimination is a cost-effective and affordable investment is essential for both Ministries of Health and potential donors as they choose among competing health needs. While several cost-effectiveness analyses have been conducted and one estimate for MDA in high prevalence areas gives a range of $4-8 per disability adjusted life year (DALY) averted, [bib_ref] The economic burden of lymphatic filariasis in India, Ramaiah [/bib_ref] [bib_ref] Treatment costs and loss of work time to individuals with chronic lymphatic..., Ramaiah [/bib_ref] [bib_ref] Cost-effectiveness of the use of vector control and mass drug administration, separately..., Krishnamoorthy [/bib_ref] [bib_ref] The Leogane, Haiti demonstration project: decreased microfilaremia and program costs after three..., De Rochars [/bib_ref] [bib_ref] Advancement of global health: key messages from the Disease Control Priorities Project, Laxminarayan [/bib_ref] more cost and cost-effectiveness data are lacking. The persistent need for detailed costing of mass drug administration (MDA) programs led to the present initiative to develop and implement a common cost-analysis protocol. Countries participating in this multi-country study were selected to represent the different stages and scope of national LF elimination efforts, different economic conditions of endemic countries, and different geographic regions affected by the disease. Health expenditures per capita in the participating countries for 2001-2002 ranged from $12 in Tanzania and Ghana, to $22 in Haiti, $27 in Burkina Faso, $30 in the Philippines, $46 in Egypt and $153 in the Dominican Republic.The cost analysis objectives were to 1) estimate total annual costs of the LF program for a specific year or more (depending on the availability of information), 2) calculate the average cost per person treated, 3) identify the relative contributions of the endemic country and external partners, 4) provide data that could be used for program evaluation and analysis, 5) understand how the results differed among countries, and 6) build a framework for the development and implementation of cost studies elsewhere.
Cost analyses aggregated in this report are based on studies in Burkina Faso, the Dominican Republic, Egypt, Ghana, Haiti, the Philippines, and Tanzania. Detailed analyses from two of these countries have already been published elsewhere. [bib_ref] The Leogane, Haiti demonstration project: decreased microfilaremia and program costs after three..., De Rochars [/bib_ref] [bib_ref] Defining the cost of the Egyptian lymphatic filariasis eliminination program, Ramzy [/bib_ref]
# Methods
## Study teams
Data collection and analysis were carried out by research teams in each of the participating countries. The cost analysis teams in Burkina Faso, Tanzania, and Ghana were part of the Ministry of Health. In Egypt the study was a joint effort between the Ministry of Health and Population and the Ain Shams University Research and Training Center on Vectors of Diseases. The Philippines' research team was based at the National Institutes of Health-Philippines. The Dominican Republic researchers at the Centro de Gerencia Social, Instituto Tecnologico de Santo Domingo worked with the Ministry of Health LF program. External consultants in Haiti worked with Ministry of Health LF program staff to collect and analyze cost study data.
## Study protocol
Researchers from the participating countries collaborated with the Emory University LF Support Center in the development of a protocol (http://www.taskforce.org/lfsc/professionals.html) which served as the tool to identify, organize, analyze, summarize and present the cost data. The protocol and accompanying data collection instrument created a systematic process for data collection and analysis so that the cost estimates would be comparable across a variety of country settings, programmatic approaches and program sizes.
The cost analysis identified both economic costs (i.e., the value of all resources used in the program, including donated items, such as medications for the MDA) and financial costs (i.e., the actual cash disbursements for a program including resources provided by the national government and local communities). Economic costs are useful to evaluate the allocation of program resources and their opportunity costs, e.g., whether these resources could be used more productively elsewhere. Financial costs are helpful to program managers to measure actual program expenditures and assess affordability.(See for the classification of costs.) Financial costs include all costs with the exception of donated materials. Capital items were annualized to reflect costs incurred in one year for the project. They do not include bicycles because these were the property of the volunteer drug distributors, and thus considered a donation. Volunteer training per diems are classified as Financial and Economic costs, but the value of volunteer unpaid time is not included in either Financial or Economic costs. Because the capital costs were annualized, the cash expenditures described reflect those incurred in one year for the project.
The protocol adopted a national program perspective because most resources dedicated to the LF MDAs were channeled through Ministry of Health (MOH) programs. Participants agreed to gather MDA costs, beginning with the year 2000, the first year of the LF global elimination effort. Costs were calculated in local currencies and converted to US dollars for the final analysis, based on average exchange values for the years being analyzed (base year 2002).
Most of the countries were still in the early stages of expanding their MDAs when the cost analysis studies began. The exception was Egypt where the MDA targeted almost 90% of the population at risk the first year and 100% the second year studied.
Project inputs defined in the protocol were: personnel, supplies and drugs (medications for the MDA and for treatment of adverse reactions), as well as capital and recurrent costs for equipment, transportation, and facilities. Each input was allocated to one or more categories of activities involved in accomplishing the MDA: training, mapping, mobilization and education, drug distribution,
## Author summary
Lymphatic filariasis (LF), commonly known as elephantiasis, is a profoundly disfiguring parasitic disease caused by thread-like nematode worms. This disease can often be disabling, thus reducing the potential productivity of the affected individuals. The WHO places the number of people at risk in 83 countries at 1.307 billion. This study was undertaken in seven countries-Burkina Faso, Ghana, Egypt, Tanzania, the Philippines, the Dominican Republic, and Haiti-using a common protocol to determine the costs of mass drug administration (MDA) programs to interrupt transmission of infection with LF, because there is lack of sufficient information about the costs of these programs. The results demonstrate that LF MDA is affordable and relatively inexpensive when compared to other public health programs. In the context of initiatives for integrating programs for the control and elimination of neglected tropical diseases, this study adds specifically to the relatively scarce body of information about the costs of MDA programs for LF. It also adds to the general knowledge about the application of methods that can be used to estimate the costs and cost-effectiveness of an integrated approach.
adverse reaction monitoring, surveillance/laboratory, and administration . Capital costs were defined as one-time investments in physical goods that have a life longer than one year and generally cost more than $1000. These costs were annualized, using a formula that accounted for years of useful life, scrap value, and a discount rate set at 3%. Recurrent costs included those items that were consumed on a regular basis; i.e., personnel time, office and laboratory supplies, fuel, and drugs. Recurrent costs for maintenance of donated capital items such as bicycles were also included. In cases where facilities or equipment costs were not available, costs of similar facilities or rentals were used as a proxy. For the purpose of Economic costs, the value of the donated drugs was set as $0. [bib_ref] Community-directed treatment with ivermectin in two Nigerian communities: an analysis of first..., Onwujekwe [/bib_ref]
## Study sites
Study sites were chosen purposively to be representative of the LF-endemic regions in the countries [fig_ref] Table 3: Country Background Information [/fig_ref]. National estimates were developed using the data from these representative study sites.
Tanzanian investigators studied four districts located on the eastern coast, Kilwa, Mafia, Mkuranga and Masasi for the years 2000-2003. Data collection for the first three years was retrospective.
Data collected retrospectively in Burkina Faso documented the first two years of the MDA in the region of Gaoua in the first year and expansion into Tenkodogo in the second year.
The Philippines team chose to study seven municipalities and one city in the province of Sorsogon. These were selected based on filariasis endemicity using microfilaria rates, accessibility and population size. Half had high microfilaria rates (MF), half were easily accessible, while one quarter had large populations. In each municipality, one sentinel barangay or village was selected based on high MF rates and one to three adjacent barangays were also studied.
Cost data collection in Haiti was conducted in Leogane, where the first MDAs took place. Located 30 km west of Port-au-Prince in the south, Leogane is one of the highest risk areas in the country. Except for lower coverage in year-2 (2001) of the MDA program (because of side reactions that occurred after the first MDA) the program expanded not only in Leogane in 2002 but also to Milot, another high risk area just south of Cap Hatien in the north.
Researchers in the Dominican Republic collected cost data for the first two MDA campaigns. These were launched in Barahona on the southwest coast of the country.
The Ghana study costed three districts from the two epidemiological zones in the country, north and south. The districts were selected to reflect differing levels of program assistance. Builsa in the upper eastern region of the country received government support only, while Lawra in the upper west and Ahanta West on the west coast also received NGO support. . MDA Activities and Other Cost Categories.
## Activity definition
Training Instruction of MOH personnel to carry out the administrative and functional activities of the MDA and instruction of volunteers to develop skills required for the MDA.
## Mapping
Testing to establish microfilaria and/or antigenaemia prevalence in communities.
## Mobilization and education
Media campaigns and community activities to increase MDA participation.
## Drug distribution
Logistic aspects of management of the drugs as well as administration of the drugs to the population.
## Adverse reaction monitoring
Observation and treatment of persons suffering adverse reactions due to the MDA.
## Surveillance and laboratory
Tracking of community members in MDA area, laboratory work for testing, case identification, etc.
## Administration
Supervisory work and paperwork to support the MDA.
## Lf non-mda costs
Costs related to LF, but not the MDA (excluded from cost calculations).
Other The cost analysis in Egypt covered the eight governorates along the Nile affected by LF. Costs were obtained for one district in each governorate and applied to the rest of the affected areas in the governorate based upon information about numbers of persons at risk, participating government personnel and quantities of medication distributed. [bib_ref] Defining the cost of the Egyptian lymphatic filariasis eliminination program, Ramzy [/bib_ref] Data Collection Data collection was both retrospective and prospective. The number of rounds of MDA costed per country ranged from one to four, with most countries costing two rounds. Countries developed national estimates for the program using data collected from a sampling of sites representative of the program [fig_ref] Table 3: Country Background Information [/fig_ref]. Coverage rates are defined as the number of individuals reported to have ingested the antifilarial drugs divided by the total at-risk population in the program area. Those excluded from treatment included pregnant women, lactating women in the 1 st week post-partum, the very sick, children under two years of age in countries where DEC plus albendazole is the MDA regimen, and children under 90 cm in height (generally under 5 years of age) where albendazole is administered with Mectizan.
The data were collected from national, regional and district levels of the health care system via pre-tested questionnaires and spreadsheets, sometimes with the assistance of other agencies such as Ministries of Agriculture and Information. As LF is but one of many population-based health programs, most inputs (including personnel time, facilities, equipment, supplies, vehicles and fuel) were often shared by more than one program, and costs were apportioned accordingly. To capture the actual costs and percentage of the resources dedicated to LF, the teams reviewed program records and interviewed LF program administrators and personnel about allocations of personnel and resource time per year. Government tax fees such as customs tax on the drugs and the road tax were excluded.
# Results
## Country mda costs
As indicated in [fig_ref] Table 4: Financial and Economic per Person Treated [/fig_ref] , the Financial costs per person treated ranged from $0.06 to $2.23 while Economic costs varied between $0.40 and $5.87. MDA coverage rates in the study populations ranged from 53% to 91%. While cost per person at risk can be easily calculated, cost per person treated is the more useful summary of costs for the purposes of planning and operations.
Of the several trends that can be seen in [fig_ref] Table 4: Financial and Economic per Person Treated [/fig_ref] , the most notable is that for those countries for which there is more than one year of data, cost per person treated decreased after the first year of the program, especially as the number of persons treated increased (see . In addition, the Financial costs per person treated for Burkina Faso, Ghana, Tanzania and the Philippines, all of which used volunteers, were the lowest among the seven countries participating in the study.
## Activities and inputs
Identification of resource allocation by activity and input is a useful and informative outcome of this study.
The use of resources for different activities varied among countries, and [fig_ref] Table 5: Costs by Activity -Percentage of Financial Costs for a 'Non-Start-Up' MDA Round [/fig_ref] identifies the proportion of the average national Financial costs expended for each activity in a 'non-start-up' MDA round (since yearly costs tend to stabilize after the 'start-up' year). Drug distribution generally represented the largest proportion of financial expenditures (average of 46%), with social mobilization/ education and administration being next most prominent.
Analysis of financial costs by input, particularly useful for projecting budgets and for gauging the need for additional program support, again gave results varying appreciatively by country [fig_ref] Table 6: Costs by Input -Percentage of Financial Costs for a 'Non-Start-Up' MDA Round [/fig_ref]. In all but one country, Egypt, the input that consumed the largest proportion of financial resources was personnel, averaging 53%, followed by supplies, equipment/ facilities and transportation.
# Funding sources
These cost analyses not only identified how program resources were allocated, but data also were used to identify sources and amounts of funding for 5 of the 7 study countries. Funding was categorized as coming from national governments (excluding external donations for LF), international organizations (IDAs, NGOs, WHO), pharmaceutical companies and local communities [fig_ref] Table 7: Sources of Funding for LF Elimination Costs [/fig_ref]. As expected, the drug donations represent a large proportion of contributions to MDA programs, so that when MOH and partner contributions are examined from the perspective of Economic costs, the drug donations can make up over 90% (range 9%-99%) of the costs. This was particularly true in countries like Burkina Faso and Tanzania where both drugs used in the MDA (albendazole and Mectizan) were donated. When Financial costs are analyzed, however, it is clear that contributions from national governments represent a significant portion of the resources used to implement the MDAs (average = 56%, range 9%-99%). These relationships can be seen graphically in [fig_ref] Figure 2: Sources of Funding for the MDAs Conducted by National LF Elimination Programs... [/fig_ref] which presents the combined average of Financial and Economic funding sources for the Burkina Faso, the Dominican Republic, Egypt, the Philippines and Tanzania programs detailed in [fig_ref] Table 7: Sources of Funding for LF Elimination Costs [/fig_ref].
## Sensitivity analyses
Sensitivity analyses were conducted on the personnel input (data not shown) to gauge how much an increase in personnel costs would impact overall costs. Personnel was selected for sensitivity analysis because, with the exception of Egypt, this input represented the largest source of costs in the participating countries. Sensitivity analyses conducted on personnel input for
# Study limitations
There were three principal study limitations. First, investigators frequently encountered problems in estimating the proportion of time and money allocated specifically to the LF MDA programs. Staff had not previously tracked time apportioned to LF-specific activities, and since LF programs are small in relation to other MOH initiatives (consuming anywhere from 1% to 5% a year in most countries), recall was problematic at times. Similar situations were encountered with respect to other resources such as
# Discussion
Good cost analysis is essential to provide an evidence-based rationale for investing in LF elimination. The key target figure is the cost per person treated. Few results have been available to date, [bib_ref] The Leogane, Haiti demonstration project: decreased microfilaremia and program costs after three..., De Rochars [/bib_ref] [bib_ref] Defining the cost of the Egyptian lymphatic filariasis eliminination program, Ramzy [/bib_ref] and little has been documented about the principal determinants of cost per person treated. This study has defined key activities and input categories and gathered cost information via the standardized cost-analysis studies of seven national LF MDA programs.
One of the most striking findings of the present studies is the variability in the Financial cost per person treated. Differences in programmatic structure can account for some of these differences because each country tailors the LF elimination strategy for a best fit to local conditions. But beyond the program-specific differences, the principal determinants underlying the variability appeared to be: 1) the newness of the country program (i.e., the number of years it had been running), 2) the use of volunteers, and 3) the size of the population treated. typically cover a limited geographic area and a relatively small population in their first year. As the population covered expands after the first year, the cost per person treated tends to drop.
## Newness or round of mda
In addition to extra expenditures in a start-up year as compared to other years, costs can also decrease over time because new costsavings strategies are identified and implemented. For example, in the first MDA year in the Dominican Republic, health workers from non-governmental organizations participated in the MDA; these individuals were paid per diem for all of the time they participated, in some cases higher than the wages of workers in the health system. MDA integration into the health system contributed to the 50% decrease in the second year Program cost per person treated. Similarly in Haiti, after the first year cost analysis revealed that 22% of resources were dedicated to adverse reactions, the adverse reaction protocol was revised to an equally effective but less costly strategy. In the second year only 4% of resources were used for adverse reactions.
## Use of volunteers
The use of volunteers had the greatest impact on costs. In Burkina Faso, Tanzania, Ghana, and the Philippines -where Program costs are lowest -health workers are employed down to the sub-district level, and volunteers, who are compensated very little in the LF program, work at the village level. Volunteers in these countries contribute a large proportion of the time dedicated to the MDA (frequently receiving per diem only for days in training) and at times provide their own transportation (e.g., Burkina Faso). However, while there is a strong relationship between the use of volunteers and lower cost per person treated, this does not suggest that any country choosing to use volunteers would see savings of 85%. Country-specific conditions that lead to the use of volunteers may also determine lower costs overall.
The use of volunteers leads to the question of how to accurately value the time volunteers 'donated' to each program. One traditional approach to valuing volunteer time is to apply the wage from the volunteer's normal paid employment and value their volunteer time accordingly. The problem with this, as pointed out by McFarland et al.in a report on the costs of onchocerciasis MDA, is that many volunteers are subsistence farmers who do not participate in the formal labor market. One alternative is to decide the fair market value of the time, i.e. the amount the volunteer would be paid if the program had to hire individuals for the work, [bib_ref] Community-directed treatment with ivermectin in two Nigerian communities: an analysis of first..., Onwujekwe [/bib_ref] or another, by using an estimate from prior studies in a similar setting.In the country programs included in this study, a diverse group of individuals, comprising students, teachers, farm laborers, and elderly retirees, served as volunteers. Owing to this fact and the economic conditions in the participating countries, applying traditional methodologies for costing volunteer time may not be appropriate. After several countries explored volunteer participation in MDAs and the earning capacity of volunteers in their regular pursuits, a decision was made by the investigators in all participating countries not to include these costs. Therefore, while the study included the per diem paid to volunteers during training, it did not account for the opportunity cost of the volunteer's time dedicated to the MDA itself. The evaluation of the community contribution from countries which used volunteers is very definitely underestimated.
Program managers can control, to some extent, the use of volunteers and might explore this strategy in resource-constrained environments. Beyond the monetary savings, there is a benefit from connecting the program with volunteers who are opinion leaders from different community sectors. So while initially decisions to utilize volunteers might be financially based, countries ultimately can benefit not only from volunteers' labor but also from their connections to the communities in which they serve. However, program managers must bear in mind the competition for volunteers from other health programs which sometimes pay volunteers more for their efforts. The sensitivity analyses on personnel time devoted to the MDA point to the possibility that it may be worthwhile exploring opportunities to increase remuneration for volunteers. Indeed, both valuing and best utilizing volunteer time merit continued examination.
## Size of population treated
The third principal source of variation in cost per person treated was the size of the population treated, an element that can be controlled by program managers during program expansion. When programs scale-up, the cost per person treated drops, primarily because most of the overhead costs are associated with start-up costs at the national and district levels. Such findings emphasize the need to keep current programs adequately funded so that these programs can expand and increase the number of persons treated and thereby capture the savings resulting from the economies of scale.
Once the principal determinants of the cost per person treated are identified, they can be used to manage program costs either by making internal changes within a program, such as scaling-up and increasing efficiency, or by taking advantage of existing external resources, as through integration. LF elimination programs can be integrated into the existing health system, as in the Dominican Republic, or with other preventive chemotherapy programs. For the LF elimination program, costs per person treated are within the range of those estimated for other similar disease control and elimination programs; namely soil-transmitted helminths ($0.25 per treatment, trachoma ($0.50 per treatment [bib_ref] The global elimination of blinding trachoma: progress and promise, Kumaresan [/bib_ref] and onchocerciasis ($0.58 per treatment. The potential programmatic overlap of activities among these and other public health initiatives includes administration, drug distribution, monitoring, surveillance, and social mobilization, so attractive cost-saving opportunities can be envisioned through integration. While estimates place the cost savings from integrated delivery between 25% and 47%, [bib_ref] Achieving the Millennium Development Goals.see commentcomment, Fenwick [/bib_ref] [bib_ref] Projected benefits from integrating NTD programs in sub-Saharan Africa, Brady [/bib_ref] the LF cost analysis protocol utilized in the present cost study can provide a useful tool by which to document the potential savings from integrating some or all programmatic activities of these initiatives. Indeed, countries that have already completed the LF cost analysis are well placed to estimate costs of integrated programs, given that many of the costs common to all programs have already been identified.
Also of particular interest from this study was the documentation of the substantial contributions (i.e., program ownership) by national governments [fig_ref] Figure 2: Sources of Funding for the MDAs Conducted by National LF Elimination Programs... [/fig_ref]. On average 56% of Financial costs of the LF elimination programs were financed by governments. The Egyptian government contributed 80% of Program costs, including participation of the Ministries of Health and Population, of Agriculture, and of Information. This is consistent with similar programs for TB and malaria which estimate that 70% of Financial costs are paid for by national governments.Furthermore, LF MDA programs do not require new or additional funding for all inputs. For example, the portion of the salary of a District Medical Officer (DMO) spent on the project would be included in the analysis as a program cost, but the DMO's salary would have been funded regardless of whether the time was spent on LF MDA activities or other programs and may not require additional financial outlays. Governments have a choice as to where they use their resources, and the percentages noted in this study emphasize the commitment these national governments have made to LF elimination.
## Resource allocation choices
Surveillance and mapping activities represented a significantly higher proportion of resource requirements in Egypt than in other countries (between 10% and 12% at both the national and governorate levels), since Egypt has selected small MDA implementation units, i.e. the village level, and hence focused on monitoring potential 'at-risk' areas closely in a large number of implementation units.
Tanzania's national program chose to invest heavily in social mobilization and education to assist the districts in raising awareness in the general population for current and future MDAs. The districts placed more emphasis on funding drug distribution and personnel training.
# Conclusion
The principal aim of this study was to provide critical information regarding the cost of implementing MDA for the prevention of lymphatic filariasis. The study demonstrates that the costs of MDA programs for LF elimination are comparable to those estimated for other similar disease control and elimination programs. While some programs had costs per person treated of over a dollar, it was quite straightforward to identify those factors most affecting program costs.
Such findings can be used on a national scale for program planning, development and fundraising, and on a global scale for calculating current global costs, predicting scale-up costs and calculating savings from integration with other programs. These results also will form the basis for guiding cost-effectiveness and cost-benefit analyses, as more information on the effectiveness of MDAs in the study countries becomes available. Additionally, the analytic tool used in this study will be valuable for further studies of the LF elimination program including costing of LF disability alleviation activities and the process certifying LF elimination.
A further finding of particular importance was documentation of the impact that the use of volunteers has on program costs. Further research on how best to utilize volunteerism for such public health programs could contribute appreciably to ensuring success of the MDA-based programs and to anchoring them in the communities they seek to protect.
Finally, implementation of this study produced a number of ancillary program benefits. The process of defining and reviewing costs allowed for a review of operations at all levels (inputs, processes and outputs) that was also frequently used to assess efficiency. The findings provide the opportunity for the development of cost-effective implementation models built on best practices from each country; hopefully, these can be adopted and adapted by new programs from the start, particularly in Africa where almost 30 more endemic countries still need to initiate LF elimination programs.
[fig] Figure 2: Sources of Funding for the MDAs Conducted by National LF Elimination Programs in Burkina Faso, Dominican Republic, Egypt, the Philippines and Tanzania (Financial Costs and Economic Costs). doi:10.1371/journal.pntd.0000067.g002 [/fig]
[table] Table 3: Country Background Information. [/table]
[table] Table 4: Financial and Economic per Person Treated. [/table]
[table] Table 5: Costs by Activity -Percentage of Financial Costs for a 'Non-Start-Up' MDA Round.Figure 1. Financial Costs and Population Treated by Country and MDA Year. doi:10.1371/journal.pntd.0000067.g001 [/table]
[table] Table 6: Costs by Input -Percentage of Financial Costs for a 'Non-Start-Up' MDA Round. [/table]
[table] Table 7: Sources of Funding for LF Elimination Costs. [/table]
|
Transient Global Amnesia and Brain Tumour: Chance Concurrence or Aetiological Association? Case Report and Systematic Literature Review
# Introduction
Syndromes of transient amnesia may broadly be conceptualised as idiopathic, epileptic or psychological in origin. Although there are some clinical features which may be helpful in distinguishing these syndromes, in clinical practice it may sometimes be difficult to tease them apart. Abnormalities on brain imaging may also complicate the interpretation.
Transient global amnesia (TGA) was first reported as such in the 1950s [bib_ref] Syndrome of isolated episode of confusion with amnesia, Bender [/bib_ref] [bib_ref] Les ictus amnésiques, Guyotat [/bib_ref]. The typical clinical features are short-lasting (<24 h) attacks of anterograde amnesia with otherwise normal cognition in the absence of any demonstrable focal neurological deficit. The differen-tial diagnosis of TGA includes transient epileptic amnesia (TEA), characterised by brief (ca. 1 h) attacks of impaired memory, most commonly occurring on waking from sleep [bib_ref] The syndrome of transient epileptic amnesia, Zeman [/bib_ref]. Diagnostic criteria for TGA have been suggested (table 1) [bib_ref] Syndromes of transient amnesia: towards a classification. A study of 153 cases, Hodges [/bib_ref].
The aetiology of TGA is still unclear. One theory suggests that TGA is a migraine-type phenomenon, another that it reflects a brief interruption of vascular perfusion of memoryeloquent structures in the medial temporal lobe, for which some evidence has accrued from neuroimaging studies finding focal hippocampal signal changes occurring over a time scale similar to that of vascular lesions [bib_ref] Transient global amnesia: functional anatomy and clinical implications, Bartsch [/bib_ref]. In the vast majority of cases, TGA is not associated with structural abnormalities on standard brain imaging techniques [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref]. However, occasional cases of TGA associated with brain tumour have been reported.
## Case report
A 66-year-old right-handed man was referred to the cognitive neurology clinic following 4 episodes of transient amnesia over a 6-month period. Each episode occurred within hours of strenuous physical exercise. In the first episode, he returned home from a bicycle ride confused about the route he had taken. The second event occurred following a walk up a steep incline. The third event occurred the day after a strenuous bike ride when the patient awoke in the morning confused as to where he was and what the plan for the day was. This confusion recurred the same day following a post-prandial nap. All the events were witnessed by the patient's wife who noted repetitive questioning to be a feature in each. All lasted between 30 min and 2 h with complete recovery. No other accompanying focal neurological symptoms were noted during the attacks.
The patient's past medical history was unremarkable: he had undergone two hip replacements but was otherwise fit and well. There was no history of recent head injury or epilepsy. He consumed 12 units of alcohol per week and was a lifelong non-smoker. He had previously worked as a project manager for a bank and regularly enjoyed cycling, golf and walking. There was no family history of any cognitive disorder.
Neurological examination in the clinic was unremarkable. On cognitive testing, he scored 0/28 on the Six-Item Cognitive Impairment Test (normal) and 28/30 on the Mini-Mental State Examination.
A diagnosis of TGA, concordant with suggested diagnostic criteria [bib_ref] Syndromes of transient amnesia: towards a classification. A study of 153 cases, Hodges [/bib_ref] , was thought likely on the basis of the history, at least for the first 2 episodes. However, because of the recurrent nature and brevity of the attacks and events emerging on waking from sleep, all features more suggestive of TEA [bib_ref] The syndrome of transient epileptic amnesia, Zeman [/bib_ref] , further investigations were undertaken.
Magnetic resonance (MR) brain imaging showed some very subtle but unequivocal signal change in the right amygdala , the nature of which was uncertain. Diffusionweighted imaging was normal with no restriction of diffusion and no disruption of limbic white matter tracts or adjacent temporal fibre bundles. Single-voxel MR spectroscopy of the affected area showed elevated choline resonance (NAA:Cho ratio = 0.64). Repeat MR brain imaging performed 2, 6 and 12 months later showed unchanged appearances. The imaging findings were thought most likely to indicate low-grade neoplasia, or possibly dysplasia or gliosis. Initial concerns that the imaging findings might reflect limbic encephalitis or frequent epileptic seizures ('epileptic crisis') were refuted by the clinical phenotype and the stability of the imaging changes, without evidence of focal atrophy.
The standard electroencephalogram (EEG) was within normal limits, but sleep-deprived EEG showed excess slow waves over the right temporal region and one prolonged run of slow waves followed by brief high-amplitude sharp wave bursts.
Because of the clinical and radiological stability of the patient, a biopsy of the lesion was not thought to be indicated. The patient continues to be followed up with regular surveillance MR brain imaging and has had further occasional amnesic episodes on waking from sleep. For these episodes, prophylactic anti-epileptic drug therapy has been offered but the patient has declined this on the basis of the infrequency of the events and concerns about possible adverse effects of medication. Repeat Mini-Mental State Examination 12 months after initial presentation was stable (score 29/30).
In light of our experience with this case, we undertook a systematic review to identify reported cases of TGA and brain tumour. (Although the diagnosis of tumour was not definitively established in our case, it was our provisional diagnosis since it was the most likely on the basis of the investigation findings, and the system of surveillance initiated was that employed at our centre for suspected tumour cases. [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref] [bib_ref] Ictus amnésique symptomatique d'un glioblastome intéressant le trigone, Aimard [/bib_ref] [bib_ref] A case of transient global amnesia due to a pituitary tumor, Hartley [/bib_ref] [bib_ref] Gliome du système limbique postérieur, révélé par une amnésie globale transitoire. Observation..., Boudin [/bib_ref] [bib_ref] Transient global amnesia due to a dominant hemisphere tumor, Lisak [/bib_ref] [bib_ref] Transient global amnesia due to a glioma in the dominant hemisphere, Shuping [/bib_ref] [bib_ref] Transient global amnesia associated with a single metastasis in the non-dominant hemisphere...., Findler [/bib_ref] [bib_ref] Transient global amnesia and meningioma, Meador [/bib_ref] [bib_ref] Amnesia global transitoria asociada a un meningioma del lobulo frontal, Riva [/bib_ref] [bib_ref] Meningioma and transient global amnesia: another report, Collins [/bib_ref] [bib_ref] Cranial CT scan in transient global amnesia, Matias-Guiu [/bib_ref] [bib_ref] Transient global amnesia and falcotentorial meningioma -a case report, Araga [/bib_ref] [bib_ref] Ethmoidal meningioma revealed by transient global amnesia, Cattaino [/bib_ref] [bib_ref] Transient global amnesia associated with a right sphenoid ridge meningioma: a case..., Po [/bib_ref] [bib_ref] Transient amnesic syndrome after spontaneous haemorrhage into a hypothalamic pilocytic astrocytoma, Sorenson [/bib_ref] [bib_ref] Hemorrhagic pituitary adenoma manifesting as transient global amnesia, Honma [/bib_ref] [bib_ref] Transient amnesia in a patient with left temporal tumor. Symptomatic transient global..., Huang [/bib_ref] [bib_ref] Transient global amnesia as a revealing sign of giant transtentorial meningioma. Case..., Dinca [/bib_ref]. This has extended the previous narrative reviews of TGA and tumour by Dinca et al. [bib_ref] Transient global amnesia as a revealing sign of giant transtentorial meningioma. Case..., Dinca [/bib_ref] and Daniel. No cases of transient amnesia associated with either dysplasia or gliosis were identified, although pathological evidence of gliosis may be found in some cases of epilepsyrelated amnesia; however, the latter is persistent rather than transient [bib_ref] Epilepsy-related long-term amnesia: anatomical perspectives, Butler [/bib_ref].
The only study reporting data which might address the question of the frequency of TGA associated with brain tumour was that of Agosti et al. [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref] : in a series of 130 patients meeting diagnostic criteria for TGA and undergoing investigation with modern brain imaging modalities, 13 patients had structural brain lesions, of whom 2 (1.5%) had falx meningiomas. No case of brain tumour was encountered in the series (n = 114) of Hodges and Warlow [bib_ref] Syndromes of transient amnesia: towards a classification. A study of 153 cases, Hodges [/bib_ref]. In a consecutive series of 24 TGA cases seen in our clinic [bib_ref] Transient global amnesia in the district general hospital, Larner [/bib_ref] [bib_ref] Amnesia as a sex-related adverse event, Larner [/bib_ref] [bib_ref] Transient amnesia: epileptic or global? A differential diagnosis with significant implications for..., Ung [/bib_ref] , this is the first case encountered in which a brain tumour has been found.
Many of the reports [bib_ref] Ictus amnésique symptomatique d'un glioblastome intéressant le trigone, Aimard [/bib_ref] [bib_ref] A case of transient global amnesia due to a pituitary tumor, Hartley [/bib_ref] [bib_ref] Gliome du système limbique postérieur, révélé par une amnésie globale transitoire. Observation..., Boudin [/bib_ref] [bib_ref] Transient global amnesia due to a dominant hemisphere tumor, Lisak [/bib_ref] [bib_ref] Transient global amnesia due to a glioma in the dominant hemisphere, Shuping [/bib_ref] [bib_ref] Transient global amnesia associated with a single metastasis in the non-dominant hemisphere...., Findler [/bib_ref] [bib_ref] Transient global amnesia and meningioma, Meador [/bib_ref] [bib_ref] Amnesia global transitoria asociada a un meningioma del lobulo frontal, Riva [/bib_ref] [bib_ref] Meningioma and transient global amnesia: another report, Collins [/bib_ref] [bib_ref] Cranial CT scan in transient global amnesia, Matias-Guiu [/bib_ref] [bib_ref] Transient global amnesia and falcotentorial meningioma -a case report, Araga [/bib_ref] [bib_ref] Ethmoidal meningioma revealed by transient global amnesia, Cattaino [/bib_ref] [bib_ref] Transient global amnesia associated with a right sphenoid ridge meningioma: a case..., Po [/bib_ref] predated widely accepted clinical diagnostic criteria for TGA [bib_ref] Syndromes of transient amnesia: towards a classification. A study of 153 cases, Hodges [/bib_ref] , and for this reason, some cases might be excluded as not conforming to criteria. For example, in one case, the amnesic episode lasted more than 24 h [bib_ref] Ictus amnésique symptomatique d'un glioblastome intéressant le trigone, Aimard [/bib_ref] , and in another, progressive memory problems followed a generalised tonic clonic seizure [bib_ref] Transient global amnesia due to a glioma in the dominant hemisphere, Shuping [/bib_ref] ; in yet another case, the patient was reported to have 6 episodes of TGA and on examination had bilateral papilloedema [bib_ref] Cranial CT scan in transient global amnesia, Matias-Guiu [/bib_ref] (criticised as unlikely to be TGA by Hodges. Caplan [bib_ref] Transient global amnesia: criteria and classification, Caplan [/bib_ref] previously criticised the case reported by Meador et al. [bib_ref] Transient global amnesia and meningioma, Meador [/bib_ref] on the grounds that the reported clinical features (two short-lasting and unobserved episodes of loss of awareness) did not suggest TGA.
The clinical features of TGA cases with and without neuroimaging evidence of structural brain lesions are said to be the same [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref]. However, in the current case and one previous report [bib_ref] Transient amnesia in a patient with left temporal tumor. Symptomatic transient global..., Huang [/bib_ref] , both associated with medial temporal lobe pathology, there have been clinical features suggestive of both TGA and TEA.
Various tumour locations were reported. Some were located distant from memoryeloquent structures, such as falx meningiomas [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref] [bib_ref] Transient global amnesia and falcotentorial meningioma -a case report, Araga [/bib_ref] and pituitary tumour [bib_ref] A case of transient global amnesia due to a pituitary tumor, Hartley [/bib_ref] , one associated with pituitary haemorrhage [bib_ref] Hemorrhagic pituitary adenoma manifesting as transient global amnesia, Honma [/bib_ref]. On the other hand, tumours located within or in proximity to memory-eloquent structures, such as the temporal lobe [bib_ref] Transient global amnesia and meningioma, Meador [/bib_ref] [bib_ref] Cranial CT scan in transient global amnesia, Matias-Guiu [/bib_ref] [bib_ref] Transient amnesia in a patient with left temporal tumor. Symptomatic transient global..., Huang [/bib_ref] [bib_ref] Transient global amnesia as a revealing sign of giant transtentorial meningioma. Case..., Dinca [/bib_ref] , limbic system [bib_ref] Gliome du système limbique postérieur, révélé par une amnésie globale transitoire. Observation..., Boudin [/bib_ref] , hippocampus [bib_ref] Transient global amnesia due to a glioma in the dominant hemisphere, Shuping [/bib_ref] or amygdala (current case), were also reported.
Most reported cases of TGA and brain tumour had either histologically proven primary brain tumours or were presumed to be primary brain tumours where histological analysis was not performed. Two cases were secondary tumours, from a bladder transitional cell carcinoma [bib_ref] Transient global amnesia associated with a single metastasis in the non-dominant hemisphere...., Findler [/bib_ref] and a lung primary [bib_ref] Transient amnesia in a patient with left temporal tumor. Symptomatic transient global..., Huang [/bib_ref] , respectively.
If the absence of EEG abnormalities is accepted as a TGA diagnostic criterion (EEG is very seldom recorded during an amnesic event) [bib_ref] Transient amnesia: epileptic or global? A differential diagnosis with significant implications for..., Ung [/bib_ref] , 6 of the reported cases might be excluded on the basis of abnormalities on interval EEG studies . (EEG was not reported, presumably not undertaken or normal in at least 6 other cases [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref] [bib_ref] Amnesia global transitoria asociada a un meningioma del lobulo frontal, Riva [/bib_ref] [bib_ref] Transient global amnesia associated with a right sphenoid ridge meningioma: a case..., Po [/bib_ref] [bib_ref] Transient amnesic syndrome after spontaneous haemorrhage into a hypothalamic pilocytic astrocytoma, Sorenson [/bib_ref] [bib_ref] Hemorrhagic pituitary adenoma manifesting as transient global amnesia, Honma [/bib_ref] It is possible that some of these cases might have been instances of TEA rather than of TGA, although EEG abnormalities are only seen in about one third of TEA cases on interval EEG [bib_ref] The syndrome of transient epileptic amnesia, Zeman [/bib_ref]. In his review, Danielstated his belief that the cases reported by Hartley et al. [bib_ref] A case of transient global amnesia due to a pituitary tumor, Hartley [/bib_ref] , Shuping et al. [bib_ref] Transient global amnesia due to a glioma in the dominant hemisphere, Shuping [/bib_ref] and Honma and Nagao [bib_ref] Hemorrhagic pituitary adenoma manifesting as transient global amnesia, Honma [/bib_ref] are likely to be instances of TEA rather than of TGA, and possibly also the case of Meador et al. [bib_ref] Transient global amnesia and meningioma, Meador [/bib_ref].
# Discussion
What is the clinical significance of these reports of TGA and brain tumour? Some reports fail to meet the clinical criteria for TGA whereas others have features which might be more in keeping with TEA. In some cases, the finding of a tumour is unlikely to be any more than chance concurrence, based on tumour location distant from memory-eloquent structures, whereas other tumour locations are more obviously of possible pathophysiological relevance. Such localised neoplastic lesions might result in abnormal electrical activity within medial temporal lobe networks, hence producing TEA-like events, and also compromise medial temporal lobe blood supply, for example in association with exercise, and hence predispose to typical TGA-like events. In other words, the tumour might lower the threshold for the clinical manifestation of vascular and/or epileptiform events. Alternatively, it is recognised that a subset of 'pure TGA' patients, largely identifiable because of atypically brief or recurrent attacks, develop epilepsy and/or respond to antiepileptic medication [bib_ref] Syndromes of transient amnesia: towards a classification. A study of 153 cases, Hodges [/bib_ref]. These features -short duration of symptoms (<2 h), stereotyped semiology, and high recurrence rate -are all typical of TEA as opposed to TGA. Hence, all the reported cases of tumourrelated TGA might in fact be instances of TEA.
In summary, the evidence suggests that the concurrence of brain tumour and TGA is extremely rare. Although some authors accept that brain tumour may be associated with or imitate TGA [bib_ref] Transient global amnesia: look at mechanisms not causes, Roach [/bib_ref] , on the basis of our experience and the systematic literature review, we doubt if a tumour is ever aetiologically significant in TGA, since the reported cases could all be TEA, in some instances 'masquerading' as TGA [bib_ref] The aetiology of transient global amnesia. A case-control study of 114 cases..., Hodges [/bib_ref]. Although some authors advocate brain imaging in all TGA cases to exclude a symptomatic lesion [bib_ref] Transient global amnesia and brain lesions: new hints into clinical criteria, Agosti [/bib_ref] , on the basis of this review, we would argue that neuroimaging should be reserved for those TGA patients with atypical clinical features, such as recurrent episodes and/or episodes clinically more suggestive (in terms of their brevity and stereotyped nature) of TEA. Attacks must be witnessed and information available from a capable observer who was present for most of the attack -There must be clear-cut anterograde amnesia during the attack -Clouding of consciousness and loss of personal identity must be absent and the cognitive impairment limited to amnesia (i.e., no aphasia, apraxia, etc.) -There should be no accompanying focal neurological symptoms during the attack and no significant neurological signs afterwards -Epileptic features must be absent -Attacks must resolve within 24 h -Patients with recent head injury or active epilepsy (i.e., remaining on medication or with 1 seizure in the past 2 years) are excluded
[table] Table 1: Diagnostic criteria for definite TGA[4] [/table]
|
Leveling the playing field: The utility of coronary artery calcium scoring in cardiovascular risk stratification in South Asians
# Introduction
South Asian (SA) individuals make up a diverse ethnic milieu of individuals that trace their roots back to the Southern part of Asia including the Indian subcontinent. The term SA refers collectively to individuals with ancestry from the countries of Bangladesh, Bhutan, India, the Maldives, Nepal, Pakistan, and Sri Lanka. SAs are a rapidly growing ethnic group within and outside of the United States due to Western migration. Despite the growth and socioeconomic success of this group, Western acculturation has uncovered new challenges with regard to cardiometabolic health.
As an ethnicity, SA individuals are at high risk for atherosclerotic cardiovascular disease (ASCVD). The first reports of elevated ASCVD risk came from Singapore in 1959 [bib_ref] Ethnic group differences in coronary heart disease in Singapore: an analysis of..., Danaraj [/bib_ref]. Over the last several decades, a growing body of knowledge on ASCVD in SA individuals has pointed to a higher burden, earlier onset, and greater mortality of ASCVD within the SA community when compared to Non-Hispanic Whites (NHWs) [bib_ref] Patterns of Asian and non-Asian morbidity in hospitals, Donaldson [/bib_ref] [bib_ref] First myocardial infarctions in Asian and white men, Hughes [/bib_ref] [bib_ref] Lessons from the study of immigrant mortality, Marmot [/bib_ref] [bib_ref] Spectrum of cardiovascular diseases inAsian-American racial/ethnic subgroups, Holland [/bib_ref].
Notably, SA individuals, particularly those that reside in the United States and other Westernized countries, are at an even greater risk for ASCVD and mortality related to ASCVD. The mortality burden from ischemic heart disease among SA individuals is reflected by the proportional mortality rates. The proportional mortality rates in Asian Indian men and women are 1.43 and 1.12 respectively and are substantially elevated, when compared with NHW men (1.08) and women (0.92) [bib_ref] Cardiovascular disease mortality in Asian Americans, Jose [/bib_ref]. Acknowledging this burden of heart disease, SA ethnicity is now listed as a 'risk-enhancing' factor, a factor that is thought to confer an elevated risk for ASCVD, in the 2018 American College of Cardiology (ACC)/ American Heart Association (AHA)/ Multi-society Guideline on the management of Blood Cholesterol [bib_ref] The Use of Risk Enhancing Factors to Personalize ASCVD Risk Assessment: Evidence..., Agarwala [/bib_ref].
While the ASCVD burden among SA individuals is multifactorial, the greatest need lies in appropriate risk stratification, prevention, and management of disease. A number of challenges lie in clinical risk stratification in these individuals using currently available risk calculators, namely due to either the overestimation or underestimation of risk. In this White paper, we will discuss the utility of noninvasive imaging with coronary artery calcium (CAC) scoring and coronary computed tomography (CT) to assist in "leveling the playing field" with regard to assessment of ASCVD risk.
## Drivers of ascvd among south asians
The drivers of the elevated ASCVD burden among SA individuals residing in the United States include a broad variety of biological mechanisms and non-biological social and health-related behaviors. Insulin resistance, impaired glucose tolerance, metabolic syndrome (MetS), and type II diabetes mellitus collectively are some of the most common risk factors among SA individuals when compared to NHWs. Type II diabetes prevalence is at least 2-fold higher among SA individuals when compared to NHW [bib_ref] Type 2 diabetes in South Asians: similarities and differences with white Caucasian..., Gujral [/bib_ref]. Overall, the prevalence of type II diabetes and prediabetes is estimated at 29% and 37% respectively among Asian Indians [bib_ref] Prevalence and correlates of diabetes in South asian indians in the United..., Kanaya [/bib_ref]. Obesity, particularly abdominal adiposity and increased waist circumference is highly prevalent among SAs and contributes to increased cardiometabolic risk. The World Health Organization and American Diabetes Association have recommended lower body mass index (BMI) cutoffs for individuals of SA ethnicity such that a normal BMI is 18.5-22.9 kg/m 2 , an overweight BMI is 23.0-27.4 kg/m 2 , and an obese BMI is ≥27.5 kg/m 2. The lower cut points encourage healthy eating and weight management earlier on and trigger screening for cardiovascular disease risk factors, which become more prevalent at a lower BMI among SA individuals [bib_ref] Less favorable body composition and adipokines in South Asians compared with other..., Shah [/bib_ref]. Dyslipidemia including elevated triglycerides, low HDL-C, potentially dysfunctional HDL-C, and smaller, lower density LDL particles are also notable contributors to the increased ASCVD risk [bib_ref] Lipoprotein abnormalities in South Asians and its association with cardiovascular disease: Current..., Bilen [/bib_ref]. Higher serum levels of lipoprotein(a) have also been reported among SA individuals and reflect a genetically inherited component to ASCVD risk [bib_ref] Differences in risk factors, atherosclerosis, and cardiovascular disease between ethnic groups in..., Anand [/bib_ref]. Other ASCVD risk factors that elevate the risk include hypertension, which is more prevalent amongst SA individuals compared to NHW, and smoking [bib_ref] Prevention of atherosclerotic cardiovascular disease in South Asians in the US: A..., Kalra [/bib_ref]. While traditional tobacco use (i.e. cigarette smoking) is lower among SA individuals, clinical questionnaires and assessments often do not take into account other forms of cultural tobacco use that may be more commonly used (i.e. bidis, paan, paan masala, gutka, hookah, for example) and which influence ASCVD risk [bib_ref] Smokeless tobacco use by south Asians, Siddiqi [/bib_ref] [bib_ref] Awareness and Use of South Asian Tobacco Products Among South Asians in..., Hrywna [/bib_ref].
Non-biological mechanisms are also though to play a significant role in ASCVD risk and the development of its risk factors. Western acculturation of traditional dietary patterns to incorporate a higher proportion of foods that are higher in saturated fats and sugars and lower in nutritional value and low physical activity rates compared to other racial and ethnic minorities further elevate the risk of ASCVD through non-biological mechanisms [bib_ref] Prevention of atherosclerotic cardiovascular disease in South Asians in the US: A..., Kalra [/bib_ref] [bib_ref] Correlates of lifestyle: physical activity among South Asian Indian immigrants, Daniel [/bib_ref].
## Current guideline recommendations for coronary artery calcium scoring
The 2018 ACC/AHA Cholesterol Guideline and 2019 ACC/AHA Primary Prevention Guideline recommends the use of the Pooled Cohort Equations (PCEs) to stratify patients into low, borderline, intermediate and high ASCVD risk [bib_ref] PCNA Guideline on the Management of Blood Cholesterol: A Report of the..., Grundy [/bib_ref] [bib_ref] ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease, Arnett [/bib_ref]. Screening recommendations using the PCE begin at the age of 40 years. Per the guidelines, CAC scoring may be considered in individuals at borderline or intermediate risk in whom the benefit of statin use may be unclear. In younger adults, ages 20-39 years, a lifetime risk assessment is recommended and statin therapy may be considered in those with a strong family history of ASCVD or in those with significantly elevated LDL-C (including familial hypercholesterolemia).
## Limitations of traditional risk stratification algorithms and current guideline recommendations in south asians
Risk assessment is a fundamental aid for ASCVD risk reduction counseling. The 2018 ACC/AHA Guideline on the Management of Blood Cholesterol listed SA as a high-risk ethnicity and a risk-enhancing factor, thereby recommending that SA individuals at borderline (5-<7. 10year ASCVD risk) or intermediate (≥7.5%-<20% 10-year ASCVD risk) risk or higher per the PCEs as candidates for consideration of statin initiation or intensification [bib_ref] PCNA Guideline on the Management of Blood Cholesterol: A Report of the..., Grundy [/bib_ref]. While these recommendations and other risk prediction tools acknowledge the elevated risk at a population level the elevated risk among SA individuals, there are limitations to risk prediction, discrimination, and calibration in these individuals (summarized in . First, available guidelines recommend risk algorithms have not been derived from or prospectively validated in SA adults; second, SA (and other ethnicities) are recommended to be classified as 'White' race (underestimating risk using the Pooled Cohorts Equations); third, crude adjustment factors to calculate ASCVD risk fail to accurately capture risk (using the UK based QRISK2 and QRISK Lifetime3rd Joint British Societies' Risk Calculator); fourth, limited considerations have been given for native versus migrant populations; and finally, there is paucity of disaggregated data, which masks meaningful ASCVD health differences in SA subgroups. Overall, the PCEs have reasonable functionality in SA individuals (majority Asian Indian population) in the US who are at the extremes of risk (<5% and >20%), however, systematic consideration of statin treatment in these individuals at borderline and intermediate risk may result in overtreatment with statin therapy [bib_ref] Implications of the 2019 American College of Cardiology/American Heart Association Primary Prevention..., Haque [/bib_ref]. More refined means of risk stratification are needed for SA individuals, particularly for those who carry higher ASCVD risk compared to Asian Indians and who are underrepresented in current research studies such as those of Pakistani and Bangladeshi origin.
There is robust evidence demonstrating significant ASCVD burden in younger SA individuals, particularly before the age of 45 years [bib_ref] Atherosclerotic Cardiovascular Disease in South Asians in the United States: Epidemiology, Risk..., Volgman [/bib_ref]. Per current screening recommendations, many younger SA individuals may be misclassified as low-risk by the PCEs and would remain inadequately screened and treated [bib_ref] Primary Prevention Management of Elevated Lipoprotein(a), Agarwala [/bib_ref]. Comprehensive screening with cardiovascular imaging and blood biomarkers did not reduce the risk of death in a majority White population, leaving the role of population level cardiovascular screening in SA individuals less clear, and unlikely to be recommended at this time.
## Coronary calcium scoring in south asians
Substantial variability exists in the CVD prevalence, incidence, risk, and health-seeking behavior among SA subgroups [bib_ref] Spectrum of cardiovascular diseases inAsian-American racial/ethnic subgroups, Holland [/bib_ref] [bib_ref] Cardiovascular disease mortality in Asian Americans, Jose [/bib_ref] [bib_ref] Call to action: cardiovascular disease in Asian Americans: a science advisory from..., Palaniappan [/bib_ref] [bib_ref] Heterogeneity of Cardiovascular Disease Risk Factors Among Asian Immigrants: Insights From the, Koirala [/bib_ref] [bib_ref] Burden of cardiovascular risk factors and disease in five Asian groups in..., Satish [/bib_ref] [bib_ref] Heterogeneity in cardio-metabolic risk factors and atherosclerotic cardiovascular disease among Asian groups..., Satish [/bib_ref]. In addition to frequent and universal screening for traditional risk factors and metrics, the use of CAC, a highly specific marker of subclinical atherosclerosis, therefore may be a useful test to improve risk stratification in SA individiuals [bib_ref] Discordance between 10-year cardiovascular risk estimates using the ACC/AHA 2013 estimator and..., Rifai [/bib_ref]. CAC testing is a cost-effective, highly reproducible, and specific marker of subclinical atherosclerosis, proven to improve ASCVD risk assessment across all racial/ethnic groups, thereby serving as a guide for initiating or deferring preventive therapies Limitations of the Pooled Cohort Equations Among South Asians. [bib_ref] Coronary Calcium Score and Cardiovascular Risk, Greenland [/bib_ref]. An elevated CAC score has been shown to be independently associated with ASCVD and further stratifies ASCVD risk while CAC score of 0 does not further stratify ASCVD risk [bib_ref] Independent Association of Lipoprotein(a) and Coronary Artery Calcification With Atherosclerotic Cardiovascular Risk, Mehta [/bib_ref].
Among asymptomatic Asian Indians without diabetes who were borderline risk in the MASALA cohort, 54% had an absence of CAC (score of 0). Among Asian Indians without diabetes who were at intermediate risk, 30% had an absence of CAC, but 70% had a CAC score that was >0. The prevalence of CAC was higher among Asian Indians at intermediate risk than at borderline risk [bib_ref] PCNA Guideline on the Management of Blood Cholesterol: A Report of the..., Grundy [/bib_ref] [bib_ref] Implications of the 2019 American College of Cardiology/American Heart Association Primary Prevention..., Haque [/bib_ref]. In these individuals at borderline or intermediate risk who are asymptomatic and free of diabetes, CAC may provide additional value in ASCVD risk stratification. It is important to note that this cohort was enriched in Asian Indians and had very few Pakistani and Bangladeshi Americans, both of which are higher risk SA subgroups. Additional considerations for the use of CAC to further assess ASCVD risk include individuals who are at low risk by the PCE (who would otherwise not qualify for statin therapy) but have a family history premature ASCVD. In these patients, the use of CAC, and particularly early screening beginning around the ages of 30 or 40 years may identify higher risk individuals in whom preventive therapies may be initiated earlier on despite the lack of traditional risk factors assessed by the PCEs. However, it is important to note that the absence of CAC particularly in younger adults with an elevated lifetime risk does not diminish the importance of healthy lifestyle behaviors [bib_ref] Association of Coronary Artery Calcium With Long-term, Cause-Specific Mortality Among Young Adults, Miedema [/bib_ref]. Prediabetes and MetS are highly prevalent among SA indiviiduals and increase ASCVD risk. In a study conducted using the MESA cohort, the addition of CAC improved risk classification of individuals with MetS, suggesting an additional potential role of CAC screening [bib_ref] Coronary Artery Calcium Score for Long-term Risk Classification in Individuals With Type..., Malik [/bib_ref].
Racial and ethnic differences: With regard to CAC progression, SA men have greater CAC progression when compared to Chinese, Black and Latino men after adjustment for age, diabetes mellitus, hypertension, and statin use, but no difference was noted SA and White men [bib_ref] Incidence and Progression of Coronary Artery Calcium in South Asians Compared With..., Kanaya [/bib_ref]. SA women appear to have a similarly elevated CAC burden when compared to other racial and ethnic groups (African America, Latino, and Chinese Americans), but older SA women, ≥70 years, were noted to have a higher CAC burden compared to other racial and ethnic groups [bib_ref] Comparing coronary artery calcium among U.S. South Asians with four racial/ethnic groups:..., Kanaya [/bib_ref]. In one study of Asian-Indians, the presence of CAC in individuals <50 years was 46%, which was higher than other racial and ethnic groups [bib_ref] Comparison of prevalence and severity of coronary calcium determined by electron beam..., Hatwalkar [/bib_ref]. CAC was found among 97% of Asian-Indian individuals >60 years pointing to the high burden of CAC among this ethnic group [bib_ref] Comparison of prevalence and severity of coronary calcium determined by electron beam..., Hatwalkar [/bib_ref]. These findings correspond with the high prevalence of CAD in this group.
The presence of a family history of coronary heart disease (CHD) in SA individuals compared to those of other race/ethnicities is associated with CHD (OR 1.71 [95% CI 1.21, 2.42]; p = 0.002) and severe subclinical atherosclerosis (CAC >300) [bib_ref] Family History of CHD Is Associated With Severe CAC in South Asians:..., Patel [/bib_ref] [bib_ref] Inflammation and coronary artery calcification in South Asians: The Mediators of Atherosclerosis..., Mehta [/bib_ref] With respect to biomarkers [Lp(a)], tumor necrosis factor-alpha, adiponectin, and leptin], no association was found with CAC presence or severity in the MASALA Study, suggesting markers of inflammation and cardiometabolic risk are independent of CAC in SA individuals, similar to other race/ethnicities [bib_ref] Hypertension guidelines and coronary artery calcification among South Asians: Results from MASALA..., Patel [/bib_ref]. CAC may help identify those SA adults who would best qualify for aggressive lifestyle optimization and anti-hypertensive pharmacotherapy [bib_ref] Hypertension guidelines and coronary artery calcification among South Asians: Results from MASALA..., Patel [/bib_ref] [bib_ref] ACC/AHA/AAPA/ABC/ACPM/ AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High..., Whelton [/bib_ref]. For example, more than half of all SA adults qualified for lifestyle modifications and 17% were recommended anti-hypertensive pharmacotherapy by 2017 ACC/AHA Hypertension Guideline compared to 8% by JNC7 (chi-square p<0.001) [bib_ref] Hypertension guidelines and coronary artery calcification among South Asians: Results from MASALA..., Patel [/bib_ref]. It is important to note that many of the studies that assessed the use of CAC among SA adults were highly enriched in Asian Indians. While CAC has been shown to be a reliable marker among other racial and ethnic groups, additional studies are needed among SA individuals beyond just Asian Indians.
Taken together, when compared to available risk markers, CAC scoring is a useful noninvasive modality for further risk refinement and guide other preventive therapy in SA adults.
## Coronary computed tomography angiography in south asians
Coronary CT angiography has been used to identify high risk patients. While current data in SA adults using coronary CT angiography is limited, quantitative analysis of the coronary arteries by CT showed that SA adults had smaller normalized proximal LAD luminal diameters when compared with Caucasians and also had more severe CAD including a higher number of diseased vessel segments and greater mean percent stenosis in the proximal LAD and RCA [bib_ref] Quantitative angiography in South Asians reveals differences in vessel size and coronary..., Hasan [/bib_ref]. Notably, SA individuals with type II diabetes mellitus have a higher prevalence and extent of asymptomatic coronary artery disease when compared to matched Caucasian patients [bib_ref] Comparison by computed tomographic angiography-the presence and extent of coronary arterial atherosclerosis..., Roos [/bib_ref]. Furthermore, SA individuals may have differential non-calcified plaque given the higher prevalence of elevated Lp(a) and MetS. Future studies should evaluate this as potential tool for diagnosis and treatment.
Another important use of coronary CT is to assess for differences in plaque characteristics. Plaque characteristics beyond the traditional CAC score may be helpful when assessing ASCVD risk. SA adults have a higher number of vessels with calcified plaque compared to other race/ ethnicities [bib_ref] Comparison by computed tomographic angiography-the presence and extent of coronary arterial atherosclerosis..., Roos [/bib_ref] [bib_ref] Distribution of calcium volume, density, number, and type of coronary vessel with..., Rifai [/bib_ref]. Significant coronary artery disease (CAD) (>50% stenosis) was more frequent in the left anterior descending coronary artery among asymptomatic SA adults compared to NHW adults with diabetes mellitus [bib_ref] Comparison by computed tomographic angiography-the presence and extent of coronary arterial atherosclerosis..., Roos [/bib_ref]. Compared to NHW, SA have significantly lower CAC volume but higher CAC density when accounting for traditional risk factors [bib_ref] Distribution of calcium volume, density, number, and type of coronary vessel with..., Rifai [/bib_ref]. As noted by the authors, higher CAC density may reflect stable plaque (less likely to rupture), which is influenced by statin use (pro-calcific effects) [bib_ref] Distribution of calcium volume, density, number, and type of coronary vessel with..., Rifai [/bib_ref] [bib_ref] Calcium density of coronary artery plaque and risk of incident cardiovascular events, Criqui [/bib_ref]. Both CAC volume and density appear to be correlated with ASCVD risk [bib_ref] Association of Coronary Artery Calcium Density and Volume With Predicted Atherosclerotic Cardiovascular..., Rifai [/bib_ref]. The long-term clinical and prognostic significance of this is currently unknown and remains an area for future investigation.
## Beyond statins
Beyond categorizing individuals as statin candidates and those who are not in a binary fashion, the detection and assessment of plaque using CAC and coronary CT can assist with personalized allocation of preventive cardiometabolic therapies tailored to individual needs. CAC testing and specifically CAC>100 Agatston may guide Aspirin utilization in asymptomatic adults in primary prevention cohorts [bib_ref] Value of Coronary Artery Calcium Scanning in Association With the Net Benefit..., Ajufo [/bib_ref] [bib_ref] Use of coronary artery calcium testing to guide aspirin utilization for primary..., Miedema [/bib_ref]. Additionally, more intensive lifestyle counseling and enhanced used of other cardiometabolic therapies that reduce the risk of ASCVD events, i.e. Glucagon-like peptide-1 receptor agonists (GLP-1 RA), Sodium-glucose cotransporter-2 (SGLT2) inhibitors, and icosapent ethyl (IPE) may be offered to those who may not otherwise qualify for these therapies.
## Clinical implications
The evaluation and management of traditional risk ASCVD risk factors among SAs remains important. Targeted strategies are needed to comprehensively reduce ASCVD risk among SA individuals residing in Westernized countries including the United States as well as those living in their native countries. These include strategies such as enhancing access to basic preventive care with enhanced early detection and treatment of conditions that are highly prevalent among SA individuals (i.e. diabetes mellitus, dyslipidemia, and MetS), increasing health literacy and education, nutrition and physical activity counseling.
Outside of clinic settings, interventions focused on tackling poverty and reducing air pollution are important ways to help reduce ASCVD burden. Local and national cultural community organizations may also be leveraged to assist with spreading awareness of heart disease and its risk factors and to help facilitate access to care. It is equally important to address unique challenges and barriers to care that have arisen in diaspora countries that adversely impact access to care: socioeconomic status, language barriers, Islamophobia, racism, low access to care, lack of insurance, and stressors of migration.
Currently, there exist notable limitations to currently available clinical risk stratification algorithms that may inadequately estimate ASCVD risk in SA individuals. CAC testing is a cost-effective, actionable, and reproducible tool that can accurately detect subclinical atherosclerosis and thereby guide preventive therapies. CAC scoring in patients at intermediate risk of ASCVD in whom the decision about statin use is uncertain has been listed as a high-value care test [bib_ref] Strategies to Reduce Low-Value Cardiovascular Care: A Scientific Statement From the American..., Kini [/bib_ref]. Beyond traditionally recommended indications, CAC screening may be considered in individuals who are at a higher risk than that estimated by traditional risk stratification tools including those classified as low risk, but with a family history of premature ASCVD. These individuals as such may not qualify for statin and other preventive therapies based on stratification by the PCEs alone, but CAC scoring may further refine their classification. Additionally, individuals with medical comorbidities that are highly prevalent drivers of ASCVD such as prediabetes and MetS may be considered as candidates for CAC scoring to help identify those in whom a primary prevention statin and other preventive cardiometabolic therapies may be considered.
While traditional CAC scoring recommendations begin at the age of 45 years, the ASCVD burden among SA individuals begins roughly 10 years earlier than in other racial and ethnic groups and argues for earlier screening. As such, after risk assessment, implementation of CAC testing, particularly in those with significant risk factors or a family history of CHD may aid in the detection and management of subclinical atherosclerosis [bib_ref] Association of Coronary Artery Calcium in Adults Aged 32 to 46 Years..., Carr [/bib_ref] [bib_ref] Yield of screening for coronary artery calcium in early middle-age adults based..., Okwuosa [/bib_ref] [bib_ref] Clinical indications for coronary artery calcium scoring in asymptomatic patients: Expert consensus..., Hecht [/bib_ref] [bib_ref] Interplay of Risk Factors and Coronary Artery Calcium for CHD Risk in..., Mortensen [/bib_ref]. [fig_ref] Table 2: Key Recommendations for Coronary Artery Calcium Scoring among South Asians [/fig_ref] summarizes the key recommendations for CAC scoring in SA patients. There is a need for additional validation of CAC percentiles specific to SA individuals and those <45 years. summarizes risk stratification among SA individuals, which has also been previously described (https://www.acc.org/latest-in-cardiology/a rticles/2021/12/06/13/32/how-many-south-asians-does-it-take) [bib_ref] Prevention of atherosclerotic cardiovascular disease in South Asians in the US: A..., Kalra [/bib_ref] [bib_ref] Lipid Association of India Expert Consensus Statement on Management of Dyslipidemia in..., Puri [/bib_ref].
Future ACC/AHA guidelines and recommendations may consider novel indications for CAC scoring among SA adults and other high-risk groups as well as earlier screening ages to help allocate preventive cardiometabolic therapies. Comprehensive screening with cardiovascular imaging and blood biomarkers did not reduce the risk of death in a majority White population, leaving the role of population level cardiovascular screening in SA adults less clear, and unlikely to be recommended at this time [bib_ref] Five-Year Outcomes of the Danish Cardiovascular Screening (DANCAVAS) Trial, Lindholt [/bib_ref]. However, currently available clinical data among SA individuals underscores the need for appropriate risk stratification and management in individuals at an elevated ASCVD risk [bib_ref] Implications of the 2019 American College of Cardiology/American Heart Association Primary Prevention..., Haque [/bib_ref] [bib_ref] Coronary Artery Calcium Score for Long-term Risk Classification in Individuals With Type..., Malik [/bib_ref].
Use of the Multi-Ethnic Study of Atherosclerosis (MESA) 10-year CHD risk calculator with coronary artery calcification may be considered for SA men. While a SA specific calculator does not currently exist, in the MESA calculator, SA men may be represented as "White" race given that CAC scores are similar in SA men and NHW men.
Cost and insurance reimbursement considerations are also important issues to consider for CAC scoring. Traditionally, CAC scoring has not been well covered by insurance. While out of pocket costs may range from $49-$400 per test nationally, this may be prohibitively expensive for individuals of lower socioeconomic status or may not be readily available to those with limited access to healthcare services. This may further increase disparities by not making this screening tool affordable or accessible to higher-risk, lower-income SA individuals.
Growing awareness of the unique challenges in SA cardiovascular health has led to the creation of specialty clinics around the United States uniquely focused on lowering the burden of heart disease among SA individuals.
# Conclusion
Comprehensive and actionable risk stratification tools have great utility in tackling the burden of ASCVD among SA adults. Current data suggest that the use of CAC scoring is a cost effective and reproducible metric to detect subclinical atherosclerosis and further enhance ASCVD risk stratification in select individuals. While we have begun to understand the differences in the types and prevalence of risk factors, there is a need for further studies among SA individuals to understand the nuances and heterogeneity of risk even within this diverse group and standardized clinical guidelines that incorporate cultural practices and traditional beliefs to optimize care. Greater efforts are needed in raising awareness of heart disease and its risk factors using informative and culturally focused messaging from healthcare providers as well as through the community and cultural organizations as well as eliminating barriers to care (language, socioeconomic status, literacy etc.). Further studies are needed in disaggregated SA subgroups that have been traditionally underrepresented in Westernized studies, specifically Bangladeshi and Pakistani Americans.
## Key recommendations for the use of coronary artery calcium scoring in south
Asian adults ■ Individuals at borderline and intermediate ASCVD risk in whom the decision about statin use is uncertain. ■ Individuals who are at a higher risk than that estimated by traditional risk stratification tools i.e. those classified as low risk, but with a family history of premature ASCVD and/or two or more risk enhancing factors ■ Individuals with medical comorbidities that are highly prevalent drivers of ASCVD such as prediabetes and metabolic syndrome ■ Consider earlier CAC screening (<45 years) in individuals with multiple traditional risk factors, familial hypercholesterolemia, or a family history of (premature) CHD. . Proposed biometrics and risk stratification of South Asian Individuals.
# Authorship contribution
Anandita Agarwala and Jaideep Patel drafted the manuscript. Michael Blaha, Miguel Cainzos-Achirica, Khurram Nasir, and Matthew Budoff edited and provided key feedback on the manuscript.
# Funding
None.
## Declaration of competing interest
None of the authors have any conflicts of interest relevant to this paper.
[table] Table 2: Key Recommendations for Coronary Artery Calcium Scoring among South Asians. [/table]
|
Actions of Two Bi-Articular Muscles of the Lower Extremity: A Review
The extremities of the human body contain several bi-articular muscles. The actions produced by muscles at the joints they cross are greatly influenced by joint moment arms and muscle length. These factors are dynamic and subject to change as joint angles are altered. Therefore, to more completely understand the actions of such muscles, the angles of both joints must be manipulated. This report reviews investigations, which have explored the actions of two biarticular muscles of the lower extremities (gastrocnemius and rectus femoris) as the joints they cross are moved into various combinations of angles. The findings have both clinical and physical performance ramifications. Figure 4. Mean hip flexion SM, PM, MM values for all knee angles at each hip angle [7].
# Introduction
Bi-articular muscles are commonly found in the upper and lower extremities of the human body. These muscles generally cross two joints and influence movement at both. The rectus femoris (RF) spans the hip and knee, and the gastrocnemius (GA) crosses the knee and ankle. The actions of these muscles at their primary joints have been known for well over 100 years. The RF is an extensor of the leg, and the GA is a powerful plantarflexor. The descriptions of these particular actions have been relatively unchanged for many years and appear in most anatomy textbooks. However, these muscle action descriptions do not consider the influence the second joint may have on the muscle's action at the primary joint, or vice versa. For example, considering the GA action at the ankle, how does the plantarflexion (PF) torque it generates change as the angles of the knee and ankle change? At what combination does muscular insufficiency arise? Advances in technology have made it possible to answer questions of this type, resulting in more detailed descriptions of the bi-articular muscles of the extremities. In our previous papers, we discussed in some detail issues such as muscle tissue, joint moments and moment arms [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] and their effects on bi-articular muscle actions.
## Ga
The GA is one of 14 muscles that act upon the knee, and nine of these, including the GA are bi-articular. However, the GA is only one of these nine that acts on both the knee and ankle, and the others cross the knee and hip. At the knee, the GAs' actions oppose those of the quadriceps femoris, acting synergistically with the other primary knee flexors (biceps femoris, semitendinosus, semimembranosus, gracilis, popliteus, and sartorius). But it also belongs to another group of muscles that cross the ankle. It opposes the action of the dorsiflexors (e.g., tibialis anterior, extensor digitorum longus and hallucis longus) and it is a powerful plantarflexor working with other posterior leg muscles (e.g., soleus, tibialis posterior, flexor digitorum and hallucis longus). PF forces can be quite high. It is estimated that young males can generate PF torque ranging between 1,000 to 1,780 N.
The GA consists of two heads arising from the posterior aspects of the femoral condyles. These merge into a common belly that rides on the proximal half of the sural aspect of the leg. It shares an insertion with the soleus on the calcaneus via the achilles tendon. The GA and soleus are collectively referred to as the triceps surae and the innervation is supplied by the tibial nerve entering the proximal segments of the muscles. These two muscles provide approximately 80% of force of PF, which is a principal component to a large portion of the gait cycle and essential to nearly all forms of human locomotion. Di Nardo et al [bib_ref] Assessment of the activation modalities of gastrocnemius lateralis and tibialis anterior during..., Nardo [/bib_ref] explored the GA's role in PF using sEMG to measure activation patterns in the lateral portion of the GA when walking. These investigators found that the lateral portion of the GA was active in the stance phase when transitioning from flat foot contact to toe-off and then again in the final portion of the swing phase.
Riemann et al [bib_ref] The Effects of Sex, Riemann [/bib_ref] investigated the GA's contribution to the stability of the ankle across three ankle angles. These in- vestigators measured the stiffness values of the GA as the subject's ankles were passively moved from 10° of dorsiflexion through the neutral position and into 10° of PF, while the knee was alternately positioned at 0 and 90°. The results showed that the 0° knee angle produced higher stiffness values across the three ankle positions than the 90° knee angle. This shows that both joints influence the forces the GA produces. The stiffness index increased as the ankle moved from PF to dorsiflexion and actually began early in the neutral position. In our initial project with the GA [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] , we investigated how the knee and ankle joint angles interacted in the development of knee joint flexion moments. We used 17 subjects (10 females and seven males) in this project and measured the knee flexion moments produced by the GA at 24 knee and ankle joint positions. Knee angles of 0 (anatomical position), 15, 30, 45, 60, 75, 95, and 105° (Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] labeled the knee angles from the anatomical position as follows: 180, 165, 150, 135, 120, 105, 90, and 75°. In Landin et al, they used the more common clinical sequence starting with 0 for the anatomical position, and then progressing through 15, 30, 45, 60, 75, 90, and 105° of knee flexion. The Li et al sequence was converted to avoid confusion.) were crossed with three ankle positions (neutral, 15° dorsiflexed and plantarflexed). A Biodex System 3 dynamometer controlled the knee angles and removable casts were used to hold the ankle in each of its three positions.
Three dependent measures were of interest: 1) passive moment (PM), which was the joint moment without stimulation, and obtained by finding the mean of the joint moment before and after the stimulation period; 2) maximum moment (MM), which was the joint moment during the stimulation; and 3) stimulated moment (SM), which was the joint moment induced by the stimulation and was obtained by taking the difference between the MM and PM. Except where unique results were obtained, this review will deal with the MM findings since it represents the moments most applicable to the activities of daily living.
Results [fig_ref] Figure 1: MM across knee and ankle joint combinations[8] [/fig_ref] showed that the MM reduced linearly as knee flexion increased from the anatomical (0°) position. This reduction became greater as the ankle joint moved from dorsi-flexion to PF. Moving the ankle from dorsiflexion to the neutral position induced a greater reduction than the move from neutral to PF. Furthermore, MM production was influenced more by ankle joint changes with the knee between 0 and 30° compared to the lesser degrees of knee angle.
The most striking observation was that the knee flexion moment decreased approximately 50% as knee flexion moved from 0 to 30°. The reduction in the MM slowed across the 45 -75° knee angles. When the knee was flexed beyond 75°, the stimulation had little effect on the knee flexion moment. In other words, the GA's greatest contribution to knee flexion occurred with the knee in full extension. This was observed across the three ankle joint angles.
In a recent follow-up to the Li et al's [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] paper, Landin et al [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] used similar methods to measure the GA production of a PF joint moment. [fig_ref] Figure 2: Maximum plantarflexion moment across joint combinations[6] [/fig_ref] illustrates the principal finding regarding the MM. Significant effects were obtained for the knee. A fully extended knee (0°) and a dorsiflexed ankle (+15°) created the highest (24.95 ± 10.1 Nm) torque values, while the 105° knee angle and a plantarflexed ankle (-15°) produced the smallest values (13.02 ± 4.9 Nm). Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] reported that the GA produced the greatest knee flexion moment in the same joint combination, although the force was not as high (approximately 9.5 Nm), as we noted for PF (approximately 25 Nm). We contend that this is most likely due to the disparity in the mechanical advantage of the knee and ankle lever systems. At the knee, the GA works in a thirdclass lever, which produces little mechanical advantage. At the ankle, however, the GA is part of a second-class lever, which has considerable mechanical power. Based on our results and those of Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] , it is the 0° knee position and a dorsiflexed ankle that creates the optimal length for the GA to work at either joint. It is apparently the mechanical properties of the lever system that causes the disparity in the GA's knee and ankle joint moments.
Regarding the lowest torque, the results reported in our more recent paper are somewhat in contrast to Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref]. Their results showed that flexing the knee to 90° or more, regardless of the ankle position, did not significantly reduce torque any further. More precisely, the knee flexion moment did not significantly decrease as the knee was flexed to 105° with the ankle plantarflexed, neutral, or dorsiflexed. In our more recent project [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] , we found that the PF joint moment decreased linearly from the 0/+15 joint combination, through the 105/-15 combination, at which point the GA was in its most shortened position. It is interesting to note that even in the 105/-15 position, the PF MM (13.02 Nm) exceeded the knee MM reported by Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] at the 0/+15 position. The GA produced a greater PF torque in the position of shortest length than it produced for knee flexion in its most elongated position. As argued above, this discrepancy is most likely due to the mechanical properties of a second-class compared to a third-class lever. This lever effect also likely explains the fact that Li et al found the smallest knee flexion moment before reaching the shortest position for the GA. The mechanical properties of a third-class lever may have rendered the GA insufficient before it reached its shortest length, whereas the second-class lever for the GA at the ankle allowed it to generate a PF torque force throughout the range motion used in these studies. Further research using knee flexion angles beyond 105° may provide further information on this issue. Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] argued that their findings were clinically important due to the GA's perceived role in knee joint stability during various lower extremity movements. Following knee joint traumas such as ACL ruptures, the GA's synergistic work with the quadriceps assists in stabilizing the knee during weight bearing motions [bib_ref] Anterior positioning of tibia during motion after anterior cruciate ligament injury, Kvist [/bib_ref]. Their results suggested that flexed knee positions (> 30°), where the GA has less influence should be avoided, particularly in the early stages of rehabilitation. Rehabilitation exercises that maintain the knee in a position of stability (< 30°) appear to be best. Maintaining this knee angle while stimulating co-contraction of all the dynamic stabilizers at the knee, which includes the GA, enhances stability through increased tibiofemoral joint compression, joint stiffness, and proprioceptive influences as suggested by Wilk et al [bib_ref] Rehabilita-Actions of Two Bi-Articular Muscles, Wilk [/bib_ref].
The clinical implications of Landin et al [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] suggest that having a patient seated at the end of an examination table with the foot subject to gravity, while useful for evaluating the triceps surae, is a poor position from which to assess the GA as a plantarflexor. This position places the knee at greater than 60° of flexion and maintains a somewhat plantarflexed foot, both of which lead to significant declines in GA torque production. Even if the clinician holds the patient's ankle in a dorsiflexed position, the flexed knee will still mask the GA's true strength.
Manual testing of the GA in isolation should be performed, whenever possible, with the knee extended and the ankle dorsiflexed to potentially elicit the maximum PF torque from the GA. The recent work with the GA reveals that the knee angle is the main factor in the development of both a knee flexion and PF moment of the GA. The GA is a much more powerful plantarflexor than it is a knee flexor, due largely to the mechanical advantage of a second-class lever. However, it is also possible that the 30 total degrees of the ankle ROM used in Li et al [bib_ref] The function of gastrocnemius as a knee flexor at selected knee and..., Li [/bib_ref] and Landin et al [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] was too restrictive. Total ankle ROM (dorsiflexion through PF) can approach 85°. Moving the ankle closer to the extreme ends of the ROM may have altered the findings. There are difficulties with attaining the full ankle ROM. The dorsiflexion range is reduced as the knee moves closer to full extension, and therefore, cannot be achieved in all joint combinations. Still, future investigations should employ as much of the ankle ROM as possible.
## Rf
The RF is the sole bi-articular muscle of the quadriceps group. A direct head and a reflected head form its proximal attachments. The direct attachment arises from the anterior inferior iliac spine (AIIS) of the pelvis, while the reflected attachment emerges from the rim of the acetabulum and the fibrous capsule of the hip joint. The two proximal attachments quickly blend into a common belly that coalesces into the quadriceps tendon and, in conjunction with the vasti muscles, then inserts on the tibial tuberosity.
The knee extension work of the quadriceps group is widely known and well documented, with a 60 -90° knee angle appearing optimal [bib_ref] Eccentric and concentric torquevelocity characteristics, torque output comparisons, and gravity effect torque..., Westing [/bib_ref]. Although the function of the RF as a hip flexor is largely unknown, anatomists have traditionally included it with the hip flexor group. At the hip, the RF works in a third-class lever, meaning that the application of the force lies between a proximal fulcrum and a distal resistance. As such, the hip is specialized for range and speed of motion, but high demands are placed on the muscles moving the lever. Consequently, the RF is heavily involved in all swinging actions of the lower extremity. In Landin et al [bib_ref] Knee and Ankle Joint Angles Influence the Plantarflexion Torque of the Gastrocnemius, Landin [/bib_ref] , we cited research with kicking for an often used example. Hsu et al [bib_ref] Proximal rectus femoris avulsions in national football league kickers: a report of..., Hsu [/bib_ref] described this action as placing unique demands on the RF for it is the only sport-related movement that requires substantial hyperextension of the hip at the outset. This passively stretches the RF, then while assisting (to some unknown extent) with the initial hip flexion motion, the RF also quickly becomes involved with extending the leg. As a result, the RF is working simultaneously across two joints and becomes subject to distal strains or tears.
From a clinical perspective, the role of RF in hip flexion pertains to muscle imbalance issues. Muscles can react to stress different ways, and an injury, volitional inactivity, or both can produce stress. Postural muscles tend to tighten, while their antagonistic muscles (the phasic group) tend to weaken. When postural muscles (e.g., iliopsoas, RF, and quadratus lumborum) around the pelvis tighten, the phasic muscles (e.g., rectus abdominis, gluteals, vastus lateralis, medialis, and intermedius) weaken and imbalances arise [bib_ref] Abdominal muscle training in sport, Norris [/bib_ref].
Jandacoined the phrase "lower crossed syndrome" (LCS) to describe a muscle imbalance across the pelvis. In LCS, the gluteals and abdominals lengthen and weaken, while the iliopsoas and RF shorten and tighten. This often leads to an anterior pelvic tilt, which in turn causes a variety of musculo-skeletal maladies and poor posture. What makes the LCS particularly troublesome is that sitting, even with good posture, creates conditions favorable to its development. Furthermore, rehabilitation protocols following a variety of lower extremity injuries strengthen the RF and the three vasti muscles are a primary focus of the exercises. Quadriceps strength is a key element in making the knee joint more stable and in slowing the progression of osteoarthritis, particularly in partially menisectomized knees [bib_ref] Knee strength and knee adduction moments following arthroscopic partial meniscectomy, Sturnieks [/bib_ref]. However, since the RF is one of the postural muscles that tend to tighten, the potential exists for an exacerbation of a muscle imbalance and LCS. If the RF plays a significant role in hip flexion, then increasing its strength in a quadriceps regimen presents a quandary. Should the RF not substantially contribute to hip flexion, then the use of quadriceps protocols in rehabilitation can be employed without concern for creating or exacerbating an existing LCS.
In our project with the RF and hip flexion, we used 16 female volunteers from a university undergraduate population. The equipment and procedures generally followed the descriptions provided in previous sections. In this project, the Biodex chair controlled hip angles and the removable casts controlled the knee angles. Six hip angles (85, 70, 55, 40, 25, and 10°) were crossed with four knee angles (0 the anatomical position, 30, 60, and 90°) to create 24 joint combinations. At the hip, the 85° (upright) and the 10° (reclined) positions represented the limits of the Biodex chair. [fig_ref] Figure 3: Maximum hip flexion moment across joint combinations[7] [/fig_ref] displays the MM of the hip flexion function of the RF across the various joint combinations. The knee produced the only significant effect on the hip flexion moment of the RF. As knee flexion increased from 0 to 90°, the hip flexion moment rose from 10.44 to 15.3 Nm. This can be seen on [fig_ref] Figure 3: Maximum hip flexion moment across joint combinations[7] [/fig_ref] at the 10° hip angle.
An intriguing result of this project was the contribution of the PM to the MM. As can be seen in , a large portion of the MM was produced by the PM. We reported that this ranged from 74% to 78% across the joint combinations. Con- sequently, the stimulation, as shown by the SM line in contributed little to the MM. The hip is so thoroughly supported by various muscle and connective tissue components that in a passive state simply extending the hip can create a substantial hip flexion torque. Moving the hip to the 10° position used in Landin et almay not have produced an optimal length [bib_ref] The variation in isometric tension with sarcomere length in vertebrate muscle fibres, Gordon [/bib_ref] for the RF during stimulation, but clearly created tension in the other soft tissue structures surrounding the joint. These data show that contraction of the RF accounts for approximately 25% of the total hip flexion torque, indicating that chronic shortening of the phasic muscles and related fascial components can, even without strengthening of the RF, disrupt the postural/phasic muscle balance. This finding makes clear how the LCS described by Jandacan be problematic. One avenue of future work would be investigating the effects stretching routines, either chronic or acute, have on the PM. Shortening of the myofascial components surrounding the hip has been discussed in the literature for some time, and findings from over 20 years ago show that elongating these tissue elements can improve gait [bib_ref] The effects of two stretching procedures on hip range of motion and..., Godges [/bib_ref]. Furthermore, single bouts of stretching have a deleterious effect on power production [bib_ref] Static stretching impairs sprint performance in collegiate track and field athletes, Winchester [/bib_ref] , so there may be an effect on a PM, but it would not necessarily be detrimental regarding LCS issues. A recent study of muscle activation patterns during gait showed that the RF displays high activity at the start of the swing phase and at stance phase initiation [bib_ref] Assessment of the activation modalities of gastrocnemius lateralis and tibialis anterior during..., Nardo [/bib_ref]. Our resultsdo not contradict these, but rather indicate that, at the hip angles we used, contraction of the RF does not contribute much to hip flexion. Schadel et al [bib_ref] Relation of anterior pelvic tilt during running to clinical and kinematic measures..., Schache [/bib_ref] demonstrated that considerable variability in hip extension exists between runners. Motion analysis of 14 elite runners performing a treadmill run at 21 km/h showed an average hip extension of -11.7°, with a range of about 20°. Due to limits of the Biodex chair we were unable to extend the hip beyond +10°, which is just shy of the anatomical position. A second direction for future work should include hip extension angles that approximate those reported by Schadel et al [bib_ref] Relation of anterior pelvic tilt during running to clinical and kinematic measures..., Schache [/bib_ref]. It may well be that, since the 10/90 hip and knee position used in our project may have elongated the RF beyond its optimal length, it seems unlikely that greater lengthening would be beneficial. Still, further investigation is needed.
[fig] Figure 1: MM across knee and ankle joint combinations[8]. [/fig]
[fig] Figure 2: Maximum plantarflexion moment across joint combinations[6]. [/fig]
[fig] Figure 3: Maximum hip flexion moment across joint combinations[7]. [/fig]
|
Antimicrobial Activity of a Triple Antibiotic Combination Toward Ocular Pseudomonas aeruginosa Clinical Isolates
Citation: Mei JA, Johnson W, Kinn B, Laskey E, Nolin L, Bhamare P, Stalker C, Dunman PM, Wozniak RAF. Antimicrobial activity of a triple antibiotic combination toward ocular pseudomonas aeruginosa clinical isolates. Transl Vis Sci Technol. 2022;11(5):26, https://doi.org/10.1167/tvst.11.5.26 Purpose: Pseudomonas aeruginosa is a leading cause of corneal infections. Recently, we discovered an antimicrobial drug combination, polymyxin B/trimethoprim (PT) + rifampin, that displayed impressive efficacy toward P. aeruginosa in both in vitro and in vivo studies. As such, this combination was further evaluated as a potential keratitis therapeutic through testing the combination's efficacy against a diverse set of P. aeruginosa clinical isolates.Methods: Minimum inhibitory concentrations (MICs) of moxifloxacin, levofloxacin, erythromycin, tobramycin, PT, polymyxin B (alone), trimethoprim (alone), and rifampin were determined for 154 ocular clinical P. aeruginosa isolates, 90% of which were derived from corneal scrapings. Additionally, the efficacy of PT + rifampin was evaluated utilizing fractional inhibitory concentration (FIC) testing.Results: While 100% of isolates were resistant to erythromycin (average MIC 224 ± 110 μg·mL −1 ) and trimethoprim (alone) (206 ± 67.3 μg·mL −1 ), antibiotic resistance was generally found to be low: moxifloxacin (2% of isolates resistant; average MIC 1.08 ± 1.61 μg·mL −1 ), levofloxacin (3.9%; 1.02 ± 2.96 μg·mL −1 ), tobramycin (1%; 0.319 ± 1.31 μg·mL −1 ), polymyxin B (0%; 0.539 ± 0.206 μg·mL −1 ), PT (0%; 0.416 ± 0.135 μg·mL −1 ), and rifampin (0%; 23.4 ± 6.86 μg·mL −1 ). Additionally, FIC testing revealed that PT + rifampin eradicated 100% of isolates demonstrating additive or synergistic activity in 95% of isolates (average FIC index 0.701 ± 0.132).Conclusions:The drug combination of PT + rifampin was effective against a large panel of clinically relevant P. aeruginosa strains and, as such, may represent a promising therapeutic for P. aeruginosa keratitis.Translational Relevance: This work furthers the preclinical development of a novel antibiotic combination for the treatment of corneal infections (bacterial keratitis).
# Introduction
Bacterial keratitis (corneal infection) is a serious disease requiring urgent topical antimicrobial treatment to mitigate ocular tissue damage and preserve vision. While a wide variety of micro-organisms have been implicated in bacterial keratitis, Pseudomonas aeruginosa stands out as the leading Gram-negative pathogen implicated in this disease, particularly in contact lens wearers. [bib_ref] The persistent dilemma of microbial keratitis: global burden, diagnosis, and antimicrobial resistance, Ung [/bib_ref] [bib_ref] Pseudomonas aeruginosa and microbial keratitis, Hilliam [/bib_ref] [bib_ref] Infectious keratitis: an update on epidemiology, causative microorganisms, risk factors, and antimicrobial..., Ting [/bib_ref] In fact, some studies have indicated that among all organisms responsible for contact lens-associated keratitis, the prevalence of P. aeruginosa can be as high as 70%. [bib_ref] Overview of mechanisms of antibiotic resistance in Pseudomonas aeruginosa: an ocular perspective, Subedi [/bib_ref] [bib_ref] Clinical presentation and morbidity of contact lens-associated microbial keratitis: a retrospective study, Hoddenbach [/bib_ref] [bib_ref] Antibiotic susceptibility patterns of pseudomonas corneal ulcers in contact lens wearers, Mohammadpour [/bib_ref] With a current estimated 38.5 million contact lens wearers in the United States 7 and the rising popularity of multifocal, toric, and novelty contact lenses, P. aeruginosa keratitis has become a major health care and ophthalmic concern.
Currently, topical ophthalmic fluoroquinolones are widely used in the treatment of P. aeruginosa bacterial keratitis given their broad-spectrum activity, excellent tissue penetration, and patient tolerability. Fortunately, in the United States, circulating antibiotic resistance among ocular P. aeruginosa isolates toward fluoroquinolones as well as other ophthalmic antibiotics has remained low. [bib_ref] Antibiotic resistance among ocular pathogens in the United States: five-year results from..., Asbell [/bib_ref] [bib_ref] Antibiotic resistance among pediatricsourced ocular pathogens: 8-year findings from the Antibiotic Resistance..., Alter [/bib_ref] However, reports are now emerging describing significant P. aeruginosa resistance globally. For example, P. aeruginosa resistance to moxifloxacin, a commonly utilized fourth-generation fluoroquinolone as well as the aminoglycoside gentamicin, has been reported upward of 50% in large patient series based in India. [bib_ref] Emerging moxifloxacin resistance in Pseudomonas aeruginosa keratitis isolates in South India, Oldenburg [/bib_ref] [bib_ref] Susceptibility trends of pseudomonas species from corneal ulcers, Smitha [/bib_ref] Unfortunately, the clinical consequences of resistant infections are significant and include increased disease severity and worse visual outcomes. [bib_ref] Relationship of in vitro susceptibility to moxifloxacin and in vivo clinical outcome..., Lalitha [/bib_ref] [bib_ref] Bacterial susceptibility to topical antimicrobials and clinical outcome in bacterial keratitis, Kaye [/bib_ref] However, despite this worrisome trend of emerging antibiotic resistance, there is a paucity of commercial alternatives to fluoroquinolones in the treatment of keratitis.
We recently described the synergistic antimicrobial activity of a novel drug combination, polymyxin B/trimethoprim (PT) + rifampin toward P. aeruginosa and Staphylococcus aureus, another leading cause of keratitis, in both in vitro and in vivo studies. [bib_ref] A novel, broad-spectrum antimicrobial combination for the treatment of Pseudomonas aeruginosa corneal..., Chojnacki [/bib_ref] [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] While PT alone is commonly used for the treatment of mild bacterial conjunctivitis, its use in more serious corneal infections is limited due to weak antimicrobial potency and inadequate tissue penetration compared to fluoroquinolones. [bib_ref] Speed of bacterial kill with a fluoroquinolone compared with nonfluoroquinolones: clinical implications..., Lichtenstein [/bib_ref] [bib_ref] Penetration of topically administered ofloxacin and trimethoprim into aqueous humor, Price [/bib_ref] [bib_ref] A multicenter comparison of polymyxin B sulfate/trimethoprim ophthalmic solution and moxifloxacin in..., Granet [/bib_ref] However, we have demonstrated that the combination of PT + rifampin overcomes these liabilities and displays antimicrobial efficacy in a murine model of bacterial keratitis that equals or exceeds that of fourth-generation fluoroquinolones. [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] Importantly, PT + rifampin exhibits in vivo efficacy toward fluoroquinolone-resistant S. aureus and P. aeruginosa strains, suggesting that it may be an effective treatment option for infections that otherwise might fail currently available options. [bib_ref] A novel, broad-spectrum antimicrobial combination for the treatment of Pseudomonas aeruginosa corneal..., Chojnacki [/bib_ref] [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] To further investigate the incidence of antibiotic resistance and expand understanding of the potential therapeutic value of PT + rifampin for bacterial keratitis, we performed antimicrobial activity assays on a contemporary collection of 154 P. aeruginosa clinical ocular isolates. The entire strain set was first evaluated for antibiotic resistance to a panel of antibiotics that are commonly used for ocular treatment, including erythromycin, PT, levofloxacin, moxifloxacin, and tobramycin. Second, we evaluated the antimicrobial efficacy of PT + rifampin as well as rifampin (alone), trimethoprim (alone), and polymyxin B (alone) toward the entire isolate collection. Third, we evaluated whether the antimicrobial performance of PT + rifampin is a consequence of the combination's synergistic effects. Collectively, our results demonstrate that with the exception of erythromycin and trimethoprim, resistance levels remain low among US ocular P. aeruginosa isolates. Additionally, we establish that the combination of PT + rifampin displays potent antimicrobial activity with 100% of the isolate collection susceptible to the combination, including strains resistant to fluoroquinolones. Further, we show that the antimicrobial effectiveness of PT+ rifampin is associated with the combination's synergistic effects. Thus, PT + rifampin may represent a potential novel therapeutic to treat this blinding disease.
# Materials and methods
## Bacterial strains and growth conditions
A total of 154 P. aeruginosa clinical ocular isolates were commercially obtained from International Health Management Associates (IHMA) (Schaumburg, IL, USA). Individual isolates were grown overnight and subcultured in fresh Mueller-Hinton (MH) media shaking with aeration at 37°C to early exponential phase. A range of OD 600 = 0.2 to 0.4 was found to correspond to approximately 10 8 colony-forming units (CFU)/mL, and subcultures were then diluted 1:100 in fresh media for subsequent minimum inhibitory concentration (MIC) and fractional inhibitory concentration (FIC) testing. The pan-sensitive laboratory strain PAO1 was used as a control for this study.
## Mic
Each clinical isolate was tested for susceptibility to a panel of eight clinically relevant antibiotics, including erythromycin, PT, rifampin, moxifloxacin, levofloxacin, trimethoprim, tobramycin, and polymyxin B, using the standard MIC testing guidelines. [bib_ref] Speed of bacterial kill with a fluoroquinolone compared with nonfluoroquinolones: clinical implications..., Lichtenstein [/bib_ref] Individual wells of a 96-well microtiter plate were prepared by adding 88 μL of fresh MH broth, 2 μL of increasing concentrations of each antibiotic, and 10 μL of the appropriate bacterial culture to achieve a final concentration 10 4 CFU/well and incubated for 16 hours at 37°C. The MIC value for each antibiotic was determined as the lowest concentration of antibiotic that inhibited bacterial growth as visualized by the naked eye. Resistance was characterized as having an MIC value of ≥4 μg/mL for levofloxacin, ≥16 μg/mL for tobramycin, and ≥4 μg/mL for polymyxin B per the 2021 Clinical and Laboratory Standards Institute (CLSI).Resistance for the remaining antibiotics (PT, rifampin, and moxifloxacin) was characterized as having MIC values ≥4× the MIC value for a susceptible laboratory strain, PAO1. Erythromycin and trimethoprim were used as controls due to the known insusceptibility of P. aeruginosa to these antibiotics. 20
## Fic testing of pt + rifampin
A standard checkerboard assay was done to determine the efficacy of the drug combination (PT plus rifampin) on the clinical strain set according to CLSI guidelines.In total, 10 μL of the indicated 10 6 CFU/mL bacterial culture was added to 88 μL MH media and 2 μL of antibiotics in individual wells of a 96-well microtiter plate with each column containing twofold increasing concentrations of PT (0.0-2.5 μg/mL) and each row containing twofold increasing concentrations of rifampin (0.0-60 μg/mL). PT + rifampin drug combinations were prepared separately in 200 μL to minimize pipetting errors. The plates were then incubated for 16 hours at 37°C and visually inspected for growth. The fractional inhibitory concentration index (FICI) was calculated using the following formula: FICI = (MIC of PT in combination / MIC of PT alone) + (MIC of rifampin in combination / MIC of rifampin alone). The averaged FICI from three biological replicates was defined as either synergistic (FICI <0.5), additive (0.5-1), indifference (1-4), or antagonistic (FICI >4). 21
# Results
## Strain set characteristics
In total, 154 ocular P. aeruginosa isolates were obtained from IHMA between 2016 and 2020. [fig_ref] Table 1: Characteristics of 154 Ocular Pseudomonas aeruginosa Clinical Set [/fig_ref] provides the characteristics of the strain set. Fortysix percent (n = 71) of isolates were collected from male patients and 54% (n = 83) from female patients. Ages of patients at the time of isolate collection ranged from 1 to 104, with the majority of patients aged 40 to 59 (31%) and 60 to 79 years (27%). Ninety percent of the isolates were isolated from corneal scrapings, and 10% were broadly categorized as from eyes that could include corneal, conjunctival, intracameral, and/or intravitreal samples. The geographic representation included 139 (90%) of isolates from North America, 13 (8%) from Europe, 1 (1%) from Latin America, and 1 (1%) from Asia. Among isolates collected from the United States, 15 states were represented: Alabama (n = 5), California (n = 46), Colorado
## Antibiotic resistance profiles of isolates toward commercially available ophthalmic antibiotics
MIC testing was performed on the entire strain set in triplicate to measure the effectiveness of five commonly used ophthalmic antibiotics: moxifloxacin, levofloxacin, erythromycin, tobramycin, and PT. Additionally, to further support the preclinical development of the novel drug combination PT + rifampin as a potential keratitis therapeutic, MIC testing was completed for the individual components: trimethoprim (alone), polymyxin B (alone), and rifampin. Resistance was characterized as having an MIC value above the 2021 CLSI 19 break point when available or an MIC value ≥4× the MIC value for a susceptible laboratory strain, PAO1.
MIC testing revealed that overall resistance was very low among this set of clinical isolates, with only six strains (3.9%) resistant to one antibiotic and one strain that demonstrated multidrug resistance as defined by resistance to three or more classes of antibiotics [bib_ref] Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim..., Magiorakos [/bib_ref] [fig_ref] Table 1: Characteristics of 154 Ocular Pseudomonas aeruginosa Clinical Set [/fig_ref]. More specifically, as expected, 100% of isolates were resistant to erythromycin (average MIC 224 ± 110 μg·mL −1 ) and trimethoprim (206 ± 67.3 μg·mL −1 ) due to known
## Potent antimicrobial activity of pt + rifampin
Previous studies have demonstrated that the combination of PT + rifampin is a synergistic, broadspectrum antimicrobial. Moreover, PT + rifampin has also been shown to effectively eradicate both P. aeruginosa and S. aureus murine keratitis infections caused by antibiotic-resistant strains. [bib_ref] A novel, broad-spectrum antimicrobial combination for the treatment of Pseudomonas aeruginosa corneal..., Chojnacki [/bib_ref] [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] Thus, in order to further investigate the therapeutic promise of PT + rifampin, we tested the antimicrobial susceptibility of the entire strain set to the combination.
The triple antibiotic combination, PT + rifampin, demonstrated synergistic or additive antimicrobial activity in 146 isolates (95%) with FICI values ranging from 0.446 to 1.005 (average 0.701 ± .132) [fig_ref] Table 2: Antibiotic Resistance among the 154-Member Clinical Strain Set [/fig_ref]. Specifically, PT + rifampin displayed synergistic activity toward three isolates (average FICI 0.457 ± 0.012) and additive toward 143 isolates (FICI 0.690 ± 0.111), and the combination had a neutral effect toward the remaining eight isolates (FICI = 1 ± 0.002). In line with these findings, we found a corresponding reduction in the MICs for both PT and rifampin when the drugs were tested against the clinical strain set in combination versus alone. For example, on average, there was a twofold reduction in the MIC for PT when measured alone versus in combination with rifampin (0.416 ± 0.135 μg/mL vs. 0.180 ± 0.045 μg/mL) and a fourfold reduction in the MIC of rifampin when in combination with PT versus alone (23.4 ± 6.86 μg/mL vs. 5.41 ± 3.11 μg/mL) [fig_ref] Table 3: FIC Testing of PT + Rifampin against 154-Member P [/fig_ref]. Importantly, PT + rifampin displayed this potent activity, even toward drug resistant isolates. For example, IHMA1564153, a multidrug-resistant strain with resistance to moxifloxacin, levofloxacin, and tobramycin, was found to be highly susceptible to PT + rifampin with a synergistic FICI value of 0.458 ± 0.072. Moreover, the efficacy of PT + rifampin was equivalent when comparing fluoroquinolone-resistant versus fluoroquinolone-sensitive isolates. The average FICI value of all six levofloxacin-resistant isolates (three of which were also resistant to moxifloxacin) was found to be 0.681 ± 0.18, compared to 0.702 ± 0.132 for all fluoroquinolone-sensitive strains.
Taken together, these data demonstrate the potent antimicrobial activity of the combination of PT + rifampin and its ability to effectively eradicate 100%
# Discussion
Antibiotic resistance is currently one of the most pressing concerns in modern medicine, resulting in significant clinical adverse outcomes due to treatment failures in addition to escalating health care costs. [bib_ref] The impact of antimicrobial resistance on health and economic outcomes, Cosgrove [/bib_ref] [bib_ref] Comparison of mortality associated with methicillinresistant and methicillin-susceptible Staphylococcus aureus bacteremia: a..., Cosgrove [/bib_ref] [bib_ref] The negative impact of antibiotic resistance, Friedman [/bib_ref] P. aeruginosa, one of the most common causes of ocular infections, has been designated by the Centers for Disease Control and Prevention as an ESKAPE pathogen, one of a group of six organisms of particular health care concern given their ability to "escape" traditional antimicrobial therapies as well as a "serious threat" in the recently published 2019 Antibiotic Resistant Threats in the United States.Fortunately, however, our current data indicate very low circulating antibiotic resistance among P. aeruginosa ocular isolates, a finding that is supported by other large-scale surveillance studies. For example, the Antibiotic Resistance Monitoring in Ocular Microorganisms (ARMOR) study reported resistance rates of approximately 5% or lower to fluoroquinolones such as ofloxacin, ciprofloxacin, levofloxacin, and gatifloxacin in a set of nearly 700 clinical isolates collected from 2009 to 2018 in the United States. [bib_ref] Trends in antibiotic resistance among ocular microorganisms in the United States from, Asbell [/bib_ref] Despite the current reassuring low rates of antibiotic resistance among P. aeruginosa ocular isolates in the United States, reports from India describe emerging antibiotic resistance particularly toward fluoroquinolones, the mainstay of treatment of corneal infections. Given that resistance will undoubtedly rise, the development of novel antimicrobial therapeutics is essential to stay ahead of the curve. In response to this growing need, we have recently described a novel drug combination containing PT + rifampin that displays impressive antimicrobial activity toward the two most common causes of keratitis, S. aureus and P. aeruginosa. [bib_ref] A novel, broad-spectrum antimicrobial combination for the treatment of Pseudomonas aeruginosa corneal..., Chojnacki [/bib_ref] [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] Importantly, the potent activity of PT + rifampin extended to an in vivo keratitis model, where it was shown to successfully eradicate P. aeruginosa and S. aureus corneal infections, even those caused by fluoroquinolone-resistant clinical isolates. [bib_ref] A novel, broad-spectrum antimicrobial combination for the treatment of Pseudomonas aeruginosa corneal..., Chojnacki [/bib_ref] [bib_ref] Development of a broad-spectrum antimicrobial combination for the treatment of Staphylococcus aureus..., Chojnacki [/bib_ref] Thus, to further advance our understanding of the therapeutic potential of PT + rifampin, we evaluated the combination's activity toward a diverse panel of clinically relevant ocular isolates.
Our results indicate that the combination of PT + rifampin is effective toward contemporary circulating P. aeruginosa strains, including those that are resistant to one or more currently available antibiotics. Importantly, the combination of PT + rifampin appears to be synergistic or additive toward 95% of these isolates, as reflected in the decreased concentrations of PT and rifampin required for efficacy when in combination compared to each agent individually.
Given that current standard of care necessitates intensive, frequent dosing of topical antibiotics in cases of severe keratitis, the potency of PT + rifampin may allow for reduced dosing schedules for patients and potentially decreased toxicity to the ocular surface. The effectiveness of this combination may be due, in part, to its multiple mechanisms of action. While polymyxin B acts as a detergent to disrupt bacterial cell membranes, trimethoprim inhibits bacterial DNA synthesis through the inhibition of dihydrofolate reductase, and rifampin inhibits DNA transcription through binding to bacterial RNA polymerase. [bib_ref] Polymyxins: antibacterial activity, susceptibility testing, and resistance mechanisms encoded by plasmids or..., Poirel [/bib_ref] [bib_ref] Structural mechanism for rifampicin inhibition of bacterial RNA polymerase, Campbell [/bib_ref] [bib_ref] Trimethoprim: mechanisms of action, antimicrobial activity, bacterial resistance, pharmacokinetics, adverse reactions, and..., Gleckman [/bib_ref] In summary, in an era of rising antibiotic resistance, the need for novel therapeutics is critical. This is particularly true for the treatment of keratitis, in which immediate empiric therapy that can successfully contend with circulating antibiotic resistance is necessary to prevent permanent ocular tissue damage. We have demonstrated the potency of a novel antibiotic combination, PT + rifampin, to successfully eradicate ocular clinical isolates of P. aeruginosa with varying resistance profiles, suggesting that the combination of PT + rifampin may represent a promising new therapeutic option to fill this critical need.
[table] Table 1: Characteristics of 154 Ocular Pseudomonas aeruginosa Clinical Set [/table]
[table] Table 2: Antibiotic Resistance among the 154-Member Clinical Strain Set [/table]
[table] Table 3: FIC Testing of PT + Rifampin against 154-Member P. aeruginosa Ocular Strain Set [/table]
|
Recurrent COVID‐19 pneumonia in the course of chemotherapy: Consequence of a weakened immune system?
[bib_ref] Lymphopenia predicts disease severity of COVID-19: a descriptive and predictive study, Tan [/bib_ref] [bib_ref] Renin-angiotensin-aldosterone system inhibitors in patients 198 with covid-19, Vaduganathan [/bib_ref] [bib_ref] Not just antibodies: B cells and T cells mediate immunity to COVID-19, Cox [/bib_ref] [bib_ref] Cancer patients in SARS-CoV-2 infection: a nationwide analysis in China, Liang [/bib_ref] [bib_ref] COVID-19: is it just a lung disease? A case-based review, Spuntarelli [/bib_ref] [bib_ref] Extremely severe case of COVID-19 pneumonia recovered despite bad prognostic indicators: a..., Bentivegna [/bib_ref] [bib_ref] Incidence and determinants of high-sensitivity troponin and natriuretic peptides elevation at admission..., Arcari [/bib_ref]
## Conflict of interests
The authors declare that there are no conflict of interests.
# Author contributions
Michelangelo Luciani participated in article preparation. All authors materially participated in the research and data collection.
Luciano De Biase, Paolo Martelletti, and Georgio Sesti have approved the final article.
# Ethics statement
Written informed consent was obtained by the patient for publication of this case report and accompanying images. Prot. n. 52 SA_2020 del Registro Pareri Rif. CE 5773_2020.
F I G U R E 1 HRTC at the emergency room (1a, 1b, 1c) Multiple areas of frosted glass hyperdensity are observed, mainly located in the sub-pleural area with spotted distribution. In the context of these areas there is fibrous thickening of the inter and intra lobular septa (crazy paving pattern) and increase in the size of the vessels. There are some areas of parenchymal consolidation with aerial bronchogram in the context, especially in the right upper lobe and middle lobe. The parenchymal alterations described are localized to all the pulmonary lobes. HRTC at the discharge (1d, 1e, 1f) Previous injuries described have almost completely regressed. HRTC, high-resolution chest tomography |
A Conceptual Framework for Planning Systemic Human Adaptation to Global Warming
Human activity is having multiple, inter-related effects on ecosystems. Greenhouse gas emissions persisting along current trajectories threaten to significantly alter human society. At 0.85 °C of anthropogenic warming, deleterious human impacts are acutely evident. Additional warming of 0.5 °C-1.0 °C from already emitted CO2 will further intensify extreme heat and damaging storm events. Failing to sufficiently address this trend will have a heavy human toll directly and indirectly on health. Along with mitigation efforts, societal adaptation to a warmer world is imperative. Adaptation efforts need to be significantly upscaled to prepare society to lessen the public health effects of rising temperatures. Modifying societal behaviour is inherently complex and presents a major policy challenge. We propose a social systems framework for conceptualizing adaptation that maps out three domains within the adaptation policy landscape: acclimatisation, behavioural adaptation and technological adaptation, which operate at societal and personal levels. We propose that overlaying this framework on a systems approach to societal change planning methods will enhance governments' capacity and efficacy in strategic planning for adaptation. This conceptual framework provides a policy oriented planning assessment tool that will help planners match interventions to the behaviours being targeted for change. We provide illustrative examples to demonstrate the framework's application as a planning tool.
# Introduction
Human activity is having multiple, inter-related systemic effects on the Earth's environment [bib_ref] Planetary boundaries: Exploring the safe operating space for humanity, Rockström [/bib_ref]. These effects are a function of human population size, our socially defined choices about the source and quantity of energy and materials' usage, and the energy efficiency of technology used [bib_ref] Impact of population growth, Ehrlich [/bib_ref]. It is inevitable that on a planet with finite resources and physical space, unrestrained population growth and resource use will ultimately overwhelm the capacity of planetary systems to provide and replenish [bib_ref] Will limits of the earth's resources control human numbers? Environ, Pimentel [/bib_ref]. Due to a lack of conviction to the science and entrenched socio-economic paradigms, human behaviour has not adequately responded to the environmental and societal vulnerabilities presaged by these effects, which are now rebounding on human health and prosperity.
One part of this global environmental change, the anthropogenically unbalanced greenhouse gas cycle, contributes ocean acidification and global warming to the mix. Sea level rise and climate change are two of the many consequences of warming. On current trajectories, humanity may have to cope with 2.6 °C-4.8 °C total warming this century. The potential for global warming to adversely affect human wellbeing and health is present and growing [bib_ref] Climate change: Present and future risks to health, and necessary responses, Mcmichael [/bib_ref] [bib_ref] Contribution of Working Group II to the Fifth Assessment Report of the..., Ipcc [/bib_ref]. Human health effects will be both direct and indirect, and are already occurring. More frequent extreme heat and more damaging storm events can be expected. Global warming has the most immediate potential to disrupt human society, and will act synergistically with the other aspects of global environmental change (Global environmental change refers to the multiple environmental and ecological changes comprising: biodiversity loss, deforestation and land degradation, ozone depletion, warming, ocean acidification, fresh water scarcity, persistent organic pollutants, other pollution, waste, that all together threaten fresh water, agriculture, fisheries and the ecological systems that support human society.).
The science behind the enhanced greenhouse effect which drives anthropogenic global warming is now sufficiently robust, although details about the timing and degree of outcomes at specific sites are less certain. In recent years, temperatures in three of Australia's southern capital cities, home to nearly half the Australian population, have on occasion exceeded 46 °C, Sydney (46.2 °C), Melbourne and Perth (46.7 °C)). Given Australia's naturally warm, and increasingly hot climate, uncompensable heat exposure risk is a major health concern, adaptation is a vital survival strategy [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref]. Key groups of factors (domains) for climate change adaptation include acclimatisation to heat, human and societal behavioural adaptive responses, and technological adaption. However the degree to which each of these factors are amenable to government policy influence remains poorly understood. Adaptation to heat is therefore a major public health challenge. Heat exposure provides an illustrative example of the multi-level influences and approaches that need to be considered if providing population health protection as the world continues to warm.
To avoid a grim future, the primary task required of humanity is to take urgent useful action and substantially change its operations to mitigate further warming. This is currently progressing too slowly. The complexity and novelty of the responses required to tackle the unfolding crisis, superimposed on the high level of urgency, act as a brake on action. Secondly, human society needs to plan and implement adaptations to the approximately 0.5 °C-1.0 °C further warming already embedded in the climate system, 0.3 °C of which will occur by 2030.
One important factor hampering policy development and implementation is the resistance to change occasioned by the complex relationships between climate exposures and individual and societal harm, because they generate a policy minefield. However, with informed planning, these can instead be transformed into a fertile bed for ideas and action. What remains lacking is a planning tool that facilitates examination of the upstream, downstream and intersecting cross sectoral influences that can determine policy success or failure.
Government leadership, key to motivating and driving the necessary societal changes, is currently lacking in Anglophone countries, particularly Australia. Initially an innovative world leader, the 2013 change in Australian government resulted in diametrically opposed priorities, and a reversal of progress. With its chequered history of responding to climate change, Australia provides an interesting case for analysing climate change health protection policy developments.
In this paper we offer a new method to guide policy responses to global warming. We look at the adaptation process recommended by the Intergovernmental Panel on Climate Change (IPCC) [bib_ref] Contribution of Working Group II to the Fifth Assessment Report of the..., Ipcc [/bib_ref] , and suggest that a planning phase is essential to implementing successful adaptation. We propose that adaptation is primarily, but not entirely, a matter of behaviour change at both individual and societal levels. We provide a brief snapshot of methods important to the planning phase and support the contention that focus on the process of change is essential. We pull together elements from the described methods to arrive at a conceptual model to assist policy makers and governments with adaptation planning. Finally, as a way of introducing the method, we offer some hypothetical examples of how this conceptual framework might be applied in practice to strengthen adaptation planning. We hope this will help policy makers and decision takers to more successfully plan and therefore implement effective (that is harm moderating and beneficial to human and natural systems) adaptation.
## The adaptation process
The IPCC defines adaptation as an "adjustment in natural or human systems in response to actual or expected climatic stimuli or their effects, which moderates harm or exploits beneficial opportunities", p.72). The IPCC had previously considered that climate adaptation policy refers to actions taken by governments including legislation, regulations and incentives to mandate or facilitate changes in socio-economic systems aimed at reducing vulnerability to climate change, including climate variability and extremes. Changes can be made in "practices, processes, or structures of systems to projected or actual changes in climate", p.5).
An adaptation process has a series of phases. The IPCC Working Group Two presents a three-phase model of adaptation: Scoping, Analysis and Implementation [bib_ref] Contribution of Working Group II to the Fifth Assessment Report of the..., Ipcc [/bib_ref] (see [fig_ref] Figure 1: Extended adaptation process model [/fig_ref] SPM-3). The IPCC recommends the model as a dynamic adaptive governance process that includes constant feedback and iteration in the face of uncertainty [bib_ref] Taking uncertainty seriously: Adaptive governance and international trade, Cooney [/bib_ref]. Missing from that model, however, is an overt planning phase where the analysis is translated into a plan for implementation. We suggest that this fourth phase-Planning-is essential and logically sits between the Analysis and the Implementation phases as shown in [fig_ref] Figure 1: Extended adaptation process model [/fig_ref]. The Planning phase is a crucial step for applying considered thought to translating key findings from the analysis into implementation plans. This vital phase involves a process of: selecting from the various possible appropriate tactics and instruments, designing programs which incorporate these, then testing and comparing program alternatives, to identify and prioritize those more likely to be successful, and finally, refining these options before they advance to implementation. This Planning phase maximizes efficacy and minimizes risks of unintended maladaptive (Maladaptation occurs when action taken ostensibly to avoid or reduce vulnerability to climate change impacts adversely on, or increases the vulnerability of other systems, sectors or social groups) interventions.
## A focus on adaptation as behaviour change
A societal adaptation response requires changes to societal systems. Since societal systems are complex, changing systems is also complex, and poorly planned interventions risk ineffective or even maladaptive responses. Implementation of an adaptation strategy essentially involves changing behaviour simultaneously at both individual and societal institutional levels. To date, the research effort has largely been limited to identifying and describing the problems surrounding adaptation action and suggesting solutions for these [bib_ref] A framework for analysing regional adaptive capacity assessments: Challenges for methodology and..., Juhola [/bib_ref] (see also . Whilst not without some success, this approach has failed to bring about the necessary broadscale changes to social activity. O'Brien strongly argues that, in the realm of addressing global environmental change, the translation of knowledge into action requires a shift in emphasis from knowledge about the situation (the problem) and solutions, to a focus on the process of making change itself [bib_ref] Global environmental change III Closing the gap between knowledge and action, O'brien [/bib_ref]. Furthermore, an integrated understanding of change requires critical research into multiple dimensions including politics, culture, identity and connections. Dovers and Hezri identify the "lack of attention to the mechanisms of policy and institutional change" as a critically important gap [bib_ref] Institutions and policy processes: The means to the ends of adaptation, Dovers [/bib_ref] , p.219). They argue investigation needs to turn from what needs to happen to how it can happen-the actual mechanisms. Implementation of adaptation plans therefore require a mechanism for bringing about social change. Existing social change theories stress that societal change requires a mixture of individual and institutional level adjustments to attitudes and behaviour [bib_ref] Transitioning to low carbon communities-From behaviour change to systemic change: Lessons from..., Moloney [/bib_ref]. To achieve beneficial societal outcomes, governments use a range of policy instruments. Tactics for prompting societal change include regulation, legislation, modifying institutional arrangements, providing or removing infrastructure, supporting information and education, and legitimized by providing social proof , because of the strong tendency of humans to require social acceptance, people tend to emulate behaviours in their social group. This behaviour becomes the social norm for that group. Social proof occurs when the majority of people in that community are perceived to be behaving according to those social norms). These devices inform people about problems and possible solutions; explain, guide and encourage change; create social structures that make the new behaviours the 'easy' choices; discourage and suppress unwanted behaviours by creating both disincentives and barriers but more importantly signaling that the social norm is changing. The more people adhere to and see others adhering to the new behaviours, the more proof there is that the new behaviours are the socially acceptable ones. Furthermore, the availability of relevant technology can assist or block such change; for example, abundant cheap oil and internal combustion engines discouraged electric motor vehicle development. As in all change, different arrays of these tactics will be required to stimulate that change.
However, this mix of factors confronts policy makers with a challenge. Accordingly, climate change response planning processes continue to be troubled by a lack of coherent and credible assessments of potential trade-offs and synergies, and by the value-ridden features underpinning decision making in the field [bib_ref] Advancing methodological thinking and practice for development-compatible climate policy planning, Scrieciu [/bib_ref]. In the context of this uncertainty, and increasing pressure for action, tools are needed to facilitate the planning for implementation of policy for changing behaviour on societal scales.
The necessity for societal change being propelled by global environmental change provides an impetus and an opportunity to transform our society into one that is respectful of, and sensitive to, the requirements of people, their wellbeing and livelihoods, and also the environment. Transform, because as the IPCC stresses, our situation "… suggests transformational change may be a requirement … in a changing climate; i.e., not only for adapting to the impacts of climate change, but for altering the systems and structures, economic and social relations, and beliefs and behaviors that contribute to climate change and social vulnerability"(p.89, Box TS.8), regardless of whether it is forced/reactive or deliberate, while recognizing the processes and outcomes of such transformation are uncertain. Whether humanity avails itself of this opportunity depends on the planning choices made when selecting actions to mitigate and adapt to warming.
## Adaptation planning in australia
In 2009, the Australian Government embarked on a program to improve Australia's Climate Change adaptive capacity. A series of National Climate Change Adaption Research Networks (NCCARNs) were established to explore Australia's key climate threats, and identify requisite adaptation strategies [bib_ref] Climate change and human health: Building Australia's adaptation capacity, Hanna [/bib_ref]. Hundreds of projects were funded. Within the Health Network, projects focused on: responses to heatwaves, infectious disease, effects on Indigenous communities, severe weather, food security, and many others. The framework we present here, also funded through that NCCARN scheme, is a preparatory phase for a project designed to identify heat health threats, to project future health burdens, and to identify adaptation options. In Australia and elsewhere, heat now presents a major public health threat, as increasingly intense heat waves are breaching human thermotolerances, and generating mass death events on many continents [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref] [bib_ref] Historical global analysis of occurrences and human casualty of extreme temperature events..., Lee [/bib_ref]. This paper and the conceptual framework we introduce therefore focuses on adaptation to heat, although we submit that with modification the concept may be re-designed for broader applicability.
A series of papers published in the Asia-Pacific Journal of Public Health in 2011 analysed the State of the Science and Policy with regard to climate change health threats in Australia [bib_ref] The effects of extreme heat on human mortality and morbidity in Australia:..., Bi [/bib_ref] [bib_ref] Climate change and Australian agriculture: A review of the threats facing rural..., Hanna [/bib_ref] [bib_ref] Climate change and rising heat: Population health implications for working people in..., Hanna [/bib_ref]. The series' objective was to document climate change impacts and spur public health action on developing comprehensive climate change health and adaptation policy [bib_ref] Climate change: The greatest equity issue in public health, Binns [/bib_ref]. This investigation of Australia's health policy architecture revealed that adaptation policy had been scant, and attempts at policy development have been short term, unintegrated, and readily captive to political whim [bib_ref] Climate change and human health: Building Australia's adaptation capacity, Hanna [/bib_ref]. Much international adaptation planning literature has appropriately focused on developing nations, and how to integrate adaptation into development plans. Developed nations such as Australia present a different context because of their historical economic advantage, the moral imperatives around contraction and convergence, and because consequently more profound, wide-ranging societal and economic change will be needed for adaptation [bib_ref] Making integration of adaptation and mitigation work: Mainstreaming into sustainable development policies?..., Swart [/bib_ref]. Hence we bring adaptation thinking to industrialized nations, where development is no longer the primary challenge, but the need persists for adaptation to improve public health for all. Nonetheless, this framework applies equally well to other developed countries and developing nations.
## Systems thinking, decisions and behaviour change
## Applying systems thinking to adaptation planning
Adaptation responses occur within a societal system. To achieve transformation, successful adaptations require a systems based methodology for societal change planning, such as Collaborative Conceptual Modelling developed by Newell and Proust [bib_ref] Human health and climate change: Leverage points for adaptation in urban environments, Proust [/bib_ref]. Systems based adaptation policy interventions can be gleaned through a series of models at varying scales that depict the components and their interactions at the level of complexity required to analyse and plan interventions.
Systems theory methods enable planners to choose and map a system, in order to see how the elements within that system interrelate and interact. Adopting a systemic approach to policy development has been shown to reduce the chances that policies will produce unexpected or perverse outcomes. Therefore, introducing a formalized system mapping step within the policy planning process facilitates assessment of how altering the disposition of one element, or introducing feedbacks between elements, might further influence other elements within that system. This process provides for a better understanding of the effects introduced influences might have in modifying the system. This is explained in more detail in the examples below.
## A systems approach to behaviour change
Social behaviour is a manifestation of culture, guided by the practices and norms of that society. Social norms don't operate in isolation but propel a sequence of behavioural responses and technological change, which forces further behaviour adjustment as a feedback loop [bib_ref] System Innovation and the Laundry, Shove [/bib_ref].
We suggest that successfully orchestrating behaviour and attitude change at personal and societal levels, and consequently across the domains of behaviour and technology, necessitates planning to include different policy implementation strategies, instruments and tactics. Indeed Moloney et al. identify that a range of coordinated strategies addressing change in individual practices and change of the cultural and institutional structures are required [bib_ref] Transitioning to low carbon communities-From behaviour change to systemic change: Lessons from..., Moloney [/bib_ref]. The tactics or actions needed to achieve each of these particular changes will depend on whose specific behaviour is being targeted for change, what the change is, what barriers need to be overcome, and what inducing or guiding factors are needed. We hypothesize that behaviours in similar domains at individual or societal levels will require similar individually focused behaviour change tactics and/or changes to societal institutions to encourage and support them.
## Current planning and decision support methods
Several decision support methodologies currently exist to assist planning and policy development. Collectively these are known as multi-criteria decision analyses (MDCA). They provide a sound approach at both macro and micro levels of analysis. Within MDCA, policy oriented impact assessment [bib_ref] Policy-oriented climate impact assessment: The Tennessee valley authority and Apalachicola bay, Meo [/bib_ref] permits evaluation of trade-offs within and between policy options, and provides real-time feedback to decision takers and planners. MDCA include both information and values, and are more versatile than economic cost-benefit analyses [bib_ref] The use of multi-criteria decision-making methods in the integrated assessment of climate..., Bell [/bib_ref]. Chalabi and Kovats offer a multi-criteria decision tree to assist in evaluation and comparison of health policy options for health [bib_ref] Tools for developing adaptation policy to protect human health, Chalabi [/bib_ref]. Other methodologies, such as scenario planning, have also been met with some success [bib_ref] Scenario planning: A tool for conservation in an uncertain world, Peterson [/bib_ref]. Our Framework is designed to build on these approaches.
## The place for a conceptual framework in systemic behaviour change
Fussel describes a comprehensive adaptation framework [bib_ref] Assessing adaptation to the health risks of climate change: What guidance can..., Füssel [/bib_ref]. Within the broad process of adaptation planning described therein, we aim to complement existing vulnerability and adaptation assessment guidelines by focusing on the planning phase within the adaptation policy assessment step of the adaptation pathway.
Introduced policies are more likely to be embraced when fully contextualized within their existing policy landscape. For example, to realize their full potential, climate adaptation policies should therefore build on existing local/national policies for responding to drought, extreme weather and emergency events. To aid this we introduce a novel conceptual framework that actively encourages a holistic multi-sectoral approach to climate adaptation. A clear need exists for a framework that serves as a planning tool. We draw together knowledge from policy making, systems analysis methods and MDCA to create a framework for use as a decision focused or policy oriented assessment tool to improve adaptation planning.
Our conceptual framework operates at the macro level. It overlays a systems approach which can identify and map the important elements and their relationships within broader society to which policy interventions can be most effectively targeted. The framework enables planners to arrange their model of the social system they seek to influence, at the choice of intervention stage, to select an intervention that is more likely to work for changing the identified behaviours. This extends the systems approach.
The key feature of the framework is that it allows policy makers to ascertain which specific behaviour change tools or tactics are appropriate to the behaviours (practices and norms) that have been pinpointed as potential foci for intervention at the implementation phase. Its application allows for a priori testing of the comparative likelihood of success of various possible strategy options to achieve the intended behaviour change (see examples below).
## Climate change adaptation conceptual framework
Our Climate Change Adaptation Conceptual Framework [fig_ref] Figure 2: Climate change adaptation conceptual framework [/fig_ref] provides a scaffolding to enable planners and policy makers to visualize and map the system at this broader level.
## Ccac framework outline
Adaptation to protect human health from adverse effects of climate change involves behavioural, socio-cultural and technological aspects which operate at two different scales or levels, personal and societal, and across three different domains. [fig_ref] Figure 2: Climate change adaptation conceptual framework [/fig_ref] depicts our CCAC Framework for visualizing and understanding the contextual landscape of adaptation to increased heat, which we detail below. At the personal level, three adaptation domains operate:
(1) acclimatisation, which is the physiological response the body makes to maintain long term homeostasis in response to prolonged physical change in the surrounding environment [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref] [bib_ref] Human Physiology: From Cells to Systems, Sherwood [/bib_ref] ; (2) behavioural adaptive responses; and (3) technological adaptive responses which people adopt to protect themselves from overheating, or to assist cooling.
In response to prolonged heat exposure, physiological responses to maintain thermoregulation result in acclimatisation, which may occur subsequent to natural exposure or as a consequence of institutionalisation of a purpose designed program [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref]. Behavioural responses are cognitive attempts to reduce unpleasant exposures to heat, in order to maintain thermal comfort. When these behaviors become embedded into societal practices they become part of the culture. Technological adaptations cover a range of factors: the existence, availability/accessibility, affordability of particular technologies; infrastructure and some institutional arrangements, regulations and standards can be included. It also covers such things as societal level urban planning and building design, through to availability for personal deployment of air conditioners and fans.
Climate resilience literature stresses the need for interconnecting health and institutional systems as well as infrastructure [bib_ref] Integration in urban climate adaptation: Lessons from Rotterdam on integration between scientific..., Groot [/bib_ref]. This applies to efforts to boost health resilience to climate change. Success hinges upon the capacity to incorporate that interdependence between multiple sectors, and between shared and diverse levels of impacts, risks and responses to these in a dynamic biophysical, economic, institutional and socio-political environment. The CCACF presented here, can facilitate this function. It also makes clear that each domain/level component operates as a node within an influence network; for instance acclimatisation involves a behavioural component, (for example, exercise in the heat, which is required to acclimatize), use of technology requires behaviour (using it correctly). Technology also directs and constrains behaviour. This demonstrates the dynamically interactive relationship between personal and social behaviours, and technological responses. The CCACF also strongly illustrates that the core component to adaptation planning is changing behaviour, since human behaviour is the common facet to all aspects of responding to heat. The desired behaviours must be easy choices [bib_ref] Making healthy behaviors the easy choice for employees: A review of the..., Kahn-Marshall [/bib_ref] , else uptake will be piecemeal, and the drivers to make those choices easy often exist outside the health sector, hence the need for multi-sectoral involvement. Air conditioning (A/C) provides a good example for visualizing the link between the levels and domains within the CCACF. As technology becomes widely accessible, more individuals use it (a behavioural adaptation), which over time, changes the social norms around A/C to make it more acceptable and so reinforces use. These links operate as a feedback loop. Other behavioural adaptation factors to heat include wearing temperature-appropriate clothing, resting and seeking shelter or shade, and these influence heat exposure, thermal comfort and physiological acclimatisation [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref].
When considering the CCACF in the context of responses to health threats of increasing exposure to heat, only the behavioural and technological domains can exist at the societal level. The individual physiological response, acclimatisation, cannot operate at the societal level. Excluding the few exceptions where policies might aim to specifically influence individual acclimatization, within this Framework, acclimatisation and adaptation are considered separately with adaptation defined as the behavioural and technological aspects. Together acclimatisation and adaptation could be termed accommodation to global warming.
Within each of the five main components (domains/levels) of the Framework are the various specific factors or elements that comprise the sub-systems which government may seek to influence. Influence can be achieved directly through legislation and regulation, or indirectly via information and social marketing campaigns. These government interventions are inputs into the system modelled within the Framework and the consequent changes to societal adaptive capacity are outcomes (neither are shown in [fig_ref] Figure 2: Climate change adaptation conceptual framework [/fig_ref] ; see [fig_ref] Figure 4: Social norms-house design-air conditioning use feedback loops [/fig_ref].
Adopting our conceptual framework encourages planners to group elements into natural clusters within each of the domains/levels, based on the judgement that each cluster of elements is likely to respond to similar behavior change techniques. Through this organisational clarification, those strategies and tactics more likely to be effective interventions can be identified and targeted. Inevitably, strategies will emerge that fall outside health sector responsibility. Detailed examination of the elements within the system being modified across the domains/levels, will thus highlight possibilities for, and serve to build, cross sectoral collaboration. Sharing the model, once populated with elements, with intersectoral colleagues can serve to illustrate the inter-connectedness and shared goals, and thus help optimize buy-in. This practice will also reduce activity at crossed purpose, and hence the risk of counterproductive policy development. Working synergistically can amplify effectiveness.
Elaboration of these elements and a more detailed examination of their interactions allows for construction of progressively increasing sophistication of the model, enabling points of influence to be more accurately identified. Planners may then pre-test the multiple potential outcomes of a proposed input factor as it interacts with the network of elements within each cell, and across the three domains and two levels. The multi-criteria aspect enforces consideration of important upstream influences and potential downstream effects that could ultimately either negate or potentiate outcomes by delivering multiple benefits. We give an example of this below. The drill-depth provided by applying the Framework as a planning exercise permits planners to identify specific foci for targeting interventions carrying the highest likelihood of success.
## Framework overlay to a planning model
This section first outlines and then steps through operationalizing the CCACF. We expand from the macro level to introduce additional elements within each domain's/level's component and elaborate the sub-systems within these. We then describe how each might fit together in a real world example of policy options available to government for influencing adaptive social change around the use of A/C. Influence is bi-directional. Individuals can alter the behaviour of many [bib_ref] Identifying opinion leaders to promote behavior change, Valente [/bib_ref] , and, in a cycle of reinforcing feedbacks, individual attitudes and behaviours can be powerfully shaped by societal norms [bib_ref] The constructive, destructive, and reconstructive power of social norms, Schultz [/bib_ref]. The personal and societal levels are therefore interconnected in that change at one level may dynamically either reinforce and strengthen, or counterbalance and curtail, behaviors in the other. Similarly, using the CCACF alerts planners to actively check for responses in each of the three domains that modulate activity in the others. This second order model [fig_ref] Figure 3: Climate change adaptation conceptual framework expanded model, provides an example of application... [/fig_ref] reveals the sub-systems within each domains/levels and maps some of the connections between them. A complete Collaborative Conceptual Modelling (CCM) exercise would go into much greater level of detail [bib_ref] Human health and climate change: Leverage points for adaptation in urban environments, Proust [/bib_ref]. In this illustration we want to demonstrate how using the conceptual framework as a planning overlay adds value to the process, by illuminating the specific domain and level of the elements being targeted for influence.
Choice of elements in a model is dictated by the purpose of each particular exercise and the desired outcome. It is prudent to carefully select which elements are to be included in a model. For instance personal acclimatisation operates through thermal comfort which is already captured by other elements and, in this instance, only adds unneeded detail to the model [fig_ref] Figure 3: Climate change adaptation conceptual framework expanded model, provides an example of application... [/fig_ref]. In terms of the strength of the reaction or effect, adaptation responses can feedback to reinforce or increase the action, or to provide a counterbalance or stabilize the reaction [bib_ref] Human health and climate change: Leverage points for adaptation in urban environments, Proust [/bib_ref]. In turn these changes can promote adaptation or be maladaptive. Two feedbacks have been highlighted in this example: (1) a reinforcing feedback from use of A/C to social norms on further A/C use, and (2) a reinforcing feedback of A/C use that may hinder physiological acclimatisation, undermine capacity to cope with the heat and achieve thermal comfort, so further increasing A/C use. Within the model, the use of A/C is of itself neither adaptive nor maladaptive. The CCACF only models the system to help planners identify points potentially receptive to influence. The assessment as to whether the output of the system is adaptive or maladaptive rests with the planners applying the model. In a way, this exemplifies the real world situation in that these judgements are societally bound and the frame is often political. For example, the model makers may decide that increased use of air conditioning can be considered a maladaptive adaptation, as it runs counter to attempts to address global warming by reducing energy use. Such an argument may not apply if energy involved in the manufacture, transport and lifetime usage is derived, substantially or wholly, from renewable (non-fossil) sources. The reinforcing feedback between housing design and social norms for A/C shown in [fig_ref] Figure 4: Social norms-house design-air conditioning use feedback loops [/fig_ref] , again demonstrates the importance of having a systemic and multi-sectoral approach toward, and populating the model with, the key elements relevant to the issue under consideration and the outcome desired. Social norms that unquestioningly accept A/C use as a fundamental requirement for a house to be habitable, reinforce acceptability for house design that necessitates A/C. Consequently policy makers, planners and users can become blind to alternative potential methods for attaining thermal comfort. Upstream and downstream considerations can be revealed via working through the Framework. A social norm that accepts A/C is necessary for thermal comfort and safety, also drives demand for its instillation in public housing; as such, norms extend across all socio-economic groups. Urgency to adopt architectural designs that incorporate passive cooling in public and private infrastructure diminish amid the mindset that A/C is the preferred response to hot buildings. Hence the reinforcing feedback drives installation and use of A/C, which further normalizes A/C as the preferred method for attaining comfort in heat. The power of the confluence of the availability of A/C technology and the social pressures to use it is demonstrated by its widespread adoption in the USA, and the reported 80% decline in heat deaths since 1960, attributed to its use [bib_ref] Adapting to Climate Change: The Remarkable Decline in the U.S. Temperature-Mortality Relationship..., Barreca [/bib_ref] , although this is debated [bib_ref] Heat-related mortality and adaptation to heat in the United States, Bobb [/bib_ref].
On the other hand, to attain an adaptive outcome, jurisdictional departments responsible for building codes might use passive cooling via house design as a starting point when examining and re-engineering the feedback loop. Different inputs to the feedback could drive reduction of A/C use. Placing this feedback loop into the Framework facilitates easier identification of appropriate policy instruments (three examples shown in [fig_ref] Figure 4: Social norms-house design-air conditioning use feedback loops [/fig_ref] and where they might be applied.
Adaptation is ultimately about behaviour change within the societal and individual levels. That change can then feedback to determine responses within the acclimatisation and the technological domains. Influence into these other domains can also feedback onto behaviour. Identification of the level and domain to apply the policy influence, helps clarify the likelihood of success of potential tactics and policy instruments, and thus allows for better informed planning decisions.
For instance, continuing the example above, policy makers may decide to focus on affecting social norms for energy use and urban and housing design. Planners may wish to evaluate the feasibility, compare the relative merits, or identify the required inputs of three alternative options involving social norms: (a) social norms to influence individual's decisions (such as using air-conditioning), (b) social norms to drive preferences and therefore demand for ecologically sustainable housing design, or (c) social norms designed to influence community adoption of adaptive technology, such as decisions to install and use A/C. In the social behaviour domain, the instruments and tactics to be used such as social marketing campaigns, will most likely differ when targeting social norms to influence individuals' decisions, (such as using A/C), social norms to drive preferences and therefore demand for ecologically sustainable housing design and social norms designed to influence community adoption of adaptive technology, such as decisions to install and use A/C. Of course policy planners would probably use several instruments to guide change. This example [fig_ref] Figure 4: Social norms-house design-air conditioning use feedback loops [/fig_ref] illustrates how these options can be considered within the Framework. An element in the social behaviour cell (social norms) can be identified and assessed, across domains and across levels. This structural approach reveals the differing applicability of instruments and tactics, such as social marketing campaigns or regulations, for the task. The figure also demonstrates how the Framework elucidates potential for synergies, when multi-pronged campaigns operate simultaneously across separate cells. Similarly, conflicting strategies and pressures that can operate counter-directionally and undermine progress, or desired community responses, may be identified.
Additionally, a government may decide to intervene at the social level within the technological domain, by regulating building codes to guide housing design in a way that reduces the need to buy and install A/C. Alternatively, architect or building professional bodies may review their standards and revise professional training curricula in order to reinforce ecologically appropriate house design to undergraduates. In turn, such alterations in house designs feed into decisions to install A/C, which can then impact social norms of A/C use as described above. Thus the Framework assists in the design of synchronized, multi-sectoral, whole-of-government approaches.
The crux is not A/C use per se; rather, that it is possible for decision takers to visualize the policy option landscape (in our example, around A/C use) from a 'helicopter' vantage point, to deduce what particular behaviour change strategy or tactic combinations carry the greatest potential to influence the behaviour(s) in question. In the A/C example, with the primary policy aim of reducing adverse health effects of heat exposure whilst not increasing greenhouse gas emissions, we have chosen urban and house design as the entry point to reduce the need to A/C. This option represents a government level regulatory intervention focused on changing the behaviour and attitudes of architects and builders. The social marketing campaign, aimed at the attitudes of house buyers, seeks to reinforce the effects of design standards, and to change personal protective behaviours (changing into more appropriate clothing). Not shown in our example, other interventions might be to ensure building inspectors monitor the construction of new houses to ensure compliance with the correct standard. This opens into a different system, encompassing education of inspectors and quality performance programs, price incentives on energy, accompanied by marketing campaigns aimed at householder behaviour to discourage energy use. The Framework is thus versatile in its utility and applicability, and the process of investigating interventions within their correct domain/level provides for better informed choices of tactic to elicit uptake of climate change adaptations. Quantification of the model flows and exploration of how different mixes of policy options change behaviour, can then be undertaken at the local level. Further, this model might sit within and integrate with a larger model, such as for example, planning a state-wide heat wave emergency response. Both the framework and the model scale up and down as required (see example at .
Inducing behaviour change is difficult for numerous reasons. For people to alter their behavior, they first need to understand that there is a problem, that there are solutions, that those solutions are viable, and that the new order is not too unfamiliar. It is also important that doubt and uncertainty have not been sown, and they understand that a behavioural shift does not undermine their own identity or create too great a conflict between what needs to be done and their sense of their own capacity.
When the behaviours to be changed are also community wide, then the added challenges of confronting society's worldviews, altering societal practices and institutions, and overcoming the barriers of the resilience of societal systems and the intransigence of vested interests become apparent. Clearly social psychologists and marketers have a role in planning and implementation phases. A major advantage of using this conceptual framework is its potential to help planners recognize, and subsequently reduce or remove barriers to adaptive actionby embedding the step of deliberately considering at the planning phase what behavioural strategies and tactics are to be incorporated, thereby improving the likelihood of success. Finally we hope that use of this conceptual framework will help define questions for future research and approaches to assist the critical task of adaptation planning.
As a further example of how the CCACF can be applied across scales (from local to national), we discuss the situation of preventing heat morbidity and mortality in older people. Older people, particularly women, are found to be at greatest risk from heat waves [bib_ref] Man in Extreme Environments, Problems of the Newborn and the Elderly, Robertshaw [/bib_ref].
In a local public health service area, the desired endpoint might be to reduce the number of older people succumbing to heat, being admitted to hospital or dying from heat during a heat wave event. In , the end point of the influence diagram, Heat Effects, would be the numbers so affected. The factors leading toward this end point are given, set out in appropriate domain/levels.
The level of interest for this model might be state and national governments, the regional health authority, local governments or local health authorities, or primary health care services, or combinations of the above. Approaches to the elements within the model will vary according to capacities and responsibilities of agencies involved. For instance a national government may set broad policy directions and restrict their focus to upstream issues such as: house design standards, compliance regulation and budget; as policy enablers to ensure all relevant institutions devise and enact heat response plans; and devising policy and programs to ensure factors such as poverty do not prohibit elderly people's ability to adopt personal protective behaviours.
## Figure 5.
Heat response across scales. The figure models elements that contribute to planning to prevent adverse heat effects on elderly people. Different levels of government would act at the level where their influence would be most appropriate.
A regional/state government may direct their energies to detailed examination of methods for creating early warning and response plans, provide training programs and resources for aged care personnel, removing socioeconomic barriers to personal action, and enforcing building standards.
Operating at ground level, local governments and health services connect with their communities to elaborate and deliver heat plans. It would be at this level that strategies to improve personal acclimatization might be addressed, in concert with primary health care services as part of a broader aged care assessment and functional support system.
At each level, potential interventions can be assessed, consequences mapped out, other tools such as running desk top simulations of adaptation options (see and the effects quantified. Best fit options can then be taken forward for piloting.
## Adaptation using the framework
## Conceptual framework assists
Climate change adaptation is a wicked problem, A wicked problem is one exhibiting complexity, which does not have a 'solution' as such, whose resolution is encapsulated in its problem definition, which requires a systemic approach to definition and resolution, and, because of this property, may trigger other 'problems' in other parts of the system requiring further resolution. They tend to be social and political in nature) made difficult by conflicting interests, and a maze of interrelating components. While our Climate Change Adaptation Conceptual Framework does not deal with many of these factors, it can at least help planners as a decision support tool to chart the interrelated elements of the situation, to facilitate identification of what behaviour change tools might best suit key elements amenable for manipulation in their societal models.
Adaptation needs to happen at all levels, personal and societal. Leadership can come from influential champions, governments or commercial interests. However given the enormity of the problem, we contend that governmental leadership is crucial to accelerate the process of adaptation at all levels. Governments can act rapidly across a range of policy areas, to regulate corporations and engage the public [bib_ref] Personal Responsibility and Changing Behaviour: The State of Knowledge and its Implications..., Halpern [/bib_ref] , and to orchestrate the urgently needed ongoing transformational shifts. Our framework seeks to help governments to do this.
## Can human society successfully adapt to the increasing heat?
As outlined in the introduction, hazardous global temperature rises are expected this century. The long-term habitability of much of the planet is at risk [bib_ref] Limitations to thermoregulation and acclimatisation challenges human adaptation to global warming, Hanna [/bib_ref] [bib_ref] An Adaptability Limit to Climate Change Due to Heat Stress, Sherwood [/bib_ref] [bib_ref] Four degrees and beyond: The potential for a global temperature increase of..., New [/bib_ref] [bib_ref] Planet under Pressure: State of the Planet Declaration, Stafford [/bib_ref].
As humanity's situation in a degrading environment deteriorates in future, the capacity of governments to make adequate and feasible policy responses will become seriously limited. Major mitigation effort is necessary now. Meanwhile society will have to adapt to the change that is currently inevitable (possibly 2.6-4.8 °C by 2040 [bib_ref] Greenhouse-gas emission targets for limiting global warming to 2 °C, Meinshausen [/bib_ref]. Use of frameworks such as the one we propose provides an important tool to guide appropriate and effective government adaptation responses to protect human society in the short to medium term, against a range of future climate threats, cross a spectrum of stages of economic development. Action that both mitigates and adapts to global warming delivers much needed additional co-benefits for health and society [bib_ref] Planet under Pressure: State of the Planet Declaration, Stafford [/bib_ref] [bib_ref] Public health benefits of strategies to reduce greenhouse-gas emissions: Food and agriculture, Friel [/bib_ref] [bib_ref] Public health benefits of strategies to reduce greenhouse-gas emissions: Overview and implications..., Haines [/bib_ref].
Importantly, society urgently needs to begin the conversation about the relationship between greenhouse gas emissions, population levels, and resource limits. In responding to these wicked problems, the imperative for all governments to now adopt an expansive, holistic approach to policy is becoming increasingly clear. Use of the CCACF described here facilitates the deep thinking required. This will play a part along with other planning and decision support tools and procedures to help formulate the behaviour modification tactics entailed in implementing the adaptation plans necessary to permit healthy, prosperous, eco-and socially sustainable human habitation of our planet.
To engineer the necessary transformative changes, governments will need to reassume their historical roles of safeguarding society and even to broach new policy realms. Governments must take the lead to drive behavioural change; behaviours regarding food choices, energy use, transport, work practices, leisure activities, and the seasonal or daily scheduling of these activities. Many of these are currently under powerful corporate influence, which acts to block pro-health behavioural choices. Therefore, counter to the recommendations of the Australian Productivity Commission, which confined government's role in adaptation to 'removing regulatory impediments', we argue that there is a strong case for governments instead to proactively regulate corporate behaviours relating to resource use, production and marketing. Planning for such societal change may also benefit from this framework approach to guide behaviour modifications that can deliver changed patterns in consumption of goods and services and hasten the shift towards a more ecologically sustainable economy.
One essential component to adaptation planning and policy implementation, to deliver both societal benefits and cost effectiveness, is to systematically look for, plan and incorporate synergies that result in co-benefits into all programs and strategies. Instilling a phase that enforces policy planners to undertake holistic and broad analysis of the full policy and response landscape that encompasses multi sectoral involvement that is backed up by specific analysis of exactly how the requisite behaviour changes might occur, is vital. Thus, applying our Framework to Climate Adaptation Planning can help maximize optimal outcomes, and avoid unintended shocks of maladaptation.
# Discussion/limitations
This framework intends to fill a niche position in the adaptation domain, within the planning phase. It is designed specifically with adaptation to heat in mind, but we are confident that the concept might be adaptable to other topics and domains. We have intentionally taken a behaviouralist approach based on three lines of reasoning. First, is the understanding that humans act both individually and collectively according to personal belief patterns amidst the prevailing diversity of cultural norms and practices. Norms shape all actions within the social, economic and political systems, which give rise to the organizations and structures that form the institutions which underpin societal functioning. Secondly, we upheld the assumption that social change involves changing the behaviour of people. Ripple effects emerge as this flows through into changing the practices, norms and institutions. We see no contradiction between a behavioural and a structural approach to adaptation planning. Indeed, they are complementary, and proceed hand-in-hand, feeding back on each other.
Third, based on these two previous assumptions, we contend that featuring a specific planning step that aligns behaviour change instruments to the site for policy intervention will positively enhance success. This complements the emerging, but not yet universal use of systems theory approaches to planning.
The Framework has limitations. Omission of key elements will constrain efficacy of final policy development. Application of the Framework requires a phase of background investigation in order to appropriately populate the cells with relevant elements. However, this also serves as its strength, in that it encourages the deep analysis, and cross portfolio discussions necessary for sound policy development.
A second limitation lies in its inherent complexity, such that urgent timelines may limit the incidence of uptake. However, this limitation could also be considered another strength, as detailed cross sectoral exploration and hind-casting yields benefits in imaging the desired future, and mapping the pathway to reach the desired endpoint. Operationalizing the Framework could be progressed relatively swiftly through a biphasic multidisciplinary workshop approach, with initial scoping in the first instance, to ensure all requisite disciplinary groups and elements are included for the second, detailed planning workshop. In order to optimise progress, attendees would need relevant expertise and be fully cognisant of the problem at hand, and bring sufficient level of authority to advance discussions. Whilst these requirements present perennial challenges in the policy sphere, high level commitment to the process can both streamline it and deliver cohesive effective policies.
# Conclusions
Within the broad scope of adaptation discussion, we seek to open up dialogue of proactive adaptation policy for industrialized and developing nations, in preparation for meeting the imminent climate challenges. The foremost aspect of adaptation is behavioural change, at both societal and individual levels. This paper describes a framework for conceptualizing adaptation within a social systems behaviour change perspective, and a mechanism by which governments can better plan and gauge possible effects of their policy decisions, particularly regarding the behaviour strategies employed. Societal change involves influencing technological development, and individual and societal behaviours to bring about transformative adaptive change. By dividing the landscape of implementation into these domains/levels, behavioural strategies with a higher likelihood of success in achieving desired behaviour change in each of these domain categories can be designed and prioritized by policy makers. Gleaning and prioritizing tactics will also help in testing or piloting interventions.
Applied to the example of adaptation to heat, our Framework describes three domains of adaptation: acclimatisation, behavioural changes and technological development across individual and societal levels. We have given examples of how overlaying this conceptual framework onto other social change planning methods provides a policy development tool that that furnishes policy makers with capability to identify elements amenable to manipulation, and to pre-analyse success likelihood of options for policy instruments to drive societal change by thinking more formally about the tactics to use.
We invite policy makers and other researchers to apply and modify our Framework and develop a common and systematic approach to adaptation planning. At the current rate of greenhouse gas emissions, human society is following a pathway towards a catastrophic future. We can avoid this future with urgent action on mitigation. Even with rapid mitigation, the continuing warming, climate interruptions and sea level rise that are embedded in the Earth's natural system necessitate immediate, comprehensive activity be undertaken to maximize the chance of success for human adaptation, and ultimately, protection of health and prosperity. Time is of the essence.
The Climate Change Adaptation Conceptual Framework presented here is simple, reliable, flexible, utilitarian and convenient for the needs of climate adaptation policy planning. We present this CCAC Framework in the interests in advancing effective adaptation.
[fig] Figure 1: Extended adaptation process model. This extended model includes a phase for planning between analysis and implementation. [/fig]
[fig] Figure 2: Climate change adaptation conceptual framework. This Climate Change Adaptation Conceptual Framework comprises three domains, (acclimatisation, adaptation behaviours and technology) which operate at societal or personal levels. Subsystem elements (seeFigure 3) within each of the five cells, interact with elements in other cells, across the three domains and two levels, under the influence of external factors, such as a planned policy intervention, to produce outcomes. [/fig]
[fig] Figure 3: Climate change adaptation conceptual framework expanded model, provides an example of application of this framework, showing subsystem elements within each level and domain, using the example of A/C. Arrows depict influence flows that lead to use of A/C. Numbers represent the feedback loops discussed below. [/fig]
[fig] Figure 4: Social norms-house design-air conditioning use feedback loops. Further illustration of the use of the CCAC framework, here populated only by relevant elements of a single feedback loop relating to house design and A/C, and introducing examples of strategies that could be applied (external influences). [/fig]
|
Chlamydia trachomatis infection in a sample of northern Brazilian pregnant women: prevalence and prenatal importance
Pregnant women Neonatal infection Respiratory symptoms Polymerase chain reaction Brazil a b s t r a c t There are limited data regarding prevalence of Chlamydia trachomatis infection among northern Brazilian pregnant women.Objective: The purpose of this study was to estimate the prevalence of chlamydial infection among pregnant women in their third trimester and to determine the repercussion of this infection on their offspring.Methods:In the first phase of this study 100 pregnant women receiving prenatal care in a local public university hospital were examined to assess the prevalence of genital C. trachomatis infection by polymerase chain reaction technique. In the second phase, 88 pregnant women were prospectively evaluated for premature rupture of membranes, puerperal consequences associated with chlamydial infection, and neonates were checked for low-birth weight.Results: The prevalence rate of chlamydial infection was 11%, and 72.7% of the positive participants were predominantly less than 30 years of age (p = 0.1319). A total of 36.4% of the participants had premature rupture of membranes (p = 0.9998). Neither low-birth weight infants nor preterm delivery were observed. A cohort of 16 newborn babies were followedup up to 60 days of life to ascertain outcome: 50% had respiratory symptoms. Neonates born to infected mothers had a higher risk to develop respiratory symptoms in the first 60 days of life.Conclusion:The scarcity of data about the effects of chlamydial infection on pregnancy and neonatal outcomes justified this study. Diagnosing and treating chlamydial infection during the third trimester of pregnancy may prevent neonate infection. Therefore, preventive screening should be seen as a priority for early detection of asymptomatic C. trachomatis infection as part of local public health strategies.
# Introduction
Chlamydia trachomatis infection is nowadays considered the most frequent sexually transmitted disease (STD) in the developed world. The World Health Organization (WHO) estimate that annually nearly 100 million new cases occur around the world. 1 Studies show that chlamydial infections is the commonest transmitted bacterial disease in European countries and in the United States with more than 2.8 million new cases estimated to occur each year. [bib_ref] Recent trends in the epidemiology of sexually transmitted infections in the European..., Fenton [/bib_ref] During 2009, more than 1.2 million cases of chlamydial infection were reported to Centers for Diseases Control (CDC). Prevalence of chlamydial infection was 6.8% among sexually active females aged 14-19 years.According to the CDC, C. trachomatis infection among pregnant women in the United States is the third main cause of STD, with bacterial vaginosis and herpes simplex virus 2 being first and second, respectively.In Brazil, according to STDs epidemiologic studies in specific groups of women that attended gynecology, family planning, and prenatal clinics showed a prevalence of 2.1-27.1% for genital infection by C. trachomatis.In the Amazon region, there is little data about the actual prevalence of chlamydial infections. Only AIDS, HIV in pregnant women and children, syphilis in pregnant women, and congenital syphilis have compulsory notification.In Manaus, the capital city of the state of Amazonas (AM), Brazil, Alencar et al. established a 27.1% prevalence of chlamydial infection by direct immunofluorescence technique. [bib_ref] Detecção de Chlamydia trachomatis por imunofluorescência direta em esfregaços endocervicais, Alencar [/bib_ref] On the other hand, Santos et al. and de Lima Freitas found 20.6% and 52.8% of sexually active women and infertile women, respectively, to be positive for Chlamydia using PCR technique. [bib_ref] Detection of Chlamydia trachomatis in endocervical smears of sexually active women in..., Santos [/bib_ref] [bib_ref] High prevalence detection of Chlamydia trachomatis by polymerase chain reaction in endocervical..., De Lima Freitas [/bib_ref] Many studies have provided estimates of chlamydial infection in pregnant women throughout the world. C. trachomatis disease, considered to be a treatable condition, continues to infect massive populations and the exact magnitude of this is still unknown. 9,10 Prevalence rates among pregnant women vary enormously around the world. 9,11 Jalil et al. conducted a study to estimate the prevalence of C. trachomatis infection in pregnant women from six Brazilian cities and found 9.4% using hybrid capture technique. [bib_ref] Prevalence of Chlamydia and Neisseria gonorrhoeae infections in pregnant women in six..., Jalil [/bib_ref] On the other hand, in Fortaleza, the capital city of Ceará, Brazil, Martins et al. found a prevalence of 11% in pregnant women by PCR technique.Uncomplicated C. trachomatis cervicitis in pregnancy may result in sporadic and recurrent miscarriage, preterm labor, premature rupture of membranes, low birth weight and postpartum endometritis. [bib_ref] Influence of infection with Chlamydia trachomatis on pregnancy outcome: infant health and..., Mardh [/bib_ref] Maternal C. trachomatis infections have been associated with increased morbidity in newborns and infants up to three months of age. Approximately two thirds of neonates born vaginally to infected mothers will become infected at birth. [bib_ref] Chlamydia trachomatis infections, Weinstock [/bib_ref] These infections can result in conjunctivitis, otitis media, pharyngitis and pneumonia in the newborn. Moreover, neonatal infection with C. trachomatis can cause long-term sequelae such as chronic obstructive pulmonary disease. [bib_ref] Influence of infection with Chlamydia trachomatis on pregnancy outcome: infant health and..., Mardh [/bib_ref] The primary aim of this study was to assess the prevalence of C. trachomatis in the final trimester among pregnant women who attended the low-risk prenatal clinic of a public health university hospital in Manaus, Amazonas, Brazil.
In Manaus, there are no studies related to vertical transmission of C. trachomatis, so the secondary aim of this study was to ascertain possible neonatal implications associated to maternal C. trachomatis infection. Moreover, the purpose of this study was to evaluate C. trachomatis infection rate associated with socio-economic status of the participants enrolled in the study, adverse pregnancy outcomes associated with miscarriage, preterm labor, premature rupture of membranes and low birth weight.
# Materials and methods
Endocervical samples were collected from 100 pregnant women receiving care at the low-risk prenatal clinic of Dona Francisca Mendes University Hospital in Manaus, Amazonas, Brazil, between January and June 2005. The first phase of this study was a cross-sectional evaluation of the prevalence of Chlamydia infection in the final trimester of pregnancy (as of 29 weeks of gestation). In the second phase of this study, pregnant women were prospectively followed to evaluate postpartum effects of this infection as well as upon their neonates up to sixty (60) days after birth.
This study was approved by the Human Research Ethics Committee at the Federal University of Amazonas and performed in accordance with the Declaration of Helsinki.
## Assessment of patients
Pregnant women included in this study answered a standardized questionnaire to collect information on age, marital status and medical history, number of lifetime sexual partners and socio-economic status. Family income and education level were the criteria used to determine socio-economic status. Patients were excluded if the gestational age was less than 28 weeks and six days, had infections requiring antibiotic therapy or a history of antibiotic use in the preceding 30 days. In addition, pregnant women who had sexual intercourse in the past 72 h, history of vaginal bleeding or who used vaginal cream, suspected premature rupture of the membranes, and a diagnosis of placenta previa were also excluded.
In the first phase of the study all participants were contacted by telephone so that the researchers could collect information about their maternity admission and start the second phase of the study. During the second phase of the study 88 pregnant women who met the eligibility criteria had their delivery checked and were followed-up for 60 days after giving birth. Of a total of 100 participants in the first phase of the study 12 were excluded from the study: four could not be contacted and eight due to insufficient follow-up time (less than 40 days of postpartum).
A cohort study was carried out in 88 newborns who were evaluated in the first 24 h and monitored thereafter every 15 days up to 60 days of life. However, 12 neonates were excluded due to loss of contact (four) and to follow-up time less than 60 days of life (eight). Moreover, one case of stillbirth occurred. Subsequent evaluations were conducted by telephone and if some ocular discharge or respiratory symptoms appeared in the neonate a clinical evaluation was requested. Furthermore, patient's complaints, clinical evaluation and physical exams were registered in the questionnaire and entered in Epi-Info data bank.
## Collection of the samples
Endocervical material for smears was collected using a cytobrush and subsequently re-suspended in 400 L of TE (10 mM Tris-HCl pH 8.0 and 1 mM EDTA) and stored at −20 - C until DNA extraction.
## Preparation of the samples for pcr
Each sample was supplemented with 400 L TPK solution [TE, 20% Tween, 10 mg/mL proteinase K], followed by incubation at 56 - C for 1 h and then boiling for 10 min. DNA from samples that contained blood was extracted by phenol/chloroform.
## Pcr amplification of c. trachomatis dna plasmid
The
## Dna sequencing
To confirm the presence of 241 bp PCR products, we sequenced positive samples of C. trachomatis with the MegaBACE 1000 DNA Analysis System. The amplified products were digested with EXO and SAP enzymes, followed by incubation at 37 - C and 80 - C for 1 h and 30 min, respectively. The purified products were sequenced using DYEnamic ET DYE Terminator Cycle Sequencing kit for MegaBace from GE Healthcare Lifesciences according to manufacturer's protocol. The sequencing reactions were separated and analyzed on the automatic sequencer.
# Statistical analysis
Statistical analyses were performed to calculate the absolute and relative frequency distributions for qualitative variables; as well as median and standard deviation values for quantitative variables. The data were analyzed to evaluate the association between the categorical variables and PCR results using the Pearson 2 test with the Yates correction or Fisher's exact test. To analyze the risk factors of the neonates relative risks (RR) were calculated. Student's t-test was used to compare means of normally distributed data. To estimate the prevalence of C. trachomatis in pregnant women a 95% confidence interval [CI] was calculated. Throughout this study, p < 0.05 was considered significant. Statistical analysis was
# Results
## Phase i -study
Out of 100 patients 11 (11.0%) were C. trachomatis-positive by PCR. The mean age of PCR-positive patients was 23.5 years with a standard deviation of ±4.5 years. Although we assessed associations between C. trachomatis infection with family income (45.5%) and their educational level (45.5%), no statistically significant association could be found.
Endocervical secretion and cervicitis were observed in 63.6% and 36.4%, respectively, among those women positive for C. trachomatis infection. Both endocervical secretion and cervicitis were significantly associated with C. trachomatis infection (p = 0.0411 and 0.0193, respectively) as shown in [fig_ref] Table 1 -: Speculum examination according to PCR results for C [/fig_ref].
## Phase ii study -maternal and perinatal data
There was no significant association between positivity for C. trachomatis infection and premature rupture of membranes [fig_ref] Table 2 -: Premature rupture of membranes in pregnant women according to PCR results for... [/fig_ref]. Neither preterm labor (less than 37 weeks gestational age) and nor low-birth weight (weight less than 2500 g/5.5 pounds) was observed among the PCR-positive women.
Cesarean delivery was indicated for PCR-positive primiparous pregnant woman. Chronic tubal sequelae were assessed by the surgery description. One week after birth a patient was readmitted to the hospital with diagnosis of postnatal endometritis. Another C. trachomatis positive participant gave birth to a stillborn fetus with 34 weeks gestation. After this episode, this patient developed postnatal endometritis.
## Cohort study of neonatal outcome
Of the 87 newborns, 16 (18.4%) presented respiratory symptoms. Of these newborns, eight (50%) were born to C. trachomatis-positive mothers evidencing a RR of 7.7 [fig_ref] Table 3 -: Incidence of respiratory symptoms in a cohort of 87 neonates exposed according... [/fig_ref]. Moreover, neonates born to positive mothers were more likely to develop respiratory symptoms such as nasal obstruction (RR = 6.7), coryza (RR = 7.7), cough (RR = 27.0) and dyspnea (RR = 15.4) (CI 95%) [fig_ref] Table 4 -: Incidence of nasal obstruction, coryza, cough and dyspnea in a cohort of... [/fig_ref]. There was no statistically significant (p = 0.8312) association between C. trachomatis infection and type of delivery (cesarean section or vaginal) [fig_ref] Table 5 -: Mode of delivery and respiratory symptoms of neonates in Manaus, Amazonas, Brazil [/fig_ref].
# Discussion
Studies show that C. trachomatis infection during pregnancy may lead to serious complications. Epidemiological data suggest that pregnant women may be a specific target group for C. trachomatis screening. Antenatal screening, as recommended by the CDC,would be beneficial to decrease morbidity amongst women themselves, but also to prevent vertical (infant) and horizontal (partner) transmission. Although studies have demonstrated prevalence of C. trachomatis in pregnant women using hybrid capture technique in Brazilian cities, [bib_ref] Prevalence of Chlamydia and Neisseria gonorrhoeae infections in pregnant women in six..., Jalil [/bib_ref] we decided to undertake this study using the conventional PCR technique. However, although hybrid capture technique is widely used for the detection of pathogens, it may lack sensitivity in comparison with PCR techniques. [bib_ref] Comparison of hybrid capture and PCR for HPV detection in clinical samples, Rodrigues [/bib_ref] PCR was the first technique to be developed and because of its ability to amplify specific targets and due to superior sensitivity it remains the most widely used molecular technique. [bib_ref] New diagnostic tools, Sanguinetti [/bib_ref] The prevalence of C. trachomatis infection in this study was 11.0%. Although Jalil et al. [bib_ref] Prevalence of Chlamydia and Neisseria gonorrhoeae infections in pregnant women in six..., Jalil [/bib_ref] found a prevalence of 9.4% among Brazilian pregnant women, the rate identified in this study is in agreement with the findings of Martins et al.(11.0%). However, as mentioned above there were no data on C. trachomatis infection among pregnant women in the city of Manaus before this study. The main goal of this study was to better understand the epidemiology of Chlamydia infection in this Brazilian city. As compared to other C. trachomatis infection prevalence studies conducted in Brazil, it is possible that the prevalence rate identified in this study is associated with the population characteristic. Although we found that the majority (45.5%) of the participants with PCR-positive results had a low family income, in addition to low educational level (45.5%), and had no knowledge of the existence of C. trachomatis and its effects on women's reproductive health. A total of 81.8% of PCR-positive C. trachomatis women in this study were married or were living a cohabitating situation. Only one lifetime sexual partner was reported by 90.9% of the women; 72.7% of the PCR-positive patients were under 25 years old. The risk of presenting chlamydial infection is two times higher in under 25 years old women. [bib_ref] Prevalence and associated factors for Chlamydia trachomatis infection among undocumented immigrants in..., Jackson [/bib_ref] In most of the studies the majority of positive participants were under 25 years old, which is could possibly be explained by the unhealthy lifestyle behavior of this age group, such as unprotected sex. [bib_ref] Chlamydia trachomatis in subfertile women undergoing uterine instrumentation. Screen or treat?, Land [/bib_ref] These factors may be due to high number of unrecognized infected men which provides a reservoir for spreading the infection to women via sexual contact. [bib_ref] Assessment of Chlamydia trachomatis infection in asymptomatic male partners of infertile couples, Hamdad-Doudi [/bib_ref] These findings indicate the need for disseminating information about the risk factors for Chlamydia infection. The following factors may be responsible for the inadequate incorporation of preventive measures by women: low educational level of the population; and lack of screening for Chlamydia infection in the public health system. [bib_ref] Interventions for increasing Chlamydia screening in primary care: a review, Ginige [/bib_ref] Also, it is possible that the prevalence rate identified in this study is associated with low socio-economic status and low family income of the participants enrolled in this study. Women of low-income groups or of low educational level have no access to screening tests. Considering that 36.4% of the PCR-positive patients had cervicitis identified at gynecological examination, this is an important clinical finding for women's reproductive health because in certain situations, cervicitis may ensue infection in pregnancy, which may result in preterm labor and neonatal infections. This clinical finding is in agreement with the syndromic approach protocol for vaginal discharge and cervicitis established by CDC.Considering that chlamydial infections are usually asymptomatic and therefore difficult to screen, women without such marker would be considered to be at high risk of ascending infection, which may lead to severe complications resulting in tubal damage. [bib_ref] Cervical epithelial cells from Chlamydia trachomatis-infected sites coexpress higher levels of chlamydial..., Jha [/bib_ref] In this study, we verified a high rate of abortion. In contrast to most studies on the subject, 30 (90.9%) of these participants were negative for C. trachomatis infection. Although various studies state that C. trachomatis may cause abortion in several animal species, endocervical Chlamydia infections may or may not be associated with spontaneous abortion. [bib_ref] Dewhurst's textbook of obstetrics and gynaecology for postgraduates, Howie [/bib_ref] [bib_ref] Prevalence of antibody to Chlamydia trachomatis in spontaneous abortion and infertility, Quinn [/bib_ref] This diagnostic accuracy is important because it has the potential to help identifying cases at higher risk for spontaneous abortion.
In this study, we verified a case of a pregnant woman diagnosed with Chlamydia infection developed a stillborn fetus. The putative association with the infection could not be confirm because the fetus was investigated. Andrews et al. [bib_ref] The preterm prediction study: association of second-trimester genitourinary chlamydial infection with subsequent..., Andrews [/bib_ref] related a study where chlamydial infection detected at 24week gestation could be associated with double risk of preterm birth. Yet, infection that was detected at 28-week gestation had no significant effect of preterm birth.
Also, some studies indicate that C. trachomatis is an important etiological agent of endometritis in women with concomitant signs of salpingitis. [bib_ref] Endometritis caused by C. trachomatis, Mårdh [/bib_ref] It is possible that cases of postpartum endometritis in this study were due to C. trachomatis infection. Ascending infections of the female genital tract with sexually transmitted agents of the lower genital tract may occur during the postpartum period. [bib_ref] Microbial aetiology and diagnostic criteria of postpartum endometritis in, Temmerman [/bib_ref] In this study, 50% of positive C. trachomatis newborns presented respiratory problems. Various studies state that transmission of C. trachomatis from mothers to their newborns usually occurs at the time of delivery with passage of the newborn through the infected endocervix. According to Numazaki et al. and Schaad et al. the possibility of intrauterine infection at late pregnancy has also been reported. [bib_ref] Chlamydia trachomatis infection in early neonatal period, Numazaki [/bib_ref] Our finding showed a relation between C. trachomatis infection and respiratory symptoms. Cough symptoms showed to have an increased risk (RR 27.0) when neonates are born to infected mothers, followed by dyspnea (RR 15.4), coryza (RR 7.7) and nasal obstruction (RR 6.7). Respiratory symptoms were more frequent in neonates born to Chlamydia infected mothers. In addition, in this study, respiratory symptoms such as cough, was the most important symptom found in infants. Therefore, it is clear that clinical findings in this study in neonates born to Chlamydia infected mothers were associated with antenatal chlamydial infection in their mothers. Bébéar and Barbeyrac have also described that untreated infection acquired at birth can persist for months or years. [bib_ref] Genital Chlamydia trachomatis infections, Bébéar [/bib_ref] Moreover, the role of chlamydial infection resulting in neonatal pneumonia has also been considered. Although mortality associated to neonatal pneumonia caused by chlamydial infection is low, sequelae such as impairment in lung function may be evidenced. 14 These findings indicate the great importance of a systematic investigation for this curable STD in pregnant women. This points to the necessity of an important clinical measure to reduce vertical transmission and complications of C. trachomatis infection and other causes of curable STD.
# Conclusion
Our results demonstrate a true local reality, where women lack information about Chlamydia infection and an awareness of the possibility of gestational, perinatal and neonatal complications. Most of the time, pregnant women who are less privileged in socioeconomic and educational terms are left without any or insufficient medical assistance. It is important to emphasize the need for an equitable distribution of resources in health services. Therefore, it is essential to implement antenatal screening programs in public health service laboratories using the PCR method for detecting Chlamydia infection in this high risk population during the last month of pregnancy.
[table] Table 1 -: Speculum examination according to PCR results for C. trachomatis in pregnant women in Manaus, Amazonas, Brazil. [/table]
[table] Table 2 -: Premature rupture of membranes in pregnant women according to PCR results for Chlamydia trachomatis in Manaus, Amazonas, Brazil. [/table]
[table] Table 3 -: Incidence of respiratory symptoms in a cohort of 87 neonates exposed according to maternal Chlamydia trachomatis infection in Manaus, Amazonas, Brazil. [/table]
[table] Table 5 -: Mode of delivery and respiratory symptoms of neonates in Manaus, Amazonas, Brazil. [/table]
[table] Table 4 -: Incidence of nasal obstruction, coryza, cough and dyspnea in a cohort of neonates according to maternal Chlamydia trachomatis infection up to 60 days after birth in Manaus, Amazonas, Brazil. [/table]
|
Postoperative Adhesive Small Bowel Obstruction Presenting as Acute Urinary Retention
Postoperative adhesions are commonly seen after most abdominal surgeries. The majority of patients with intra-abdominal adhesions remain asymptomatic; however, some patients may develop symptoms ranging from mild abdominal pain to sometimes life-threatening complications. The most severe complication seen in association with postoperative adhesions is small bowel obstruction (SBO). This complication is rarely seen in the pediatric age-group. Adhesions have the potential to cause bowel obstruction. In addition, they can also affect the normal intestinal motility and transit leading to constipation. Several studies reported in literature show a strong association between constipation and the urinary disorders, such as incontinence and urinary urgency. We herein report a case of a boy who developed adhesive SBO with segmental ileal dilatation leading to constipation, urinary symptoms, and finally presenting to our hospital as acute urinary retention.
# Background
Abdominal adhesions are fibrous bands that span two or more intra-abdominal organs and/or the inner abdominal wall (i.e., peritoneal membrane). These adhesions are commonly seen after abdominal surgeries. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] They develop as part of the normal healing process occurring after peritoneal damage. The severity and extent of adhesions varies, with the majority not manifesting clinically. [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] Postoperative adhesions in adults can present as SBO (transient, partial, or complete), persistent or intermittent bloating, alteration in bowel habits (constipation or frequent loose stools), nausea, secondary infertility, chronic abdominal, and pelvic pain. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] However, adhesive intestinal obstruction is less commonly seen in the pediatric age-group. [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] We herein report a child with postoperative adhesive SBO presenting as an acute urinary retention.
## Case description
A 7-year-old boy born of non-consanguineous marriage presented with severe abdominal pain, vomiting for 1 day, and urinary retention for 18 hours. He was also constipated for 3 days at the time of presentation. The patient had a significant postnatal history and was operated on day 2 of life for ileal atresia. He underwent resection and primary anastomosis with intensive care stay for 30 days. He was apparently doing well till 3 years of age when he developed intermittent mild abdominal pain and was treated symptomatically. Later on, he developed constipation and was treated on and off by laxatives with temporary relief. However, over the last 3 months, the patient developed frequent urination and urge incontinence. He received antibiotics empirically for urinary tract infection (UTI) without any relief. Because of the recent onset urinary symptoms not attributed to UTI, the urodynamic study and the ultrasound (US) abdomen were requested; US abdomen was normal and the urodynamic study showed a small capacity overactive bladder. The child presented to us after 2 days of the urodynamic study with acute urinary retention. The possibilities considered were possible trauma causing urethral stricture during the urodynamic study and pressure effects of chronic constipation. On examination, the child was afebrile, with respiratory rate, 20/ minutes; heart rate, 90/minutes; and blood pressure, 90/60 mmHg. His weight and height were between the 25th and 50th percentiles for height/age index by gender. On the abdominal examination, there was generalized tenderness without guarding or rigidity. There was dullness in the suprapubic region suggestive of bladder distension. There was no other organomegaly, neurological, respiratory, and cardiac examinations were normal. Bladder decompression was performed by catheterization (250 mL output of clear urine), followed by marked pain relief after the procedure. Complete blood count, renal function tests, liver function tests, and inflammatory markers were normal. Urine microscopy and culture were normal. A plain abdominal radiograph was performed in supine and standing positions which showed dilated small bowel loops with air fluid levels. Fecal loading of the colon was also seen. In view of air fluid levels, barium meal study was done which showed marked segmental dilatation of the ileum distally [fig_ref] Figure 1: Barium meal with follow through showing marked segmental dilatation of the ileum... [/fig_ref]. Abdominal adhesions were suspected in view of the radiological findings and the past history of abdominal operation. Laparotomy was done which showed segmental dilatation (around 30 cm) of the distal ileum at a distance of 20 cm for the ileocecal valve. The old anastomosis was patent. There were extensive abdominal adhesions. Adhesiolysis of small bowel along with resection of the distended ileum and end-to-end anastomosis was performed. He was discharged on laxatives and is currently doing well on follow-up. The urinary symptoms and abdominal pain had subsided on follow-up.
# Discussion
The incidence of adhesive intestinal obstruction in children varies from 2.2 to 8.3%. [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] [bib_ref] Incidence of postoperative adhesion obstruction following neonatal laparotomy, Wilkins [/bib_ref] Despite several advances in surgical techniques including laparoscopy, there is little change in the epidemiology and understanding of the pathophysiology of adhesions. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] [bib_ref] Postoperative abdominal adhesions and their prevention in gynaecological surgery: II. How can..., Pouly [/bib_ref] The timing from the initial operation to the presentation remains variable and can range from 7 days to 12 years. [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] In Janik et al. series, adhesive intestinal obstruction developed within 2 years in 80% of their patients. [bib_ref] An assessment of the surgical management of adhesive small bowel obstruction in..., Janik [/bib_ref] Abdominal adhesions have no specific laboratory features and are not readily visible by currently available imaging methods. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] Thus, patients may experience protracted symptoms, delayed diagnosis, and adverse medical outcomes. Our patient was operated on day 2 of life and gradually developed abdominal adhesions which started manifesting clinically at 3 years of age. Initially, he had vague symptoms of abdominal pain which progressed to constipation with urinary symptoms (urge incontinence and frequent micturition) and finally acute urinary retention. During this period, he was investigated and treated several times for constipation and UTI but remained undiagnosed for adhesive disease. The majority of children with intra-abdominal adhesions are asymptomatic. 1,2 Some of the risk factors for the development of symptomatic "adhesive disease" include infection, hypoxia, and multiple surgeries. [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] Adhesions that develop after appendectomy and procedures involving the left side of colon and rectum have been found to have enhanced predilection for intestinal obstruction. [bib_ref] Clinical implications of postsurgical adhesions, Diamond [/bib_ref] Although a conservative treatment does not remove the cause of the obstruction, the surgery can induce new adhesions. Thus, the management of SBO caused by adhesions remains controversial and difficult. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] [bib_ref] Adhesive intestinal obstruction in infants and children: the place of conservative treatment, Al-Salem [/bib_ref] [bib_ref] Only surgical treatment to be considered for adhesive small bowel obstruction: a..., Tabchouri [/bib_ref] Adhesions have the potential to interfere with the normal intestinal motility and transit leading to altered bowel habits and constipation. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] Similarly, due to adhesions, our patient had gradually developed constipation with X-ray abdomen showing fecal loading [fig_ref] Figure 2: X-ray abdomen supine and erect showing dilated small bowel loops with air... [/fig_ref]. Although rare, constipation has been reported to cause acute urinary retention in children. [bib_ref] Acute urinary retention in a pre-school girl with constipation, Traslaviña [/bib_ref] The association of constipation with urinary disorders such as incontinence and urgency is well established. [bib_ref] Acute urinary retention in a pre-school girl with constipation, Traslaviña [/bib_ref] [bib_ref] Prevalence of lower urinary tract symptoms in schoolage children, Vaz [/bib_ref] Children with constipation have been found to have large volumes of residual urine and urinary tract dilatation, including the ureteropelvic tract. [bib_ref] Constipation is associated with incomplete bladder emptying in healthy children, Chang [/bib_ref] The pathophysiologic association of voiding disorders seen in children with constipation has been attributed to several factors. During the pelvic floor formation, there is a close embryological association (claoca) between the bladder and the rectum, sharing the same innervation, nerve roots S2 to S4. These nerve roots are responsible for controlling the motor function of the internal anal and urinary sphincters. [bib_ref] Acute urinary retention in a pre-school girl with constipation, Traslaviña [/bib_ref] [bib_ref] Constipation is associated with incomplete bladder emptying in healthy children, Chang [/bib_ref] [bib_ref] Rectal distention inhibits the spinal micturition reflex via glycinergic or GABAergic mechanisms..., Miyazato [/bib_ref] Experiments with rats showed that the distension of rectum with a balloon results in diminished bladder contractility. [bib_ref] Rectal distention inhibits the spinal micturition reflex via glycinergic or GABAergic mechanisms..., Miyazato [/bib_ref] Chronic fecal retention can also result in the involuntary contraction of the pelvic floor muscle and the external anal sphincter, thereby making bladder emptying difficult. [bib_ref] The impact of constipation on the urinary tract system, Halachmi [/bib_ref] [bib_ref] Functional constipation in children, Chase [/bib_ref] Additionally, due to the close anatomical association, the presence of impacted stool in the rectum (and in our case, also the dilated segmental ileum) reduces the bladder functional capacity. This results in a feeling of earlier bladder emptying, that is, increased frequency and hesitancy. [bib_ref] The impact of constipation on the urinary tract system, Halachmi [/bib_ref] [bib_ref] Functional constipation in children, Chase [/bib_ref] Moreover, a chronically full rectal ampulla can lead to invaginations in the posterior wall of the bladder and urethral obstruction. [bib_ref] Functional constipation in children, Chase [/bib_ref] All the above factors may have played a role in causing urinary symptoms and acute urinary retention in our patient.
## Conclusion a n d clinical significance
A past history of abdominopelvic surgery in a child and presence of symptoms such as chronic abdominal pain, constipation, and/ or urinary symptoms should alert the pediatrician to consider a
[fig] Figure 2: X-ray abdomen supine and erect showing dilated small bowel loops with air fluid levels. Fecal loading of the colon is seen [/fig]
[fig] Figure 1: Barium meal with follow through showing marked segmental dilatation of the ileum possible diagnosis of adhesive intestinal disease. A high index of suspicion will help in early detection of cases and thereby reduce the morbidity and mortality. Patients should be counseled regarding the possibility of adhesions and the associated risk during preoperative explanation of the planned procedure. orcid Syed Ahmed Zaki https://orcid.org/0000-0003-2652-4585 references [/fig]
|
A Molecular Insight into Algal-Oomycete Warfare: cDNA Analysis of Ectocarpus siliculosus Infected with the Basal Oomycete Eurychasma dicksonii
Brown algae are the predominant primary producers in coastal habitats, and like land plants are subject to disease and parasitism. Eurychasma dicksonii is an abundant, and probably cosmopolitan, obligate biotrophic oomycete pathogen of marine brown algae. Oomycetes (or water moulds) are pathogenic or saprophytic non-photosynthetic Stramenopiles, mostly known for causing devastating agricultural and aquacultural diseases. Whilst molecular knowledge is restricted to crop pathogens, pathogenic oomycetes actually infect hosts from most eukaryotic lineages. Molecular evidence indicates that Eu. dicksonii belongs to the most early-branching oomycete clade known so far. Therefore Eu. dicksonii is of considerable interest due to its presumed environmental impact and phylogenetic position. Here we report the first large scale functional molecular data acquired on the most basal oomycete to date. 9873 unigenes, totalling over 3.5Mb of sequence data, were produced from Sanger-sequenced and pyrosequenced EST libraries of infected Ectocarpus siliculosus. 6787 unigenes (70%) were of algal origin, and 3086 (30%) oomycete origin. 57% of Eu. dicksonii sequences had no similarity to published sequence data, indicating that this dataset is largely unique. We were unable to positively identify sequences belonging to the RXLR and CRN groups of oomycete effectors identified in higher oomycetes, however we uncovered other unique pathogenicity factors. These included putative algal cell wall degrading enzymes, cell surface proteins, and cyclophilin-like proteins. A first look at the host response to infection has also revealed movement of the host nucleus to the site of infection as well as expression of genes responsible for strengthening the cell wall, and secretion of proteins such as protease inhibitors. We also found evidence of transcriptional reprogramming of E. siliculosus transposable elements and of a viral gene inserted in the host genome. Citation: Grenville-Briggs L, Gachon CMM, Strittmatter M, Sterck L, Kü pper FC, et al. (2011) A Molecular Insight into Algal-Oomycete Warfare: cDNA Analysis of Ectocarpus siliculosus Infected with the Basal Oomycete Eurychasma dicksonii. PLoS ONE 6(9): e24500.
# Introduction
Like all other living organisms, algae suffer from diseases, which may range from spectacular outbreaks in natural populations to significant losses in multibillion dollar crops such as nori [bib_ref] Algal diseases: spotlight on a black box, Gachon [/bib_ref]. As aquaculture continues to rise worldwide, and with algae considered as a sustainable biofuel source, pressure is mounting to design efficient disease control methods. More generally, parasites and pathogens are increasingly being considered of equal importance with predators for ecosystem functioning [bib_ref] Parasites in food webs: the ultimate missing links, Lafferty [/bib_ref]. In aquatic as well as terrestrial environments, altered disease patterns in disturbed environments are blamed for sudden extinctions, regime shifts, and spreading of alien species. Likewise, algal pathogens exert a range of complex, profound, sometimes subtle, and often unexpected impacts in aquatic ecosytems. These range from host generation shifts to changes in biogeochemical cycling and atmosphere chemistry (e.g. [bib_ref] Cheshire Cat'' escape strategy of the coccolithophore Emiliania huxleyi in response to..., Frada [/bib_ref].
Despite mounting recognition of their importance, molecular knowledge on algal pathogens is hitherto restricted to viruses (reviewed in [bib_ref] Algal diseases: spotlight on a black box, Gachon [/bib_ref]. Likewise, the molecular characterisation of algal immune responses is in its infancy [bib_ref] Pathogen-induced defense and innate immunity in macroalgae, Weinberger [/bib_ref]. To help address these fundamental and ecological questions, we have developed a laboratory-controlled pathosystem involving the genome model seaweed Ectocarpus siliculosus [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] and the oomycete pathogen Eurychasma dicksonii [bib_ref] Infection experiments reveal broad host ranges of Eurychasma dicksonii (Oomycota) and Chytridium..., Müller [/bib_ref] [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref].
Eu. dicksonii is the most common and widespread eukaryotic pathogen of marine algae. It occurs in all cold and temperate seas worldwide [bib_ref] Differential host susceptibility to the marine oomycete pathogen Eurychasma dicksonii detected by..., Gachon [/bib_ref]. In culture, it infects virtually all species of brown algae tested (more than 40, [bib_ref] Infection experiments reveal broad host ranges of Eurychasma dicksonii (Oomycota) and Chytridium..., Müller [/bib_ref] , including representatives of all major orders of this phylum. As brown algae are the predominant primary producers of temperate and polar rocky shores, and Eu dicksonii occurs in frequent epidemics [bib_ref] Massive occurrence of the heterokont and fungal parasites Anisolpidium, Eurychasma and Chytridium..., Küpper [/bib_ref] , we infer that this pathogen contributes to shaping natural algal populations, therefore profoundly impacting ecosystem functioning [bib_ref] Ecology of lower oomycetes, Strittmatter [/bib_ref].
Importantly, molecular taxonomy unambiguously designates Eu. dicksonii as the most early-branching clade within the oomycete lineage [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref] [bib_ref] Phylogenetic affinities of two eukaryotic pathogens of marine macroalgae, Eurychasma dicksonii (Wright)..., Küpper [/bib_ref]. Oomycetes (or water moulds) are classified within the group of Stramenopiles together with diatoms, golden-brown and brown algae [bib_ref] The deep roots of Eukaryotes, Baldauf [/bib_ref]. They are secondarily non-photosynthetic organisms which exhibit either pathogenic or saprophytic lifestyles [bib_ref] Phytophthora Genome Sequences Uncover Evolutionary Origins and Mechanisms of, Tyler [/bib_ref] [bib_ref] The Biotrophic Stages of Oomycete-Plant Interactions, Grenville-Briggs [/bib_ref]. In fact, many of the most devastating agricultural and aquacultural pathogens belong to the oomycetes [bib_ref] Molecular genetics of pathogenic oomycetes, Kamoun [/bib_ref] [bib_ref] New insights into animal pathogenic oomycetes, Phillips [/bib_ref] [bib_ref] Saprolegnia parasitica, an oomycete pathogen with a fishy appetite:new challenges for an..., Van West [/bib_ref]. For example, Phytophthora infestans, the causal agent of potato late blight, is responsible for annual crop losses and pesticide costs exceeding £5 billion worldwide [bib_ref] Societal costs of late blight in potato and prospects of durable resistance..., Haverkort [/bib_ref]. Many other species cause significant environmental damage, especially when recently introduced (e.g. Phytophthora ramorum). Therefore, the best-studied oomycete species are phylogenetically highly-derived plant pathogens, of which five fully sequenced genomes are hitherto available. However, pathogenic oomycetes are taxonomically much more diverse, and infect a remarkable palette of hosts ranging from marine algae, crustaceans, plants, nematodes, fungi, insects, to fishes and mammals [bib_ref] New insights into animal pathogenic oomycetes, Phillips [/bib_ref]. From a physiological standpoint, Eu. dicksonii is typical of most basal oomycetes, inasmuch as it is an obligate biotroph, exclusively growing and propagating within a living host. After encysting at the surface of an algal cell, infectious spores inject their protoplasm into the host cytoplasm, and develop into a multinucleate coenocytic intracellular thallus. After having progressively filled the infected algal cell, the latter differentiates into a sporangium that releases new infectious zoospores [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref]. The extremely broad host range of Eu. dicksonii is unusual among biotrophic pathogens, and may reflect a low level of specialization, suggesting that it may have retained ancestral infection mechanisms.
In summary, Eu. dicksonii is of considerable experimental interest because of its basal phylogenetic position, its presumed environmental impact, and amenability to molecular studies. Here, we set out to investigate the physiology of the E. siliculosus -Eu dicksonii host-pathogen interaction by constructing cDNA libraries of axenic cultures. We aimed to address what pathogenicity determinants and strategies Eu. dicksonii uses to infect E. siliculosus and what are the consequences of infection on algal physiology. Overall, we describe here a very original dataset that underlines both the taxonomical distance between Eu dicksonii and betterstudied oomycetes, and the specialised marine lifestyle of this pathogen.
# Materials and methods
## Biological material and microscopy
Axenic cultures of the fully sequenced Ectocarpus siliculosus strain CCAP 1310/4 and of Eurychasma dicksonii CCAP 4018/1 were obtained as described by [bib_ref] Axenic clonal cultures of filamentous brown algae. Initiation and maintenance, Müller [/bib_ref]. They were maintained in half strength Provasoli medium [bib_ref] Long term and diurnal carpospore discharge patterns in the Ceramiaceae, Rhodomelaceae and..., West [/bib_ref] at 15uC, under illumination with daylight type fluorescent lamps at a 2-10 mE.m 22 .s 21 irradiance and a 12 h photoperiod, as described by [bib_ref] Infection experiments reveal broad host ranges of Eurychasma dicksonii (Oomycota) and Chytridium..., Müller [/bib_ref]. Being an obligate biotroph, Eu. dicksonii was maintained in co-culture with its E. siliculosus host and transferred into fresh medium every second week, together with some new uninfected algal host. The Eu. dicksonii CCAP 4018/1 strain was originally established from a single developing parasitic thallus, which was propagated into a clonal healthy algal host. For microscopy, infected algae were fixed for 45 min with 4% paraformaldehyde in microtubule-stabilising buffer [bib_ref] Immunofluorescence and electron microscopic studies of microtubule organisation during the cell cycle..., Katsaros [/bib_ref] , dipped into methanol for 30 s, and transferred into a DAPI solution (10 mg/mL in sterile seawater) for 5 minutes and observed under an epifluoresence microscope equipped with the following filter set: L365 FT 395, LP 420.
## Construction of cdna libraries and sequencing strategy
Fifty E. siliculosus cells each containing a developing intracellular Eu. dicksonii thallus were dissected under the stereomicroscope using a glass Pasteur pipette, and transferred into RNALaterH. Total RNAs were extracted, subjected to reverse-transcription using a poly-dT oligonucleotide, PCR-amplified, and cloned directionally into a pBluescript II sk+ vector (Vertis Biotechnologie AG, Germany). After plating, 3000 colonies were robot-picked. Their insert was further sequenced from both ends using Sanger technology (Genoscope, France).
In parallel, two densely infected E. siliculosus cultures (,20 mg FW each) were harvested in RNALaterH, without prior dissection. The ''young'' culture (4 weeks after inoculation) contained all Eu. dicksonii development stages, including young intracellular thalli, mature dehiscent sporangia and encysted spores at the surface of algal cells. In contrast, the ''old'' culture (2 months after inoculation) predominantly contained mature dehiscent sporangia and encysted spores. Total RNA was extracted and subjected to cDNA synthesis as above. For each sample, 50,000 sequence reads were obtained using 454 pyrosequencing.
# Bioinformatic analysis
Data assembly. The sequence reads were first assembled separately for each individual dataset using CAP3 (default parameters) [bib_ref] CAP3: A DNA sequence assembly program, Huang [/bib_ref] , and then further assembled, also using CAP3, (default parameters, without resorting to quality file) into the final hybrid assembly discussed throughout the manuscript. Unigenes containing poor quality sequence reads (manual assessment of quality) were removed from the final dataset. A total of 5889 unassembled EST sequences (dissected library) were submitted to the EST db at EMBL and were assigned accession numbers FR839767-FR845655. Pyrosequencing reads (old culture and young culture libraries) were submitted to the EMBL short-read assembly (SRA) database. Sequences from each of the three databases were assembled into a total of 9847 unigenes.
Sequence Analysis. Unigenes were mapped to the E. siliculosus nuclear, chloroplastic and mitochondrial genomes with the use of GenomeThreader [bib_ref] Engineering a software tool for gene structure prediction in higher organisms, Gremme [/bib_ref] to produce the 'host' dataset. Genome-Threader was applied with default parameters, except for the minimum alignment length and percent identity, which were set to 90% and 95% resp. Non-mapping unigenes were assigned to Eu. dicksonii after blastn analysis (98% to 100% sequence similarity) to remove potential contaminants, producing the 'pathogen' dataset. Open reading frames were predicted with OrfPredictor [bib_ref] OrfPredictor: predicting proteincoding regions in EST-derived sequences, Min [/bib_ref] , based on a blastx analysis run against the NCBI Genbank non-redundant (nr) database and the E. siliculosus genome database (http:// bioinformatics.psb.ugent.be/genomes/view/Ectocarpus-siliculosus). Sequence similarity searches were carried out using local blast tools [bib_ref] Basic local alignment search tool, Altschul [/bib_ref] , the NCBI nr database (28-07-2010 version) and several predicted proteomes of fully sequenced Eukaryotes. These include the proteomes of six oomycete species sequenced to date: Phytophthora infestans, [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref] ; Phytophthora sojae, [bib_ref] Phytophthora Genome Sequences Uncover Evolutionary Origins and Mechanisms of, Tyler [/bib_ref] ; Phytophthora ramorum, [bib_ref] Phytophthora Genome Sequences Uncover Evolutionary Origins and Mechanisms of, Tyler [/bib_ref] ; Pythium ultimum, [bib_ref] Genome sequence of the necrotrophic plant pathogen, Pythium ultimum, reveals original pathogenicity..., Lévesque [/bib_ref] ; Hyaloperonospora arabidopsidis [bib_ref] Signatures of Adaptation to Obligate Biotrophy in the Hyaloperonospora arabidopsidis, Baxter [/bib_ref] and Saprolegnia parasitica (draft genome, publically available at; (www.broadinstitute. org/annotation/genome/Saprolegnia_parasitica/MultiHome.html). Additional proteomes included were: the model brown alga and host species used in the current study, Ectocarpus siliculosus; the marine centric diatom Thalassiosira pseudonana; the marine pennate diatom Phaeodactylum tricornutum; the coccolith-bearing haptophyte Emiliania huxleyi; the Apicomplexan malaria parasite Plasmodium falciparum; the model plant Arabidopsis thaliana; the amoeboflagellate Naegleria gruberi; the amphibian pathogenic chytridiomycete fungus Batrachochytrium dendrobatidis and the plant pathogenic ascomycete fungus Mycosphaerella fijiensis. Predicted protein domains were identified using a standalone InterProScan program [bib_ref] InterPro: the integrative protein signature database, Hunter [/bib_ref] (default settings) and the NCBI conserved domain database [bib_ref] CDD: a Conserved Domain Database for the functional annotation of proteins, Marchler-Bauer [/bib_ref]. E. siliculosus transposable elements were identifed and named according to [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] and. Their sequences were retrieved from the following URL: http:// urgi.versailles.inra.fr/Data/Transposable-elements/Ectocarpus.
The Emboss tool infoseq [bib_ref] EMBOSS: the European Molecular Biology Open Software Suite, Rice [/bib_ref] was used to generate statistics for host and pathogen datasets independently.
Signal peptide predictions were carried out using SignalP version 3.0 [bib_ref] Locating proteins in the cell using TargetP, SignalP, and related tools, Emanuelsson [/bib_ref] with a hidden Markov model probability cutoff of 0.6, and 1160 candidate secreted proteins were obtained.
Effector protein motif predictions. Searches for RXLR and CRN motifs were carried out according to the methods of [bib_ref] Transcriptome of Aphanomyces euteiches: new oomycete putative pathogenicity factors and metabolic pathways, Gaulin [/bib_ref]. Briefly, string searches were carried out using the expression RxLR -x (1,40)-[ED] [ED] [KR] using custom build python scripts. This string search was also performed just with the RxLR expression, and was performed on all sequences, as well as those with predicted signal peptides. String searches were also performed with both the CRN motifs from Phytophthora species and Pythium ultimum. HMM profile searches for all these motifs were also carried out using the HMMer 2.2 package as described in [bib_ref] Transcriptome of Aphanomyces euteiches: new oomycete putative pathogenicity factors and metabolic pathways, Gaulin [/bib_ref].
Multiple alignments were conducted using the program CLUSTALW [bib_ref] Clustal W and Clustal X version 2.0, Larkin [/bib_ref] (default parameters) and visualised using Geneious (version 4.8;. Distance trees were constructed using Geneious with neighbour joining algorithms using 1000 bootstrap replications.
# Results
## Cytological observations of eu. dicksonii infected e. siliculosus
The morphological development and ultrastructural cytology of Eu. dicksonii infecting both E. siliculosus and the related filamentous phaeophyte alga Pylaiella littoralis has been investigated previously [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref]. Infection is initated by secondary cysts, which attach to the surface of the host cell. Upon germination, the algal cell wall is breached, and cytoplasm from the pathogenic cell is transfered into the host cytoplasm. The initial phase of parasite development as a non-walled thallus is reminiscent of fungal pathogens of higher plants, such as Plasmodiophora brassicae and Olpidium brassicae [fig_ref] Figure 1: Co-localisation of E [/fig_ref] , [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref]. The Eu. dicksonii thallus develops as a spherical intracellular syncitium, progressively filling the individual algal cell and ultimately causing its hypertrophy [fig_ref] Figure 1: Co-localisation of E [/fig_ref]. The sporangium develops by differentiation of the syncitium that releases new infectious zoospores [fig_ref] Figure 1: Co-localisation of E [/fig_ref]. The Eu. dicksonii thallus never propagates to other cells within the algal filament. During infection, the host nucleus and Golgi apparatus are closely associated with the pathogen thallus [fig_ref] Figure 1: Co-localisation of E [/fig_ref].
cDNA library quality control and statistics 3000 clones from microdissected Eu. dicksonii-infected E. siliculosus cells were Sanger-sequenced (Genoscope, France) and assembled into contigs (CAP3). This resulted in the identification of 1424 unigenes. Pyrosequencing reads, from both the 'young culture' and 'old culture,' were assembled at the NERC National Molecular Genetics Facility (Liverpool, UK), resulting in two data sets of 4778 unigenes (young culture) and 6840 unigenes (old culture). Unigenes from these three datasets were combined to produce a single dataset, and assembled to produce 9873 unigenes totalling over 3.5 Mb of sequence data [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. 28 sequences (15 from the young culture library, 12 from the old culture library, and 1 singlet from the dissected library) were discarded on the basis of poor sequence (containing regions of more than 20 bases called as Ns). 38 singlets, originating from the dissected library, were also removed on the basis of containing less than 90 nt of readable sequence. The final total hybrid assembly (discussed throughout the text) of 9873 unigenes therefore contained 9658 contigs and 215 singlets. The total 9873 unigene dataset was made up of 295 contigs with representatives from all three libraries, 1680 contigs with representatives in both the young culture and old culture libraries, 52 contigs with representatives in both the young culture and dissected libraries, 136 contigs with representatives in both the dissected and old culture library and 7710 unigenes that were only present in one of the three libraries [fig_ref] Figure 2: Venn Diagram showing numbers of overlapping sequences from each library within the... [/fig_ref].
Unigenes were mapped to the E. siliculosus genome as described in [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. 6787 unigenes were mapped to the E. siliculosus genome and, therefore, predicted to be of host origin. 3152 unigenes did not map to the E. siliculosus genome. 66 poor quality or short sequences were excluded, leaving 3086 sequences ascribed to Eu. dicksonii. This resulted in the creation of two datasets, 'host sequences' containing 6787 unigenes [fig_ref] Figure 1: Co-localisation of E [/fig_ref] , and 'pathogen sequences' [fig_ref] Figure 2: Venn Diagram showing numbers of overlapping sequences from each library within the... [/fig_ref] containing 3086 unigenes. Strikingly, unigenes from the host dataset have a mean GC of 51.9% whereas those from the pathogen dataset have a much lower mean GC of 40.5% [fig_ref] Table 2: Characteristics of host and pathogen datasets [/fig_ref].
## Highly expressed transcripts
Major up-regulation of Ectocarpus transposable elements. We used the number of sequence reads normalised over contig length as a proxy to evaluate relative gene expression levels in both the ''young'' and ''old'' pyrosequenced libraries. Unsurprisingly, the vast majority of the most abundantly expressed transcripts are host housekeeping genes. However, 16 (resp. 25) of the top 150 most expressed unigenes in the young and old libraries are E. siliculosus transposable elements (TEs). E. siliculosus TEs identified among the top most 150 highly expressed contigs in at least one library were extracted, and their expression level is plotted on [fig_ref] Figure 3: Over-representation of Ectocarpus transposable elements [/fig_ref]. We further compared the expression level of TEs annotated by [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] between the two Eu. dicksoniii-infected libraries and the Ectocarpus genome initiative EST collection [fig_ref] Figure 3: Over-representation of Ectocarpus transposable elements [/fig_ref]. The latter mostly encompasses distinct development stages of unstressed Ectocarpus. Whereas pyrosequencing read counts and EST numbers are not directly comparable between the EGI dataset [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] and ours, differential expression of some TEs is nevertheless clearly discernible. Indeed, the five most expressed TEs identified by Cock and co-authors are also well represented in both Eu. dicksonii-infected libraries. However, the retroelements RTE2, RTE3 and RTE4, the LTR element NgaroDIRS6, and the LARD element EsLARD1_ZnF, virtually silent in unstressed conditions, appear strongly induced in the ''young'' and ''old'' pyrosequenced libraries. Interestingly, RTE2, RTE3 and RTE4 also belong to the top 10 most abundant repeats in the E. siliculosus genome, suggesting stress-induced transposing activity in the E. siliculosus genome.
Highly expressed Eurychasma transcripts. The 10 most abundant sequences within the dissected library are of pathogen origin, with the exception of a conserved unknown E. siliculosus protein, demonstrating the efficacy of microdissection for enriching the sample in pathogen RNAs [fig_ref] Table 3: Top 10 most highly expressed unigenes in the Eu [/fig_ref]. The majority of the most abundant sequences are housekeeping and ribosomal genes. However, three unigenes of Eu. dicksonii origin are highly abundant in this library and are proteins with unknown functions [fig_ref] Table 3: Top 10 most highly expressed unigenes in the Eu [/fig_ref]. Contig500203 encodes a methionine and lysine rich unknown protein. Contig500580 encodes a predicted protein with similarity to a group of unknown oomycete proteins. These proteins are a group of at least 8 proteins present in the Saprolegnia parasitica genome, but absent in the genomes of other sequenced oomycetes. These appear to be uncharacterised cytoplasmic proteins, which may be ancestral proteins lost from higher oomycetes. Contig50031 appears to encode a lysine and alanine rich protein or protein fragment, with no similarity to known proteins.
## Functional annotation
The 9873 unigene set was annotated by comparison to the NCBI non-redundant (nr) protein database and the E. siliculosus genome database (http://bioinformatics.psb. ugent.be/webtools/bogas/overview/Ectsi; [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] using BLAST analyses [bib_ref] Engineering a software tool for gene structure prediction in higher organisms, Gremme [/bib_ref]. Overall, approximately 20% of the sequences in the total dataset showed similarity to previously described genes in the NCBI nr protein database, using an e value cutoff of ,1e-05, indicating that this is largely a unique dataset.
Functional annotation of host genes. The E. siliculosus unigene set (host sequences) was initially annotated by direct comparison to the E siliculosus genome sequence [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. Blastx analysis (expectation value ,1e -05) of this dataset to the E. siliculosus predicted proteome resulted in 1299 (19.2%) of sequences matching predicted E siliculosus proteins. 46% (3163) have a significant blastn hit to genomic mRNA sequences, which include the 39 UTRs, with the remainder of the sequences hitting mitochondrial, chloroplastic or intergenic sequences. Protein sequences were predicted using ORFPredictor [bib_ref] OrfPredictor: predicting proteincoding regions in EST-derived sequences, Min [/bib_ref] using the E. siliculosus blastx results and checked for agreement with the E. siliculosus genome database. Predicted protein sequences were analysed for the presence of an N-terminal signal sequence, using both Hidden Markov Models and Neural Network algorithms in SignalP [bib_ref] Locating proteins in the cell using TargetP, SignalP, and related tools, Emanuelsson [/bib_ref]. 839 (12.4%) of the host unigenes were predicted to contain a signal peptide. Predicted proteins were also analysed for functional domains using InterProScan [bib_ref] InterPro: the integrative protein signature database, Hunter [/bib_ref]. 822 unigenes (12%) of the dataset had one or more hits to functional domains using InterProScan. All the 822 predicted ORFs with InterproScan hits were within the 1299 sequences predicted to match protein-coding regions of the E. siliculosus genome. 63% of the unigenes that are derived from E. siliculosus protein coding regions, therefore, contain functional domains as identified by InterproScan. Functional annotation of host genes was checked with the HECTAR predictions for full-length sequences in the E. siliculosus genome database [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. Predicted ORFs in this dataset were assigned to functional categories based on GO annotations of the InterProScan/HECTAR predictions . The largest category was gene regulatory proteins (15% of the total), including predicted transcription factors, translation factors, DNA and RNA binding proteins and proteins containing one or more coiled coil domains. 14% of predicted ORFs were classified as involved in translation, with approximately two thirds of these being ribosomal proteins. 14% of predicted proteins were also classified as being involved in protein modification, targeting or turnover including chaperones, and proteins involved in ubiquitination and proteolysis. Functional annotation of pathogen genes. The Eu. dicksonii unigene dataset (containing sequences that did not map to the host genome) was analysed and ORFs were predicted as described above for E. siliculosus.
InterProScan analysis was performed on Eu. dicksonii predicted ORFs, in the same manner as on the host dataset. 1397 of the 3086 (44%) pathogen unigenes were predicted to contain one or more functional domain using InterProScan. As seen with the host dataset, the largest class of proteins were those predicted to contain gene regulatory domains, including those with coiled-coil domains. This class represented almost a quarter (24%) of all pathogen sequences with functional domains . 8% of sequences were classified as being involved in signalling in the pathogen dataset, compared to just 4% of host sequences. Interestingly, proteins predicted to be involved in pathogenicity and suppression of immunity were also identified in the pathogen dataset, but were absent from host sequences. A higher percentage of proteins classified as having a role in redox reactions, including those involved in detoxification and homeostasis, were identified within the host dataset (9%) in comparison to the pathogen dataset (5%). It is anticipated that several of these proteins function in cellular defence mechanisms.
In addition to blastx similarity searches with the NCBI nr database, fifteen eukaryotic proteomes (listed in the methods section) were added to the analysis [fig_ref] Figure 5: Comparison of Eu [/fig_ref]. Up to 25% of the pathogen sequences showed significant similarity (, 1e-05) to proteins from sequenced oomycetes. Interestingly about the same number (28% at ,1e -05) showed significant similarity to proteins in the NCBI nr dataset and to the algal host E. siliculosus predicted proteome (24.4% at 1e-05), indicating that this dataset is largely made up of unique, uncharacterised sequences. As expected, the Eu. dicksonii dataset showed the most similarity to the oomycete proteomes, and to E. siliculosus. A lesser degree of similarity was seen with the diatom, haptophyta and apicomplexa proteomes. Surprisingly, a higher level of similarity (up to 18%, at 1 e-05) was observed between the pathogen unigenes and the proteome of the model plant Arabidopsis thaliana, than with some of the chromalveolata, such as Plasmodium falciparum with which there was only 11% similarity (at 1e-05). Likewise, up to 15% sequence similarity was seen between Eu. dicksonii unigenes and distantly related organisms such as the amoeboflagellate Naegleria gruberi and true fungi [fig_ref] Figure 5: Comparison of Eu [/fig_ref]. 351 (11.6%) of the pathogen sequences were predicted to contain a signal peptide. 179 of these also had one or more hits to InterProScan domains, including non-specific 'seg' hits. At least 30 of those containing known domains were in fact membrane proteins, or fragments of membrane proteins that were likely picked up with the SignalP algorithms because they are not full length. In the majority of cases, it is not possible to conclude whether our contigs represent full-length sequences with real signal peptides, or simply fragments of proteins with hydrophobic or transmembrane spanning regions. However, 10 sequences from the Eu. dicksonii dataset are predicted to encode full-length, unknown, secreted proteins that do not contain transmembrane domains [fig_ref] Table 4: Predicted full-length secreted Eu [/fig_ref]. These predicted proteins either have no sequence similarity to known proteins or domains, or have an *custom database made up of proteomes described in [fig_ref] Figure 5: Comparison of Eu [/fig_ref]. doi:10.1371/journal.pone.0024500.t003 . Classification of unigenes based on identification of functional domains using InterproScan. Functional categories assigned to the 822 E. siliculosus (host) and 1397 Eu. dicksonii (pathogen) predicted ORFs (12% and 43% of total unigenes, respectively) with an Interpro hit. Numbers assigned to each category are percentages of the 822 and 1397 predicted ORFs. Amino acids includes amino acid synthesis and metabolism; cell cycle genes also include mitosis, DNA and RNA synthetic genes; CW/CHO includes cell wall biosynthesis and carbohydrate synthesis and metabolism; cytoskeleton includes transcripts involved in cytoskeletal rearrangements; defence and stress includes transcripts involved in general and specific defence or stress responses; energy/metabolism includes transcripts involved in energy production and cellular metabolism but does not include those involved specifically in respiration; gene regulation includes transcripts involved in DNA and RNA binding, transcription factors and transcripts with coiled-coil domains, lipids includes both lipid biosynthesis and metabolism; pathogenicity includes transcripts with a predicted function in pathogenicity or virulence, photosynth includes photosynthetic machinery; protein mod/turnover includes transcripts involved in protein folding, targeting, modification and degradation; prot-prot int includes transcripts predicted to be involved in protein-protein interactions; redox includes transcripts involved in homeostasis, detoxification and those classified as redox antioxidants; mitochondrial/respn transcripts are predicted to be mitochondrial and/or involved in respiration; signalling includes transcripts with a predicted role in signalling or signal transduction; small mol binding includes those transcripts which bind small molecules; translation includes those transcripts with a role in translation, including ribosomal proteins; transposons includes sequences both of transposons and those of putative retroviral origin; transcripts which could not be assigned to any one of the above categories are classified in the section termed other. doi:10.1371/journal.pone.0024500.g004
Interpro hit to the non-specific 'seg' database. There is no significant sequence similarity between the 10 predicted proteins. However, three of them have similarity to uncharacterised secreted proteins from other oomycetes. These proteins may possibly be ancient oomycete effectors, however further work is required to investigate the precise function of these proteins. Other potentially secreted pathogenicity factors, which contain conserved domains are listed in [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref].
## Insights into oomycete infection of brown algae
Pathogen cell surface proteins. Mucins are typically high molecular weight cysteine, threonine, or serine non-glycosylated proteins that act as lubricants, protectants, signal tranducers or adhesins [bib_ref] The structure and assembly of secreted mucins, Perez-Vilar [/bib_ref] and have been identified in the cell walls of higher oomycetes such as P. ramorum [bib_ref] Identification of cell wall-associated proteins from Phytophthora ramorum, Meijer [/bib_ref] and P. infestans [bib_ref] Identification of appressorial and mycelial cell wall proteins and a survey of..., Grenville-Briggs [/bib_ref]. We identified three gene fragments (contig1838, contig403969 and contig500361) with similarity to mucins or mucin-like sequences in the pathogen dataset [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. We also identified a unigene sequence, contig500361, from the dissected library with a fibronectin type-3 domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. Contig500361 is missing an initial start codon, but the entire protein fragment is predicted to be on the outside of the cell, anchored to the cell membrane, using both the new Eukaryotic subcellular predictor, Euk-mPLoc 2.0, [bib_ref] A New Method for Predicting the Subcellular Localization of Eukaryotic Proteins with..., Chou [/bib_ref] and TMHMM [bib_ref] Predicting transmembrane protein topology with a hidden Markov model: Application to complete..., Krogh [/bib_ref]. Contig500361 is not predicted to contain a signal sequence for secretion, according to SignalPhowever it is likely that the N-terminal section of the protein is missing from Contig500361. Fibronectin is a high molecular weight extracellular matrix protein with diverse functions including roles in adhesion, growth, differentiation and wound healing [bib_ref] New insights into form and function of fibronectin splice variants, White [/bib_ref]. Contig500361 is, therefore, likely to be part of a large surface protein in Eu. dicksonii.
Host and pathogen cell wall interactions. Higher oomycetes secrete cell wall degrading enzymes, to aid in penetration of host cells [bib_ref] Characterisation of 1,3-beta-glucanase and 1,3;1,4-beta-glucanase genes from Phytophthora infestans, Mcleod [/bib_ref]. Since Eu. dicksonii penetrates the host algal cell to initiate infection, we looked for cell wall degrading enzymes within the Eu. dicksonii unigenes, as well as for algal genes that may be involved in strengthening or remodelling the host cell wall to prevent penetration.
An Eu. dicksonii unigene had similarity to genes involved in the degradation of alginates, which are a key component of the brown algal cell wall matrix [bib_ref] Structure of the cell walls of marine algae and ecophysiological functions of..., Kloareg [/bib_ref]. Contig500758 was predicted by InterproScan to contain a full alginate lyase 2 domain and has weak similarity to a 222 aa alginate lyase from Pseudomonas syringae pv. phaseolicola [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. Phylogenetic analysis showed that it also had similarity to alginate lyase sequences from Chlorella virus and algal gastropods . However, it is likely that this gene is not full length, and therefore the prediction of signal peptides, specific substrates or catalytic residues is speculative without first obtaining the full length gene sequence. Interestingly, the Hyaloperonospora arabidopsidis, Saprolegnia parasitica, Phytophthora infestans, Phytophthora ramorum and Phytophthora sojae genomes do not contain homologues of Contig500758, indicating that this may be uniquely produced for the degradation of the algal host wall by Eu. dicksonii, and thus directly relate to its host spectrum. Con-tig500758 has similarity to alginate lyases from algal grazers (Haliotis species) and fungi and groups within a clade of alginate lyases from these organisms .
We also identified a mannuronan C5-epimerase, the enzyme that catalyses the final step in alginate biosynthesis in the E. siliculosus sequences (Contig290, E. siliculosus Esi0495_0002). In kelps, C5 epimerases belong to a small family, with some isoforms induced in protoplast and elicitor-treated cultures [bib_ref] Identification of stress genes from Laminaria digitata (Phaeophyceae) protoplast cultures by expressed..., Roeder [/bib_ref] [bib_ref] Expression profiling of the mannuronan C5-epimerase multigenic family in the brown alga..., Tonon [/bib_ref]. The representation of the otherwise lowly expressed Esi0495_0002 gene [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] in our dataset suggests that alginate biosynthesis may be important to strengthen the host cell wall as a response to Eu. dicksonii infection. However, none of the other Ectocarpus enzymes involved in alginate biosynthesis [bib_ref] The cell wall polysaccharide metabolism of the brown alga Ectocarpus siliculosus. Insights..., Michel [/bib_ref] were represented in our dataset.
Although not full length, two additional Eu. dicksonii contigs encode glucanases that may be involved in breakdown of host cell wall glucans or cellulose. The predicted protein sequence of Contig500983 contains a partial exo-beta-1,3-glucanase domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. It has similarity to glucan 1,3 beta-glycosidase and endo-1,3-beta glucanase from P. infestans, both of which are secreted proteins. Contig403087 is similar to the secreted putative exo-1,3-beta-glucanase from P. infestans [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref].
There are 9 putative cellulose synthase genes in E. siliculosus [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. We identified three of these in our host dataset, suggesting a role for host cellulose synthesis during infection by Eu. dicksonii. Contig1852 mapped to host gene Esi0004_0105, Contig404144 mapped to host gene Esi0185_0053, whilst Contig301614 and Contig1623 mapped to the 39 untranslated region of Esi0231_0017 [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref].
Eu. dicksonii thalli are first formed in a non-walled state, and then mature to walled thalli as infection progresses [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref]. Therefore we expect cell wall biosynthesis to be an important part of the infection process of this oomycete. Oomycetes have traditionally been described as cellulosic organisms, however, chitin and chitosaccharides have been identified in the cell walls of several oomycetes, as have chitin synthases [bib_ref] Characterisation of chitin and chitin synthase from the cellulosic cell wall fungus..., Bulone [/bib_ref] [bib_ref] Cell wall chitosaccharides are essential components and exposed patterns of the phytopathogenic..., Badreddine [/bib_ref]. We were not able to identify genes with similarity to cellulose synthases in Eu. dicksonii. However, we did identify a putative chitin synthase, namely Contig500448, which displays significant similarity to the chitin synthase 1 gene from the oomycete Saprolegnia monoica [bib_ref] CHS2 a chitin synthase gene from the oomycete Saprolegnia monoica, Mort-Bontemps [/bib_ref] [bib_ref] Chitin synthases from Saprolegnia are involved in tip growth and represent a..., Guerriero [/bib_ref] [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref].
Brown algal cell walls also contain sulfated fucans (fucoidans). Possible enzymes in the biochemical pathway for the synthesis and sulfation of fucans have recently been identified in the E. siliculosus genome [bib_ref] The cell wall polysaccharide metabolism of the brown alga Ectocarpus siliculosus. Insights..., Michel [/bib_ref] , however, we were unable to identify any of these enzymes in our E. siliculosus dataset. At least two kinds of glycosidases act on fucans, flucan sulfate hydrolases and fucoidonase [bib_ref] Sulfated fucans, fresh perspectives: structures, functions, and biological properties of sulfated fucans..., Berteau [/bib_ref] however we were unable to detect pathogen transcripts with similarity to these enzymes in our dataset.
Searching for known oomycete effector protein families. All oomycete avirulence genes cloned to date contain a signal peptide for secretion and the RXLR amino acid motif at the N terminus of the protein. Hundreds of effector molecules containing this motif have been predicted in the genomes of sequenced oomycetes [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref]. We therefore mined our dataset for RXLR-like sequences. We were unable to unambiguously identify transcripts that fulfill the criteria for putative RXLR effectors, as identified by [bib_ref] A translocation signal for delivery of oomycete effector proteins into host plant..., Whisson [/bib_ref] and [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref].
The Crinkler (CRN) protein family elicits host responses in P. infestans host interactions, and contain an LFLAK motif. These effector proteins have been identified in all Phytophthora species sequenced to date [bib_ref] Phytophthora Genome Sequences Uncover Evolutionary Origins and Mechanisms of, Tyler [/bib_ref] [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref] and in the legume pathogen Aphanomyces euteiches [bib_ref] Transcriptome of Aphanomyces euteiches: new oomycete putative pathogenicity factors and metabolic pathways, Gaulin [/bib_ref]. A group of similar proteins has also been identified in the Pythium ultimum genome, with a related, but divergent, motif [bib_ref] Genome sequence of the necrotrophic plant pathogen, Pythium ultimum, reveals original pathogenicity..., Lévesque [/bib_ref]. It is therefore possible that CRN proteins are ancestral effector proteins present throughout the oomycete lineage. However, we were unable to identify CRN proteins in our dataset. Thus, complete transcriptome or full genome sequencing will be needed to prove or disprove the absence of RXLR and CRN effectors in Eu. dicksonii.
## Genes encoding other potential pathogenicity factors in
Eu. Dicksonii. A vast array of pathogenicity factors and potential effector molecules, which may be involved during infection, or which trigger host defences, have now been identified in higher oomycete pathogens [bib_ref] Molecular genetics of pathogenic oomycetes, Kamoun [/bib_ref] [bib_ref] Recent developments in effector biology of filamentous plant pathogens, Oliva [/bib_ref]. Pathogenicity determinants are, either, presented at the cell surface, or secreted and/or actively transported into the host cell, to manipulate the host. We therefore, mined our dataset for potentially secreted proteins with similarity to known effectors or pathogenicity factors.
Contig400638 encodes a protein with a predicted protein tyrosine phosphatase-like (PTPLA) domain and a signal peptide for secretion. It has significant similarity to a conserved hypothetic protein from P. infestans and to a protein tyrosine phosphatase-like protein from Rattus norvegicus [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. PTPLA proteins are key regulatory proteins, involved in regulating signal transduction by removal of phosphate from tyrosine residues in proteins such as MAP kinases. It is, therefore, possible that Eu. dicksonii secretes a PTPLA protein to interfere with host defence signal transduction, perhaps blocking transduction of host signaling that would lead to an algal defence response.
Contig500988 encodes a protein with a signal peptide and a histone deacetylase domain. The predicted protein has significant similarity to histone deacetylases from protozoan parasites [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. Histone deacetylases may be involved in transcriptional regulation by removal of acetyl groups from the lysine residues of histones.
Sequences derived from Eu. dicksonii with similarity to cyclophilins/immunophilins, which may also be involved in pathogenicity were also identified [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref].
Eu. dicksonii genes involved in growth regulation and programmed cell death. Several Eu. dicksonii sequences, which are not predicted to be secreted, or which are not full-length, are similar to genes involved in growth inhibition or programmed cell death in other models. Singlet 8YN09FM1 encodes a protein fragment which contains a PHD domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. PHD folds into an interleaved type of Zn-finger chelating Zn ions in a similar manner to RING-finger domains. Several PHD-finger proteins have now been identified that bind modules of methylated histone H3 and thereby inhibit transcription [bib_ref] Molecular basis for site-specific readout of histone H3K4Me3 by the BPTF PHD..., Li [/bib_ref] [bib_ref] Molecular mechanism of histone H3K4me3 regulation by plant homeodomain of ING2, Peñ A [/bib_ref] [bib_ref] YNGI PHD finger binding to H3 trimethylated at K4 promotes NuA3 HAT..., Taverna [/bib_ref]. 8YN09FM1 also contains an inhibitor of growth domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref] , which binds chromatin and acts as a transcription regulator. 8YN09FM1 does not encode a full-length protein, and does not contain the predicted start of the protein. It is therefore not possible to determine if this protein is secreted to interfere with host growth during infection, or is targeted to regulate Eu. dicksonii growth, temporarily silencing genes possibly as a stealth mechanism, upon onset of infection. Alternately, this could be the product of a generalized stress response in Eu. dicksonii.
Contig400862 and contig402793 also encode fragments of Eu. dicksonii translation factor(s) possibly involved in programmed cell death [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. Neither contig is full-length and both are missing the predicted start of the gene(s), so it is not possible to predict the presence of signal peptides for secretion. The similar P. infestans protein (XP_002899252; PITG_14133) does not contain a signal peptide. Therefore Contig_400862 and contig_402793 may also represent oomycete gene(s) targeted internally rather than to the host.
Insights into pathogen physiology, stress responses and metabolism. We identified a unigene encoding a protein fragment predicted to be an hypoxia-induced protein.
Contig_404084 from Eu. dicksonii was similar to HIGi hypoxiainducible domain family member 2A from Xenopus laevis and the . A Eu. dicksonii candidate pathogenicity effector has similarity with alginate lyases of brown algal grazers and fungi. The conserved amino acid sequence region between eukaryotic, bacterial and viral alginate lyases present in the NCBI non-redundant database and the predicted protein sequence of the Eu. dicksonii gene fragment Contig500758 were aligned using MUSCLE. A dendrogram was produced using Geneious 4.8 with a neighbour-joining algorithm and 1000 bootstrap iterations. Branches with less than 50% bootstrap support were collapsed. Genbank accession numbers of the sequences used in the alignment are indicated on the tree. The Eu. dicksonii sequence groups with a clade of eukaryotic alginate lyases from fungi and abalone (Haliotis sp.), highlighted in blue. The latter also contains sequences from a green algal virus, thought to have been acquired by horizontal gene transfer. doi:10.1371/journal.pone.0024500.g006 predicted protein fragment contained two transmembrane domains and an HIG_1_N response to hypoxia domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. We also identified three sequences with thioredoxin-like domains within our pathogen dataset [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. Several gene fragments with functional domains classified as transporters were also identified in Eu. dicksonii [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. None of these sequences were full length, or predicted to contain signal peptides, however they may play an important role in the uptake of nutrients from the host.
One pathogen sequence, Contig_500199, predicted to encode a secreted lipase was identified in the dissected library, indicating host lipids may be an important nutrient source for Eu. dicksonii [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref]. We identified 13 sequences, which we were unable to determine as secreted (1.3% of those with functional domains; 0.4% of total dataset) of Eu. dicksonii origin, that are involved in lipid metabolism, and 29 (2.3% of those with a functional domain and 1% of total dataset) involved in amino acid biosynthesis and metabolism , highlighting the importance of these processes in oomycete metabolism as previously reported in higher oomycetes [bib_ref] Cell wall chemistry, morphogenesis and taxonomy of fungi, Bartnicki-Garcia [/bib_ref] [bib_ref] Elevated Amino Acid Biosynthesis in Phytophthora infestans During Appressorium Formation and Potato..., Grenville-Briggs [/bib_ref].
Host responses to Eu. Dicksonii. Contig1359 encodes the full-length putative protein inhibitor Esi0079_0059 from E. siliculosus. Esi0079_0059 contains a proteinase inhibitor I13 domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref] and is predicted to exhibit serine-type endopeptidase inhibitor activity. This protein also contains a signal peptide for secretion from host cells. Esi0079_0059, is therefore, likely to be secreted as a defence response to protect host proteins from Eu. dicksonii secreted peptidases.
Contig303625 encodes the fragment of an E. siliculosus nuclear movement domain protein (Esi0000_0481). This protein contains a p23_NUDC_like domain [fig_ref] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly [/fig_ref] , required for movement of the nucleus of other eukaryote species. Interestingly, the expression of this unigene correlates with a close association between the host nucleus and the infecting Eu. dicksonii thallus. The latter is observed at the earliest detectable stage of infection, suggesting an early migration of the host nucleus towards the infection site [fig_ref] Figure 1: Co-localisation of E [/fig_ref].
Contig2091 encodes an E. siliculosus protein (Esi0035_0036) currently un-annotated in the genome database. This protein contains a GRIM_19 domain, which promotes cell death via apoptosis in animals.
Expression of an E. siliculosus viral sequence. Intriguingly, Contig_403835 encodes a fragment of Esi0052_0171 a gene that belongs to the lysogenic virus inserted into the Ectocarpus genome. However, a comprehensive transcriptomic analysis reported by [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] revealed that the inserted viral genome is transcriptionally silent throughout the life cycle of Ectocarpus, even following the application of several stress treatments (hyperosmotic, hypoosmotic and oxidative stress). Hence, this result suggests that although lysogeny has yet to be observed in the E siliculosus strain used in the present study, the expression of inserted viral genes may still be triggered by infection from another pathogen. Unfortunately, the gene encoded by Esi0052_0171 (EsV-1-191) does not contain any conserved domains or have similarity to any functionally characterised protein, giving little clue as to what its function could be.
# Discussion
In this study we describe an extensive characterisation of an EST collection of the brown alga E. siliculosus infected with the marine oomycete Eurychasma dicksonii. Given its wide host range and worldwide distribution, Eu. dicksonii is likely to be important to the ecology of coastal habitats where brown algae are predominant primary producers. Furthermore, Eu. dicksonii belongs to the most early branching clade within the oomycete lineage, and is therefore of considerable interest to decipher the origin and evolution of pathogenicty in this group. This study represents the first large scale sequencing study undertaken outside the best studied oomycete orders (namely Pythiales, Peronosporales and Saprolegniales). The latter, however, only represent a fraction of the lineage's diversity. Thus, and perhaps not so surprisingly, only 1314 (42% of a total 3074) pathogen unigenes shared sequence similarity with known oomycete genes. 1760 of the pathogen unigenes had no similarity to any known sequence, and no defining protein domains with which to predict function. In a similar study of the Aphanomyces euteiches transcriptome, Gaulin et al [bib_ref] Transcriptome of Aphanomyces euteiches: new oomycete putative pathogenicity factors and metabolic pathways, Gaulin [/bib_ref] found that approximately 70% of EST sequences showed similarity to previously described genes in the NCBI nonredundant protein database and around 80% showed similarity to Phytophthora proteins. Within the sequences we identified of Eu. dicksonii origin, only a maximum of 28% showed similarity to previously described proteins. Therefore, our Eu dicksonii gene dataset is largely unique and indicates how far away we are from identifying the full repertoire of oomycete genes, despite the availability of several genomes from plant and fish pathogenic species. Additionally, it is clear from this study that Eu. dicksonii shares some genetic similarity with higher order oomycetes, but is in fact very unique genetically, in comparison to these other sequenced oomycetes.
Making use of the recently completed E. siliculosus genome [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref] , we were able to unambiguously map 6787 of our unigenes to the algal host. Of these sequences, only approximately 20% mapped to protein coding regions. E. siliculosus contains large 39UTR regions and mapping of EST datasets generated in previous projects has resulted in large proportions mapping to these 39 UTRS [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. Therefore the result that up to 80% of our sequences map outside of the protein coding regions of E. siliculosus genes is consistent with previously reported data and may go part-way to explaining the low number of host sequences with significant similarity in the NCBI nr protein database.
The unusually high level of expression of a broad range of TEs in E. siliculosus is also fully consistent the observations reported by the Ectocarpus Genome Initiative [bib_ref] The Ectocarpus genome and the independent evolution of multicellularity in the brown..., Cock [/bib_ref]. Importantly however, the most highly expressed TEs in our libraries only partially overlap those identified in this earlier work. Our results suggest an environmental, most probably stress-dependent, transcriptional regulation of TEs in E.siliculosus We conclude that, whilst methylation has not been detected in E. siliculosus, TE expression is probably tightly regulated in the genome by an as yet uncharacterised mechanism. Finally, the fact that the retroelements RTE2, RTE3 and RTE4 are the most highly induced in infected libraries and among the most abundant repeats in the E. siliculosus genome points to a stress-induced transposing activity, which remains to be investigated.
We found a GC content in host sequences of 51.9% consistent with the E. siliculosus genome GC content of 53.6%. However, analysis of our Eu. dicksonii dataset revealed a much lower GC content, at 40.5%. This figure appears to be representative of real protein encoding Eu. dicksonii genes sequenced so far, and is not a result of poor quality or non-protein coding sequencing. This result is in stark contrast with the GC content of other sequenced oomycete genomes, which is approximately 58% GC [bib_ref] Phytophthora Genome Sequences Uncover Evolutionary Origins and Mechanisms of, Tyler [/bib_ref] [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref] [bib_ref] Genome sequence of the necrotrophic plant pathogen, Pythium ultimum, reveals original pathogenicity..., Lévesque [/bib_ref] [bib_ref] Signatures of Adaptation to Obligate Biotrophy in the Hyaloperonospora arabidopsidis, Baxter [/bib_ref]. A breadth of possible explanations, including pathogenic lifestyle, have been formulated to account for low genome GC content [bib_ref] Mutational Patterns Cannot Explain Genome Composition: Are There Any Neutral Sites in..., Rocha [/bib_ref]. Hence, the biological significance of the GC-impoverished Eu. dicksonii transcriptome remains unclear . Interestingly, it is known that unlike other oomycetes, Eu. dicksonii uses a previously reported stop codon to encode tryptophan in mitochondrial genes [bib_ref] The development, ultrastructural cytology and molecular phylogeny of the basal oomycete Eurychasma..., Sekimoto [/bib_ref]. This bias towards the use of UGA, rather than UGG in mitochondria is consistent with the widely held view that AT-rich genomes are prone to Trp-UGA codon reassignment.
Several hundred candidate effector molecules carrying a RXLR or CRN motif have now been identified in the genomes of oomycetes pathogenic on plants [bib_ref] Molecular genetics of pathogenic oomycetes, Kamoun [/bib_ref] [bib_ref] Genome sequence analysis of the Irish potato famine pathogen Phytophthora infestans, Haas [/bib_ref] and one putative RXLReffector was found in Saprolegnia parasitica, which is a pathogen of fish [bib_ref] The putative RxLR effector protein SpHtp1 from the fish pathogenic oomycete Saprolegnia..., Van West [/bib_ref]. It is widely assumed that the RXLR motif mediates the translocation of pathogenicity effectors into host cells [bib_ref] A translocation signal for delivery of oomycete effector proteins into host plant..., Whisson [/bib_ref] [bib_ref] Towards an understanding of how RxLR effector proteins are translocated from oomycetes..., Grouffaud [/bib_ref]. It is also functionally interchangeable with the PEXEL translocation signal of apicomplexa pathogens [bib_ref] The malarial host-targeting signal is conserved in the Irish potato famine pathogen, Bhattacharjee [/bib_ref] [bib_ref] Plasmodium falciparum and Hyaloperonospora parasitica effector translocation motifs are functional in Phytophthora..., Grouffaud [/bib_ref]. Despite Eu. dicksonii being an obligate intracellular biotrophic pathogen, we were unable to unambiguously identify oomycete effector molecules in our dataset. Thus, it is possible that Eu. dicksonii does not contain RXLR proteins, and uses an alternate signal for translocation of effectors into host cells. Alternately, since the current study does not represent a complete Eu. dicksonii transcriptome, and since many of our sequences are not full length, RXLR effectors may be present, but unidentifiable using the current dataset. Furthermore, if such effector sequences are expressed at low levels we may not pick them up in EST libraries derived from infected host material. Full transcriptome or genome sequencing of Eu. dicksonii is therefore required to comprehensively search for the ancient origins of RXLR, CRN or other effectors within the oomycete lineage.
Whilst RXLR or CRN-type effectors were not identified in this study, we were able to identify sequences as potential pathogenicity factors, including both genes not previously described as oomycete pathogenicity determinants, and genes with similarity to pathogenicity factors from oomycetes and other organisms.
A novel putative pathogenicity factor is the protein encoded by Contig500758, a predicted alginate lyase. This sequence does not have homologues in sequenced oomycete genomes and appears to be most similar to alginate lyases from fungi and brown algal grazers such as abalone (Haliotis ssp., . Since alginates are the major structural component of the brown algal cell wall, we hypothesize that this Eu. dicksonii protein may be a major pathogenicity determinant. Several other sequences with similarity to cell wall degrading enzymes were also identified in the pathogen dataset. It has been reported that Eu. dicksonii has an extremely wide host range, infecting all brown algae tested so far [bib_ref] Infection experiments reveal broad host ranges of Eurychasma dicksonii (Oomycota) and Chytridium..., Müller [/bib_ref]. It is possible that this infection is achieved primarily through the pathogen's ability to degrade host alginate, and other cell wall structural components, and thereby enter the host cell. Several other proteins involved in production or degradation of both host and pathogen cell walls were also identified in this study, indicating the importance of the cell wall in oomycete-host interactions. Enzymes involved in alginate and cellulose biosynthesis were identified in the host sequence set, whilst a putative chitin synthase was identified in Eu. dicksonii. This suggests that chitin or chitosaccharides may be ancient and important components of the oomycete cell wall that may have either been lost in higher oomycetes, or which fulfil subtle functions in structuring the cell wall.
We identified three E. siliculosus putative cellulose synthases in our host dataset. Evidence for secreted cellulose degrading enzymes in Eu. dicksonii was not as conclusive. It is possible that E. siliculosus strengthens the cell wall using cellulose in response to pathogen attack and degradation of alginates. Esi0185_0053 has a CesA_CelA-like domain, and Esi0231_0017 has a cellulose synthase domain, along with a glycosyl transferase GTA family domain. However, Esi0004_0105 only contains a glycosyl transferase GTA family domain. Since the annotation of these genes is purely based on bioinformatic analysis, their exact enzymatic function is not known. Therefore, it is possible that at least one of these genes could function in the production of the related defence molecule callose, rather than cellulose.
Rearrangement of actin microfilaments to focus on the infection site, and movement of the host nucleus to the site of attack is a well-documented feature of plant-oomycete and plant-fungal interactions (reviewed in [bib_ref] Update: The cytoskeleton as a regulator and target of biotic interactions in..., Takemoto [/bib_ref]. However, little is known of the brown algal response to infection. In the present study, we identified an E. siliculosus candidate nuclear movement protein within our dataset. Furthermore, the Eu. dicksonii thallus is always observed in close proximity to the host nucleus and Gogli, ; [fig_ref] Figure 1: Co-localisation of E [/fig_ref] this study). This association between Eu. dicksonii and its host nucleus is already established at the earliest detectable stages of infection, and is conserved across the brown algal genera Ectocarpus, Macrocystis and Pylaiella. These observations suggest that migration of the host nucleus towards the parasitic thallus is a physiologically-relevant feature underpinning the infection process.
Further host responses to infection by Eu. dicksonii might be mediated by the secretion of a proteinase inhibitor, presumably targeted towards proteinases produced by the pathogen. These findings highlight the power of the production of EST libraries from infected host material to build up a picture of the dynamics of the interactions between E. siliculosus and Eu. dicksonii during infection. However, since many of the oomycete sequences identified in this study do not show similarity to previously described genes, future challenges include functional characterisation of such genes and identification of further pathogenicity determinants in this organism.
## Supporting information
[fig] Figure 1: Co-localisation of E. siliculosus nuclei and Eu. dicksonii thalli. A, B, C: Successive stages of Ectocarpus infection by Eurychasma dicksonii. The Eu. dicksonii thallus develops as a spherical intracellular syncitium (A, arrows), progressively filling any individually infected algal cell, ultimately causing its hypertrophy (B, arrow). The syncitium then differentiates into a sporangium (C, arrows, Congo red staining) that releases new infectious spores into the medium via its apical apertures (C, arrowheads). The original infectious spore at the surface of the infected Ectocarpus cell is visible in B (arrowhead). The structure designated with a brace in C is an algal plurilocular sporangium containing parthenogenetic zoospores.Bars: A, B: 20 mm; C: 50 mm. D. Eu. dicksonii syncyitium (3 nuclei visible, arrowheads) developing next to the nucleus of a highly vacuolated E. siliculosus cell (strain CCAP 1310/299). Picture: courtesy of Dr S. Sekimoto. E & F. DAPI staining (E) and corresponding Nomarski image (F) of the microscopic sexual development stage (gametophyte) of the kelp Macrocystis pyrifera infected with Eurychasma dicksonii. The left hand side algal cell contains a very young Eurychasma thallus (one nucleus visible, white arrowhead) derived from the protoplasm of the empty spore visible at its surface. Insert: merged images. Scale bar: 10 um. doi:10.1371/journal.pone.0024500.g001 [/fig]
[fig] Figure 2: Venn Diagram showing numbers of overlapping sequences from each library within the total dataset. doi:10.1371/journal.pone.0024500.g002 [/fig]
[fig] Figure 3: Over-representation of Ectocarpus transposable elements (TEs) in Eurychasma-infected libraries. A. Relative expression of Ectocarpus TEs in the pyrosequenced ''young'' and ''old'' Eurychasma-infected libraries. Total read counts were normalised over contig length in order to identify the most expressed unigenes in each library. Expression values are given for all TEs belonging to the top 150 most highly expressed unigenes in at least one library. If several contigs matched the same TE in a given library, their relative expression levels were summed. TE nomenclature is as perMaumus (2009). B. Differential expression pattern of Ectocarpus TEs between the pyrosequenced ''young'' and ''old'' Eurychasma-infected libraries vs. the Ectocarpus Genome Initiative EST collection (''EGI dataset'',Cock et al, 2010). For each library, sequence read counts were normalised over the genome coverage of each TE in order to control for the leaky transcription hypothesis. The 5 most highly expressed TEs in the EGI dataset are highlighted in orange; those circled in red belong to the 10 most abundant repeated elements identified in the Ectocarpus genome. Arrows point to TEs highly induced in the ''young'' and ''old'' infected cultures. doi:10.1371/journal.pone.0024500.g003 [/fig]
[fig] Figure 5: Comparison of Eu. dicksonii sequences to the non-redundant protein database at NCBI and proteomes of fully sequenced organisms. Unigenes were blasted using the BLASTX algorithm to the nr database and to the proteomes of Phytophthora infestans (Pinf), Phytophthora ramorum (Pram), Phytophthora sojae (Psoj), Pythium ultimum (Pult), Saprolegnia parasitica (Sapa), Hyaloperonospora arabidopsidis (Hpa), Ectocarpus siliculosus (Ectsi), Thalassisosira pseudonana (Thaps), Phaeodactylum triconutum (Phatr), Emiliana huxleyi (Emilhu), Plasmodium falciparum (Plafa), Arabidopsis thaliana (Atha), Naegleria gruberi (Naegr), Batrachochytrium dendrobatidis (Batde) and Mycosphaerella fijiensis (Mycfi). For each Evalue class the percentage of unigenes showing similarity is indicated. doi:10.1371/journal.pone.0024500.g005 [/fig]
[fig] Figure S1: Ectocarpus assembled contigs. (TXT) Figure S2 Eurychasma assembled contigs. (TXT) [/fig]
[table] Table 1: Assembly statistics, individual cDNA libraries and final hybrid assembly. [/table]
[table] Table 2: Characteristics of host and pathogen datasets. [/table]
[table] Table 3: Top 10 most highly expressed unigenes in the Eu. dicksonii-enriched microdissected library. [/table]
[table] Table 4: Predicted full-length secreted Eu. dicksonii proteins of unknown function. [/table]
[table] Table S1: Sequences providing insights into Eu. dicksonii infection of E. siliculosus. (XLS) [/table]
|
Off-Label Use of Chloroquine and Hydroxychloroquine for COVID-19 Treatment in Africa Against WHO Recommendation
COVID-19 is continuing as a big challenge for the globe and several types of research are continued to find safe and effective treatment and preventive options. Although
# Introduction
## Chloroquine and hydroxychloroquine
Chloroquine and its derivative hydroxychloroquine are old drugs used for the treatment of malaria and inflammatory diseases, and even have potential chemosensitization and radiosensitization activities. Chloroquine was discovered in 1934 by a German scientist called Bayer. [bib_ref] Quinoline compound, and process of making the same, Andersag [/bib_ref] Although the mechanism continues to be not well understood, it has shown to inhibit the parasitic enzyme heme polymerase that converts the toxic heme into non-toxic hemozoin, thereby leading to the buildup of toxic heme within the parasite. These drugs may also interfere with the biosynthesis of parasitic nucleic acids. Hydroxychloroquine was first synthesized in 1945 by introducing a hydroxyl group into chloroquine. Animal studies showed that hydroxychloroquine is much less (around 40%) toxic than chloroquine. [bib_ref] Animal toxicity and pharmacokinetics of hydroxychloroquine sulfate, Mattson [/bib_ref] Hydroxychloroquine was approved for use in 1955 for the treatment of malaria and other several diseases including arthritis and systemic lupus erythematosus. [bib_ref] Antimalarial agents: chloroquine, hydroxychloroquine, and quinacrine, Tanenbaum [/bib_ref] Chloroquine is a white or slightly yellow, odorless, and crystalline powder with a bitter taste substance. The melting point of this drug is between 87 to 92°C. It is very slightly soluble in water, but very soluble in chloroform, ether and dilutes acids. Chemically, it is 7-chloro-4-(4-diethylamino-1-methyl butyl amino) quinoline, or N4-(7-chloro-4-quinolinyl)-N1-N1-diethyl-1,4-pentanediamine as shown in . The chemical structure of hydroxychloroquine is drawn by introducing a hydroxyl group into chloroquine as shown in .
Chloroquine and hydroxychloroquine have serious adverse events. They potentially cause heart rhythm problems, and this could be exacerbated if the treatment is combined with other medicines, such as the antibiotic azithromycin, that have similar effects on the heart. Also, hypoglycemia, neuropsychiatric effects (such as agitation, confusion, hallucinations, and paranoia), interactions with other drugs, metabolic variability are some of the problems related to these drugs. The overdose of chloroquine and hydroxychloroquine are highly toxic and can cause seizures, coma, and cardiac arrest. [bib_ref] Safety considerations with chloroquine, hydroxychloroquine, and azithromycin in the management of SARS-CoV-2..., Juurlink [/bib_ref] Chloroquine and hydroxychloroquine, alone or in combination with azithromycin, have been highly advertised as possible therapies for COVID-19.Chloroquine phosphate stood out among possible personalized pharmacological therapies for COVID-19 due to the antiviral effect confirmed in preclinical studies and its high specificity towards the SARS-CoV-2 receptor: angiotensin-converting enzyme 2 (ACE-2). Large dissemination within lung tissue (the major organ affected in COVID-, and the consequences described in case series of severely ill COVID-19 patients with promising clinical outcomes after being treated with chloroquine have recommended its possible use as an appropriate pharmacological option. [bib_ref] Potential negative effects of the free use of chloroquine to manage COVID-19..., Teherán [/bib_ref] Controversy About the Use of Chloroquine/Hydroxychloroquine for COVID-19 Treatment
The pandemic coronavirus disease-19 (COVID-19) has pushed the global healthcare system to a crisis and the world continues to search out safe and effective treatment options for COVID-19. [bib_ref] A systematic review of the prophylactic role of chloroquine and hydroxychloroquine in..., Shah [/bib_ref] [bib_ref] Pharmacologic treatments for coronavirus disease 2019 (COVID-19): a review, Sanders [/bib_ref] Although there's no conclusive evidence of their benefit, the globe continues the utilization of hydroxychloroquine or chloroquine, often together with a second-generation macrolide for the treatment on COVID-19. [bib_ref] Coronavirus disease 2019 treatment: a review of early and emerging options, Mccreary [/bib_ref] Since in vitro studies and a preliminary clinical report suggested the efficacy of chloroquine for COVID-19-associated pneumonia, there is increasing interest in this old antimalarial drug. [bib_ref] Chloroquine for SARS-CoV-2: implications of its unique pharmacokinetic and safety properties, Smit [/bib_ref] Although generally safe when used for officially approved indications like autoimmune disorders or malaria, the risk and benefit of these drugs are poorly assessed in COVID-19. A multinational, observational world study on hospitalized patients with COVID-19 found that the employment of a regimen containing hydroxychloroquine or chloroquine (with or without a macrolide) was related to no evidence of benefit but instead was related to a rise in the risk of ventricular arrhythmias and death of patients in the hospital. This study recommended that these drug regimens should not be used unless urgent confirmatory findings from randomized clinical trials are obtained.Stronger evidence from well-designed robust randomized clinical trials is required before conclusively determining the role of chloroquine and hydroxychloroquine in the treatment of COVID-19. [bib_ref] An updated systematic review of the therapeutic role of hydroxychloroquine in Coronavirus..., Das [/bib_ref] The United States Food and Drug Administration (FDA) also struggles to play a critical role to seek out a solution for the COVID-19 pandemic. President Trump has directed the FDA to continue its effort to make sure the supply of probably safe and effective life-saving drugs to COVID-19 patients. The FDA has been working in collaboration with other government agencies and academic centers that are investigating the utilization of the drug chloroquine to see whether it is accustomed to treating patients with mild-to-severe COVID-19. FDA has investigated if the drug potentially reduces the duration of symptoms further as viral shedding, which might help prevent the spread of disease. FDA researches are continuing to work out the efficacy of using chloroquine to treat COVID-19.Although large-scale clinical and basic researches are still required to clarify its specific mechanism and to optimize the treatment plan, the FDA issued an emergency use authorization to use chloroquine or hydroxychloroquine for the treatment of COVID-19. According to the FDA, these drugs have shown good activity in laboratory studies and anecdotal reports suggest they may offer some benefit to hospitalized COVID-19 patients. Since there is no other better option currently, it is a hopeful practice to use these drugs to COVID-19. FDA also allowed hydroxychloroquine sulfate and chloroquine phosphate products to be distributed and used for hospitalized patients with COVID-19. The FDA has warned that these drugs should only be used in hospitals or for clinical trials because they have been connected to a risk of heart rhythm problems, especially when paired with the antibiotic azithromycin. [bib_ref] Covid-19: WHO halts hydroxychloroquine trial to review links with increased mortality risk, Mahase [/bib_ref] Most of the evidence to support the use of either of the drugs against the COVID-19 disease comes from a small trial in France.President Donald Trump has been also tweeting consistently that hydroxychloroquine and the antibiotic azithromycin could be one of the biggest game-changers in the history of medicine. The president has said he has been taking hydroxychloroquine to protect against the virus.The European Medicines Agency (EMA) has warned against the widespread use of hydroxychloroquine or chloroquine to treat COVID-19. The EMA noted that there are no studies with strong evidence about the efficacy of these two medicines in treating COVID-19. The agency stressed that patients and healthcare professionals only use chloroquine and hydroxychloroquine for their authorized use or as part of clinical trials or national emergency use programs for the treatment of COVID-19.According to the COVID-19 EMA pandemic task force, both drugs have been shown to cause heart rhythm problems due to electrical activity disruption. It could be highly fatal if the medication is taken at high doses or combined with other drugs, such as the antibiotic azithromycin, that have similar effects on the heart. The drugs have also toxic effects including liver and kidney problems, nerve cell damage that can lead to seizures, and hypoglycemia.The first randomized controlled clinical trial that tested chloroquine and hydroxychloroquine in Brazil, Spain, and Mozambique also concluded that these drugs are not safe when used in high doses to treat patients with severe symptoms of COVID-19. [bib_ref] Death threats after a trial on chloroquine for COVID-19, Ektorp [/bib_ref] This clinical trial was prematurely halted as the higher dosage of chloroquine diphosphate for 10 days was associated with more toxic effects and lethality. The study recommended that the higher chloroquine dosage should not be suggested for critically ill patients with COVID-19 because of its potential safety hazards, especially when taken concomitantly with azithromycin and oseltamivir.After all, the World Health Organization (WHO) temporarily stopped any clinical trials that use chloroquine and hydroxychloroquine to treat COVID-19 patients after the shocking result of one study which is published in the most famous medical journal known as The Lancet. This study reported that these drugs have much more harm than good as patients getting hydroxychloroquine were dying at higher rates than other coronavirus infected patients.Finally, on May 22, WHO Director-General Dr. Tedros Adhanom Ghebreyesus declared that the clinical trials, as well as the clinical use of hydroxychloroquine or chloroquine with or without macrolide, are halted internationally as they did not conclusively show in-hospital outcome benefits. [bib_ref] Covid-19: WHO halts hydroxychloroquine trial to review links with increased mortality risk, Mahase [/bib_ref] Although WHO and other international organizations recommend to do not use these drugs, both hydroxychloroquine and chloroquine with or without macrolide are frequently used in several African countries for the treatment of COVID-19. Many African countries have already permitted at the government level to use these drugs to treat COVID-19 patients by opposing WHO warnings. Even the regional organization such as the Economic Community of West African States (ECOWAS) has officially approved for the use of chloroquine to treat coronavirus infected patients.This improper use of anti-malarial-drugs may highly complicate the COVID-19 case management in Africa. In addition to the toxic effects, shortages and increased market prices of these drugs lead African countries vulnerable to substandard and falsified medical products due to their poor healthcare system. [bib_ref] Hydroxychloroquine for the prevention or treatment of COVID-19 in Africa: caution for..., Abena [/bib_ref] The objective of the study was to review the off-label use practice of chloroquine and hydroxychloroquine for the treatment and prevention of COVID-19 in African countries against WHO recommendations.
# Materials and methods
## Study area
The countries in the continent of Africa were chosen for this review. For this study purpose, the countries were classified as East Africa, North Africa, West Africa, Central Africa, and South Africa.
## Study design
The study aimed to include all published studies without limitations, which focused on the off-label use practice of chloroquine and hydroxychloroquine for the treatment of COVID-19 in African countries against WHO recommendation. The study also looked for commentaries, reviews, viewpoints, opinions, local and international news, and interviews. Clinical trials, medical case reports, and media reports that do not contain information about the use of chloroquine and hydroxychloroquine to treat or prevent COVID-19 were excluded. Studies that evaluated the prophylactic effects of chloroquine or hydroxychloroquine were also excluded. Also, studies that are published in predator journals are excluded from the study.
## Search strategy
Data on the use of chloroquine and hydroxychloroquine for the treatment and prevention of COVID-19 disease in African countries were sourced from online electronic databases of published scientific literature. The clinical case reports and clinical trials of these drugs related to COVID-19 were searched from credible sources including Scopus, PubMed, Hindawi, OVID, Google Scholar, EMBASE, Web of Science, ClinicalTrials.gov, WHO International Clinical Trials Registry Platform, and Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials [CENTRAL], and Cochrane Methodology Register) from the beginning of COVID-19 until 15 June 2020. Some studies were also identified with a manual Google search. The utilization pattern of these drugs in African countries was also sourced from different local and international media. Primary search terms were "COVID-19", "Chloroquine", "Hydroxychloroquine" "Offlabel", "Africa" and "WHO recommendation". These search terms were adapted for use with different bibliographic databases in combination with database-specific filters for studies, if available. The search strategy was used to obtain the titles and the abstracts of the relevant studies in English, and they were independently screened and subsequently retrieved abstracts, and if necessary, the full text of articles to determine the suitability. A flowchart of the literature search present in [fig_ref] Figure 3: Flow chart of literature search [/fig_ref].
## Appraisal of the selected articles
The papers were critically appraised to identify and select following the checklist of PRESS 2015. [bib_ref] PRESS peer review of electronic search strategies: 2015 guideline statement, Mcgowan [/bib_ref] No assumptions or simplifications were made during the process.
## Off-label use practice of chloroquine/hydroxychloroquine in african countries
Like that of other countries of the world, COVID-19 is becoming highly pandemic in most African countries. According to the latest data of John Hopkins University and Africa Center for Disease Control, there are more than 200,000 confirmed cases and 6000 deaths due to coronavirus in Africa until June 13, 2020. Despite the relatively lower number of COVID-19 cases in Africa, the pandemic remains a major threat to the continent's poor health systems.As Africa News reported that five African countries (Egypt, Zambia, Nigeria, Tunisia, and South Africa) are doing different clinical trials by the support of Africa Centres for Disease Control (ACDC) as part of the continent's step for treatment and prevention of the COVID-19. Egypt has 13 trials ongoing which specifically focus on COVID-19 therapeutics and around two trials on vaccines. Egyptian scientists are also doing nutritional support especially with honey and some immunotherapy trials. Zambia has one trial ongoing with hydroxychloroquine, an anti-malarial drug, which advertised as a potent remedy for coronavirus. South Africa is also looking at the efficacy of chloroquine, interferon, and remdesivir. Like in the case of Egypt, Nigeria has a trial in progress for different therapeutic agents. There are also two trials ongoing in Tunisia.Although there are several ongoing hopeful clinical trials for antiviral and immune-based therapies, there are no scientifically confirmed, clinically effective pharmacological treatments for COVID-19. The commonly used old antimalarial drugs, chloroquine, and hydroxychloroquine have become the attention of global scientific media and political focus regardless of a lack of randomized clinical trials supporting the drug's efficacy and safety. The efficacy, as well as the safety of both of these drugs, is still not defined by clinical trials as well as WHO for the treatment and prevention of COVID-19. [bib_ref] Hydroxychloroquine for the prevention or treatment of COVID-19 in Africa: caution for..., Abena [/bib_ref] The irrelevant and controversial promotions even by higher public officials, mainly by Trump, lead to the widespread use of chloroquine and hydroxychloroquine for COVID-19, mainly for African peoples. It leads to selftreatment, fatal overdoses, and extensive shortages of these drugs on the continent. For instance, Nigeria records several chloroquine poisonings cases after using it for COVID-19 treatment. As CNN reported from Lagos, Nigeria showed that three Nigerians in the country are critically ill in the hospital due to these drugs.The worldwide controversial reports that chloroquine and hydroxychloroquine may be effective against COVID-19 have received great attention, increasing the risk of the introduction of falsified versions of these medicines mainly in African countries due to their poor health system. For instance, five different types of falsified chloroquine tablets were discovered between March 31, 2020, and April 4, 2020, in Cameroon and the Democratic Republic of Congo by liquid chromatographic and mass spectrometric analysis. Such products represent a serious risk to patients. These events exemplify that once new medicines, regimens or vaccines against COVID-19 developed, its falsified product forms will enter the market immediately, especially in low-and middle-income countries. [bib_ref] Identification of falsified chloroquine tablets in Africa at the Time of the..., Gnegel [/bib_ref]
## Off-label use of chloroquine/ hydroxychloroquine in east africa
There are around 3305 confirmed COVID-19 cases and 96 Deaths in Kenya until June 13, 2020. South African Broadcasting Corporation (SABC) reported that Kenya approved the use of chloroquine to treat critical cases of patients with coronavirus. However, it is given only to critical patients as noted by Patrick Amoth, Kenyan Health Director-General. He, however, said that the sale of hydroxychloroquine and chloroquine over the counter at pharmacies is not allowed. The country's Pharmacy and Poisons Board has also banned over the counter sale of both hydroxychloroquine and chloroquine in pharmacies. On the contrary to this, Kenyan medical experts warn to do not use these drugs since they require further clinical study and might not be the cure of COVID-19.Djibouti is one of the African countries which is extremely infected with COVID-19. There are around 4207 COVID-19 cases per 100,000 people until June 13. However, as health officials of the country said that a death rate is just 0.5%. The top official of Djibouti's main COVID-19 response center said that giving the antimalarial drug, chloroquine, for COVID-19 patients is the main reason for the country's low death rate. The government approved to treat all COVID-19 patients with chloroquine by ignoring WHO caution. Ahmed Zouiten, the
## Dovepress
WHO representative for the Djibouti country office, told via VOA that while the WHO has no scientific evidence to support the utilization of antimalarial drugs to treat COVID-19, observationally their use alongside antibiotics does seem to be working to Djibouti's case. He believed that treating all COVID cases with antibiotics (azithromycin) and antimalarial drugs (chloroquine/hydroxychloroquine) has shown a positive impact. Dr. Maad Nasser Mohammed, a specialist in tropical infections who is chargeable for Djibouti's main COVID-19 response center at the Arta Regional Hospital, also confirmed that the Djiboutian approach of treating all positive COVID cases with antibiotics and anti-malarial drugs provides a positive result.Another VOA report by interviewing one of Djibouti's higher health officials showed that except for suffering from the loss of taste, cough, and fever, most of the patients did not experience any breathing problems after chloroquine treatment. He noted that chloroquine acts positively on COVID-19 treatment.The Global Times report also showed that both Ethiopia and Djibouti use an anti-malarial drug, chloroquine, and an antibiotic, azithromycin, to treat COVID-19 patients against WHO recommendations.Uganda recorded good results by treating COVID-19 patients with hydroxychloroquine or chloroquine. The Uganda Ministry of Health told The Independent that they got good results by using hydroxychloroquine and chloroquine in managing COVID-19 patients even though the drugs did not treat the disease. Uganda has used either of the two drugs in combination with azithromycin to manage COVID-19 patients at Entebbe General Hospital. Until June 13, 2020, 53 patients treated by these drugs are fully recovered. In addition to chloroquine and hydroxychloroquine, the patients have been given Vitamin C. Dr. Diana Atwine, secretary of the Ministry of Health, said that despite the recent findings, Uganda has scored good results from using these drugs. He also added that these drugs are not new for them and they know well about the side effects.
## Off-label use of chloroquine/ hydroxychloroquine in north africa
Egypt is the leading African country with COVID cases as well as deaths. Like many African countries, Egypt uses chloroquine in treating COVID-19 patients. As Egypt Today report by mentioning the Egyptian Minister of Health, Hala Zayed, showed that Egypt is using anti-malarial and HIV drugs in the treatment of coronavirus (Covid-19) disease. As he said that they use these medications because they get positive results in treating more cases. Zayed cautioned that these drugs should not be used haphazardly by people, and physicians are those only who have the authority to prescribe them. He finally reminded citizens to not use these medications without consulting a doctor.On the contrary to Hala Zayed, the Health Ministry spokesman, Khaled Megahed, told to Egypt Today that the country denies producing chloroquine to treat COVID-19 patients. He denied the news that a pharmaceutical company in the Egyptian capital, Cairo, started producing chloroquine phosphate medication primarily used to treat malaria patients. He said that the World Health Organization has not yet adopted any effective treatment for the novel coronavirus. Despite being seen as a good treatment for coronavirus by many people around the world, including US President Donald Trump, many clinical trials declared that both chloroquine and hydroxychloroquine are not yet confirmed to be effective for the treating of COVID-19 patients. According to the unproven announcement, the Arab Company for Drug Industry and Medical Appliances (ACDIMA) aims to produce around 200,000 doses of chloroquine by last May.According to the AFP report, Algeria is also one of the North African countries which use the anti-malarial drug (hydroxychloroquine) for COVID-19 treatment despite WHO dropping trials. One of the members of the scientific committee on the north African country's Covid-19 outbreak, Mohamed Bekkat, said that they have treated thousands of cases with this hydroxychloroquine very successfully in Algeria. He also said that they have not noted any undesirable reactions. The Health Minister, Abderrahmane Benbouzid, also said that Algeria had great success in using hydroxychloroquine in combination with azithromycin. He also noted that there are no adverse reactions among several thousand patients who have taken these drugs. Several Algerian doctors have also claimed the treatment is very effective.Morocco also continues the use of chloroquine for COVID-19 treatment despite the controversy and warnings raised by the WHO. They were treating Covid-19 patients with chloroquine since April 8 and appear to have no intention of stopping. As the Minister of Health of Morocco, Khalid Ait Taleb has confirmed to AFP that though the WHO has decided to momentarily suspend clinical trials with hydroxychloroquine, Morocco According to Ait Tayeb, this treatment can reduce the viral load very quickly. Moroccan Health authorities also started safeguarding a satisfactory stock of chloroquine by purchasing from an international pharmaceutical company just two weeks after the declaration of the first COVID-19 case in the country. In addition to this, the Health Ministry initiated chloroquine therapy along with complementary medicine to enhance efficacy.Another North African country, Tunisia, was initially excited to and even started manufacturing its hydroxychloroquine. However, Tunisia declared that the treatment with hydroxychloroquine has been stopped in all the country's hospitals following the statement from WHO.
## Off-label use of chloroquine/ hydroxychloroquine in west africa
The Economic Community of West African States (ECOWAS) has approved the use of chloroquine to treat coronavirus patients. ECOWAS supports the supplementary anti-viral treatment of coronavirus with hydroxychloroquine should be for mild forms of the virus. Meanwhile, Ghana's health minister, Kweku Agyeman-Manu, said that he has approved for the anti-malarial drug to be used for treating COVID-19 patients. He told journalists in Accra that the chloroquine tests at Ridge Hospital are still positive and the tests are continued against WHO caution. The health minister also noted that the country has enough stock to treat all COVID-patients on admission.The Ghanaian Times also reported that the government of Ghana decided to use hydroxychloroquine for COVID-19 case management. However, Ghana's COVID-19 case management team is thinking about whether to continue or stop using for COVID-19 in the country after WHO halting the clinical trials of hydroxychloroquine as a possible treatment for coronavirus. The Director-General of the Ghana Health Service (GHS), Dr. Patrick Aboagye, said to Accra that the case management team of Ghana has decided to stop using hydroxychloroquine for the treatment of COVID-19.The Ghana News Agency reported that Ghana denies chloroquine phosphate to cure coronavirus. The Director-General said that there is no strong evidence that chloroquine phosphate 250 mg is an effective treatment of COVID-19. They could therefore not make any recommendation for the use of the drug to treat the virus.The Ghana Medical Association (GMA) is also cautioning Ghanaians against the use and abuse of drugs such as chloroquine, hydroxychloroquine, and azithromycin in the treatment of Covid-19. According to GMA, these drugs can affect the function of the heart by causing abnormal heartbeat, making it deadly to patients with COVID-19.The report of the Anadolu Agency showed that Nigeria goes into clinical trials with hydroxychloroquine. The director of Nigeria's National Agency for Food and Drug Administration and Control (NAFDAC) said that they believe in hydroxychloroquine and they will continue clinical trials of this drug on COVID-19 patients, despite the WHO warning. The director told that they are confused by the data of WHO reports to chloroquine whether it has from the Caucasian population or the African population. The director was also quoted by Nigerian local media as saying If the information they're gazing and also the reason for suspending the trials is from the Caucasian population, then it's going to be justified. However, we do not think they got data from the African population yet because our genetic make-up is different.
She believed that if medical doctors, research scientists, pharmacists, and herbal experts work together, the clinical trials for this drug will be finished in 3-4 months. "The narrative might change later, but for now, we believe in hydroxychloroquine" she added.However, the NAFDAC has warned Nigerians against taking self-medication of chloroquine for treatment for coronavirus . Nobody should buy chloroquine and use it. Taking of chloroquine requires guidance and supervision to protect patients from damage and death related to its overdose.The Director-General of the Nigerian Centre for Disease Control (NCDC) warning to do not use chloroquine and hydroxychloroquine for the treatment as well as for the prevention of coronavirus disease. The director noted that the use of chloroquine and hydroxychloroquine for the management of COVID-19 disease has not been validated and approved by the WHO. Furthermore, Nigerians should know that selfmedication and misuse of these drugs can cause harm and may lead to death. As the Lagos state officials told CNN, three people were hospitalized in the city after taking chloroquine. One witness also told CNN that in DovePress a pharmacy near his home in Lagos, the price of chloroquine rises by more than 400% in a matter of hours after Trump endorses it for coronavirus treatment. Finally, the Director-General of NCDC concluded that there have been promising results by researchers, but until now both drugs have not been licensed to treat COVID-19 related symptoms or prevent infection.African news reported that the Senegal government also continues treating COVID-19 patients with hydroxychloroquine despite the ineffective and potentially harmful results of the recent study. They believed that the antimalarial drugs, chloroquine, and its derivative hydroxychloroquine, have been considered cheap and safe when used correctly for COVID-19 treatment. The head of Senegal's Centre for Health Emergency Operations, Abdoulaye Bousso, told the AFP that the country's hydroxychloroquine treatment program would nonetheless continue, without offering further details about the drug toxicities as well as warnings of WHO. Seydi, the Senegalese infectious-diseases doctor, also believed that hydroxychloroquine can be taken only after offered consent to patients and after performing a cardiogram by the medical staff.The senior health official of Senegal also told the Jakarta Post that Senegal is continuing to administer hydroxychloroquine to COVID-19 patients. The official said that the results of the treatments by this drug are still very encouraging. The televised statement of Moussa, the doctor in charge of treating Senegalese COVID-19 patients, said there is hopeful evidence that people treated with hydroxychloroquine recovered faster. He told AFP that about 50% of Senegal's COVID-19 patients were being treated with hydroxychloroquine. However, he did not offer for the media any further details about the drug dosage, the exact number of patients under treatment, the age-range of the patients, or whether they had any underlying medical conditions. He said that patients under his care had exhibited "no side effects" and that he and his team are trying by combining hydroxychloroquine with azithromycin to achieve better results. 47
## Off-label use of chloroquine/ hydroxychloroquine in central africa
In Cameroon, chloroquine therapy for COVID-19 patients becomes the state protocol after the French expert, Professor Didier Raoult's recommendation. A mixture of chloroquine (an antimalarial) combined with azithromycin (an antibiotic) to avoid the risks of secondary infections and to treat patients with Covid-19 has taken very seriously for Cameron. AFP and France 24 report showed the country's scientific council and Cameroon's Ministry of Health proposed the widespread use of chloroquine treatment. Although scientists lack conclusive data for their clinical trials, Cameron believes that the treatment is essential to reduce the viral load as well as the contagiousness nature of the disease. The council wished to combine the treatment, as recommended by Raoult, with azithromycin to avoid the risks of secondary infections. As the council said that the treatment protocol was validated for the management of all types of patients with positive Covid-19 tests, from asymptomatic cases to patients suffering from severe infections.Beyond a public health resolution, the choice of dual therapy represents a logistical challenge in Cameroon. Chloroquine was massively used in the country at one time and stocks were exhausted. Dr. Etoundi, director of the fight against diseases, epidemics, and pandemics of Cameroon's Ministry of Health, explained that they had to place large orders from abroad and restart the wide national industrial production of chloroquine. Cameroon's minister of scientific research and innovation, Madeleine Tchuente, said that the government has ordered the production of chloroquine to treat COVID-19 patients and they are started producing 6000 tablets per day. As she said, the President, Paul Biya, has ordered them to produce at least eight million tablets of chloroquine each day as the situation is getting worse. The black market for drugs in normal times in Cameroon is rising with the Covid-19 crisis. Several fake chloroquine products are circulated within the Cameroon health network.
## Off-label use of chloroquine/ hydroxychloroquine in south africa
South Africa is one of the highly infected African countries by COVID-19 and they join a clinical trial on COVID-19 under WHO supervision. The South African Health Products Regulatory Authority (SAHPRA) has warned health professionals to stop prescribing the medication chloroquine to prevent infection or to treat the disease, COVID-19 outside the hospitals. This warning comes after several pharmacies in South African cities have run out of chloroquine. The national health department of the country released a new guideline that allows using chloroquine only in hospitalized patients with severe COVID-19 and for consented clinical trials. This new guideline also noted that chloroquine should be distributed to all hospitals at affordable prices.The Katharine Child report to CNN also showed that chloroquine is made available and affordable for every corner of South Africa. However, science still does not know enough to back the antimalarial drug chloroquine into widespread use against the new coronavirus. A local pharmaceutical company in Johannesburg, South Africa, has received permission from the medical regulator of the country to import half a million chloroquine phosphate tablets for use in severely ill Covid-19 patients. The largest private pharmaceutical company in the country said that they donated more than 50,000 boxes of 10 tablets to the department of health for state use in a week.The health minister of Mozambique is also considering the possibility of using chloroquine to treat severe cases of coronavirus disease. The National Director of Public Health of Mozambique, Rosa Marlene, told the African press agency (APA) that the country allowed to use chloroquine for severely affected patients and the experiment of the consented clinical trial. She told that the worldwide trends of chloroquine use showed excellent results.Covid-19 complicates the tracing and treatment of malaria and compromise the wider public health programs in Zimbabwe. World Health Organization said that Zimbabwe reported that mosquito-borne diseases are significantly spiked in the country. WHO's Global Malaria Programme said that in malaria-endemic countries, COVID-19 could make malaria more deadly. Zimbabwe's ministry of health has already reported that there are three times more malaria cases than last year. As the ministry of health told that the occurrence of these elevated malarial cases is due to the similarity of symptoms, their main symptom, fever, between COVID-19 and malaria. This makes the country's malarial management very complicated. The malaria cases are much higher than current Covid-19 cases in the country. Like other African countries, antimalarial drugs, chloroquine, and hydroxychloroquine are being used to treat Covid-19 in Zimbabwe. However, Zimbabwe's doctors advised their population to do not use chloroquine as there is insufficient data on their use for that purpose and high doses of the drugs can be lifethreatening. [bib_ref] Incidence of coronavirus disease (COVID-19) and countries affected by malarial infections, Ahmed [/bib_ref] [bib_ref] Covid-19 could complicate tracing and treatment of malaria and compromise wider public..., Ryan [/bib_ref] Generally, many African countries use chloroquine in an off-label manner for the treatment as well as for the prevention of COVID-19 against WHO recommendation as shown in [fig_ref] Table 1: African Countries Who Use Chloroquine or Hydroxychloroquine in an Off-Label Manner for... [/fig_ref].
# Conclusion and recommendations
According to this review, many African countries use chloroquine and hydroxychloroquine as both a preventive measure and for treating patients with COVID-19 regardless of WHO caution and toxicity findings of several clinical trials. Falsified and substandard chloroquine and hydroxychloroquine drugs are started to emerge in some African countries by considering the COVID-19 pandemic as a good opportunity. The health branches of the African government should assess the risks and benefits before providing these drugs to their people to treat COVID-19. African countries should strongly consider implementing prescription monitoring outlines to ensure that any offlabel chloroquine/hydroxychloroquine use is appropriate and beneficial during this pandemic. The World Health Organization should intervene in this off-label use practices of antimalarial drugs in many African countries against their licensed purpose. If the existing global efforts worthful, much work is needed in Africa by the continental collaborated and coordinated production, distribution, and post-marketing surveillance aligned to the low-cost distribution of any approved COVID-19 drug to prevent use off-label and falsified drugs.
Healthcare professionals are recommended to closely monitor patients with COVID-19 receiving chloroquine or hydroxychloroquine and to consider patients with preexisting heart problems that may make them more susceptible to heart rhythm issues. They should carefully consider the likelihood of cardiac arrhythmia, particularly with higher doses, and exercise extra caution when combining treatment with other medicines like azithromycin, which will cause similar side effects on the heart. African national drug regulatory organizations and other health authorities are advised to instantly notify WHO if falsified and substandard products are discovered in their country.
# Data sharing statement
The data used to support the findings of this study are included in the article.
# Author contributions
The author conceived, designed, analyzed the data, wrote the paper, and read and approved the final manuscript.
# Disclosure
The author has declared that they have no competing interests for this work.
## Research and reports in tropical medicine
## Dovepress
## Publish your work in this journal
Research and Reports in Tropical Medicine is an international, peerreviewed, open access journal publishing original research, case reports, editorials, reviews and commentaries on all areas of tropical medicine, including: Diseases and medicine in tropical regions; Entomology; Epidemiology; Health economics issues; Infectious disease; Laboratory science and new technology in tropical medicine;
Parasitology; Public health medicine/health care policy in tropical regions; and Microbiology. The manuscript management system is completely online and includes a very quick and fair peer-review system. Visit http://www.dovepress.com/testimonials.php to read real quotes from published authors.
[fig] Figure 1, Figure 2: The chemical structure of chloroquine. The chemical structure of hydroxychloroquine. [/fig]
[fig] Figure 3: Flow chart of literature search. Dovepress Belayneh Research and Reports in Tropical Medicine 2020:11 submit your manuscript | www.dovepress.com [/fig]
[table] Table 1: African Countries Who Use Chloroquine or Hydroxychloroquine in an Off-Label Manner for the Treatment and Prevention of COVID-19 Against WHO Recommendation [/table]
|
Inhibition of translation termination by small molecules targeting ribosomal release factors
The bacterial ribosome is an important drug target for antibiotics that can inhibit different stages of protein synthesis. Among the various classes of compounds that impair translation there are, however, no known small-molecule inhibitors that specifically target ribosomal release factors (RFs). The class I RFs are essential for correct termination of translation and they differ considerably between bacteria and eukaryotes, making them potential targets for inhibiting bacterial protein synthesis. We carried out virtual screening of a large compound library against 3D structures of free and ribosome-bound RFs in order to search for small molecules that could potentially inhibit termination by binding to the Rfs. Here, we report identification of two such compounds which are found both to bind free RFs in solution and to inhibit peptide release on the ribosome, without affecting peptide bond formation.Bacterial ribosomes with their auxiliary translation factors are major targets for drug intervention by a wide range of different antibiotics. Such compounds are often derived from natural products and inhibit different phases of protein synthesis, including initiation, peptide chain elongation and mRNA-tRNA translocation, or severely increase the error rate of the translation process 1 . Many of these antibiotics bind directly to the ribosome and thereby interfere with specific stages of protein synthesis. There are also examples, such as kirromycin and fusidic acid, which act by binding to elongation factors (EF-Tu and EF-G, respectively) and thereby blocking their release from the ribosome 2 . Among inhibitors that target the ribosome itself, many of them have been shown to bind either near the peptidyl transferase center (PTC), the mRNA decoding site or in the peptide exit tunnel 1,3 . Both the PTC, where the peptide bond formation reaction takes place, and the decoding site, where cognate aminoacyl-tRNAs are matched with the mRNA codon, are also highly conserved throughout the kingdoms of life. This makes it difficult to find molecules that selectively can block only bacterial protein synthesis. Among inhibitors of the elongation factors EF-Tu and EF-G, the only one used clinically is fusidic acid which is administered topically against skin infections 4 .Interestingly, however, no specific small molecule inhibitors of translation termination have ever been reported. Termination of bacterial protein synthesis occurs when a stop codon is presented in the ribosomal A-site and is recognized by a class I release factor, RF1 or RF2. These release factors (RFs) have different but overlapping specificities, where RF1 reads UAA and UAG and RF2 reads UAA and UGA, with strong discrimination against sense codons 5,6 . The RFs are multi-domain proteins, where binding and stop codon recognition by domain 2 at the decoding site causes the universally conserved GGQ motif of domain 3 to insert into the A-site of the PTC, some 80 Å away from the decoding site 7 . This event triggers hydrolysis of the peptidyl-tRNA bond in the P-site of the PTC, and the nascent peptide chain can then be released via the ribosomal exit tunnel(Fig. 1A). In eukaryotes and archaea, on the other hand a single omnipotent RF reads all three stop codons. Although the mechanism of translation termination is basically the same, there is neither sequence nor structural homology between the bacterial RFs and the eukaryotic eRF1, apart from the universally conserved GGQ motif. Hence, it could be expected that putative inhibitors of bacterial RFs and termination, may also selectively inhibit bacteria.The only known small molecule inhibitor that affects termination is the natural product Blasticidin S (MW = 422), which acts on both prokaryotic and eukaryotic cells and inhibits both peptidyl-tRNA hydrolysis and to a lesser extent peptide bond formation 8 . However, this drug is basically a PTC inhibitor that has been shown to deform the CCA end of the P-site tRNA 8 and it binds to the ribosomal 50S subunit in a region shared
www.nature.com/scientificreports www.nature.com/scientificreports/ with several other antibiotics [bib_ref] Structures of five antibiotics bound at the peptidyl transferase center of the..., Hansen [/bib_ref]. Hence, its mode of action is interference with the P-site tRNA rather than direct interaction with RFs. In contrast, the 18-residue antimicrobial peptide Api137 (i.e., not a small molecule) is a specific inhibitor of termination, with no effect on peptide elongation. It has recently been shown to bind in the peptide exit tunnel and establish a direct interaction with the GGQ motif of RF1 [bib_ref] An antimicrobial peptide that inhibits translation by trapping release factors on the..., Florin [/bib_ref]. The mode of action here is particularly interesting in that Api137 was shown to enter the exit tunnel after the nascent polypeptide chain has been released, but before the RF has dissociated from the ribosome, thereby trapping the RF on it. The net effect is thus that the pool of free RF1 and RF2 is depleted, causing ribosome stalling at the stop codons [bib_ref] An antimicrobial peptide that inhibits translation by trapping release factors on the..., Florin [/bib_ref]. RF1 and RF2 have very high affinity for stop codon programmed ribosomes [bib_ref] Release of peptide promoted by the GGQ motif of class 1 release..., Zavialov [/bib_ref] , with a K d ~ 10 −10 M, and their dissociation is catalyzed by the auxiliary release factor RF3. It is thus conceivable that this high intrinsic affinity could make the class I RFs vulnerable to becoming trapped on the ribosome.
We wanted to explore the possibility that small drug-like molecules might also be able to specifically inhibit translation termination by directly binding to bacterial RFs. To search for such inhibitors, we first carried out computational structure-based screening of a virtual library of commercially available compounds against 3D structures of RF2, which is the major release factor in most bacteria [bib_ref] The concentration of polypeptide-chain release factor-1 and factor-2 at different growth-rates of..., Adamski [/bib_ref]. Hit molecules from this screen were then acquired and tested for their ability to (i) bind to free RF2 in solution, (ii) inhibit peptide release from the ribosome in an in vitro termination assay and (iii) arrest bacterial cell growth. Herein, we identify and report two compounds with MW ~ 440 that fulfill these three criteria. These molecules have no effect on peptide chain elongation and are thus the first specific small molecule termination inhibitors, which may open the path for development of a new class of antibiotics.
# Results
Computational analysis of RF2 as a possible drug target. The release factors RF1 and RF2 acquire an open conformation [fig_ref] Figure 1: The bacterial 70S ribosome termination complex with RF2 [/fig_ref] on the 70S ribosome [bib_ref] Structural basis for translation termination on the 70S ribosome, Laurberg [/bib_ref] [bib_ref] Insights into translational termination from the structure of RF2 bound to the..., Weixlbaumer [/bib_ref] , which is distinctly different from the closed conformation observed in crystal structures of free RFs [bib_ref] Bacterial polypeptide release factor RF2 is structurally distinct from eukaryotic eRF1, Vestergaard [/bib_ref]. The conformational equilibrium of the free RFs in solution, as revealed by SAXS experiments shows that this open conformation is dominating at about 80%. We explored both conformations of RF2 from E. coli and T. thermophilus in search of putative binding sites with a series of molecular dynamics (MD) simulations in mixed solvents [bib_ref] Molecular simulations with solvent competition quantify water displaceability and provide accurate interaction..., Alvarez-Garcia [/bib_ref]. By using a water/ethanol solvent mixture of 80/20% mole fraction, this approach allows for identification of preferred hydrophobic and hydrophilic binding regions while also exploring protein flexibility. The MD simulations identified a putative druggable cavity between domains 1 and 4, approximately lined by the first and third α-helix of domain 1 and the C-terminal α-helix of domain 4 [fig_ref] Figure 1: The bacterial 70S ribosome termination complex with RF2 [/fig_ref]. This site was selected as the target for virtual screening of 3.4 million drug-like compounds from commercial libraries. These have been filtered with respect to the Lipinski rules, drug-like properties and chemical diversity to yield a structurally diverse selection of lead-like compounds. The virtual screening procedure involves molecular docking of this virtual compound library to a given 3D structure of the target protein. To account for protein flexibility, three representative conformations were explored in parallel by automated docking with the Glide program [bib_ref] Glide: a new approach for rapid, accurate docking and scoring. 1. method..., Friesner [/bib_ref]. These correspond to the open ribosome-bound structure [bib_ref] Crystal structure of the 70S ribosome bound with the Q253P mutant form..., Santos [/bib_ref] [bib_ref] Uniformity of peptide release is maintained by methylation of release factors, Pierson [/bib_ref] , the closed crystal structure of isolated RF2 [bib_ref] Bacterial polypeptide release factor RF2 is structurally distinct from eukaryotic eRF1, Vestergaard [/bib_ref] [bib_ref] Release Factors 2 from Escherichia coli and Thermus thermophilus: structural, spectroscopic and..., Zoldak [/bib_ref] and an intermediate conformation that was found to be stable in the MD simulations. The top 200 ranked compounds were then inspected manually and 60 of them were purchased, based on their chemical diversity and repetitive pattern of interactions within the binding site .
phenotypical screening of bacterial cell growth. To select molecules for further characterization of possible RF binding and termination inhibition, the 60 compounds were initially tested for their minimum inhibitory concentration (MIC) on bacterial growth. This was done using both wild type Staphylococcus aureus strain ATCC29213 (Gram + ) and Escherichia coli strain MG1655 (Gram − ), for which the RF2 protein shows about 45% identity. The initial screening results are given in for the eight compounds that were selected for further characterization of their potential interaction with the RFs. All of these [fig_ref] Figure 1: The bacterial 70S ribosome termination complex with RF2 [/fig_ref] and www.nature.com/scientificreports www.nature.com/scientificreports/ 161) were found to inhibit growth of S. aureus and two of them also prevented E. coli growth (118 and 129) at concentrations in the range of 3570 to 3900 mg l −1 . Each compound was analyzed at different concentrations depending on their solubility in DMSO and their propensity to precipitate in cell culture medium.
## Ligand interaction with rfs in solution.
To determine whether the eight selected compounds could bind to RF2 in vitro, we tested the compounds in a binding assay with free E. coli RF2 in solution. Each compound was titrated against 60 nM of the labeled RF2 protein using microscale thermophoresis (MST). Compounds 129, 137 and 138 did not exhibit binding in the MST assay. Compounds 102, 118 and 128 on the other hand, showed precipitation at higher concentrations of DMSO than could be tolerated by the RF2 protein (15%) and prevented us to proceed with the experiment. However, two compounds (115 and 161) changed the thermophoretic mobility and were found to bind the protein with dissociation constants (K d ) in the μM range. The K d determined for ligands 115 and 161 were 4.2 ± 1.7 μM and 37 ± 14 μM, respectively [fig_ref] Figure 2: Binding of compounds 115 and 161 to free RF1 and RF2 in... [/fig_ref]. These two molecules have a common scaffold, but diverge in terms of their substituents , and both can apparently bind to free RF2 in solution. To also compare the affinity of the two best compounds to the homologous RF1 from E. coli, we performed the same assay by titrating the compounds against 10 nM of labeled protein. Here, the K d 's were found to be 96.5 ± 36 μM and 44.6 ± 32 μM, for ligands 115 and 161, respectively [fig_ref] Figure 2: Binding of compounds 115 and 161 to free RF1 and RF2 in... [/fig_ref].
The inhibitors block turnover of RF2 but do not affect peptide bond formation. The fact that the two ligands 115 and 161 evidently can bind bacterial RFs makes it particularly interesting to examine their effect in cell-free translation systems. Based on the MST results, we therefore measured inhibition of translation termination by these two compounds using a fully reconstructed in vitro translation system with purified components from E. coli. A peptide release complex (RC) was prepared containing mRNA programmed 70S ribosomes carrying [ 3 H]fMet-tRNA fMet in the P-site and a UGA stop codon in the A-site. Addition of RF2 to this mix allows release of [ 3 H]fMet. The two compounds 115 and 161 were tested with two concentrations of RF2 and with a fixed concentration of RC. In the first single-round peptide release assay, the RF2 concentration was 10-fold higher than that of RC. Under this condition, [ 3 H]fMet from all RCs should be released at once and recycling of RF2 will not be needed. No inhibition of single-round peptide release was, however, observed even at a concentration of www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ 1 mM of either of the two compounds [fig_ref] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide... [/fig_ref]. This indicates that neither 115 nor 161 were able to inhibit the initial association of RF2 with the RC or catalysis of the peptide release reaction.
In the second multiple-round peptide release assay, the concentration of RF2 was 50-fold lower than of that RC, which means that multiple turnover of RF2 is required to release [ 3 H]fMet from all RC. Interestingly, in this case, inhibition of peptide release was observed [fig_ref] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide... [/fig_ref] and the IC 50 value for compound 115 was determined to be 47.5 μM [fig_ref] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide... [/fig_ref]. From this, the corresponding K i was estimated to 3.3 μM, which is very close to the K d value determined from MST experiments (K d = 4.2 μM). Compound 161 was found to have a higher IC 50 value of 180 μM [fig_ref] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide... [/fig_ref] , which corresponds to a K i of 12 μM, again relatively close to the K d value of 37 μM obtained from the MST measurements [fig_ref] Figure 2: Binding of compounds 115 and 161 to free RF1 and RF2 in... [/fig_ref]. These results strongly suggest that both molecules 115 and 161 bind to RF2 on the 70S ribosomal termination complex, thereby inhibiting dissociation and recycling of RF2, which is required for new rounds of translation. Their effect on RF1 catalyzed peptide release was also tested in the multi-round assay, but in that case no significant inhibition was observed [fig_ref] Figure 1: The bacterial 70S ribosome termination complex with RF2 [/fig_ref].
In order to exclude the possibility that the two compounds might not be specific inhibitors of termination, but instead general inhibitors of the PTC, we tested them also in a di-and tripeptide formation assay [bib_ref] Molecular mechanism of viomycin inhibition of peptide elongation in bacteria, Holm [/bib_ref]. In this experiment, a 70S initiation complex carrying an mRNA coding for Met-Leu-Leu-Stop, and [ 3 H]fMet-tRNA fMet in the P-site was rapidly mixed at 37 °C with an elongation mix containing the ternary complex of EF-Tu-Leu-tRNA Leu and EF-G. The compounds 115 and 161 were premixed with the 70S ribosome at varied concentrations. As a control, the solvent DMSO was tested alone. As shown in [fig_ref] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide... [/fig_ref] ,F, no inhibition of dipeptide or tripeptide formation could be seen even at the highest concentration (500 μM) of 115 and 161. This result confirms that these two compounds are not targeting ribosomes for its peptidyl transferase activity and translocation, but are specific inhibitors of RF2 turnover in the termination process.
Effect of inhibitors on cell growth. Compounds 115 and 161 were again specifically tested for their effect on bacterial growth using wt S. aureus, wt E. coli and the two E. coli mutants lpxC and ΔtolC, where minimum inhibitory concentrations (MIC) was determined. For S. aureus, the MIC for compound 115 was 250 mg l −1 (57 μM) and 125 mg l −1 (28 μM) for compound 161 [fig_ref] Table 2: MIC values for compounds 115 and 161 in different strains of E [/fig_ref]. For wt E. coli the MIC values of compound 115 and 161 were above the highest concentration analyzed, 500 mg l −1 . Higher concentrations were not possible to test for these specific compounds (115 and 161) due to their precipitation in the cell culture media above 500 mg l −1 . In contrast, the MICs for the E. coli drug-hypersensitive lpxC mutant and the efflux-defective ΔtolC mutant were considerably lower. The lpxC mutant showed MICs of 64 mg l −1 (14 μM) for compound 115 and 125 mg l −1 for compound 161, while the ΔtolC mutant has a MIC of 250 mg l −1 for compound 115 and 125 mg l −1 for compound 161. These results suggest that the outer membrane of E. coli is a significant barrier for uptake of the compounds, as indicated by the lpxC mutant defective in lipid A biosynthesis, and that both compounds are substrates for TolC-dependent efflux. As a control, the MIC values for kirromycin were also determined for S. aureus and wt E. coli and found to be 100 mg l −1 and 50 mg l −1 , respectively.
A time-kill assay was further used to determine if the compounds were bacteriostatic or bactericidal. Here, both compounds were found to be bactericidal, that is, the cell number decreased significantly for S. aureus and the tolC-defective E. coli strain during the first four hours of incubation [fig_ref] Figure 4: Inhibitors are bactericidal in S [/fig_ref]. However, after about eight hours of incubation, the number of live cells increased in all time-kill assays regardless of bacterial strain or concentration of the inhibitors 115 and 161. The reason for this regrowth is unclear, but could result from instability of the compounds, emergence of resistant mutants, or regrowth of physiologically adapted cells.
# Discussion
While the ribosomal translation system is the target for many different classes of antibiotics, no small-molecule inhibitors have ever been reported that specifically target the bacterial release factors (RF1 and RF2), which are required for correct and efficient termination of protein synthesis and release of newly synthesized proteins from the ribosome. In principle, the bacterial RFs would appear as interesting drug targets, since their eukaryotic counterpart eRF1 has little sequence homology and a very different 3D structure from RF1 and RF2 [bib_ref] Fidelity at the molecular level: lessons from protein synthesis, Zaher [/bib_ref] [bib_ref] Structural basis for translation termination on the 70S ribosome, Laurberg [/bib_ref] [bib_ref] Insights into translational termination from the structure of RF2 bound to the..., Weixlbaumer [/bib_ref] [bib_ref] Bacterial polypeptide release factor RF2 is structurally distinct from eukaryotic eRF1, Vestergaard [/bib_ref] [bib_ref] The crystal structure of human eucaryotic release factor eRF1-mechanism of stop codon..., Song [/bib_ref] [bib_ref] Decoding mammalian ribosome-mRNA states by translational GTPase complexes, Shao [/bib_ref]. The only known potent small-molecule inhibitor of bacterial termination is the natural product Blasticidin S produced by Streptomyces bacteria, which with an apparent K i of about 30 nM, effectively blocks peptide release 8 . However, this compound binds to the PTC of the ribosome 9 and has been shown to deform the 3'-end of the P-site tRNA such that the tRNA conformation induces steric hindrance for RF binding to the ribosomal A-site [bib_ref] Blasticidin S inhibits translation by trapping deformed tRNA on the ribosome, Svidritskiy [/bib_ref]. The inhibitory effect of Blasticidin S in termination is thus indirect and does not involve binding to the RFs themselves. Logically, therefore, the molecule also has an inhibitory effect on peptide bond formation (elongation), but with www.nature.com/scientificreports www.nature.com/scientificreports/ a 6-fold higher K i value than for the termination reaction. Moreover, it is not selective for bacteria and it also inhibits translation in mammalian cells [bib_ref] Inhibition of protein synthesis by Blasticidin S, Yamaguchi [/bib_ref] , which likely reflects the conserved structure of the P-site on the large ribosomal subunit.
We sought here to identify small drug-like molecules from a large virtual library with binding affinity for the free RFs in solution or bound to the ribosome, with the idea that such compounds could possibly have an effect on translation termination. The strategy was then to combine computational screening (docking) with RF binding experiments and biochemical measurements of the effects of selected compounds on translation termination in a reconstituted cell-free system. All compounds from the virtual screening that were acquired were initially screened for their effect on bacterial growth in S. aureus and E. coli strains. This was done in order to filter out a smaller subset of potentially interesting molecules for subsequent in vitro binding and inhibition experiments. Obviously, this phenotypic screening may yield both false positives and negatives with regard to the ability of the molecules to inhibit termination. However, the same would be true for initial screening of binding to the free RFs in solution since both the target site and the actual target RF conformation are unknown. The strategy employed here was, nevertheless, successful in picking out compounds that both can bind to RF2 and inhibit peptide release. The most potent compounds 115 and 161 have both binding affinities (K d ) to free RF2 and inhibition constants on the ribosome (K i ) in the multiple round peptide release assay in the low μM region. While K i values in this range are clearly higher than those typical for tight binding drug molecules, they are unusually good for hits from a first round of virtual screening [bib_ref] Virtual screening and bioassay study of novel inhibitors for the dengue virus..., Luzhkov [/bib_ref]. Moreover, neither of these molecules have any effect on peptide bond formation or tRNA translocation, which shows that they indeed are specific for termination where they most likely interact directly with the RF.
The fact that neither of the two hits 115 and 161 show any inhibition of peptide release in the single round experiments, but only under multiple round conditions when recycling of RF2 is needed, indicates that they are not able to effectively compete with RF2 binding to the ribosome. This is perhaps not so surprising considering that the affinity of the RFs for the stop codon-programmed ribosome is very high, with K d in the sub-nM region [bib_ref] Release of peptide promoted by the GGQ motif of class 1 release..., Zavialov [/bib_ref] , compared to the μM K d 's measured for the small molecules. Instead it appears that the interaction of the inhibitors with the termination complex is such that RF2 becomes trapped on the ribosome and unavailable for multiple turnovers. This situation is similar to that found for the antibacterial peptide Api137 studied by Wilson, Rodnina and coworkers [bib_ref] An antimicrobial peptide that inhibits translation by trapping release factors on the..., Florin [/bib_ref] [bib_ref] Visualization of translation termination intermediates trapped by the Apidaecin 137 peptide during..., Graf [/bib_ref] , where binding in the exit tunnel after peptide release traps the RF on the ribosome via a direct interaction with it. It is thus likely that our inhibitors bind to and stabilize this state on the ribosome and thereby prevent conformational changes of the RFs required for their dissociation after peptide release.
It should be emphasized that the present study is not aimed at the development of new antibiotics, but rather at the (in vitro) exploration of the bacterial RFs as possible drug targets for such efforts. In particular, we were interested in exploring whether small synthetic molecules would be able to block translation termination through a mechanism involving direct binding to the RFs, which do not appear to have received much attention in antibiotics development so far. Since the results obtained from our combined computational and biochemical approach now show that small drug-like molecules can indeed inhibit termination via such a mechanism, this can be regarded as a proof-of-principle. The fact that the MIC values in the bacterial growth assays are high, as well as the actual mode of action in growth inhibition, is thus not relevant for our present purposes since these experiments were primarily used as a screening tool. A first step in further inhibitor development would rather be to try to improve in vitro affinities and inhibition by medicinal chemistry in order to arrive at intrinsically more potent compounds that could eventually be tested for effects on bacterial growth. That is, while the inhibitors discovered here are only moderately potent by normal drug standards, they could serve as useful lead compounds for exploring chemical modifications around this scaffold by organic synthesis. www.nature.com/scientificreports www.nature.com/scientificreports/ Methods protein target site search and virtual screening. Crystal structures of free RF2 in the closed conformation with PDB codes 1GQE 15 and 2IHR 21 , and structures of the open conformation in complex with the ribosome with PDB codes 5DFE 20 and 4V9N 19 , were used as input for MD exploration in mixed solvents with the MDmix software [bib_ref] Molecular simulations with solvent competition quantify water displaceability and provide accurate interaction..., Alvarez-Garcia [/bib_ref] [bib_ref] Binding site detection and druggability index from first principles, Seco [/bib_ref]. Each structure was immersed in a pre-equilibrated solvent box of 20% ethanol and 80% water and counter-ions were added to neutralize the system. Molecular dynamics (MD) simulations were performed with the Amber14 program and force field using periodic boundary conditions and the particle mesh Ewald method for long-range electrostatic interactions [bib_ref] The Amber biomolecular simulation programs, Case [/bib_ref] [bib_ref] ff14SB: improving the accuracy of protein side chain and backbone parameters from..., Maier [/bib_ref]. The structures were slowly heated for 5 ns to the target temperature (300 K) in the NPT ensemble, followed by an additional 5 ns MD equilibration in the NVT ensemble, which was also used in the subsequent production phase. Five replicates of each system were run for 100 ns each, and frames were collected every 10 ps for subsequent analysis with MDmix, which estimates the probability of binding spots based on the calculated density of ethanol molecules. The output from these calculations yields a series of grids with binding free energy values for the ethanol hydroxyl (hydrophilic) and methyl (hydrophobic) probes, as well as the water occupancy.
A collection of 3.4 M compounds was kindly provided by Prof. Xavier Barril (University of Barcelona). This collection was generated from commercial libraries of six preferred vendors (Specs, Enamine, Life Chemicals, Princeton BioMolecular, Key Organics and Asinex) and further filtered to retain non-reactive and drug-like compounds. The library was prepared for docking using Schrödinger's Ligprep program [bib_ref] Epik: a software program for pK a prediction and protonation state generation..., Shelley [/bib_ref] [bib_ref] Towards the comprehensive, rapid, and accurate prediction of the favorable tautomeric states..., Greenwood [/bib_ref] and a grid of 10 × 10 × 10 Å was built to comprise the binding site determined by the MDmix calculations. The virtual screening workflow (VSW) in the Schrödinger package (Version 15.3) was used for docking with Glide 18 and final rescoring with the MM-GBSA method [bib_ref] The VSGB 2.0 model: a next generation energy model for high resolution..., Li [/bib_ref]. The top 200 compounds were visually inspected to yield a final selection of 60 compounds to be purchased. The tested anti-RF2 chemical compounds from the virtual screen were purchased from Enamine Ltd., Life Chemicals Inc. and Asinex Corporation (Supplementary . Ligands were dissolved in dimethyl sulfoxide (DMSO) to an initial working concentration of 5-200 mM depending on the compound.
Binding assays. A thermal shift assay using the Bio-Rad CFX Connect Real-Time PCR system was used to characterize the proteins (RF1 and RF2) and determine their stability and solubility using the JBS buffer screening kit. With this approach we determined the tolerance of the proteins against 1%, 5%, 8%, 10%, 15%, 20% and 25% concentrations of DMSO. RF-ligand binding measurements were performed using the MicroScaleThermophoresis (MST) Monolith NT.115 equipment (Nano-Temper Technologies). The binding experiments were performed at 20 °C in 0.1 M HEPES buffer and 150 mM NaCl at pH 6.5. The RF2 protein was then fluorescence-labeled using the monolith Protein Labeling Kit RED-NHS (amine reactive). Twelve ligand dilutions between 500 nM and 500 μM were prepared in the same buffer as RF2. Each dilution was mixed with equal volume of l0 μl of 60 nM RF2, with a final DMSO concentration under 3%. The fluorescent molecules were excited with a red laser (650-700 nm) in the MST instrument to monitor the spatial distribution of molecules in the capillary. Thermophoresis was measured in each capillary by locally heating the sample with an infrared laser at 40% excitation for 30 s. Since bound and unbound molecules have a different response, the change in depletion in the presence of the inhibitor can be plotted and used to calculate the bound protein fraction. The dissociation constant (K d ) was then obtained by fitting the results to a binding isotherm, using the Origin software.
peptide release assay. The RC was formed in two steps. First E. coli 70S ribosome (2 μM), IF1, IF2 and IF3 (all 2 μM), XR7 Met-UGA mRNA (10 μM) with sequence UAAGGAGGUAUUAAAUGUGA (Shine-Dalgarno sequence underlined; coding sequence in bold) was incubated at 37 °C for 5 min in HEPES-polymix buffer at pH 7.5 34 containing 1 mM GTP, 1 mM ATP, 2 mM magnesium acetate and 0.1 U/mL Ribolock solution. Next, 2 μM [ 3 H]fMet-tRNA fMet was added to this mix and incubated for 10 min at 37 °C. An additional 4 mM magnesium acetate was added to stabilize the complex. The complex was then applied to a 1.1 M sucrose cushion (prepared in HEPES-polymix buffer pH 7.5) and centrifuged at 55000 rpm for 3 hours at 4 °C in a S55s rotor in SORVALL M150GX ultracentrifuge. The RC pellets were washed and dissolved in HEPES-polymix buffer (pH 7.5) and stored at −80 °C after shock freezing.
For the peptide release assay, RC and RF2 were pre-incubated for 10 min at 37 °C and then mixed together in the presence or absence of the compounds 115 and 161. For the multi-round peptide release assay, 125 nM RC was mixed with 5 nM RF2 (active concentration) and the anti-RF2 compounds at 0.003−1 mM, for 5 min at 37 °C. For the single-round peptide release assay, 125 nM RC was mixed with 1.250 μM RF2 (active concentration) and anti-RF2 compounds (0.003-1 mM) for 10 seconds at 37 °C. RF3 was not included in these assays and control experiments were performed in the absence of compounds and in the presence of 2% DMSO, which is the solvent for the anti-RF2 compounds. The reactions were quenched by an adding equal volume of 50% formic acid. The amount of [ 3 H]fMet released from the RC was determined by scintillation counting of the supernatant after centrifugation for 15 minutes at 14000 rpm at 4 °C. tripeptide formation assay monitoring peptide bond formation and translocation. Ribosomal activity for peptide bond formation and translocation was tested by di-and tripeptide formation assays [bib_ref] Molecular mechanism of viomycin inhibition of peptide elongation in bacteria, Holm [/bib_ref]. For that, an initiation complex and an elongation mix were first prepared in HEPES-Polymix buffer (pH 7.5). The initiation complex contained E. coli 70S ribosome (1 μM), IF1, IF2 and IF3 (all 2 μM), XR7 mRNA encoding Met-Leu-Leu (10 μM) with sequence UAAGGAGGUAUUAAAUG CUGCUGUAA and 1 μM [ 3 H]-fMet-tRNA fMet . The elongation mix was prepared by mixing 10 μM EF-Tu, 10 μM EF-Ts, 10 μM EF-G, 10 μM tRNA Leu , 0.5 mM Leu amino acid and 1 unit of Leu-tRNA synthetase. All reactions contained 1 mM GTP and ATP, 10 mM phosphoenol pyruvate, 50 mg/ml pyruvate kinase and 2 mg/ml myokinase. To test the effects, 0.5 mM of the anti-RF2 compounds 115 and 161 was added to both IC and EC together with DMSO as control. Both the mixes were incubated www.nature.com/scientificreports www.nature.com/scientificreports/ separately at 37 °C for 15 min. The reaction was started by mixing equal volumes of the IC and EC at 37 °C in a quench flow apparatus and was quenched by mixing 17% formic acid after different time points. The peptides were isolated from the ribosome by KOH treatment and applied to reverse phase HPLC separation with in-line radioactivity detection as described earlier Bacterial strains and minimum inhibitory concentration determinations. The four different strains analyzed were wild type (wt) Staphylococcus aureus (strain ATCC 29213), wt Escherichia coli (strain MG1655, K12), a drug-hypersensitive E. coli lpxC mutant D22 (strain MG1655) and an efflux-defective E. coli ΔtolC mutant (strain MG1655). These were grown in cation-adjusted Mueller Hinton II media (MH II, Nordic Biolabs), the standard medium for antibiotic susceptibility testing. For overnight cell cultures one colony from a MH II (Nordic Biolabs) plate was inoculated in 1 ml MH II media in a 10 ml tube and the bacteria was incubated at 37 °C shaking at 190 rpm. Cells from the overnight culture was used for MIC determinations and time-kill experiments.
The minimal inhibitory concentrations (MIC) of the drugs were measured in 96-well round-bottom microtiter plates using cation-adjusted Mueller-Hinton II (MH II) media. All measurements were done with at least three biological replicates. Briefly, bacterial colonies from a non-selective agar plate were re-suspended in 0.9% NaCl to 0.5 McFarland (≅ . × 1 5 10 8 cfu/ml). The cell suspension was further diluted to ≅ × 5 10 5 cfu/ml (50 µl of bacterial suspension to 10 ml of MHII) and then 90 µl of cell suspension (≅ × 5 10 5 cfu/l) was added to each well. For the initial screen of all the compounds different concentrations was used depending on the solvability of the compound in 50% DMSO and their propensity to precipitate in the cell culture media. For all compounds the MIC measurements were done in a two-fold dilution series and for compound 115 and 161 the concentration was from 500 mg l −1 to 4 mg l −1 by adding 10 µl of compound 115 or 161 dissolved in 50% DMSO to water giving a final concentration of DMSO of 5% at the highest compound concentration (500 mg l −1 ) analyzed. The positive control (without compound) also contained 5% DMSO to determine if the DMSO concentration influenced bacterial growth. Plates were covered and incubated without shaking at 37 °C for 16-20 h. The MIC value was determined visually as the lowest concentration of drug that inhibited growth.
## Time-kill experiments.
To determine if the compounds are bacteriostatic or bactericidal, time-kill assays were performed with wt S. aureus and a ΔtolC efflux-defective E. coli. The initial inoculums were diluted from triplicate biological replicates of overnight cell cultures inoculated with one colony in 1 ml MH II medium. The cell inoculum was diluted 1:1000 in 1 ml MH II medium and the cell culture was allowed to recover for 30 minutes before the compounds were added at a final concentration of 1x MIC (250 mg l −1 for compound 115 and 125 mg l −1 for compound 161 for both bacterial strains) or 2x MIC. As a control of the possible effect of DMSO on cell growth, the positive growth controls (three biological replicates) contained the same DMSO concentration as the cells with compound added (2x MIC). Samples (20 µl) were withdrawn before the compound was added and also after 1, 2, 4, 8 and 24 h after addition. The cells were diluted in 0.9% NaCl before they were plated on non-selective agar plates using glass beads. Colonies were counted after approximately 20 h.
## Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
www.nature.com/scientificreports www.nature.com/scientificreports/
[fig] Figure 1: The bacterial 70S ribosome termination complex with RF2. (A) View of the ribosome termination complex with E-and P-site tRNAs (brown), mRNA (green) and RF2 (dark blue). (B) Close-up view of the hinge region of RF2 between domains 1 and 4 used for virtual screening, where the putative binding region is indicated by a docked ligand (red). [/fig]
[fig] Figure 2: Binding of compounds 115 and 161 to free RF1 and RF2 in solution. Measurement of affinity between E. coli release factors -RF2 (A,B), RF1 (C,D) and the compounds 115 (A,C) and 161 (B,D) by microscale thermophoresis. The resulting binding curves are shown (average of n = 3 ± s.d) from which K d values are estimated. [/fig]
[fig] Figure 3: Effects of compounds 115 and 161 on peptide release and di-and tripeptide formation in vitro. (A,B) The effect of the compounds 115 (A) and 161 (B) on single-round (gray circle) and multiple-round (black square) peptide ([ 3 H]fMet) release assays. (C,D) Estimation of half maximal inhibitory concentration (IC 50 ) for the compounds 115 (C) and 161 (D) from the results of the multi-round peptide release assay. (E) The effect of compounds 115 (orange) and 161 (green) at 0.5 mM and 2% DMSO on fMet-Leu dipeptide and fMet-Leu-Leu tripeptide formation from the quench flow experiment. Apparent rate constants k a for dipeptide (about 14 s −1 ) and tripeptide (about 4 s −1 ) bond formation are similar in the presence of the 115 and 116 compounds (F). All experiments were done in triplicates (average of n = 3 ± s.d). [/fig]
[fig] Figure 4: Inhibitors are bactericidal in S. aureus and tolC defective E. coli. Time-kill assays for compound 115 (A) and 161 (B) for wt S. aureus and an E. coli ΔtolC mutant strain. Cells were incubated with or without compound 115 or 161 for 24 hours. 1x MIC (Minimal Inhibitory Concentration) of compound 115 and 161 are for E. coli and S. aureus 250 mg l −1 and 125 mg l −1 , respectively. Samples for determination of colony forming units were withdrawn from the time-kill experiment before and 1, 2, 4, 8 and 24 hours after the compound was added. All experiments were done with triplicate biological samples (average of n = 3 ± s.d). [/fig]
[table] Table 2: MIC values for compounds 115 and 161 in different strains of E. coli and S. aureus determined by broth micro-dilution. a 500 mg l −1 was the highest concentration tested due to precipitation in the cell culture media at higher compound concentrations. [/table]
|
Interstitial pneumonia in a glassblower: think to chronic beryllium disease!
Chronic beryllium disease (CBD) is an occupational illness with varying severity. In this report, we describe a 27 year old man, glassblower, who developed a fatal CBD after six months of unknown Beryllium's exposure. The diagnosis was suspected on histological examination and then consolidated by confirmation of Beryllium's exposure at the working area. Physicians should be aware of the potential risk to develop CBD in glassblowers. These workers should benefit from early medical surveillance using the Beryllium lymphocyte proliferation test (BeLPT) and therefore from suitable management.
# Introduction
Beryllium (Be) is a natural element used in several industries like automotive electronics, telecommunications, computers, aerospace and defense equipment, dental alloys and fluorescent lamps [bib_ref] An official American thoracic society statement: diagnosis and management of Beryllium sensitivity..., John [/bib_ref]. Workplace Be exposure can be unknown by employers, workers and even occupational physicians, thus leading to Chronic beryllium disease (CBD) misdiagnosis. We report a case of fatal CBD in a 27 years old man working in glass industry.
## Patient and observation
We report a 27 year old man, 10 pack-years smoker, with a onemonth history of dry cough, shortness of breath and general status's deterioration. A lung X ray showed bilateral alveolar opacities . Blood gases analysis in ambient air showed a pO2 of 57 mmHg, a pCO2 of 38.6 mmHg, a HCO3-of 29.2 mmHg. Because of the considerable deterioration in his respiratory condition and the need for high oxygen flow, he was referred to intensive care unit (ICU) with the diagnosis of an acute respiratory failure (ARF). Chest computed tomography (CT) showed diffuse ground-glass opacities [fig_ref] Figure 2: CT scan [/fig_ref] , [fig_ref] Figure 2: CT scan [/fig_ref] and mediastinal lymphadenopathies. All microbiological infectious investigations, especially tuberculosis, were negative. He had hypergammaglobulinemia of 21.9 g/l. Other immunological tests were otherwise negative. Transthoracic echocardiography was performed finding a hyperkinetic left ventricle, a small dilation of right ventricle with PAPs 48 mmHg and an intrapulmonary shunt revealed by contrast test. The bronchial endoscopy showed an inflammatory bronchial tree and the bronchoalveolar lavage (BAL) concluded to lymphocytic alveolitis with a ratio CD4 (helper T cells) over CD8 (suppressor T cells) (CD4/CD8) of 0.5. The bronchial biopsies showed inflammatory bronchial mucosa without any specificities. Despite the absence of Koch's bacillus in sputum analysis, we decided to begin a trial antituberculosis treatment because of the severity of lung involvement added to high endemicity of tuberculosis in our country. Two weeks later, the patient presented an allergic skin reaction. As no recovery's signs were noted, we decided to stop the treatment. Corticosteroids were initiated considering idiopathic diffuse infiltrating pneumonitis. Unfortunately, the patient did not show any improvement and the chest scan performed after 27 days of treatment showed worsening interstitial infiltrates and fibrosis.
Clinically, he required high flow oxygen (14 l/mn) alternating with non invasive ventilation sessions. Histological examination of the open lung's biopsy then performed showed diffuse lung parenchyma damage. It was altered by a mutilating fibrosis that was widening the alveolar septa. This fibrosis contained an abundant inflammatory infiltrate made of lymphocytes, plasmocytes and fewer neutrophils and eosinophils. Epithelioid and giant cell granulomas were also observed without either caseous or fibrinoïd necrosis. This mutilating fibrosis induced an enlargement of some alveoli resulting in the formation of a honeycomb aspect. Fibroblastic foci were also noticed. Some alveoli contained hyaline membranes that were sometimes incorporated in the septa. There were no mineral particles observed with polarized light [fig_ref] Figure 3: Microscopic exam [/fig_ref]. Regarding these features, the diagnosis of usual interstitial pneumonia (UIP) associated to lesions of sarcoidosis or berylliosis with aspects suggestive of an acute exacerbation was retained. Since sarcoidosis was not very likely according to CD4/CD8 ratio in BAL, we investigated risk factors for chronic beryllium disease, particularly occupation's patient history. He was working as glassblower in an artistic glass factory since six months. Further to our request, occupational department's experts inspected the workplace and confirmed beryllium existence in compounds of colored glass. After confirmation of CBD, steroids were maintained and oxygen requirement slowly decreased. The patient was finally discharged at home after 55 days steroid treatment with long term oxygen therapy (2 l/minute). Pulmonary function tests at discharge showed a severe restrictive impairement. Steroid doses were slowly decreased and stopped after 10 months. A pre-transplant assessment was also made. One year after discharge, the patient felt somewhat better and pulmonary function tests showed a slight improvement. The CT scan performed found a partial regression of lung consolidations and ground glass opacities, but increase of distortion signs, cavities appearance of which one filled by aspergilloma. He received oral voriconazole 200 mg per day during 3 months. After a two-year period from the date of CBD diagnosis, the patient was again admitted in ICU for ARF. An enhanced CT scan performed showed further deterioration of pulmonary parenchyma [fig_ref] Figure 2: CT scan [/fig_ref] , [fig_ref] Figure 2: CT scan [/fig_ref] and pulmonary embolism with fatal issue.
# Discussion
CBD is an occupational hypersensitivity disorder making a multisystem granulomatosis. It is caused by dust, fumes or mists of Be metal leading to lesions which occur primarily in the lungs. Because of to the wide use of Be, exposure can occur in environments other than the traditional areas of Be manufacturing. As reported by the international agency for research on cancer (IARC) since 1993, additives to glass are considered as products with high potential Be exposure. Among Be compounds, Beryllium-Oxide (BeO) is particularly used as an additive to glass, ceramics and plastics [3]. Thus, artistic glass industry is a potential situation to develop berylliosis. However, known lung diseases associated to glassblowing are pneumoconioses due to inhalation of inorganic dust like asbestos or silica [bib_ref] Scientific Glassblowing Safety, Robertson [/bib_ref]. To our knowledge, this is the first case of CBD reported in a glassblower and we think that similar cases should exist but are misdiagnosed. Physicians must be aware of this risk and glassblowers should benefit from early screening using the beryllium lymphocyte proliferation test (BeLPT)that confirms Be hypersensitivity preceding the disease. Indeed, at this stage, worker should be removed from his workplace, even if there is no evidence that Be exposure cessation firmly avoid progression of Be sensitization to CBD [bib_ref] Systematic review: progression of Beryllium sensitization to chronic Beryllium disease, Seidler [/bib_ref]. There are two clinical forms of Be related lung disease: an acute chemical pneumonitis due to a direct toxic injury of the lung, usually induced by brief exposure to high concentrations of Be, and completely reversibleand CBD which is a multisystem immunologic granulomatosis which can resemble sarcoidosis or idiopathic diffuse interstitial fibrosis. It occurs after exposure to lower levels of Be with a long delay (several months to 20 years) between initial exposure and patent disease. This observation highlights many particularities of clinical behavior: first, many factors leaded us to suspect the acute form of berylliosis: severity of initial clinical presentation, shortness of exposure delay, presence of hyaline membranes in some alveoli reminiscent of diffuse alveolar damage and direct exposure of the patient's respiratory tract. Indeed, glass-blowing offers special opportunities for exposure to metallic compounds in glass, which may migrate up the pipe into the worker's mouth and subsequently the bronchial tree . Nevertheless, granulomas's presence at histological examination firmly established CBD diagnosis, and very short delays (4 months) had already been reported. Second, we think that the first presentation was an acute exacerbation of usual interstitial pneumonia associated with CBD in a diffuse alveolar damage form at the fibrotic organization stage. Histological findings at the time of diagnosis of CBD have a prognosis value [bib_ref] Beryllium disease: the relation of pulmonary pathology to clinical course and prognosis..., Freiman [/bib_ref] which emphasize the importance of early medical surveillance using BeLPT.
# Conclusion
CBD diagnosis is usually based on unspecific findings. Furthermore, unsuspected Be exposure leads to misdiagnosis or to diagnosis and treatment delays. Glassblowers should undergo specific medical surveillance using the BeLPT to indicate their removal from Be exposure at the appropriate moment.
## Competing interests
The authors declare no competing interest.
# Authors' contributions
Amira Jamoussi contributed to the writing. Amira Jamoussi, Takoua Merhabene, Jalila Ben Khelil and Mohamed Besbes contributed to the review. Mona Mlika and Mohamed Faouzi EL Mezni contributed to the description of biopsies. Henda Neji contributed to CT scan descriptions. All authors read and approved the final manuscript. : Chest X ray: bilateral alveolar opacities
# Figures
[fig] Figure 2, Figure 1: CT scan (lung windowing): a & b) bilateral lung consolidations and ground glass opacities with bronchial distortion; c & d) CT scan at 2 years: partial regression of lung consolidations and ground glass opacities and increase of distortion signs and cavities appearance Figure 3: Microscopic exam: a) epitheloid granuloma; b) incorporation of hyaline membranes into the alveolar septa (arrow) (HE x 400); c) trichrom stain highlighting the honeycomb aspect (x 250) Chest X ray: bilateral alveolar opacities [/fig]
[fig] Figure 2: CT scan (lung windowing): a & b) bilateral lung consolidations and ground glass opacities with bronchial distortion; c & d) CT scan at 2 years: partial regression of lung consolidations and ground glass opacities and increase of distortion signs and cavities appearance [/fig]
[fig] Figure 3: Microscopic exam: a) epitheloid granuloma; b) incorporation of hyaline membranes into the alveolar septa (arrow) (HE x 400); c) trichrom stain highlighting the honeycomb aspect (x 250) [/fig]
|
Psychosocial effects of social media on the Saudi society during the Coronavirus Disease 2019 pandemic: A cross-sectional study
The Coronavirus Disease 2019 (COVID-19) pandemic has been posing a substantial challenge to human survival and well-being, which rely on the actions and behaviors of individuals. It is essential that accurate information is distributed; however, misinformation has been spread via social media. Consequently, the resulting panic has to be addressed while putting essential public health measures in place.
# Introduction
False information amplifies the challenges that humanity faces, such as the Coronavirus Disease 2019 (COVID-pandemic. Within weeks of the appearance of the virus in China, misleading rumors and conspiracy theories about its origin spread globally, coupled with stress out that the best way to achieve a sense of psychosocial security during a crisis, including COVID-19, is to confront false information and rumors spread through social media by flooding the social media with accurate scientific information that is understood and shared easily, and by having certified specialists answer the questions of the public. In Saudi Arabia, the society has witnessed a rise in social media sites, especially due to the increasing use of modern communication technologies. By January 2020, approximately 4.54 billion, 3.8 billion, and 5.19 billion people globally were using the internet, social networking sites, and mobile phones, respectively. Moreover, 2.78 billion individuals used their phones to surf social networking sites. Saudi Arabia ranks 17th in the rate of internet usage globally , and is at the forefront of active social media users internationally. In particular, 32.23 million individuals among its population of 34.218 million are Internet users. Of these, 25 million use social networks for an average of two hours and fifty minutes daily, mostly YouTube (76%), followed by WhatsApp (71%), Instagram (65%), Facebook (62%), and Twitter (58%) . Social media has offered many benefits during closures and health isolation (or rather "social exclusion"). It continues to be the easiest method of accessing information and has played an important role in individuals' ability to work from home and organize/join meetings, seminars, and conferences. However, we need to consider the risks it presents through misleading facts and rumors, especially in times of crises such as the current pandemic, as it may hold dire psychological, social, and health consequences for the society .
Considering the psychological, social, and behavioral effects of social media on the society during the COVID-19 pandemic, many countries, including Saudi Arabia, have implemented procedures such as taking legal action to counter the spread of rumors and misinformation. The Kingdom's public prosecution has warned against spreading news from unknown sources and rumors through social media, especially those linked to the COVID-19 crisis; anyone who provides misleading information is subject to penal accountability .
The mere fact that social media is used widely in Saudi Arabia makes it necessary to identify the exact impact of misleading information on its citizens' psychological and social well-being, in particular, on their levels of anxiety, depression, and social isolation. Similar research has not yet been undertaken in the Saudi society in relation to the pandemic. Therefore, we aimed to identify the level of generalized anxiety, depression, and social isolation among participants during COVID-19. Additionally, we examined the relationship of exposure to misleading information through social media with psychological and social well-being during the pandemic. We explored the following research questions: Q1: What is the level of generalized anxiety, depression, and social isolation among participants during COVID-19?
Q2: Is exposure to misleading news in social media related to the level of generalized anxiety, depression, and social isolation?
# Materials and methods
## Participants
Three hundred seventy-one participants (male = 272, female = 99; age: 16-60 years) participated in this cross-sectional study in Najran, Saudi Arabia. Participants were invited to complete an online questionnaire one month after the spread of COVID-19 in Najran. The timeline of this study was concurrent with the lifting of restrictions applied nationally. Also, participants were randomly selected using the simple random sampling to provide equal and independent opportunity of selection for the sample.
## Procedures
This cross-sectional study was online conducted over six days (May 10-15 May 2020) after cases in Saudi Arabia reached almost two thousand and while curfew and social distancing measures were implemented by Government authorities. We obtained approval for our quantitative study from the Dean of Scientific Research at Najran University. The department of public ration's and university media at Najran University has contacted the ministry of health in Saudi Arabia to distribute the survey link to all Saudi citizens. The ministry of health send messages with the survey link to Saudi citizens over six days. All participants completed the voluntary consent section in the questionnaire and were assured confidentiality. They completed the questionnaire on the website. The authors expected to obtain 450 responses within six days, but only 386 questionnaires were returned, of which 15 were excluded because they did not follow the questionnaire instructions. Consequently, the total number of questionnaires used for analysis was 371.
## Measurements
We administered the following three reliable scales with minor adaptions: the Generalized . Respondents are asked how often they have been upset in the past two weeks, experiencing any of the seven main generalized anxiety disorder symptoms. The answer options are "Not at all," "Several days," "More than half the days," and "Almost every day" and are recorded as 0, 1, 2, and 3, respectively. Therefore, GAD-7 scores range from 0 to 21, with scores of 5, 10, and 15 representing mild, moderate, and severe anxiety symptoms, respectively. The scale has demonstrated excellent internal consistency (Cronbach α = .92) and reliability (intraclass correlation = .83) . Several studies have confirmed its internal consistency, test-retest reliability, convergence, construction, standards, and validation . For the present study, the scale was translated from English to Arabic and then back-translated. It was presented to specialists in English and Arabic linguistics to ensure the accuracy of the translation and linguistic formulation. Furthermore, the reliability of the data collected with the Saudi version of the GAD-7 was good (Cronbach α = .87), consistent with previous studies in which it ranged from .74 to .94 .
Centre for epidemiological studies depression scale. The CES-D-10 is a 10-item Likert scale questionnaire that measures the occurrence of depressive symptoms over the week preceding the assessment [48]. It includes 10 elements, with three on depressive affect, five on physical symptoms, and two on positive affect. Among the 10 elements, eight assess the positive symptoms and two (Item 5 and 8) assess the negative symptoms of depression. The answer options for each item are presented on a four-point Likert scale ranging from "rarely or none of the time" (0) to "all the time" (3). Moreover, Item 5 and 8 are reverse scored as they are worded positively. Total scores range from 0 to 30, with higher scores indicating a greater severity of symptoms.
For the present study, the scale was translated (English to Arabic) and subsequently backtranslated. We presented it to specialists in English and Arabic linguistics to ensure the accuracy of the translation and linguistic formulation. In the present study, the Saudi version showed adequate reliability and viability, with good internal consistency (Cronbach α = .81).
De Jong Gierveld loneliness scale. The six-item DJGLS aims to assess the social (example statement: "there are many people I can completely trust") and emotional (example statement: "I feel a general sense of emptiness") aspects of isolation between individuals, with three elements each. Response options are presented on a four-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). The original version's Cronbach α was 0.76, which was consistent with many previous studies . In the present study, the scale was translated from English to Arabic, and back-translated. It was presented to specialists in English and Arabic linguistics to ensure the accuracy of the translation and linguistic formulation. The Saudi version of the DJGLS exhibited adequate reliability and viability in the present study, with good internal consistency (Cronbach α = .78).
Covariates. The following covariates were included in this study: gender, Marriage (Married and No married), age , 36-50, 50-60 and More than 60), educational level (Less than secondary, Diploma / Secondary, Bachelor, Postgraduate), Living area(urban and rural), self-rated health (Excellent, Good, Sick).
# Data analysis
We applied descriptive and inferential statistics to analyze the data. The descriptive statistics included frequency, percentage, average and standard deviation; these were analyzed using SPSS 21. To address the first research question, we conducted a univariate analysis to compare the differences between participants' characteristics based on social media exposure. Subsequently, we used multivariate analysis to examine the influence of social media exposure on the participants' anxiety, depression, and social isolation. We used linear regression to examine whether the exposure to misleading news in social media is related to the level of generalized anxiety, depression, and social isolation.shows the frequency and percentage of participants' social media exposure during the COVID-19 pandemic. Participants were predominantly male (73%) and approximately 60% were married. The groups with the highest representation were those aged 36-50 years (45.3%), those pursuing postgraduate education (48%) and those living in urban areas (80%). Contrarily, people aged over 60 years, those without secondary education, and those living in rural areas were the least represented groups in this study. Furthermore, most of the participants were in excellent health.
# Results
We conducted univariate analyses to compare the variation among participants' characteristics based on their exposure to social media. Social media exposure was higher among women (77.77%) than men (73.16%), and higher among married (71.23%) than unmarried (51.97%) participants. Moreover, middle-aged (aged 36-50 years) had higher social media exposure than older (aged 50-60 years) and younger (aged 26-35 years) individuals. Furthermore, participants with lower levels of education (middle and high school) had less exposure to social media than did those with higher education (undergraduate and graduate). Finally, participants from rural areas (87.37%) reported a higher rate of social media exposure than did those from urban areas (64.64%). Healthy participants (72%) were significantly more likely than others to have exposure to social media.presents the prevalence of anxiety, depression, and social isolation, which was 47.82%, 47.57%, and 46.42%, respectively. Furthermore, multivariate analyses indicated that the possibility of anxiety, depression and social isolation was lower among males. Further, there were no age-related differences in the incidence of anxiety and social isolation among males. However, the depression rate was higher among those aged 16-25 years as compared to those aged over 60 years. Moreover, depression, anxiety, and social isolation were higher among those who did not have high school education. Participants in rural areas had a higher rate of anxiety, depression, and social isolation than did those in urban areas. However, those with lower levels of health had higher rates of anxiety, depression, and social isolation than did those with excellent health. Furthermore, it was found that participants with frequent exposure to social media had a higher level of anxiety, depression, and social isolation as compared to those with less frequent social media exposure. In this study, social media exposure was correlated with anxiety, depression, and social isolation.
To find the relationship between the social media exposure and anxiety, depression and social isolation, we used regression analysis and illustrates this correlation. reveals the value of the correlation coefficient between social media exposure and anxiety, depression, and social isolation, which were (0.368), (0.355), and (0.342), respectively, and all correlation coefficients were significant at P-value (0.01). The table also shows that social media exposure has a relationship with the dependent variables (anxiety, depression, and social isolation) with a percentage variance of (0.135), (0.126), (0.117).
The results of simple linear regression analysis showed that the regression model was significant between the independent variable and the dependent variables. The level of significance for the P-value was less than (0.05). social media exposure has a variance ratio of (0.135), (0.126), (0.117), and this can also be deducted from the value of t and its significance on anxiety, depression, and social isolation.
# Discussion
The last national survey in Saudi Arabia indicated a prevalence of 23%, 6%, and 5.6% for anxiety, depression, and social isolation, respectively. In comparison, our cross-sectional study on a Saudi sample during the COVID-19 outbreak found higher prevalence rates for these variables (38.14%, 39.93%, and 39.6%, respectively). These results are consistent with those of previous studies, indicating that not only general health issues but also pandemics or epidemics (including COVID-19, Ebola, and SARS outbreaks, can aggravate mental health issues in individuals. Social media has been one of the primary sources of up-to-date COVID-19 information; a substantial proportion of our respondents (64%) were frequently exposed to social media. This current finding similarly assure previous finding which Saudi population is ranked within top 20 population using internet and social media compared to other population .This is often associated with a high risk of anxiety, depression, and social isolation .
Findings of single studies related to the effects of social media on mental health in reality were contradictory. It is possible that the use of social media in general may not have an impact on the mental health. But, during the outbreak of the pandemic, misleading information and false reports spread via social media. It raised concerns and anxiety among many usersand may have confused and distracted people, while causing harm to their mental health. Moreover, the spreading of misleading information is the most infectious disease resulting to more health problems including anxiety, depression and loneliness. Besides, several individuals expressed negative feelings (such as anxiety, stress, fear, and nervousness) via social media, which may have increased similar feelings in others. Hence, the WHO's Informatics team have been collaborating with communications divisions worldwide to provide accurate information to a broader audience. Furthermore, the steps taken by the Saudi Arabia-such as the closure of schools, universities, Umrah, Hajj, domestic and international aviation, and other precautionary measures-were beneficial in limiting the spread of the virus and such measures gathered praise from the WHO. However, increasingly stringent measures in the Kingdom may induce more serious mental health problems among the Saudi population.
# Conclusions
Our findings showed a significant prevalence of mental health problems within the Saudi society due to the COVID-19 pandemic; moreover, these issues are positively associated with the significant and frequent exposure to social media during the outbreak. Our results indicated that the Saudi Arabian government should pay more attention to the general population's mental health during the COVID-19 pandemic. The Saudi government was the first to implement precautionary measures to protect its citizens, which contributed significantly to the low levels of active cases or mortality as compared to several other developed countries. Moreover, the Saudi government has been providing mental health services through various channels, including electronic applications, hotlines, online counseling and courses, and outpatient consultation.However, depression, anxiety, and social isolation need to be addressed more concertedly in order to prevent their effects on infected individuals' wellbeing. The Saudi government has also imposed numerous financial penalties and imprisoned those who published inaccurate and misleading information on social media. We suggest that, in addition to upscaling the legal actions related to the publishing of misleading COVID-19 information, the Kingdom needs to monitor social media, filter out false information, and promote the spread of accurate information through cooperation with the WHO.
## Limitations of the study
Our study has a few limitations. First, it is a cross-sectional study, which presents challenges in terms of clarifying the causal relationships between social media exposure and mental health.
This can be addressed in future longitudinal studies, including case studies and cross-control studies. Furthermore, this is an intermediate sample; the survey was conducted online, which enabled rapid assessment. However, the bias created by the limited representation of elderly citizens could have affected our results. Finally, the study's findings can only be generalized to similar sample populations and societies.
## Supporting information
|
CD123 and More: How to Target the Cell Surface of Blastic Plasmacytoid Dendritic Cell Neoplasm
Citation: Bôle-Richard, E.; Pemmaraju, N.; Caël, B.; Daguindau, E.; Lane, A.A. CD123 and More: How
# Introduction
Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare and aggressive leukemia characterized by a clonal expansion of plasmacytoid dendritic cells (pDCs). There has not historically been a standard-of-care therapy for these patients, and they were treated with leukemia or lymphoma chemotherapies. However, conventional chemotherapy is largely inadequate in BPDCN, since although many patients initially respond, the responses are short-lived and relapsed BPDCN is quite chemoresistant. This has prompted the evaluation of alternative therapies such as targeted cytotoxins and immunotherapy. Recently, tagraxofusp became the first drug approved for BPDCN in the US and Europe. Tagraxofusp is an interleukin 3-diphtheria toxin fusion protein that targets the IL3 receptor alpha subunit, or CD123, which is highly expressed on the surface of BPDCN cells. Some patients do not tolerate tagraxofusp or it may not be available, and therefore standard chemotherapy is still used. The most efficient regimens are broadly separated into these sub-types: acute leukemia treatment (AML or ALL-like), lymphoma-like treatment, and asparaginase/methotrexate-based treatments that are also used in aggressive leukemias and lymphomas. Retrospective data imply a better rate of response with first-line therapy that includes asparaginase/MTX and in ALL-like compared to AML-like or lymphoma-like treatment. However, this assumption would ideally be confirmed in prospective cohorts. Beyond those "classical" chemotherapies, innovative approaches have been developed through the integration of data on the physiopathology and oncogenesis of BPDCN. Proof of concept of the efficacy of targeted therapies such as bortezomib or venetoclax has been recently provided with in vitro models and pre-clinical data [bib_ref] The Lenalidomide/Bortezomib/Dexamethasone Regimen for the Treatment of Blastic Plasmacytoid Dendritic Cell Neoplasm, Marmouset [/bib_ref] [bib_ref] Molecular Profiling of Blastic Plasmacytoid Dendritic Cell Neoplasm Reveals a Unique Pattern..., Sapienza [/bib_ref]. The recent increased use of BCL-2 inhibition in myeloid malignancies and its approval in acute myeloid leukemia along with azacitidine lead to the assessment of the venetoclax alone or in combination in ongoing clinical trials enrolling patients with BPDCN.
Even if most patients respond to initial tagraxofusp or chemotherapy, relapse is common without consolidation by hematopoietic cell transplantation (HCT). In the setting of HCT, overall survival can reach 40% after a follow-up of 5 years [bib_ref] How Should We Diagnose and Treat Blastic Plasmacytoid Dendritic Cell Neoplasm Patients?, Garnache-Ottou [/bib_ref] [bib_ref] Blastic Plasmacytoid Dendritic Cell Neoplasm: Is Transplantation the Treatment of Choice?, Dalle [/bib_ref] [bib_ref] Blastic Plasmacytoid Dendritic Cell Neoplasm: Diagnostic Criteria and Therapeutical Approaches, Pagano [/bib_ref] [bib_ref] Stem Cell Transplantation Can Provide Durable Disease Control in Blastic Plasmacytoid Dendritic..., Roos-Weil [/bib_ref] [bib_ref] Allogeneic Hematopoietic Cell Transplantation for Patients with Blastic Plasmacytoid Dendritic Cell Neoplasm..., Bashir [/bib_ref]. However, relapse still occurs in at least 2/3 of allograft patients. Transplantation in first complete remission (CR1) is associated with a better outcome compared to patients not in CR1. However, HCT is associated with a significative toxicity that limits its indication to fit and younger patients. Altogether, the data reporting outcomes with chemotherapy and HCT demonstrate a large medical need for refractory patients and/or elderly patients that may not be able to undergo intensive treatments. The long-lasting remissions that have been obtained after HCT demonstrate the curative potential of adoptive therapies or immunotherapies and support the design of such novel approaches.
Anti-tumor immunotherapy is a therapeutic strategy based on the principle of immunosurveillance of cancers: since the immune system is naturally capable of recognizing cancer cells and destroying them, immunotherapy consists of artificially mobilizing immune cells to recognize and eliminate malignant cells [bib_ref] Cancer Immunoediting: Integrating Immunity's Roles in Cancer Suppression and Promotion, Schreiber [/bib_ref]. Thanks to the various clinical successes that have occurred, immunotherapy has been considered a major breakthrough in terms of cancer treatment. Today, there are different types of immunotherapy: passive immunotherapy based on monoclonal antibodies, active immunotherapy using cytokines such as IFNα, IL-2, or TNF, and adoptive immunotherapy that uses effectors of the immune response such as T cells. Indeed, over the last twenty-five years, new cellular therapies against cancer based on the ex vivo manipulation and re-infusion of autologous or allogeneic immune cells have been widely tested in the clinic. These different immunotherapy approaches have shown great potential for the treatment of cancers, especially those resistant to conventional therapies (surgery, radiotherapy, and chemotherapy) [bib_ref] The Optimal Antigen Response of Chimeric Antigen Receptors Harboring the CD3ζ Transmembrane..., Bridgeman [/bib_ref].
## Targeted therapies with results or in evaluation for bpdcn
## Tagraxofusp
CD123, the interleukin (IL)-3 receptor alpha chain, is overexpressed in 100% of BPDCN cases, and not expressed or weakly expressed on normal cells other than normal plasmacytoid dendritic cells and basophils [bib_ref] Transfer of Chimeric Receptor Gene Made of Variable Regions of Tumor-Specific Antibody..., Mezzanzanica [/bib_ref] [bib_ref] Expression of a Hybrid Immunoglobulin-T Cell Receptor Protein in Transgenic Mice, Becker [/bib_ref]. CD123 was identified as a promising therapeutic target for BPDCN patients. Frankel et al. launched the first pilot studies investigating a CD123-targeted agent which featured a novel recombinant protein drug consisting of a modified diphtheria toxin payload that was fused to recombinant human IL-3 [bib_ref] Chimeric Immunoglobulin-T Cell Receptor Proteins Form Functional Receptors: Implications for T Cell..., Goverman [/bib_ref]. The unique drug construct, originally known as "DT-IL3", was tested initially in patients with MDS and AML, and the trial also included a small number of patients with BPDCN [bib_ref] Expression of Immunoglobulin-T-Cell Receptor Chimeric Molecules as Functional Receptors with Antibody-Type Specificity, Gross [/bib_ref]. Although the drug was found to have modest single-agent activity in AML and MDS in these early clinical trials, what stood out was the early efficacy signal in BPDCN [bib_ref] Expression of Chimeric Receptor Composed of Immunoglobulin-Derived V Regions and T-Cell Receptor-Derived..., Kuwana [/bib_ref]. Therefore, Frankel et al. embarked on a pilot early phase study specifically focusing on patients with BPDCN with the same targeted therapy, which was renamed SL-401 (Stemline Therapeutics, New York, NY, USA) [bib_ref] Specific Activation and Targeting of Cytotoxic Lymphocytes through Chimeric Single Chains Consisting..., Eshhar [/bib_ref]. In this clinical trial, a total of 11 patients, all male with a median age of~70 years, were enrolled. Most patients were only able to receive one cycle (5 doses, days 1-5) of therapy. Among the 11 patients, 9 were deemed eligible for evaluation. The authors reported an overall response rate (ORR) of 78% (7 of 9 patients responding), including five complete responses (CRs) and two partial responses (PRs). Toxicity was notable for the vascular leak syndrome (VLS), which is now known as the capillary leak syndrome (CLS). The authors noted that CLS was found to be manageable overall and tracked with low or decreased serum albumin and weight gain [bib_ref] Specific Activation and Targeting of Cytotoxic Lymphocytes through Chimeric Single Chains Consisting..., Eshhar [/bib_ref]. The accompanying editorial by Fitzgerald noted the novel agent as a potential breakthrough in the rare disease field of BPDCN [bib_ref] Expression of Chimeric Receptor Composed of Immunoglobulin-Derived V Regions and T-Cell Receptor-Derived..., Kuwana [/bib_ref].
Building upon the momentum generated by these early results, Pemmaraju et al. sought to further investigate this agent in a larger population of patients with BPDCN. They conducted the first prospective, multi-institutional study of a targeted agent in BPDCN using this agent, SL-401 (now known as tagraxofusp). Including both frontline (FL) and relapsed/refractory (R/R) subjects, this four-stage study consisted of: (1) a dose-escalation 3 + 3 design (FL and R/R) for safety; (2) an expansion stage (FL and R/R); (3) a pivotal confirmatory stage focusing only on FL patients (n = 13); and (4) an expanded access fourth stage. The first three stages were reported in The New England Journal of Medicine by Pemmaraju and Lane et al. in 2019 demonstrating that among the first 29 patients enrolled and treated at the target dose of 12 µg/kg/day dosing, in the FL setting, SL-401 monotherapy yielded a 90% overall response rate including 72% rate of CR/CRc (CR or CR "clinical"-CR in all sites except minimal residual skin abnormality). They found that the median overall survival (OS) at 2 years was 52%, and 45% of patients were bridged to stem cell transplantation in the first-line treatment setting. In the R/R setting, a 67% ORR was observed among the 13 patients treated, with a median OS of 8.5 months. CLS was the most important toxicity, which was reported in approximately 20% of patients and was the cause of two deaths.
Based on these data, SL-401 (tagraxofusp) was granted US FDA approval on 21 December 2018, for patients with previously untreated or relapsed/refractory BPDCN ages 2 and older, and then later in the EU for adults for first-line treatment in January 2021 [bib_ref] The T-Body Approach: Potential for Cancer Immunotherapy, Eshhar [/bib_ref]. The CLS toxicity appropriately was designated as a "black box warning" on the package label insert. This approval marked an important milestone in the field, as it was not only the first targeted agent approved specifically for patients with BPDCN, but also the first ever CD123-targeted agent approved in hematology/oncology [bib_ref] Chimeric Antigen Receptor (CAR)-Engineered Lymphocytes for Cancer Therapy, Ramos [/bib_ref]. Tagraxofusp is now being tested in combination with azacitidine and venetoclax in clinical trials for patients with BPDCN, AML, and myelodysplastic syndrome (MDS) with early reports of safety and efficacy reported at the American Society of Hematology Meeting in 2021.
## Monoclonal or conjugated antibodies
Given the high and uniform expression of CD123 in all BPDCNs, several antibodybased drugs targeting CD123 have been developed. Early efforts were composed of iterations targeting CD123 via "naked" monoclonal antibodiesor antibody-drug conjugates (ADCs)in patients with AML, and plans to apply the same agents to BPDCN. Unfortunately, most did not reach the stage of testing in patients with BPDCN in clinical trials. However, from a disease-specific trial that is still ongoing, we now have data in patients with BPDCN using the CD123-ADC IMGN632 (ImmunoGen, Waltham, MA, USA).
IMGN632 is a humanized IgG1 monoclonal antibody specific for CD123 with a payload consisting of an indolinobenzodiazepine pseudodimer (IGN) with a peptide linker. IMGN632 binds to CD123-expressing cells, is internalized, and releases FGN849, which is a potent DNA alkylating agent. Treatment of CD123+ AML cells with IMGN632 causes DNA damage, S-phase cell cycle arrest, and apoptosis [bib_ref] In Vivo and in Vitro Sensitivity of Blastic Plasmacytoid Dendritic Cell Neoplasm..., Angelot-Delettre [/bib_ref]. IMGN632 potently kills the BPDCN cell line CAL1 and is active in vivo in patient-derived xenograft (PDX) models of BPDCN. Laboratory studies suggest that, by virtue of their lower CD123 expression, normal hematopoietic stem and progenitor cells are not as sensitive to IMGN632 as CD123+ BPDCN and AML leukemia cells [bib_ref] Extended Diagnostic Criteria for Plasmacytoid Dendritic Cell Leukaemia, Garnache-Ottou [/bib_ref]. IMGN632 is being tested as a single agent in patients with previously untreated or relapsed/refractory (R/R) BPDCN (ClinicalTrials.gov (accessed on 21 March 2022) Identifier: NCT03386513). Results in the first 23 patients with R/R BPDCN were presented at the American Society of Hematology meeting in December 2020. Seven of twentythree patients had an objective response (2 CR, 2 CRc, 1 CRi, and 2 PR) for an overall response rate of 30% (95% CI, 13-53%) and a composite complete remission rate of 22%. The duration of response for the four CR/CRc patients was between 3 and 9 months, all without receiving stem cell transplantation. The drug was well tolerated with no grade 3 or higher adverse events in more than one patient. The most common grade 1-2 events were nausea, peripheral edema, and infusion-related reactions. In contrast to tagraxofusp, no capillary leak syndrome (CLS) was observed. After these data were reported, the study was expanded to include previously untreated patients with BPDCN and is ongoing in the US and Europe.
As a result of this encouraging preliminary evidence of activity, IMGN632 received breakthrough therapy designation (BTD) from the US Food and Drug Administration in October 2020 specifically for treatment of BPDCN. This grants priority review to the agent and manufacturer for future evaluation by the agency. IMGN632 is also being tested in combination with azacitidine, venetoclax, or both, in a Phase 1b/2 study for patients with AML (ClinicalTrials.gov (accessed on 21 March 2022) Identifier: NCT04086264). Based on safety and efficacy data from this study, IMGN632 combinations might also be explored in BPDCN in the future.
## Bispecific antibodies
Bispecific antibodies are protein drugs that have two different antigen binding sites. In general, most bispecific antibodies bind to a target on tumor cells (e.g., CD123) and a target on T cells (e.g., CD3; forming a BiTE or Bispecific T-cell Engager) to bring the immune cell in proximity to the tumor cells. This results in more specific tumor cell killing [bib_ref] Phase I Clinical Study of Diphtheria Toxin-Interleukin 3 Fusion Protein in Patients..., Frankel [/bib_ref]. Modifications of the bispecific antibody constructs to improve activity or to recruit different immune cells, such as natural killer cells, have alternative names such as dual-affinity retargeting antibodies (DARTs) or bi-specific killer engager antibodies (BiKEs). Blinatumomab is an approved bispecific antibody for the treatment of B-cell acute lymphoblastic leukemia (B-ALL) that engages CD19 on leukemia cells and CD3 on T cells. There are no currently approved bispecific antibodies that target BPDCN surface antigens.
Results from trials using bispecific antibodies have not yet been reported in patients with BPDCN. However, encouraging results from trials testing CD123 bispecifics in AML suggest they have the potential to be active in both diseases. Flotetuzumab (MacroGenics; Rockville, MD) is an anti-CD123 x CD3 DART that has been evaluated in the setting of primary induction failure or early relapsed/refractory AML. Among 88 patients treated in an open-label phase 1/2 study, 30 patients who received the recommended phase 2 dose achieved a complete remission (CR)/CR with a partial hematologic recovery (CRh) rate of 26.7%. In patients who achieved CR/CRh, median overall survival (OS) was 10.2 months (range 1.87-27.27). A 10-gene expression signature predicted CR/CRh to flotetuzumab and correlated with bone marrow immune cell infiltration at baseline. Adverse events were mostly infusion-related reactions and cytokine release syndrome (CRS), largely grade 1-2 [bib_ref] Targeted Diphtheria Toxin to Treat BPDCN, Fitzgerald [/bib_ref]. Capillary leak syndrome was not a prominent toxicity. Flotetuzumab is currently being tested in a basket trial for relapsed/refractory CD123-positive malignancies, including BPDCN (NCT04681105), but data from this trial have not yet been reported.
Other bispecific antibodies targeting CD123 × CD3 that are being tested in patients with AML include the BiTEs vibecotamab (XmAb14045) [bib_ref] Activity of SL-401, a Targeted Therapy Directed to Interleukin-3 Receptor, Frankel [/bib_ref] and APVO436 [bib_ref] Approval of Tagraxofusp-Erzs for Blastic Plasmacytoid Dendritic Cell Neoplasm, Pemmaraju [/bib_ref] (that have been reported only in abstract form to date). Both have demonstrated activity in relapsed/refractory AML and have similar safety profiles as flotetuzumab, with relatively frequent but low-grade CRS. These agents may also be tested in BPDCN in the future.
## Car-t cells
A CAR is a chimeric antigen receptor composed of three domains: (i) an extracellular domain that determines specificity-this is the scFv (single-chain variable fragment) of a monoclonal antibody specific to a tumor antigen; (ii) an intracellular signaling domain derived from a T-cell signaling molecule; (iii) a transmembrane domain (hinge or spacer) that links the two preceding domains and plays a role in the conformation and accessibility of the receptor to its target [bib_ref] Novel Therapeutic Approaches in Blastic Plasmacytoid Dendritic Cell Neoplasm (BPDCN): Era of..., Wilson [/bib_ref]. This strategy allows for direct recognition of tumor antigens expressed on the cell surface, independent of major histocompatibility complex presentation to T-cell receptors. Furthermore, by using the scFv of an antibody, CARs can be used to recognize a wide range of structures including proteins and non-protein structures, such as carbohydrate antigens [bib_ref] Targeting CD123 in Hematologic Malignancies: Identifying Suitable Patients for Targeted Therapy, Patnaik [/bib_ref].
The first CARs were developed in the late 1980s and corresponded to the variable region of a monoclonal immunoglobulin for the extracellular portion and to the regions of a TCR for the intracellular region [bib_ref] Targeting CD123 in AML, Lane [/bib_ref] [bib_ref] Characterization of SGN-CD123A, A Potent CD123-Directed Antibody-Drug Conjugate for Acute Myeloid Leukemia, Li [/bib_ref]. However, the first "true" CAR was developed in 1993. This so-called "T-body" construct consisted of an scFv fused to the CD3ζ chain [bib_ref] IMGN779, a Novel CD33-Targeting Antibody-Drug Conjugate with DNA-Alkylating Activity, Exhibits Potent Antitumor..., Kovtun [/bib_ref]. This first generation of CARs provided proof of concept despite limited clinical effect [bib_ref] A CD123-Targeting Antibody-Drug Conjugate, IMGN632, Designed to Eradicate AML While Sparing Normal..., Kovtun [/bib_ref]. Indeed, cellular therapies modified with this type of CAR showed poor expansion and limited persistence. These weaknesses reflect a failure of T-cell activation by CARs in the absence of co-stimulatory molecules such as CD80 or CD86. The interaction with these co-stimulatory molecules is part of the three signals required for full activation of a CAR-T cell. Second-and third-generation CARs, including several co-stimulatory domains, have been developed. These CARs that include 4-1BB or CD28 or both domains have been evaluated as a mechanism to promote tonic signaling and enhance in vivo persistence [bib_ref] Flotetuzumab as Salvage Immunotherapy for Refractory Acute Myeloid Leukemia, Uy [/bib_ref] [bib_ref] Preliminary Results from a Phase 1 Study of APVO436, a Novel Anti-CD123..., Watts [/bib_ref].
The expected promise of CARs has been highlighted with the success of anti-CD19 CARs in ALL (acute lymphocytic leukemia) and NHL (non-Hodgkin's lymphoma), where complete remissions have been induced in numerous patients resistant to multiple lines of chemotherapy [bib_ref] CD19-Targeted T Cells Rapidly Induce Molecular Remissions in Adults with Chemotherapy-Refractory Acute..., Brentjens [/bib_ref] [bib_ref] Efficacy and Toxicity Management of 19-28z CAR T Cell Therapy in B..., Davila [/bib_ref] [bib_ref] Chimeric Antigen Receptor-Modified T Cells for Acute Lymphoid Leukemia, Grupp [/bib_ref] [bib_ref] Chemotherapy-Refractory Diffuse Large B-Cell Lymphoma and Indolent B-Cell Malignancies Can Be Effectively..., Kochenderfer [/bib_ref]. The first published clinical trial used a second-generation CD19specific CAR (CD28/CD3ζ) for the treatment of ALL in relapsed adult patients [bib_ref] CD19-Targeted T Cells Rapidly Induce Molecular Remissions in Adults with Chemotherapy-Refractory Acute..., Brentjens [/bib_ref]. In this trial, the authors evaluated 32 patients and observed a 91% response rate. Numerous clinical studies using CARs targeting CD19-positive malignancies then followed. Although each trial had its own criteria for patient recruitment and conditioning, as well as the configuration of the CAR, similar results were obtained with a response percentage between 70 and 100% in patients with leukemia and non-Hodgkin lymphoma [bib_ref] CAR Therapy: The CD19 Paradigm, Sadelain [/bib_ref] [bib_ref] Tisagenlecleucel in Children and Young Adults with B-Cell Lymphoblastic Leukemia, Maude [/bib_ref]. Since 2017, two CD19 CARs (Tisagenlecleucel [Kymriah], Novartis; and Axicabtagene ciloleucel [Yescarta], Kite Pharma/Gilead) have been approved in the USA first, then subsequently in Europe and several other countries.
Since 2013, several groups have developed and published preclinical studies using CD123-directed CAR-T cells [bib_ref] Cells Expressing CD123-Specific Chimeric Antigen Receptors Exhibit Specific Cytolytic Effector Functions and..., Mardiros [/bib_ref] [bib_ref] Rapidly Switchable Universal CAR-T Cells for Treatment of CD123-Positive Leukemia, Loff [/bib_ref] in AML and BPDCN [fig_ref] Figure 1: Therapies that target cell surface antigens in patients with BPDCN, highlighting tagraxofusp... [/fig_ref]. These CARs are often second generation, but some of them are third generation constructs based on the costimulatory and other domains used. Today, 28 phase 1 or 2 CD123 CAR trials are listed in clinicaltrials.gov (accessed on 21 March 2022), mainly in AML. Most trials are open in China using second-or third-generation CARs, but none have communicated results to date.
In the USA, several CD123 CAR-T trials are ongoing, some of which include patients with BPDCN [fig_ref] Table 1: Clinical trials using CAR-T cells targeting CD123 in patients with leukemia including... [/fig_ref]. Cellectis conducted a clinical trial in BPDCN using allogeneic engineered T cells expressing an anti-CD123 chimeric antigen receptor. This is a secondgeneration CAR (CD123 scFv-41BB-CD3ζ) with a safety switch based on the RQR8 depletion ligand (confers susceptibility to the anti-CD20 monoclonal antibody rituximab). The expression of the endogenous T-cell receptor αβ (TCRαβ) is abrogated through the inactivation of the TCRα constant (TRAC) gene, using Cellectis' TALEN gene-editing technology. In a pre-clinical study, they demonstrated a specific cytotoxic effect on BPDCN cells and prolonged survival in mouse xenograft models of BPDCN. In the phase 1 trial, one patient with BPDCN was treated with a dose of 6.25 × 10 5 CAR-T cells/kg (NCT03203369). Unfortunately, this patient had severe cytokine release syndrome (CRS) and died on day 9 after cell infusion. This trial has been closed for patients with BPDCN. However, the trial is ongoing in AML (NCT03190278) and is currently in dose escalation, with goals to evaluate the safety and clinical activity of UCART123v1.2 and to determine the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D). In the USA, several CD123 CAR-T trials are ongoing, some of which include patients with BPDCN [fig_ref] Table 1: Clinical trials using CAR-T cells targeting CD123 in patients with leukemia including... [/fig_ref]. Cellectis conducted a clinical trial in BPDCN using allogeneic engineered T cells expressing an anti-CD123 chimeric antigen receptor. This is a secondgeneration CAR (CD123 scFv-41BB-CD3ζ) with a safety switch based on the RQR8 depletion ligand (confers susceptibility to the anti-CD20 monoclonal antibody rituximab). The expression of the endogenous T-cell receptor αβ (TCRαβ) is abrogated through the inactivation of the TCRα constant (TRAC) gene, using Cellectis' TALEN gene-editing technology. In a pre-clinical study, they demonstrated a specific cytotoxic effect on BPDCN cells and prolonged survival in mouse xenograft models of BPDCN. In the phase 1 trial, one patient with BPDCN was treated with a dose of 6.25 × 10 5 CAR-T cells/kg (NCT03203369). Unfortunately, this patient had severe cytokine release syndrome (CRS) and died on day 9 after cell infusion. This trial has been closed for patients with BPDCN. However, the trial is ongoing in AML (NCT03190278) and is currently in dose escalation, with goals to evaluate the safety and clinical activity of UCART123v1.2 and to determine the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D). Mustang Bio developed an autologous CD123 CAR-T cell product called MB-102. This CAR has a second-generation construct composed of a CD123 single-chain variable fragment, an optimized IgG4 CH2CH3 linker, with CD28 and CD3ζ signaling domains. Previously, this CAR demonstrated a strong activity on CD123+ AML cells in vitro and in vivo [bib_ref] Cells Expressing CD123-Specific Chimeric Antigen Receptors Exhibit Specific Cytolytic Effector Functions and..., Mardiros [/bib_ref] without ablation of normal hematopoietic progenitors (CFU and BFU) from cord blood. A safety switch composed of EGFRt is incorporated in the construct, targetable by the anti-EGFR monoclonal antibody cetuximab. Today, two clinical trials are ongoing using this cell product (NCT02159495 and NCT04109482). The first one is a phase 1 trial aimed to study anti-tumor activity and safety of administration of ex vivo-expanded genetically modified T cells with two cohorts (AML and BPDCN). The second one is an expanded phase 1/2 study to assess the safety and efficacy of MB-102 in patients with relapsed or refractory BPDCN. To date, two patients with BPDCN have been treated with 100 × 10 6 cell doses. Complete remissions have been observed with well-tolerated sideeffect profiles, including only reversible and expected toxicities seen (e.g., all CRS was ≤grade 3). Mireya Paulina Velasquez's team in St. Jude Children's Research Hospital is also evaluating a CD123-specific CAR in patients with recurrent/refractory CD123+ disease (AML, B-ALL, T-ALL, or BPDCN) as a bridge-to-transplant phase 1 clinical study (CATCHAML, NCT04318678). This is a second-generation CAR with a CD28 H/TM region and CD28ζ signaling domain and includes a CD20 sequence as a safety switch [bib_ref] The Art and Science of Selecting a CD123-Specific Chimeric Antigen Receptor for..., Riberdy [/bib_ref]. The University of Pennsylvania is conducting two phase 1 clinical trials using second-generation CD123 CARs based on 4.1BB and CD3ζ with different delivery methods (RNA electroporated CAR, NCT02623582; lentivirally-transduced, NCT03766126). Another trial is recruiting at the Children's Hospital of Philadelphia (NCT04678336) to evaluate CD123 CAR-4.1BB-CD3z in pediatric AML. No results have been published to date.
In Europe, groups in France and Germany are developing CAR-T cells targeting CD123. The German group of Cartellieri et al. developed a rapidly switchable universal CAR-T platform called UniCAR that was redirected against CD123+ leukemia cells to allow a highly controlled and dose-dependent activation of the CAR-T cells [bib_ref] Rapidly Switchable Universal CAR-T Cells for Treatment of CD123-Positive Leukemia, Loff [/bib_ref]. This strategy is proposed to control off-target toxicities as an alternative to needing a suicide gene. This strategy is under clinical evaluation in AML, B-ALL, and BPDCN, with CD123 positivity of more than 20% of blasts is required for study entry (NCT04230265). To date, three patients with AML have been treated and no DLTs were observed. Mild and expected adverse events have been observed (grade 1 CRS, grade 1 fever) and all patients treated have shown a clinical response (two complete remissions with incomplete hematologic recovery, one partial response).
In France, Garnache-Ottou and team investigated the anti-leukemia efficacy and safety of a third generation lentiviral CD28/4-1BB CAR-T cell product targeting CD123 (CAR123), and provided strong preclinical rationale for the clinical assessment of this autologous cell therapy. This CAR is currently undergoing clinical translation to meet good manufacturing practice (GMP) requirements using a closed automated system. A phase 1/2 clinical trial to validate the clinical proof of concept of CAR123 for patients with BPDCN will be the next step and is planned.
## Promising new targets
[formula] 2.5.1. CD38 [/formula]
CD38 is a cell surface antigen that is expressed or overexpressed in several different hematologic malignancies including multiple myeloma (MM), T-cell acute lymphoblastic leukemia (T-ALL), and others including some BPDCNs. In BPDCN, CD38 is not usually included as part of the standard diagnostic workup, but of interest, several groups have demonstrated that it can be expressed in BPDCN, with perhaps 50% or more of cases being positive [bib_ref] Blastic Plasmacytoid Dendritic Cell Neoplasm with Leukemic Presentation: 10-Color Flow Cytometry Diagnosis..., Deotare [/bib_ref]. This finding is not only of potential diagnostic relevance, but also may be considered therapeutically important. The monoclonal antibody agent known as daratumumab is an already available, approved targeted therapy directed against CD38, currently used in patients with MM. Iverson et al. demonstrated monotherapy activity with daratumumab in a 70-year-old patient with untreated CD38+ BPDCN [bib_ref] Daratumumab for Treatment of Blastic Plasmacytoid Dendritic Cell Neoplasm. A Single-Case Report, Iversen [/bib_ref]. Another recent report by Mirgh et al. demonstrated efficacy of daratumumab in combination with bortezomib in a 75-year-old patient with BPDCN, who had extensive CD38+ disease involvement that was relapsed/refractory after standard treatments [bib_ref] Daratumumab-Based Therapy after Prior Azacytidine-Venetoclax in an Octagenerian Female with BPDCN (Blastic..., Mirgh [/bib_ref]. Therefore, it would be of interest for the BPDCN field to perform further studies both into expression of CD38+ in BPDCN and development of formal clinical trials for monotherapy and combination studies with daratumumab or other novel immunotherapies in CD38+ BPDCN.
## Ha-1h
CARs directed at cell-surface targets associated with other more common tumor types may also be adapted for treating BPDCN. For example, the minor histocompatibility antigen HA-1 is exclusively expressed on hematopoietic cells and is presented to the immune system in the context of HLA-A*02:01 [bib_ref] Optimization of the HA-1-Specific T-Cell Receptor for Gene Therapy of Hematologic Malignancies, Van Loenen [/bib_ref]. The antigenic HA-1H isoform is relatively common in the general population and TCR gene transfer can create HA-1H antigen-specific T cells, including in the setting of allogeneic stem cell transplantation. Therefore, a basket CAR-T cell trial is underway for patients with multiple types of leukemias, including BPDCN, that express HLA-A*02:01 and who have persistent or relapsed disease after an allogeneic SCT from an HLA-A*02:01 or HA-1H negative donor (clinicaltrials.gov (accessed on 21 March 2022): NCT03326921). No data have been released to date.
## Cd56
As CD56 is also expressed on BPDCN blasts, and this antigen could be a target to eliminate leukemia cells. Some strategies using antibody or CAR-T cells targeting CD56 are under evaluation in other pathologies, suggesting these strategies could also be evaluated in BPDCN. For example, a new CD56-targeting monomethyl auristatin E-conjugated antibodydrug conjugate is active in preclinical models of Merkel cell carcinoma [bib_ref] Adcitmer ® , a New CD56-Targeting Monomethyl Auristatin E-Conjugated Antibody, Is a..., Esnault [/bib_ref]. Similarly, a CD56-targeted CAR-T is active in models of small-cell lung cancer and neuroblastoma. These preclinical data indicate that CD56-targeted therapies merit further investigation as a potential treatment for CD56+ hematologic malignancies such as BPDCN. However, CD56 is also expressed on NK cells and a subset of T cells. Therefore, CD56-directed therapy, particularly if long-lasting such as via anti-CD56 CAR-T cells, may need to be used as a bridge to hematopoietic stem cell transplantation, which would eliminate residual CAR-T cells and restore CD56+ lymphocytes.
## Ilt3
Immunoglobulin-like transcript 3, ILT3, encoded by the gene LILRB4, is an important cell surface regulator of dendritic cell function [bib_ref] Immunoglobulin-like Transcript 3: A Crucial Regulator of Dendritic Cell Function, Vlad [/bib_ref]. ILT3 is highly expressed on normal pDCs and monocytes, and similarly is expressed on most BPDCNs and monocytic (FAB subtype M4/M5) AMLs. An ILT3/LILRB4-directed CAR-T cell preclinical model was active against monocytic AML cells and was not toxic to normal progenitors derived from normal CD34+ umbilical cord blood in vitro or humanized hematopoiesis in a mouse model [bib_ref] A Novel Anti-LILRB4 CAR-T Cell for the Treatment of Monocytic AML, John [/bib_ref]. These data suggest that an ILT3-directed immunotherapy could also be tested in patients with BPDCN.
# Conclusions
Despite the significant increase in disease-specific research on BPDCN in recent years, patients are still in need of better treatments. The success of tagraxofusp demonstrated the potential for targeting CD123 and for the discovery and evaluation of BPDCN-specific therapies as a feasible drug-development endeavor. This strategy may be particularly effective in the long-term when there are target antigens, like CD123, that are shared with other cancers. This can expand the reach of novel agents to more patients while also directly helping those with BPDCN. Several other classes of immunotherapeutic agents that also target CD123 are in development for other hematologic malignancies, as outlined here. We hope that design of preclinical experiments and clinical trials will include BPDCN where possible, to extend the toolbox of therapies in this still orphan disease. We also propose that there are even advantages to making the initial focus of an agent's development on an orphan disease indication, such as BPDCN, as "proof-of-concept." This is particularly helpful if the orphan disease is relatively homogenous between patients and if the target is highly expressed and/or essential for the biology of the malignant cell. Dramatic responses only require small patient numbers to support early clinical evaluation. For these reasons, we believe that several additional targets and therapies being tested in other malignancies also hold promise for co-development in BPDCN. A thoughtful approach in this way is essential to bring novel therapies to rare diseases that would otherwise be less likely to be selected for drug development on their own.
[fig] Figure 1: Therapies that target cell surface antigens in patients with BPDCN, highlighting tagraxofusp as the only approved therapy as well as others in clinical trials or agents and targets in earlier preclinical development. [/fig]
[fig] Author: Contributions: E.B.-R. and A.A.L. designed the study. E.B.-R., N.P., B.C., E.D. and A.A.L. wrote the manuscript, E.B.-R. drew the figure. All authors have read and agreed to the published version of the manuscript.Funding: This study was supported by Programme de Recherche Translationnelle INCa (PRTK N • PRT-K-20-107). This study was supported in part by the MD Anderson Cancer Centre Support Grant (CCSG) CA016672 and the MD Anderson Cancer Center Leukemia SPORE CA100632. AAL is a Scholar of the Leukemia & Lymphoma Society. Conflicts of Interest: N.P.: Board of Directors: Dan's House of Hope; Consulting: AbbVie, Aptitude Health, Astellas Pharma US, Inc., Blueprint Medicines, Bristol-Myers Squibb, Celgene Corp, Cimeio Therapeutics AG, ClearView Healthcare Partners, CTI BioPharma, Dava Oncology, Immunogen, Incyte, Intellisphere, LLC., Novartis AG, Novartis Pharmaceuticals Corp, OncLive (Owned by Intellisphere, LLC), Patient Power, PharmaEssentia, Protagonist Therapeutics, Sanofi-aventis, Stemline Therapeutics, Inc., Total CME; Financial Relationship (e.g. Stock, Royalty, Gift, Employment or Business Ownership): Karger Publishers; Scientific/Advisory Committee Member: Cancer.Net, CareDx, CTI BioPharma, EUSA Pharma, Inc., Novartis Pharmaceuticals Corp, Pacylex, PharmaEssentia; Speaker/Preceptorship: AbbVie, Aplastic Anemia & MDS International Foundation, Curio Science LLC, Dava Oncology, Imedex, Magdalen Medical Publishing, Medscape, Neopharm, PeerView Institute for Medical Education, Physician Education Resource (PER), Physicians Education Resource (PER), Postgraduate Institute for Medicine, Stemline Therapeutics, Inc. A.A.L.: Consulting: Qiagen; Research Support: AbbVie, Stemline Therapeutics. Other authors have no conflicts of interest. [/fig]
[table] Table 1: Clinical trials using CAR-T cells targeting CD123 in patients with leukemia including BPDCN and active non-CAR-T cell BPDCN trials (extracted from clinicaltrials.gov (accessed on 21 March 2022)). [/table]
|
Perceptions of treatment for tics among young people with Tourette syndrome and their parents: a mixed methods study
Background: Tourette syndrome (TS) among young people is associated with psychosocial difficulties and parents play an important role in the management of the condition. Clinical guidelines have been developed for the treatment of TS and tics, but little is known about how young people and their parents perceive their treatment options or their desired outcomes of treatment. The aim of this study is to explore perceptions of treatments for tics among young people with TS and their parents. Methods: In-depth interviews with 42 young people with TS and a mixed-methods, online survey of 295 parents of young people with TS. Participant recruitment was conducted through Tourettes Action (TA): a non-profit UK organisation for the support of people with TS. Interview transcripts were analysed using thematic analysis and responses to survey open-ended questions were analysed using content analysis. Triangulation of qualitative and quantitative data from the parents' survey and qualitative data from the interviews with young people was used to increase the validity and depth of the findings. Results: A strong theme was the perception that health professionals have limited knowledge of TS and its treatment. Medication was a common treatment for tics and both young people and parents described benefits of medication. However, adverse effects were frequently described and these were a common reason for stopping medication among young people. Aripiprazole was viewed most positively. Access to behavioural interventions for tics was limited and 76% of parents wanted this treatment to be available for their child. Some young people had reservations about the effectiveness or practicality of behavioural interventions. Reduction and abolition of tics were desired outcomes of treatment, but both parents and young people also identified the importance of increasing control over tics and reducing anxiety-related symptoms. For young people, managing the urge to tic was an important outcome of treatment. Conclusions: The results suggest a need for more training in the identification and management of TS and wider availability of behavioural treatments. Clinical trials could explore the effectiveness of Aripiprazole used in combination with psycho-educational interventions to reduce anxiety and promote a sense of control.
# Background
Previously thought of as a rare condition, the prevalence of Tourette syndrome (TS) in young people aged five to 18 years is now estimated to be between 0.4% and 3.8% with an overall prevalence of 1% [bib_ref] The prevalence and epidemiology of Gilles de la Tourette syndrome. Part 1:..., Robertson [/bib_ref]. TS is defined by multiple motor tics and the presence of at least one vocal tic over a period of one year [bib_ref] Müller-Vahl K. Tourette syndrome and other tic disorders in childhood, adolescence and..., Ludolph [/bib_ref]. Some young people with TS have been found to experience emotional difficulties, to feel different or abnormal because of tics [bib_ref] Exploring the impact of chronic tic disorders on youth: results from the..., Conelea [/bib_ref] and to report problems relating with peers. They are also likely to face educational problems and [bib_ref] The presence of attention-deficit hyperactivity disorder (ADHD) and obsessive-compulsive disorder worsen psychosocial..., Debes [/bib_ref] parents of young people with TS report difficulties with their child's behaviour [bib_ref] Parenting stress and related factors in parents of children with Tourette syndrome, Lee [/bib_ref] , with both of these difficulties exacerbated by co-occurring conditions such as attention deficit hyperactivity disorder [bib_ref] Tourette syndrome, parenting aggravation, and the contribution of co-occurring conditions among a..., Robinson [/bib_ref] [bib_ref] The presence of attention-deficit hyperactivity disorder (ADHD) and obsessive-compulsive disorder worsen psychosocial..., Debes [/bib_ref].
Some young people with TS require treatment to help them manage their tics and the associated psychosocial impairment [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part II:..., Roessner [/bib_ref]. Medication is the most common form of treatment for tics [bib_ref] A survey of Tourette syndrome patients and their families: the 1987 Ohio..., Bornstein [/bib_ref] [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] , particularly risperidone, clonidine and aripiprazole, among children and adolescents [bib_ref] Treatment practices in Tourette syndrome: the European perspective, Rickards [/bib_ref]. Behavioural interventions for tics such as habit reversal are used more rarely [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] although current clinical guidelines recommend that behavioural interventions for tics are given as a first-line treatment [bib_ref] Canadian guidelines for the evidence-based treatment of tic disorders: pharmacotherapy, Pringsheim [/bib_ref] [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref] [bib_ref] Current controversies on the role of behavior therapy in Tourette syndrome, Scahill [/bib_ref] and medication when tics are severe and distressing [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref]. Current guidelines also emphasise the importance of providing information about the condition to patients and family [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref] [bib_ref] Canadian guidelines for the evidence-based treatment of tic disorders: behavioural therapy, deep..., Steeves [/bib_ref] [bib_ref] Tic disorders, Leckman [/bib_ref]. Despite data on treatment utilisation patterns and recent clinical recommendations, little is known about how young people and key carers feel and think about different treatments for tics.
The few studies that have explored young people's and parents' perceptions of treatments for tics have mainly focussed on the diagnostic process or on barriers to individual treatments. A previous study of young people with TS found a median delay of 2.8 years from age of tic onset to age of diagnosis of TS [bib_ref] Limited knowledge of Tourette syndrome causes delay in diagnosis, Debes [/bib_ref]. Young people with TS and their parents recently interviewed in a Spanish study described difficulties receiving a diagnosis of TS that were associated with a lack of knowledge regarding tics among health professionals [bib_ref] The diagnosis of Tourette's Syndrome: communication and impact, Rivera-Navarro [/bib_ref]. Furthermore, a survey study found that a common barrier for treatment of tics was difficultiy finding knowledgeable treatment providers [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref]. There is some evidence suggesting that parents of young people with TS have concerns about adverse effects of medication for tics such as antipsychotics [bib_ref] Gilles de la Tourette syndrome: patient's knowledge and concern of adverse effects, Kompoliti [/bib_ref].
The aim of this study was to explore how young people with TS and parents of young people with TS perceive different treatment strategies for tics, and to explore similarities and differences between the views of parents and young people. We integrated qualitative data from in-depth interviews with young people with TS, with quantitative and qualitative data from an online survey of parents of young people with TS. Combining this evidence together increased the depth of the findings and permitted a better understanding of how young people with TS and their parents feel and think about their treatment options. This research was part of a larger project examining the benefits and risks of different types of treatment for tics in young people with TS (Health Technology Assessment reference 10/142/01).
# Methods
## Participants and recruitment
The study had research ethics committee approval from the Medical School Research Ethics Committee of the University of Nottingham (reference number H07062012).
## Survey of parents
Survey participants were parents or carers of young people with TS or any tic disorder (referred to as "parents" in the remainder of this paper). Exclusion criterion was being a parent or carer of a young person aged over 17 years. Participants were recruited through study adverts that were posted on the UK Tourettes Action (TA) web siteand on its electronic newsletter. In addition, invitation emails that included a link to the survey were sent by TA to its members, and study leaflets were distributed at TA events such as conferences and group meetings. Participants gave online consent to participate, the survey was anonymous and it was available for six consecutive months (from December 2012 to May 2013).
## Interviews with young people
Participants were children and young people with TS between 10 and 17 years of age (referred to as "young people" in the remainder of this paper). Recruitment was conducted through the online survey of parents, study announcements posted in TA's website and social media, and through study leaflets distributed at TA's informational events. Young people whose parents did not participate in the survey were still eligible to take part in the interviews.
Young people between 16 and 17 years of age, and parents of young people aged 10 to 15 years, were contacted to discuss the study in more detail, sent the information sheet in print, a consent form and a freepost envelope. Interviews were carried out at least one week after the initial contact. On the day of the interview, parents of young people aged 10 to 15 years were asked to provide verbal informed consent, which was digitally recorded with their permission. In addition, after answering any questions about the study, their children were asked to provide verbal consent to take part in the study and this was digitally recorded. Young people between 16 and 17 years of age provided verbal informed consent on the day of the interview.
The target sample size for the interviews was flexible and aimed to include up to 50 young people depending on the number of participants needed to achieve saturation of themes (i.e., the point in the study when no new themes are identified in the data). As past research has identified treatment barriers for tics among young people with TS [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] , participants who were in the process of receiving treatment or who have not received treatment at the time of the interview were included in the analysis. This allowed exploration of young peoples' perceptions of access to care and perceived barriers towards the use of different treatments for tics.
## Data collection survey of parents
Target sample size was 300 participants based on Parker et al.. The survey was designed to collect information about parents' perceptions of treatment for their children's tics. Survey questions were developed based on existing literature [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part II:..., Roessner [/bib_ref] [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref] [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] [bib_ref] Limited knowledge of Tourette syndrome causes delay in diagnosis, Debes [/bib_ref] [bib_ref] The diagnosis of Tourette's Syndrome: communication and impact, Rivera-Navarro [/bib_ref] [bib_ref] Gilles de la Tourette syndrome: patient's knowledge and concern of adverse effects, Kompoliti [/bib_ref] and revised and piloted by experts working with this population and service users. The survey included forced choice tick box responses (e.g., "yes" or "no"), rating scales, ranking scales and text boxes for open comments. Automatic skip patterns were used to move participants past questions that were not applicable. Areas covered by the survey included: medication for tics; behavioural interventions for tics; and desired outcomes of treatment. On average, the survey took approximately 30 minutes to complete.
For the child's clinical characteristics, parents were asked to report the child's age of tic onset, diagnosis of tic disorders and diagnosis of coexisting conditions. Open-ended questions probed responses to access to treatment and diagnosis of TS. Parents and carers rated the impact of tics on their child's self-esteem, social relationships and academic performance using the overall impairment scale of the Yale Global Tic Severity Scale (YGTSS) [bib_ref] The Yale Global Tic Severity Scale: initial testing of a clinician-rated scale..., Leckman [/bib_ref] , adapted for online self-report.
Parents' views on commonly used drugs (clonidine, risperidone, aripiprazole, sulpiride, haloperidol, olanzapine, quetiapine, pimozide, clonazepam and lorazepam) were sought, plus an option to comment on non-specified drug treatment. The 10 drugs were selected based on a recent survey of choice of medication for tics among members of the European Society for the Study of TS (ESSTS) [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part II:..., Roessner [/bib_ref] and our expert panel reports of current prescribing practices in the UK.
Participants were presented with a brief description of the comprehensive behavioural intervention for tics (CBIT) derived from the therapist guide for behavioural intervention for children and adults. The description included components of habit reversal and function-based intervention (i.e., contingency management).
If this treatment was not reported, parents were asked if they would like their child to be offered behavioural intervention for tics. Parents had the option to use text boxes for additional comments throughout the survey.
Finally parents were asked to rank, from most important to least important, seven potential positive outcomes of treatment for tics. In addition, parents were presented with two open-ended questions: i) "What would you most hope a treatment for tics would do for your child?"; and ii) "Are there any other benefits from treatment that you would like to see?".
## Interviews with young people
An interview schedule was developed in consultation with the expert panel and used to guide exploration of young people's thoughts and experiences concerning services and treatments for tics. For young people who had not received medication and/or behavioural intervention for tics, the interview schedule included a short verbal description of each of these treatments.
Interviews were conducted by telephone or, where practical, face-to-face. Where the young person had been unable to recall specific information (e.g., name of a drug), permission was sought to check with the parent. Interviews, which lasted, on average, 40 minutes (range 20 to 73) were digitally recorded and transcribed verbatim following completion. Any personal identifiers were removed from the transcript.
# Data analysis
To integrate the qualitative data of the interviews of young people with TS with the quantitative and qualitative data of the survey of parents a process of triangulation was used, where the data from interviews and the data from the survey were analysed separately and the results drawn together where appropriate [bib_ref] Three techniques for integrating data in mixed methods studies, O'cathain [/bib_ref]. This permitted the examination of agreement, complementarity or disagreement [bib_ref] Three techniques for integrating data in mixed methods studies, O'cathain [/bib_ref] between the findings from the interviews with young people with TS and the survey of parents.
The study reported here adhered to the standards of reporting qualitative research (RATS) [bib_ref] How to peer review a qualitative manuscript, Clark [/bib_ref].
## Parents' survey
The survey data were analysed using SPSS v19. The numerical data reported were calculated based on the number of responses to each item and exclude missing values.
Text data from open-ended questions were analysed using an inductive content analysis approach. Content analysis allows organisation of a large amount of text into fewer and meaningful categories [bib_ref] Three approaches to qualitative content analysis, Hsieh [/bib_ref]. The categories were developed by reading the participants' responses to gain an overall sense of the data; developing codes to capture key concepts; labeling the codes, organising the codes based on commonalities; and finally defining the categories [bib_ref] Three approaches to qualitative content analysis, Hsieh [/bib_ref].
# Interview analysis
The interview data were analysed using thematic analysis. The purpose of the analysis was to explore young people's needs and perceptions of treatments for tics. The themes were developed inductively because little is known about young people's perceptions of different treatment strategies for tics. This inductive approach is used when past research in a topic is scarce and it is not possible to develop theory or research driven themes. To become familiar with the data, the interview transcripts were read repeatedly and ideas about key aspects of the data were noted down to develop initial codes. These codes helped to organise the data into broader categories and they were revised and combined to identify possible themes. The themes and subthemes were discussed and refined following discussion with a qualitative expert leading the study (CG). In order to establish reliability a code book was developed which, for each theme, listed: i) the descriptive label; ii) the definition of the theme; iii) pointers to look for when identifying the theme; iv) exclusions to the theme; and v) a sample extract. To establish the reliability of the coding one or two extracts from each of the themes and subthemes (total n = 26) were presented, unlabelled, together with the code book to a qualitative researcher (RW) who was not involved in the study. The researcher, who was blind to the original coding, used the codebook to code themes and subthemes. Her ratings were compared to the original coding by the study research fellow (JC) and agreement was excellent (92.3%, 24/26). Analysis of the interview data revealed eight superordinate themes relating to the young people's experiences of treatment and care for TS of which seven had subthemes.
# Results
## Characteristics of parent sample
A total of 358 respondents consented to participate in the survey of whom 297 answered at least one question on treatment utilisation. Two entries were identified as repeat respondents and were removed from the data analysis, resulting in 295 participants with usable data. Of the 295 participants, 276 (93.6%) reported that their child had received a diagnosis of TS, eight (2.7%) reported a diagnosis of one or more tic disorders (3 vocal tic disorder, 2 transient tic disorder and only 3 reported one or more tic disorders) and 11 (3.7%) did not report a diagnosis of TS or tic disorder or had missing values on these items. All 11 reported age of onset for tics and were included in the analysis.
As shown in [fig_ref] Table 1: Characteristics of survey participants [/fig_ref] , the majority of participants were biological mothers (92.2%, n = 237) and the mean age of the sample was 44 years (SD = 6.3). Parents and carers described young people's demographic and clinical characteristics. Young people (234 males, 61 females) ranged in age from 5 to 17 (M = 12.4, SD = 3.0). The mean age of tic onset was 5.9 years (SD = 2.8, range 1 to 17) and the mean age at diagnosis of TS was 9.1 (SD = 2.7, range 3-17). On average, parents reported that in the last seven days their child's tic-related impairment was mild (M = 2.3, SD = 1.5, range 0 to 5). Sixty-six percent had one or more co-occurring conditions. Frequently reported co-occurring conditions were obsessive compulsive disorder (35.9%, n =106), ADHD (30.2%, n = 89), anxiety (24.7%, n = 73), autism spectrum disorder (20.3%, n = 60), depression (10.8%, n = 32) and learning disability (9.8%, n = 29).
## Characteristics of young people's sample
Eighty four potential participants provided contact details and 61 replied to the initial contact. Of these, four were excluded; three because service or treatment for tics was provided outside the UK and one was too young for inclusion. Of the remaining 57, five declined participation when contacted to arrange an interview, six could not identify a convenient time and four failed to keep their interview appointment. These 15 cases did not differ significantly in age and gender from the 42 young people who were interviewed (40 by telephone, 2 faceto-face). Of the 42 young people with TS who took part in the interviews, the majority were males (76.2%, n = 32) and the median age of the sample was 13 years (range 10 to 17) [fig_ref] Table 2: Characteristics of interview participants [/fig_ref]. Twenty-three (54.8%) participants had taken medication for tics (median number of drugs taken = 1, range 1 to5). Eight (19.0%) young people had received some form of behavioural treatment for tics (median Analysis of qualitative data from interviews with young people is presented below and linked, where appropriate, to survey responses and common themes which emerged from analysis of text box data from the parent survey.
## Need for access to informed and expert care
Young people spoke about the importance of receiving care from health professionals with a good understanding of TS, and about the challenges accessing specialist care. This was quite a strong theme expressed by 22 young people. Views clustered in three subthemes.
## Perceived lack of understanding of ts among health professionals
This subtheme captures how some young people felt that health professionals have insufficient knowledge to recognise TS and limited training in providing adequate treatment. Some young people described attending different health services to receive treatment for tics because many of the services they had visited had a poor understanding of their condition.
"…most of the places we have been to about my Tourettes like it seems like no one actually knows about it, like we know more than them… when we go there they usually ask us about it more than we ask them" [P 40]
## Difficulties accessing or maintaining specialist care
This subtheme describes how some young people felt that there were difficulties in receiving specialist treatment such as delays, cancellations or insufficient funding. This subtheme also captured young people's perceptions that they had received little or no information about TS and tics from their health professionals.
"…After I got diagnosed two years after that, that's when I started to get information about it. …I would have liked it to have been just more of a frequent thing when I actually needed the help more. Some sort of like guidance on things I could do, instead of just finding out that I had to deal with it all on my own." [P 32]
## Importance of receiving informed specialist care
This subtheme reflects how young people particularly valued receiving care from health professionals who had a good understanding of TS and tics. In relation to this, young people spoke about feeling understood and more confident, and they considered that knowledgeable health professionals could provide useful information to better understand their condition.
These views were strongly echoed in the survey data with parents. Text boxes probed parents' perceptions of access to care and the diagnosis of TS and the resulting content analysis showed that almost 32% (94/295) of parents spontaneously reported that it was difficult to get a referral to specialist treatment and to access appropriate care. Of these parents, 30 commented on the limited understanding of TS and tics among general practitioners (e.g., "We were told by our GP he would grow out of this habit and we were given antihistamines") [see Additional file 2]. Parents' comments also revealed that 16.3% (48/295) felt that receiving a diagnosis of TS was a long and difficult process. Moreover, 14.9% (44/295) of parents described receiving little information or support when the child received a diagnosis of TS.
## Concerns and limitations about taking medication for tics
This theme captured young people's negative perceptions of medication for tics based on their experience of taking medication or on their understanding of what this treatment involves. This was a major theme endorsed by 29 young people with three subthemes emerging.
## Perceived adverse effects of medication
This subtheme captures how young people who took medication for tics felt that it caused adverse effects. These effects were associated with different types of medication and typically included drowsiness, tiredness, self-reported depression, nightmares, weight gain and a sense of not being oneself. For some young people adverse effects were the main reason for stopping taking medication or for changing to a different drug, even if they perceived an improvement in tics.
## Perceived limited or lack benefit of medication for tics
This subtheme reflects how some young people who took medication felt that it left the tics unchanged, had a positive effect on tics for a limited period of time or that it worsened their tics.
"I was on a medication called risperidone for a while, which was helpful to start off with, it certainly had a noticeable effect but once I had got the level, my risperidone level sort of steady after about a year or after about six months it stopped being so effective so I went off it…" [P 29]
## Concerns about taking medication for tics
This subtheme describes how young people who had not taken medication for tics perceived potential difficulties about taking drug treatment; including potential adverse effects, difficulties to remember to take the drug, the need to take medication for a prolonged period of time and the unpleasant taste of medication. In general, young people queried whether taking medication would harm their health. Parents' responses showed they shared young people's concerns regarding adverse effects of medication. The survey showed that 54.7% (152/278) of parents mentioned that their child had taken medication for tics. As shown in [fig_ref] Table 3: Medication and behavioural interventions for tics as reported by parents [/fig_ref] , for most of the drugs examined, more than 46% of parents felt that the drug had moderate or severe adverse effects. Parents' free text comments also suggested that parents were concerned about a wide variety of adverse effects including, most frequently, sleepiness, tiredness or drowsiness and weight gain [see Additional file 3]. The level of tic-related impairment was significantly higher among the group of parents whose child had taken medication for tics (M = 2.6, SD = 1.6) compared with the group of parents whose child had not (M = 1.9, SD = 1.2), t(273) = 4.2, p < .001.
## Positive experiences of medication for tics
This theme describes how young people who had taken medication for tics felt that it helped them to reduce their tics or to have better control over them. For some young people, medication for tics allowed them to feel less self-conscious about their tics and to disguise them better when in public. The responses of 18 young people endorsed this theme.
More than a third (n = 7) of young people who endorsed this theme commented specifically on aripiprazole.
"I find aripiprazole helps quite a lot actually I am not as bad as I was. I don't have many tics during the day or anything anymore." [P 10]
In the online survey, parents also identified that medication could be helpful for the child's tics. However, parents' ratings of perceived helpfulness of medication were generally moderate. An exception to this was aripiprazole (see [fig_ref] Table 3: Medication and behavioural interventions for tics as reported by parents [/fig_ref]. In the interviews, of the 10 young people who had taken aripiprazole, six also described adverse effects, the most frequent being drowsiness or sleepiness (two young people reported dizziness, one shakiness and one drooling) but only one child had stopped taking it and this was because it had little positive effects rather than because of side effects.
## Behavioural interventions for tics is a 'natural' intervention that could be incorporated into daily life
Positive perceptions of behavioural interventions for tics were captured by this theme, expressed in two subthemes.
## Positive experiences of behavioural interventions for tics
This subtheme describes how young people who had received behavioural interventions for tics felt that it helped them to better recognise, control or manage their tics. Of the eight young people who had received behavioural interventions for tics, six interviews included this subtheme. Although the treatment process was often described as long, young people felt that it was generally helpful.
"…it took me a while but now it works more… I don't have to think oh I must bite my lip I just do it without thinking, but at the time I was just like oh this isn't going to work but it got better." [P 5]
## Perceived potential helpfulness of behavioural interventions for tics
This subtheme reflects how young people who had not received behavioural interventions for tics generally held positive views about this treatment. Of the 34 young people who had not received behavioural interventions for tics, 15 were represented in this subtheme. Young people felt that this treatment involved learning and practising behaviours that were similar to some of the behaviours they have used to control their tics, and they were described as "natural".
## Limitations of behavioural interventions for the treatment for tics
This theme captures young people's negative perceptions of behavioural interventions for tics based on their direct experience of this intervention or on their understanding of what this treatment involves. This theme consists of two subthemes.
## Negative experiences of behavioural interventions for tics
This subtheme describes how young people who had received behavioural interventions for tics felt that it was unhelpful and difficult to engage with. Of the eight young people who received behavioural interventions for tics, the responses of six endorsed this theme. Some young people felt that it was difficult to execute competing responses for motor or vocal tics, and one young person felt that behavioural interventions was changing "this to that" and remembered developing a tic from a competing response.
## Perceived potential difficulties of behavioural interventions for tics
This subtheme reflects how young people who had not received behavioural interventions for tics felt that this treatment would take long, considerable effort from their part or much support from others. Of the 34 young people who had not received behavioural interventions for tics, 14 perceived at least some limitations. Some young people felt that it would be difficult to remember to do the treatment exercises, that they could interfere with daily activities or be ineffective for "strong" tics.
In the online survey, 25.9% (74/286) of parents reported their child had received a behavioural intervention but in around half of these cases (36/69) the child received fewer than five sessions. Compared with parents whose child received fewer than five sessions and those who reported between five and 10, parents with more than 10 sessions had the highest scores on perceived helpfulness of behavioural interventions, yet these differences did not reach significance at p < .05. A higher proportion of parents whose child had received behavioural interventions for tics reported that their child had taken medication, χ 2 (1, n = 278) = 8.9, p < .01.
Eight percent (6/74) of parents commented on the helpfulness of the treatment and a similar number of parents identified limitations [see Additional file 4]. Parents (9.5%, 7/74) also commented on how the young person's age and tic severity influenced treatment engagement and outcome . Of note, 6.8% (5/74) of parents described difficulties receiving behavioural interventions for tics from knowledgeable treatment providers (e.g., "More training needed for professionals… and I speak as both parent and professional!").
In the interviews with young people, those who had not received behavioural interventions for tics had both negative and positive views on this treatment. In contrast, in the online survey most parents whose child had not received behavioural intervention for tics seemed to have positive perceptions of this treatment. Among parents whose child had not received behavioural interventions, the majority (76.3%, 148/194) wanted their child to be offered this treatment and felt that knowledge and availability of this treatment were limited (22.3%, 33/148) (e.g., "My daughter has asked time and again for this kind of help but it isn't available in X") [see Additional file 5]. Moreover, 10.8% (16/148) of parents expressed an interest in behavioural interventions, particularly as an alternative to medication. Among parents who did not desire behavioural interventions for their children, 19.6% (9/46) felt that their child did not need this treatment because the tics were currently minimal or manageable [see Additional file 5]. Another 19.6% (9/46) felt that behavioural interventions would not suit their child and 8.7% (4/46) commented that these interventions were unavailable when the child needed treatment.
Desire for a treatment to stop or reduce tics and the urge to tic Two thirds of young people (n = 28) spoke about what they hoped a treatment could offer them. Although commonly young people wanted a treatment to reduce tics, reducing the urge to tic emerged as a second important subtheme.
## Wanting to stop or reduce tics
This subtheme describes how young people would like their tics to be reduced, stopped, taken away or eliminated.
## Wanting to reduce urges to tic
Reflects how for some young people the urge to tic was perceived as uncomfortable or painful, and reducing it was seen as an important outcome of treatment.
"I would like something that would kind of stop the urge to do it all the time…" [P 34] Interestingly parents did not identify controlling the urge to tic as one of their desired outcomes of treatment.
## Need to manage emotional responses associated with tics
During the interviews young people also spoke about wanting a treatment that could help them to reduce or manage negative emotions associated with tics. This theme was endorsed by 19 young people and it consists of two subthemes.
## Negative emotions as contributing to tics
This subtheme describes how young people spoke of feeling worried, anxious, or stressed in relation to their tics. Some young people remembered experiencing these emotions in different situations, particularly social situations, and considered that these emotions worsened their tics.
"…it makes me worried and then because I am worried I do it [tic] more and then because I am doing it more and people are looking I do it even more than that." Interest and attempts to reduce negative emotions associated with tics This subtheme captures young people's efforts to reduce self-reported anxiety-related emotions that were associated with tics. For some young people, these emotions were described as more troublesome than tics and they spoke about trying to find ways to reduce them. Some young people described wanting a treatment that could help them to feel relaxed and calmed.
## The importance of gaining a sense of control over ts
Half of the young people (n = 21) spoke about experiencing little or no control over tics and identified a sense of control as an important outcome of treatment.
## Perceived lack of control over ts and tics
This subtheme reflects how some young people felt that their tics could "take over"; as if these had their own will or personality.
## Need to gain control over ts and tics
This subtheme describes how young people tried or would like to stop their tics at their instigation. They felt that this would allow them to engage in other activities, such as school activities, "without having to think my tics are going to get in the way" [P 23].
In the survey, parents ranked "reduction in severity and frequency of your child's tics" as the most important desired outcome of treatment [see Additional file 6]. This was confirmed in approaching half the text comments [see Additional file 7]. Again, in line with the young people's priorities, almost a quarter (64/295) of parents hoped that treatment for tics would help the child to reduce negative emotions associated with tics, which were often described as worries, anxiety and frustration (e.g., "…stop him feeling anxious… Let his body have some peace and be able to relax without ticking"), and 19.3% (57/295) felt that a treatment should help the child to control the tics [see Additional file 7].
# Discussion
This study explored perceptions of treatments for tics by conducting in-depth qualitative interviews with young people with TS and a mixed-methods survey of parents' views on their child's treatment for tics. There were many commonalities between the views of the separate groups of parents and young people. Young people and parents expressed strong concerns that health professionals have limited knowledge of TS and this was often associated with difficulties accessing and receiving evidence-based treatment. Young people and parents described the benefits of medication for tics, aripiprazole being favourably perceived, yet both groups also described concerns about adverse effects of medication and these were a common explanation for stopping medication among young people. Access to behavioural interventions for tics seemed to be limited and young people and parents had mixed views on these interventions, yet they were generally welcomed as having few adverse effects. Reducing or stopping tics was a desired outcome of treatment for young people and parents, but so was gaining a sense of control over tics and reducing or managing negative emotions which families attributed to being associated with tics. Young people particularly highlighted the need to manage the discomfort and distress associated with the urge to tic but this was not recognised by parents.
The findings of this study suggest that young people with TS and parents have had difficulties getting a diagnosis of TS, getting referred to specialist treatment and receiving appropriate treatment for tics. Consistent with past research that has found delays in the diagnosis of TS [bib_ref] Limited knowledge of Tourette syndrome causes delay in diagnosis, Debes [/bib_ref] , the survey of parents identified an average delay of three years between age of tic onset and age of diagnosis of TS. Parents' comments further suggested that getting a referral to specialist treatment was a long and difficult process. In many cases health professionals, particularly at primary care level, were perceived as having a poor understanding of tics and there were many examples of insensitive or ill-informed communications. In relation to this, in the interviews young people described visiting many services in attempts to receive treatment for tics and perceived that their condition was not well understood among health professionals. A previous study showed that approximately a quarter of health professionals lacked general knowledge of TS and only 52% were aware of its diagnostic criteria [bib_ref] What do those who know, know? Investigating providers' knowledge about Tourette's syndrome..., Marcks [/bib_ref]. Moreover, a recent case note study of children with TS presenting at a UK non-specialist service found that at first consultation 76.5% of GPs thought that the symptoms did not indicate TS or tic disorders, and only 12.5% of GPs referred the child to a specialist service [bib_ref] Tourette's: 20 CAMHS cases, Chilvers [/bib_ref]. Current clinical guidelines recommend that health professionals work closely with patients, parents and other key carers in providing information about tics and their treatment [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref] [bib_ref] Canadian guidelines for the evidence-based treatment of tic disorders: behavioural therapy, deep..., Steeves [/bib_ref] [bib_ref] Tic disorders, Leckman [/bib_ref]. However, in addition to difficulties with diagnosis and access difficulties, the findings of the present study indicate that some young people who had received medication or behavioural interventions for tics perceived that health professionals had insufficient training in providing treatment.
Consistent with previous survey studies, medication was a frequently reported treatment for tics among young people [bib_ref] A survey of Tourette syndrome patients and their families: the 1987 Ohio..., Bornstein [/bib_ref] [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] and risperidone, clonidine and aripiprazole were the three most commonly reported drugs [bib_ref] Treatment practices in Tourette syndrome: the European perspective, Rickards [/bib_ref]. Some young people and parents had positive perceptions of medication. For parents, aripiprazole was generally perceived as helpful for the child's tics and with few adverse effects. Children also perceived that aripiprazole was effective and it also appeared to be well tolerated although, since the doses of each target drug were not known, it is not possible to determine the impact of dosage. It may be that aripiprazole is better tolerated than other antipsychotic medication. There is some evidence suggesting that aripiprazole might be an effective treatment in reducing tics [bib_ref] Open-label study comparing the efficacy and tolerability of aripiprazole and haloperidol in..., Yoo [/bib_ref] and this has been confirmed in a recent randomised trial [bib_ref] A multicenter, randomized, double-blind, placebo-controlled study of aripiprazole in children and adolescents..., Yoo [/bib_ref]. Young people had varied experiences of medication for tics and benefits of medication were generally described in terms of a reduction of tics, better control of tics and feeling less self-conscious about tics. Adverse effects, however, were commonly described by young people and parents. For young people adverse effects varied, were experienced with different drugs and were a common reason for stopping medication which highlights the importance of informed choice with regard to medication.
The study suggests that access to behavioural interventions is limited and that young people and their parents have mixed views about this treatment; however, a common theme for young people and parents was the lack of adverse effects. A previous US survey of parents of young people with TS suggested only 23.6% received behavioural treatment [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref] : very close to the 25.9% reported by parents in the present study. In addition to limited availability the results of the parent survey and the interviews with young people suggest that many may not receive sufficient "dose" of behavioural intervention, with parents reporting that most of those who receive this treatment get fewer than five sessions and young people reported a median of four sessions. Eight or more sessions have been recommended for CBITand 12 for exposure and response prevention (ER) which is an extension of habit reversal [bib_ref] Exposure with response prevention versus habit reversal in Tourettes's syndrome: A controlled..., Verdellen [/bib_ref]. It is unclear why relatively few participants in the present study had received behavioural interventions for tics but it has been suggested [bib_ref] Health care needs of children with Tourette syndrome, Bitsko [/bib_ref] that this could be due to health professionals' lack of knowledge regarding implementation [bib_ref] What do those who know, know? Investigating providers' knowledge about Tourette's syndrome..., Marcks [/bib_ref] and difficulties finding treatment providers [bib_ref] Behavior therapy for Tourette's disorder: utilization in a community sample and an..., Woods [/bib_ref]. Parents' comments indicated that 22% of those who wanted behavioural interventions for their child's tics felt that health professionals lacked awareness or training in this approach.
Some parents and young people felt that behavioural interventions were helpful for managing tics, whereas others reported difficulties engaging with this intervention and a few parents and young people described negative reactions such as attributing a new tic to treatment exercises. Young people's negative perceptions of behavioural interventions seemed to be linked to the duration, motivation and effort needed to practise this treatment, rather than to adverse effects. Using data from two large CBIT trials of children and adults, a recent study examined frequently reported concerns of CBIT including the claim that this intervention requires considerable effort for patients and that it contributes to tic substitution [bib_ref] Current controversies on the role of behavior therapy in Tourette syndrome, Scahill [/bib_ref]. The study suggests that CBIT may not require considerable effort for patients since attrition in the trials was low (9.5% to 13.9%) and comparable to placebocontrolled medication trials [bib_ref] Current controversies on the role of behavior therapy in Tourette syndrome, Scahill [/bib_ref]. In addition, the study suggests that CBIT does not result in tic substitution since in both trials there was an improvement in the YGTSS scores [bib_ref] Current controversies on the role of behavior therapy in Tourette syndrome, Scahill [/bib_ref]. Since the present study examined perceptions of treatment for tics, its findings suggest that these concerns exist in some young people with TS and their parents, and that health professionals may need to address these concerns when providing behavioural interventions. There is evidence that behavioural interventions for tics can be effective in reducing tics [bib_ref] Behavior therapy for Tourette syndrome: a systematic review and meta-analysis, Wile [/bib_ref] and current clinical guidelines suggest such interventions as a first line treatment [bib_ref] Canadian guidelines for the evidence-based treatment of tic disorders: pharmacotherapy, Pringsheim [/bib_ref] [bib_ref] European clinical guidelines for Tourette syndrome and other tic disorders. Part III:..., Verdellen [/bib_ref] [bib_ref] Current controversies on the role of behavior therapy in Tourette syndrome, Scahill [/bib_ref]. To increase the accessibility of behavioural interventions for tics other forms of delivery could be further investigated, such as videoconference for which there is some preliminary evidence of its effectiveness [bib_ref] A randomized pilot trial comparing videoconference versus face-to-face delivery of behavior therapy..., Himle [/bib_ref].
A novel feature of these results was the importance that both parents and young people placed on managing the negative emotions associated with TS which were generally described as worries, anxiety and stress. Past research suggests that anxiety, stress and tension are linked to tic exacerbation [bib_ref] The influence of contextual factors on tic expression in Tourette's syndrome: a..., Conelea [/bib_ref]. The present study seems consistent with this, but further suggests that young people and parents hope that a treatment for tics helps them to manage anxiety-related symptoms. Treatment outcome measures could incorporate the evaluation of anxiety and stress related to tics. Gaining a sense of control over tics was also mentioned by young people and parents. In relation to this, a previous qualitative study with young people with TS identified that young people felt that they needed to control their tics. For young people, help to deal with the urge to tic was important but parents did not recognise this outcome as important. In this regard, ER is one behavioural intervention for tics that has shown to reduce the urge to tic by teaching the child to resist the urge as long as possible [bib_ref] Habituation of premonitory sensations during exposure and response prevention treatment in Tourette's..., Verdellen [/bib_ref]. Research is needed to explore the scope of tailored behavioural interventions to deliver treatment outcomes that are important to young people.
# Strengths and limitations
The survey had a large sample size (n = 295) which closely approximated the minimum of 300 recommended by Parker et al.to ensure valid results. By using an anonymous online survey, parents had the opportunity to describe both positive and negative perceptions of treatment for tics, whereas the individual interviews with young people might have reduced the potential influence of other young people's opinions if data were collected in groups. Although only 20% of the children who were the subject of the survey and 23% of young people who were interviewed were female this is a close reflection of the 20%-25% female prevalence observed in representative samples of children with TS [bib_ref] Submit your next manuscript to BioMed Central and take full advantage of:..., Knight [/bib_ref]. Participants were recruited through Tourettes Action (TA); membership or communication with TA may reflect a proactive population with common views on specific treatments for tics and the findings, based on a UK population, may not generalise to all young people with TS. Although we would have liked to recruit more participants who had direct experience with behavioural interventions for tics, the relatively small proportion of parents and young people who received this treatment might reflect its limited availability in a UK population. Recall bias may have influenced parents' and young people's perceptions of treatments for tics, particularly if the young person received treatment several years before data collection. Some young people chose to be interviewed with their parent present which might have influenced responses. However, young people seemed to talk openly and to consult the parent only when they were unable to remember details of treatment, such as a drug name or the time when they received an intervention. Interviews were conducted and analysed by one researcher (JC) who had previous experience of working with children and young people but who was not an expert on TS. As someone who not involved in the setting of the original research questions or in clinical care, arguably interactions with participants were not strongly driven by the interviewer's preconceptions about treatment.
# Conclusions
Many young people and their parents in the present study perceived that TS and tic treatments are generally not well understood among health professionals. Because they described difficulties accessing specialist treatment and appropriate care, it is in the interest of young people and their parents that professional training expands knowledge and awareness of TS and tic treatments among health professionals, particularly at primary care level. Since medication was a common treatment for tics and many parents and young people perceived adverse effects of medication, young people and their families may value receiving clear information regarding the rationale for using medication and its potential adverse effects. Behavioural interventions for tics were generally perceived
[table] Table 1: Characteristics of survey participants (n = 295) [/table]
[table] Table 2: Characteristics of interview participants (n = 42) [/table]
[table] Table 3: Medication and behavioural interventions for tics as reported by parents (n = 295) [/table]
|
Plasma GDF15 level is elevated in psychosis and inversely correlated with severity
Accumulating evidence suggests that GDF15 is a biomarker for ageing and morbidity of many somatic disorders such as cancer and inflammatory disorders. Recently, elevated serum GDF15 level was proposed as a marker for mood disorder. However, psychosis severity was not investigated in relation to plasma GDF15 levels. In the present study we measured GDF15 levels in plasma of 120 psychosis patients compared to 120 age and gender matched healthy controls. Within the patient cohort GDF15 levels were evaluated for association with age, gender, lifestyle factors, C-reactive protein levels, psychosis severity and metabolic disorder. Psychosis patients had elevated GDF15 levels compared to controls (median Psychosis = 744 ng/mL, median controls = 516 ng/mL, p < 0.001). Within the psychosis cohort, GDF15 levels, when corrected for age, metabolic health and lifestyle factors, were negatively correlated with psychosis severity (β = −0.218, p = 0.012). While GDF15 levels were elevated in patients versus healthy controls, the negative correlation between psychosis severity and GDF15 suggests a loss of anti-inflammatory GDF15 mediated functionality in severe psychosis. Study replication in larger cohorts will be necessary to assess the potential of GDF15 as a prognostic biomarker in psychosis.Psychosis is a pathological trait present within psychiatric disorders such as the schizophrenia spectrum of disorders (SSD) 1 , delusional disorder 2 and bipolar disorder (BD) 3 , which incapacitates sufferers from ascertaining reality 4 . SSD and BD share etiological determinants 5 , and are often accompanied by life-long social impairment 6 , decreased level of functioning and metabolic comorbidities, such as increased plasma fasting glucose levels and waist circumference 7, 8 . Elevated inflammatory markers have been reported in the peripheral blood of psychosis patients. In a meta-analysis, drug-naive first-episode psychosis (FEP) patients were found to have higher levels of interleukin-1β (IL-1β), interleukin-6 (IL-6), tumor necrosis factor alpha (TNF-α) 9 and cardio-metabolic disturbances 10-12 . C-reactive protein (CRP), secreted by the liver in response to IL-1β, IL-6, and TNF-α has also been found to be elevated in psychotic disorders such as schizophrenia (SZ) and BD 13, 14 . The CRP levels in those groups were reported to be further increased by female gender, higher body mass index (BMI) and elevated blood glucose levels[14][15][16][17][18].The link between peripheral inflammation and metabolic comorbidity in psychotic disorders is unclear. Peripheral inflammation has been implicated in the activation of inflammatory networks in the central nervous system (CNS) 19 . Correspondingly, increased levels of pro-inflammatory cytokines and kynurenine metabolites were found in cerebrospinal fluid (CSF) and serum of chronic SZ patients in independent investigations 20, 21 . The extent to which peripheral inflammation influences CNS immune activation and vice-versa in different conditions is yet to be explored. However, irrespective of its origin, there is a growing body of support for inflammation having a significant role in the pathology and overall prognosis of psychotic disorders 22, 23 . Thus to better understand psychosis it is crucial to accurately describe the alteration of inflammatory markers.Growth Differentiation Factor-15 (GDF15) is a member of the transforming growth factor-β (TGF-β) superfamily of cytokines, upregulated in response to injury with defined roles in regulating inflammation and apoptosis 24 . It is suggested to have an overall anti-inflammatory effect and is elevated in inflammatory scenarios associated with cancer, cardiovascular disease, insulin resistance and obesity 25 , as a compensatory Published: xx xx xxxx OPEN www.nature.com/scientificreports/ 2 SCIeNtIfIC REPORtS | 7: 7906 |
mechanism. GDF15 levels have been reported to correlate with CRP levels in certain inflammatory disorders. In addition to its growing relevance as a marker of inflammationand all-cause mortality, GDF15 has been described in the literature to have neurotrophicand anti-apoptotic properties. A proteomics investigation for biomarkers in mood disorders conducted by Frye et al., described GDF15 as a putative biomarker for mood disorders, particularly BD type I. Four additional proteins had increased levels in BD type I compared to controls, but GDF15 had the strongest effect size. The increased GDF15 levels were proposed to be driven by the disorder rather than by psychotropic medication. However, the authors stated that an assessment of psychotic features, which were particularly frequent in the BD type I group, could have provided greater understanding of the elevated GDF15 levels in BD type I. Therefore, we have investigated GDF15 in the context of psychotic disorders with a focus on relating the GDF15 levels to psychosis severity. Given the previously reported relationship between circulating GDF15 levels and metabolic disorder, also detected in Frye et al., we also considered metabolic comorbidity in the study.
The aims of the present study were to determine if plasma GDF15 levels (i) were different in psychosis patients compared to healthy controls, (ii) associated with severity of psychotic illness, and (iii) associated with degree of metabolic comorbidity and CRP levels in the patients. We hypothesized that plasma GDF15 levels would be (i) increased in psychosis, (ii) elevated by metabolic comorbidity, (iii) associated with CRP levels in patients.
# Materials and methods
Psychosis patient group. All patients in regular clinical treatment from several specialized psychosis clinics, mainly in Stockholm County, were as part of a general health screening asked to participate in the Swedish Study of Metabolic Risks in Psychosis (SMRP), approximately 1000 patients with psychosis were recruited. Clinical diagnoses of psychiatric illness were made according to the diagnostic and Statistical Manual of Mental Disorders, 4th Ed. (DSM-IV). Patients were assessed using psychiatric questionnaires with data on diagnosis, severity of mental illness reported by the clinician according to the Clinical Global Impression-Severity Index (CGI), duration of illness, type of medications and duration of treatment. Medication and dosage were confirmed by reference to medical records. Somatic health, including smoking status and alcohol consumption, was assessed with a self-reported questionnaire in the clinic under the supervision of a nurse. Body mass index and waist circumference were measured. The presence of psychiatric disease in first degree relatives was reported. Anti-psychotic pharmacotherapy along with all other variables used in analyses are detailed in . Venous blood, after an overnight fast, was collected. Whole blood and serum were stored at −80 °C until analysis. Three hundred of the patients participated in a follow-up, with identical protocol as at baseline, one year later. For the present investigation, 120 consecutively recruited patients, between 2005 and 2009, out of the 300 that had given blood samples at the one-year follow-up, were included in the study and data collected at the one year follow up were used . All participants gave written informed consent. Ethical approval for the study was obtained from the Stockholm Regional Ethics Review Board. All methods were performed in accordance with relevant guidelines and regulations stipulated by Karolinska Institutet and Stockholm Regional Ethics Review Board.
GDF15 levels in patient plasma were tested for association with age, gender, CGI, smoking status, alcohol use [g/week], as well as with variables describing the metabolic profile, levels of low-density-lipoprotein (LDL), high-density-lipoprotein (HDL), fasting glucose and insulin sensitivity marker homeostatic model of assessment for insulin resistance (HOMA-IR). CRP was used to indicate inflammation. Gender and smoking were dichotomous variables while the other variables were continuous. Non-smokers were those who did not smoke at all and all others were considered as smokers.
Healthy controls. Control subjects were selected from the population based Stockholm Diabetes Prevention Program (SDPP) which comprised ~5700 participants recruited during 1992-1998 who were followed-up eight to ten years later . A total of 120 unrelated Swedish SDPP subjects were selected by matching with the patients for age, gender and BMI. All subjects had normal glucose tolerance at baseline and follow-up and a family history of diabetes corresponding to the population average in Sweden. Of the 120 control subjects 62 (51.7%) were males and 58 (48.3%) were females. The median and interquartile range (IQR) of age of the controls was 53.5 (48.0-61.3) years. The BMI median and IQR of the controls was 28.4 (25.3-31.2) kg/m 2 . Of the 120 control subjects 29 (24.1%) had family history of diabetes at baseline and 30 (25.0%) had family history of diabetes at follow-up. The samples from healthy controls were collected at the same time point as patient samples, at the one year follow up of the study and treated in an identical manner.
## Measurement of routine metabolic and inflammation markers in the sz patients. serum lipid
profiles (HDL and LDL), plasma glucose and insulin levels were determined according to standard GLP protocols. High-sensitive C-reactive protein (hsCRP) levels were determined in serum using near infrared particle immunoassay (NIPIA). All analyses were done at the Karolinska University Hospital (KUH) Laboratory in accordance with KUH stipulated good clinical practices and regulations. Reference range forhsCRP was as follows, Low risk: less than 1.0 mg/L. Average risk: 1.0 to 3.0 mg/L. High risk: above 3.0 mg/L.
## Measurement of gdf15.
The plasma samples were assayed for GDF15 protein levels with the Quantikine ® ELISA Human GDF15 immunoassay (R&D Systems) according to the manufacturer's instructions. A subset of the samples were assayed with technical replicates. The optical density of sample wells was determined using the Microplate Reader by Thermo Labsystemsm Inc, and the software Multi-skan Ascent TM for processing the data. The GDF15 concentration of each sample was calculated using the measured absorbance against a standard curve. All samples had been freeze-thawed only once prior to analysis and all samples were run within the same week.
Statistical analyses. Analysis of covariance of GDF15 levels in plasma was carried out to ascertain the effect of psychiatric diagnosis on GDF15 levels when corrected for age and lifestyle factors known to affect GDF15 levels. Two-group comparison (psychosis patients versus controls) of GDF15 levels was performed using non-parametric Mann-Whitney U test. CGI and GDF15 levels were normalized using natural logarithmic transformation (ln). Correlation analyses to test for association between GDF15 levels and the independent variables age and hsCRP were performed using Spearman's rank correlation test.
To study putative predictors of plasma GDF15 levels, Ln GDF15 was assigned as a dependent variable and linear regression was performed using a stepwise method. The stepwise model is a semi-automated process of building a regression model based on the probability value of the F-statistic (p value) of the predictor regression coefficients. We applied a stepwise iterative method of regression modelling starting with no variables in the model, testing the addition of each variable with an entry requirement of p < 0.05 and a loss of significance at p > 0.. The distributions of regression residuals were checked to ensure normality and avoid multicollinearity.
Three linear regression models were constructed to ascertain the effects of 1) demography, psychosis and lifestyle 2) metabolic factors and 3) drug treatment on plasma GDF15 levels of the patients. The equation looking at demographic, psychosis and life style factors was as follows:
[formula] Ln GDF15 = b 0 + b 1 (age) + b 2 (gender) + b 3 (lnCGI) + b 4 (smoking) + b 5 (alcoholfollows: Ln GDF15 = b 0 + b 1 (age) + b 2 (gender) + b 3 (waist) + b 4 (LDL) + b 5 (HDL) + b 6 (glucose) + b 7 (Log HOMA-IR). [/formula]
The third equation which analyzed the effect of medication on GDF15 levels was constructed as follows: Ln GDF15 = b0 + b1(age) + b2 (gender) + b n (drug), where the presence or absence of n = 11 different anti-psychotic medications which patients of the cohort were treated with were tested for effects on GDF15. The number of patients on each drug is detailed in .
To compare the effects of metabolic factors to the effects of lifestyle, demographic and psychosis-severity factors we imported the significant variables from the three models to a fourth regression model and exposed it to the forward stepwise method. The equation of the model was as follows: Ln GDF15 = b0 + b1 (age) + b2 (gender) + b3 (lnCGI) + b4 (smoke) + b5 (HDL). Analyses were carried out using IBM SPSS Statistics version 23, (IBM Corporation, USA).
# Results
Characteristics of the psychosis patient group are shown in . Psychiatric diagnose categories in the psychosis patient group were schizophrenia, schizoaffective disorder, delusional disorder, psychosis unspecified, bipolar disorder and other. We assessed the effect of diagnosis on GDF15 levels using an analysis of covariance correcting for covariates proposed to influence GDF15 levels, that is age and smoking. There was no difference in mean GDF15 levels between the psychosis diagnosis groups (p = 0.260), therefore the psychosis patient group was treated as one group in the following analyses.
Psychosis patients had significantly higher plasma GDF15 levels compared to healthy controls: median patients = 744 ng/mL, median controls = 516 ng/mL, p < 0.001 . Forty psychosis patients (33%) had hsCRP levels above high risk reference, greater than 3.0 mg /L, median Ps = 2.05 mg/L [IQR: 0.923-4.00 mg/L]. No correlation was observed between GDF15 and hsCRP levels (Spearman's ρ = 0.148, p = 0.11, n = 120, Standardized β = 0.123). In the psychosis group and healthy controls, the GDF15 plasma levels were significantly correlated with age at sampling. Psychosis patients: Spearman's ρ = 0.496, β Standardized = 0.494, p < 0.001. Healthy controls: Spearman's ρ = 0.342, β Standardized = 0.338, p < 0.001. Association between GDF15 levels and severity of psychotic illness was assessed within the psychosis group through the severity score CGI. A multiple linear regression model constructed in a stepwise way, including the predictors: age, CGI, gender, alcohol use and smoking, returned all predictors but alcohol use as statistically significant. The model explained 37.8% of the variance in GDF15 levels (adjusted R 2 = 0.378, F = 14.8, n = 91). Higher CGI scores (more severe psychosis) were associated with lower GDF15 levels (β standardized = −0.221, t = −2.58, p = 0.012). Increasing age was associated with higher GDF15 levels (β standardized = 0.491, t = 5.49, p < 0.001). Smoking was associated with higher GDF15 levels (β standardized = 0.313, t = 3.75, p < 0.001). The male gender was associated with increased GDF15 levels (β standardized = −0.176, t = −2.00, p = 0.049).
GDF15 levels were tested for association with metabolic profile through waist circumference, LDL, HDL, fasting blood glucose and HOMA-IR levels. The multiple linear regression model constructed in a stepwise method to identify putative metabolic predictors of GDF15 levels, was adjusted for age and gender. Age and HDL levels were returned as significant variables in a model that explained 30.9% of the variance in GDF15 levels (adjusted R 2 = 0.309, F = 22.5, n = 96). Higher age was associated with higher GDF15 values (β standardized = 0.590, t = 6.661, p < 0.001). Higher HDL levels were associated with lower GDF15 levels (β standardized = −0.234, t = −2.643, p = 0.01). The variables gender, waist circumference, LDL, glucose and HOMA-IR levels were not associated with GDF15 levels.
The effect of anti-psychotic medication on circulating GDF15 levels was assessed through multiple linear regression constructed in a stepwise method. The model was constructed to include age, gender, smoking status and drug treatment. Significant variables in this model that explained 33.6% of the variance in GDF15 levels (adjusted R 2 = 0.336, F = 25.5, n = 97) included age (β standardized = 0.490, t = 5.92, p < 0.001), and smoking . Metabolic characteristics of the psychosis patients (n = 120). IQR: interquartile range. HDL: highdensity-lipoprotein. LDL: low-density-lipoprotein. Log HOMA-IR: log 10 of homeostatic model of assessment for insulin resistance.
status (β standardized = 0.338, t = 4.09, p < 0.001) but there was no significant contribution from variables indicating drug treatment. A final model was created to compare metabolic effects with lifestyle, demographic and psychosis severity factors in their contribution to GDF15 level variance. The model included the factors: age, smoking, gender, CGI and LDL and we found that the variables age, smoking, gender and lnCGI could explain 35.9% of the variance in GDF15 levels (adjusted R 2 = 0.359. ANOVA: F = 13.8, n = 95, p < 0.001). Higher age was associated with higher GDF15 levels (β standardized = 0.514, t = 5.81, p < 0.001). Smoking was associated with higher GDF15 values (β standardized = 0.269, t = 3.22, p < 0.001). Higher CGI, an indication of more severe psychosis, was associated with lower GDF15 levels (β standardized = −0.188, t = −2.55, p = 0.012). HDL levels did not improve subsequent model building and thus were omitted by the regression method from the final model. The results from all regression analyses are summarized in.
# Discussion
GDF15 has been implicated in the pathology of various disorders, with both adverse and beneficial effects depending on the cells' state and microenvironment. Its elevated level in inflammatory states is suggested to be due to a compensatory anti-inflammatory effect. In the present study, we show for the first time that GDF15 is elevated in plasma of psychosis patients when compared to healthy age and gender matched controls. In addition, we show in our psychosis patient cohort, that GDF15 levels are lower in patients with a more severe psychosis. A potential description of its role as a protective agent was previously proposed in the context of cardiovascular disease, obesity and inflammatory response in vivo. GDF15 has been investigated in the context . GDF15 and high-sensitive C-reactive protein (hsCRP) levels in psychosis patients and healthy controls **p < 0.001 of neurodegeneration, dementia, BD type I and II and unipolar depression, and found to be elevated in plasma of these patient groups compared to healthy controls. Some patients with BD, especially BD type I, have episodes of psychosis, but GDF15 levels have not previously been studied specifically in relation to psychosis. Psychosis is associated with accelerated cellular ageing, and the life expectancy in psychosis patients is decreased by 15-17 years 51 . Our finding, that GDF15 is increased in psychosis patients, is in line with the notion that GDF15 concomitantly increases with morbidity and all-cause mortality.
In order to be able to propose GDF15 as a potential prognosis biomarker in psychosis, further studies delineating the role of GDF15 in CNS pathways, especially where they intersect with the pathogenesis of psychosis, are required. In mammals it is known that GDF15 is produced by the placenta, liver, kidney, heart and CNS. Additionally it is secreted by the choroid plexus (ChP) into the CSF by the ChP where it serves as a neurotrophic, anti-apoptotic and anti-inflammatory molecule.
The serotonergic and dopaminergic systems have been implicated in psychosis, therefore it is of specific interest that Gdf15 promotes survival and differentiation of embryonic rat dopaminergic neurons and serotonergic raphe neurons. Strelau et al., have shown that Gdf15 is neurotrophic and neuroprotective to rat neurons in vitro and in vivo. They described that Gdf15 rescued intoxicated dopaminergic neurons of the 6-hydroxydopamine (6-OHD) model of Parkinsonism. In a mouse study using Gdf15 knockout mice they showed that Gdf15 is neurotrophic to sensory and motor neurons. More recently Macahado et al., have described the importance of Gdf15 in the survival of dopaminergic neurons through a similar model of Parkinsonism as in showing that Gdf15 knockout mice are more severely affected than the wild-types with a greater loss of dopaminergic neurons in response to 6-OHD intoxication.
A growing body of evidence suggests that inflammation plays a role in psychosis etiology and disease progression. An increased expression of astrocyte markers was observed in the grey matter of SZ patients suggestive of neuroinflammation. Neuronal deficit caused by increased apoptosis, precipitated by microglial activation and pro-inflammatory cytokine release, has been linked to the neurodegenerative effects observed in SZ patients such as white matter disorders. The association between SZ related neurodegeneration, and cognitive decline has been established by a long-term follow up investigation. Interestingly, GDF15 levels have been found to be elevated in cognitive decline and age-related dementia. GDF15 levels have been reported to be elevated in multiple inflammatory scenarios and found to be concomitantly increased with astrocytic gliosis. Lastly, GDF15 regulates leukocyte recruitment to areas of inflammation through regulating leukocyte integrin activation. Taken together this could suggest that a role of GDF15 in SZ pathogenesis is related to preventing inflammation and neuro-degeneration.
The observation of a negative relationship between GDF15 and psychosis severity (CGI) within our psychosis cohort, is seemingly at odds with our finding that GDF15 was elevated in the patients compared to healthy controls, and the consensus of literature that higher GDF15 level is associated with increased disease morbidity. A plausible explanation for this could be that GDF15 is anti-inflammatory and thus higher levels of GDF15 could contribute to milder psychosis related co-morbidity and improved outcomes. GDF15 has been reported to have a beneficial effect in in vitro and in vivo. Gdf15 is released upon myocardial infarction, inhibiting polymorphonuclear leukocyte recruitment and inflammatory response which reduced the rate of pathology after myocardial infarction in rodents, and increased production of Gdf15 led to smaller atherosclerotic lesions in the ApoE −/− mouse model of atherosclerosis. Transgenic mice expressing human GDF15 were found to have reduced white adipose tissue and inflammatory response, a potential indication of protection against the inflammatory metabolic syndrome. Apart from the dopaminergic and serotonergic neurons, Gdf15 has been shown to have a neurotrophic effect on the motor and sensory neurons. Our finding that GDF15 levels were lower in the more severely afflicted psychosis patients could reflect the loss of anti-inflammatory and neuroprotective capabilities in these patients. While in this study we have analyzed patient plasma for GDF15 and correlated this with psychosis severity, analyzing the CSF could prove to be a more attractive biological sample for the purpose of understanding the role of GDF15 in psychosis. In a study looking at glioblastoma, Shnaper S et al. have shown that elevated levels of GDF15 in the CSF is associated with worse patient outcomes indicating that our findings in patient plasma needs to be contrasted with CSF GDF15 levels. In a separate study showed that Gdf15 delivered through the CSF promotes hippocampal neurogenesis and synaptic activity in an in vivo Alzheimer's disease model. While we found that plasma GDF15 levels were lower in the patients who had higher psychosis severity scores, the CSF GDF15 levels could in fact be elevated in the worst outcomes by ChP mediated filtration of GDF15 from systemic circulation to the CNS marking a shift in its relative prevalence aimed at conferring a protective and compensatory function to the CNS.
CRP is an indicator of inflammation which can be present at chronic low level in psychosis patients especially in those with metabolic comorbidity, which is common particularly in those who undergo certain anti-psychotic drug treatments. In the present study we did not detect a significant correlation between GDF15 levels and CRP levels within the psychosis patients. Multiple meta-analyses have shown that CRP levels are moderately elevated in SZ and BD, and, the elevated levels are increased by age, BMI and hyperglycemia, as GDF15 levels are. However, sex dependence seems different between CRP and GDF15 plasma levels, with lower CRP levels and higher GDF15 levels in males. This, together with that plasma CRP is secreted mainly by hepatocytes, that is, cells different from those secreting plasma GDF15, possibly contributes to the poor correlation between GDF15 and hsCRP. Also acute inflammation may blunt an association between GDF15 and CRP levels. Accordingly, the limited studies that looked at both GDF15 and CRP levels in the disorders chronic obstructive pulmonary disease, patients undergoing hemodialysis, subclinical atherosclerosis 63 , anaemiaand rheumatoid arthritishave arrived at discrepant results, where the GDF15-CRP correlations vary from strongly statistically significant in the former two studies to non-significant in the latter studies. To our knowledge our study is the first to compare these inflammation related markers in psychosis patient plasma. Age, gender, smoking and metabolic disorder have been associated with increased GDF15 levels in plasma or tissue specific GDF15 secretion. Accordingly, we found positive associations for GDF15 levels to age, male gender and smoking. However, in our cohort the metabolic factors were relatively less important than the psychosis severity index in explaining GDF15 levels. In agreement with findings in mood disorder patients, we detected no effect of type of antipsychotic medication on the GDF15 levels. However, the sample size was small for certain drugs. Thus, our findings propose that plasma GDF15 levels capture a process partly different from inflammation indicated by CRP levels or metabolic comorbidity.
Although the patient cohort in this study is heterogeneous as their DSM-IV-based diagnoses suggest, all patients were recruited from psychosis-specialized outpatient clinics in Sweden. The patients had all been diagnosed with a severe psychosis and impaired functionality. Our choice to include all these psychosis patients followed the Research Domain Criteria (RDoC) paradigm 67 which traverses clinical diagnoses by looking at domains of human behavior and/or functioning. We found no statistically significant differences in GDF15 levels between the DSM-IV-based diagnoses, however this may have been due to an inadequate sample size. In place of a symptom severity scale such as the positive and negative symptom scale 68 (PANSS), which was unavailable to us in this study, we used the clinician-rated CGI-S, an indication of the patients' mental health, psychosis and wellness. The lack of a symptom severity scale like PANSS is a main limitation of this study.
To conclude, we report that plasma GDF15 levels are elevated in chronic psychosis patients compared to healthy controls, and that the GDF15 levels are negatively correlated with psychosis severity. An assessment of plasma GDF15 levels in relation to psychosis has not previously been reported, thus our findings warrant replication efforts as well as an investigation of GDF15 levels as a prognostic marker for psychosis. |
Molecular mechanisms of zika virus pathogenesis: An update
[bib_ref] Pathogenesis of flavivirus infections: Using and abusing the host cell, Fernandez-Garcia [/bib_ref] [fig_ref] Figure 1: Genetic organization of Zika virus [/fig_ref] [bib_ref] Flaviviruses: Braking the entering, Pierson [/bib_ref] [bib_ref] Flaviviruses: Braking the entering, Pierson [/bib_ref] [bib_ref] Flaviviruses: Braking the entering, Pierson [/bib_ref]
## Infection of zika virus
The principal mode of ZIKV transmission is through mosquito bite particularly by Aedes aegypti and Aedes albopictus. ZIKV transmission can occur through unprotected intercourse and vertical transmission has been observed from affected mother to child during pregnancy, parturition and breastfeeding. ZIKV has been detected in various body fluids such as breast milk, urine, secretions of female genital tract, amniotic fluid, saliva, cerebrospinal fluid, aqueous humour, conjunctival fluid and nasopharyngeal area hence, transmission can occur by exchange of any body fluids between affected and unaffected individuals. The [fig_ref] Figure 2: Vertical transmission and congenital Zika Syndrome [/fig_ref] shows vertical transmission of ZIKV. During initial 2-7 days of the incubation time, infected human patients serve as a source of ZIKV with low viral load and may transmit the virus through body fluids [bib_ref] Flavivirus structural heterogeneity: Implications for cell entry, Rey [/bib_ref].
Flaviviruses deposit on skin epidermis after the mosquito bite where they encounter cells such as keratinocytes and Langerhans cells, which are highly permissive to virus. It has been reported that the entry of the virus is through the interaction of E protein with TAM (Tyro3, Axl, and Mer), TIM (T cell/transmembrane, immunoglobin, and mucin), AXL (AXL receptor tyrosine kinase) or DC-SIGN (Dendritic cell-specific intracellular adhesion molecule-3-G rabbing N on-integrin) receptors [bib_ref] Role of skin immune cells on the host susceptibility to mosquito-borne viruses, Briant [/bib_ref]. Dendritic cells are the first cells targeted by flaviviruses. Once infected, these migrate to different lymphoid organs where viral replication occurs, thereby allowing the flavivirus dissemination into the circulation and internal organs [bib_ref] Flaviviruses: Braking the entering, Pierson [/bib_ref]. ZIKV in pregnant women infects placental cells, causing their apoptosis and damage to vasculature causing placenta more permeable, facilitating entry into subsyncytial cytotrophoblasts [bib_ref] Zika virus damages the human placental barrier and presents marked fetal neurotropism, De Noronha [/bib_ref]. Post-breaching of placental barrier, it replicates in various cell types including foetal macrophages and endothelial cells and infects the Hofbauer cells present in the villi of foetal capillaries. These infected cells can function as a reservoir of ZIKV replication and finally disseminate the virus into foetal blood [bib_ref] Zika virus productively infects primary human placenta-specific macrophages, Jurado [/bib_ref]. ZIKV has the ability to target human foetal brain causing cell death in various regions of the brain leading to microcephaly. The presence of ZIKV has been shown in cortical neural progenitors, radial glial cells and newly differentiating neural progenitors of developing mice [bib_ref] Recent African strains of Zika virus display higher transmissibility and fetal pathogenicity..., Aubry [/bib_ref] [bib_ref] Vertical transmission of Zika virus targeting the radial glial cells affects cortex..., Wu [/bib_ref]. It also infects microglial and astroglial cells [bib_ref] Expression analysis highlights AXL as a candidate Zika virus entry receptor in..., Nowakowski [/bib_ref] [bib_ref] Zika virus infects human fetal brain microglia and induces inflammation, Lum [/bib_ref]. In vitro studies suggest that ZIKV can infect human microglial cell line (CHME3), human primary astrocytes, human neural progenitor cells, human brain organoids, mouse astrocytes and neural stem cells (NSCs) efficiently [bib_ref] Zika virus productively infects primary human placenta-specific macrophages, Jurado [/bib_ref] [bib_ref] MicroRNA and mRNA dysregulation in astrocytes infected with zika virus, Kozak [/bib_ref] [bib_ref] Differential virulence between Asian and African lineages of zika virus, Simonin [/bib_ref] [bib_ref] Disruption of glial cell development by Zika virus contributes to severe microcephalic..., Li [/bib_ref] [bib_ref] Zika virus dysregulates the expression of astrocytic genes involved in neurodevelopment, Shereen [/bib_ref]. The capability of virus in infecting almost all the cell types of the brain results in severe consequences on the foetal brain.
## Receptors involved in zikv entry
The virus entry into the host cells is facilitated by the interaction of ZIKV structural E protein with host receptors. Several cell surface receptors are implicated in flavivirus entry into the various host cells types [bib_ref] Flavivirus entry receptors: An update, Perera-Lecoin [/bib_ref]. Among the primary candidate receptors, best characterized are α v β 3 integrins, C-type lectin receptors, phosphatidylserine receptors TIM and TAM [bib_ref] Zika virus infects human fetal brain microglia and induces inflammation, Lum [/bib_ref] [bib_ref] West Nile virus discriminates between DC-SIGN and DC-SIGNR for cellular attachment and..., Davis [/bib_ref] [bib_ref] The TIM and TAM families of phosphatidylserine receptors mediate dengue virus entry, Meertens [/bib_ref]. Flaviviruses are known to produce a wide variety of structurally different virions, and perhaps this is the reason for their possible infectivity to various tissues, and hence the ability to infect a variety of cell types in its hosts [bib_ref] Flavivirus structural heterogeneity: Implications for cell entry, Rey [/bib_ref].
Various receptors are suggested to be involved in ZIKV entry including DC-SIGN, AXL, Tyro and TIM-1/4 entry/adhesion factors. In a study conducted by Hamel et al [bib_ref] Biology of Zika virus infection in human skin cells, Hamel [/bib_ref] , expression of candidate receptors like DC-SIGN or AXL showed enhanced viral infection in HEK293T cell line. Similarily, Tyro3-expressing HEK293T cells were also found to be highly permissive for ZIKV at 24 h post-infection (hpi), whereas expression of TIM-1 or TIM-4 showed marginal effect on ZIKV entry into HEK293T cells. There are contradictory studies regarding the role of AXL receptor in ZIKV entry into the brain cells. Meertens et al [bib_ref] Axl mediates ZIKA virus entry in human glial cells and modulates innate..., Meertens [/bib_ref] reported that the AXL receptor was responsible for virus entry into human microglia and astrocytes. The proposed mechanism suggests that the ZIKV upon interaction with AXL receptor ligand Gas6 activates its kinase activity, which results in downregulation of interferon (IFN) signalling and facilitates infection. This study also described that ZIKV infection in human glial cells could be inhibited by blocking the AXL receptor activity, for example, MYD1, a synthetic AXL decoy receptor and AXL kinase inhibitor R428 (1) ZIKV is spread from bite of mosquitoes and potentially crosses placental barrier and infects developing brain cells of the foetus. (2) ZIKV infects BMECs that form the blood-brain barrier. The BMECs release interferon β (IFNβ), IFN λ and other interferon-stimulated genes leading to the disruption of blood-brain barrier. (3) The virus infects NSCs, the primary cell type in the developing brain and induce apoptosis, UPR and alter self-renewal of NSCs. (4) Infected NSCs have altered neurogenesis and differentiating NSCs enter apoptosis in the presence of ZIKV or its proteins. (5) Zika virus also infects oligodendrocytes and mature neurons but with low propensity. (6) Infected astrocytes release MCP1 and IFN α in response to ZIKV. The virus also induces UPR and autophagy in astrocytes. (7) Microglial cells infected with ZIKV respond robustly to release various cytokines and chemokines such as IFN α, IL-6, MCP1, TNFα, IL-1β, IL 8, MIP-1α and inducible nitric oxide synthase (iNOS). The release of various inflammatory and proinflammatory molecules from astrocytes and ZIKV suggests a severe inflammatory scenario in the developing brain.
successfully inhibited the ZIKV infection. In another study by Hastings et al [bib_ref] TAM receptors are not required for Zika virus infection in mice, Hastings [/bib_ref] , it was suggested that TAM receptors-Tyro3, Axl and Mertk were not required for ZIKV entry into the various organs of the mouse. The study involved multiple knockout mice, namely AXL−/−, Mertk−/−, AXL−/−Mertk−/− and AXL−/− Tyro3−/− and showed that all these mutants could be infected efficiently with ZIKV suggesting that TAM receptor including AXL might not be required for ZIKV entry and infection in rodents, also indicating that these receptors could have relevance only to human and nonhuman primates. Besides, induced pluripotent stem cells (iPSCs) are resistant to ZIKV infection despite expressing high levels of TYRO3, a TAM receptor [bib_ref] Genetic ablation of AXL does not protect human neural progenitor cells and..., Wells [/bib_ref].
In another independent study to identify the role of AXL in virus entry, Chen et al [bib_ref] AXL promotes Zika virus infection in astrocytes by antagonizing type I interferon..., Chen [/bib_ref] showed that human astrocytoma cell line U-251MG expressed high levels of AXL receptors and was more susceptible to ZIKV, whereas T98G, glial cell line, known to express low levels of AXL, showed low susceptibility to ZIKV. The virus induces cytopathogenic effects and apoptosis in U-251MG cell line. AXL knockout U-251MG cell line was resistant to ZIKV infection. However, these experiments suggested that ZIKV entry was independent of the AXL receptor. The physiological function of AXL receptor activation is to block IFN signalling; hence, its ablation causes enhanced IFN signalling in AXL knockout cells, which is responsible for resistance to ZIKV. The authors described an explicit finding that the AXL signalling promoted ZIKV infection but was not necessary for its entry [bib_ref] AXL promotes Zika virus infection in astrocytes by antagonizing type I interferon..., Chen [/bib_ref]. It is essential to identify the candidate receptors to decipher mechanisms of ZIKV neurotropism and its pathogenesis to discover new approaches toward designing therapeutics and vaccines against the virus.
## Complications in zikv infection
Zika fever: The incubation period of flavivirus is around 2-7 days after the mosquito bite. In adults, one in five patients infected with ZIKV show symptoms [bib_ref] Zika virus outbreak on Yap Island, Federated States of Micronesia, Duffy [/bib_ref]. In contrast with dengue and chikungunya fever, the onset of symptoms is difficult to assess in zika fever since there is no abrupt clinical onset [bib_ref] Detection of Zika virus in urine, Gourinat [/bib_ref] [bib_ref] Detection of Zika virus in saliva, Musso [/bib_ref]. The mild intensity illness lasts from a span of days till weeks in case of adult infections. No haemorrhagic events associated with zika fever have been reported to date, distinctive in other flavivirus fever episodes.
Guillain-Barré syndrome: Guillain-Barré syndrome (GBS) is an acute self-limited polyneuropathy. GBS is a demyelinating disease of the peripheral nervous system, also referred to as acute inflammatory demyelinating polyneuropathy. It represents a range of symptoms from a very mild case with a brief weakness to devastating paralysis, leaving the person unable to breathe independently. The likely cause of the onset of GBS pertains to an autoimmune process [bib_ref] Guillain-Barré syndrome, Donofrio [/bib_ref].
GBS has also been observed in other arboviral infections, such as Dengue virus (DENV) and Chikungunya virus (CHIKV) [bib_ref] Dengueassociated neuromuscular complications, Garg [/bib_ref]
## Congenital zika syndrome (czs):
In a study of pregnant women with ZIKV infection, abnormally high rates of early miscarriages and intrauterine foetal demise were reported. In about 20 per cent of ZIKV-infected pregnancies, impaired foetal growth and placental dysfunction have also been reported [bib_ref] Detection of Zika virus infection in Thailand, Buathong [/bib_ref] [bib_ref] Zika virus infection and stillbirths: A case of hydrops fetalis, hydranencephaly and..., Sarno [/bib_ref]. Foetuses born with ZIKV infection have congenital zika syndrome (CZS).
CZS comprises a recognizable pattern of structural anomalies and functional disabilities in central and peripheral nervous system. By definition, severe microcephaly represents head circumference below the third percentile for sex and age. High risk of severe microcephaly cases has been reported from 1 to 15 per cent in confirmed prenatal ZIKV infections [bib_ref] Zika virus and the risk of developing microcephaly in infants: A systematic..., Antoniou [/bib_ref] [bib_ref] Maternal infection with Zika virus and prevalence of congenital disorders in infants:..., Nithiyanantham [/bib_ref]. Manifestations of ZIKV-induced severe microcephaly are predominantly neurological. Imaging studies have also described cortical disorders, ventriculomegaly, lissencephaly, cerebral calcifications, periventricular, cortical and subcortical zones calcifications, cortical development errors, hydrocephaly and pachygyria/ agyria [bib_ref] Zika virus associated with microcephaly, Mlakar [/bib_ref] [bib_ref] Congenital brain abnormalities and Zika virus: What the radiologist can expect to..., Soares De Oliveira-Szejnfeld [/bib_ref]. Craniofacial disproportion is also observed in the affected newborn [bib_ref] with congenital Zika virus infection without microcephaly at birth -Brazil, Van Der Linden [/bib_ref]. Neurological anomalies are present in affected newborns later in life with normal head circumference also [bib_ref] with congenital Zika virus infection without microcephaly at birth -Brazil, Van Der Linden [/bib_ref].
In a clinical study, 141 children (age 1-14 yr) who suffered congenital ZIKV infection and in all the infants, structural brain abnormalities were detected by neuroimaging studies. These children were assessed for the prevalence of epilepsy. The incidence of epilepsy was high (67%), and the average age of infants at the onset of epilepsy was 4.9 months. In 74 per cent of infants, the seizures started to occur during the first six months of life. The main seizure types were epileptic spasms (72%), focal motor seizures (21%) and tonic seizures (4%) [bib_ref] Epilepsy profile in infants with congenital Zika virus infection, Van Der Linden [/bib_ref]. Another follow up study conducted by the National Center on Birth Defects and Developmental Disabilities (NCBDDD), Centers for Disease Control (CDC) described that 19 children were born with severe microcephaly, positive for ZIKV infection and most of them suffered from severe functional limitations such as motor impairment, seizure disorders, hearing and vision abnormalities and sleep difficulties (age group 19-24 months post-ZIKV infection) [bib_ref] Health and development at age 19-24 months of 19 children who were..., Satterfield-Nash [/bib_ref].
## Susceptibility of central nervous system (cns) cells to zikv
Neural stem cells (NSCs): The ventricular zone (VZ), sub-VZ (SVZ) and hippocampal dentate gyrus are well-defined germinal niche of mitotically active NSCs. Given adequate cues, NSCs differentiate into neurons, astrocytes or oligodendrocytes and maintain NSC population by the process of self-renewal (symmetric cell division) [bib_ref] Mammalian neural stem cells, Gage [/bib_ref]. Any effect on proliferation and differentiation properties of NSCs challenges neurodevelopmental processes. Numerous in vitro and in vivo studies have proved that NSCs are highly susceptible to ZIKV infection [bib_ref] Zika virus disrupts neural progenitor development and leads to microcephaly in mice..., Li [/bib_ref] [bib_ref] Zika virus infects human cortical neural progenitors and attenuates their growth, Tang [/bib_ref] [bib_ref] Dysregulation of Ephrin receptor and PPAR signaling pathways in neural progenitor cells..., Thulasi Raman [/bib_ref]. The a summary of the effects of ZIKV on various brain cell types. Recent literature describes that induced pluripotent stem cells (undifferentiated form) are less permissive to ZIKV infections 55 ; however, the cells differentiated from iPSCs such as NSCs are permissive to ZIKV [bib_ref] Biology of Zika virus infection in human skin cells, Hamel [/bib_ref]. Micrographs of infected neurospheres showed ZIKV bound to the plasma membrane, endoplasmic reticulum membranes, in mitochondria and vesicles of cells [bib_ref] Zika virus impairs growth in human neurospheres and brain organoids, Garcez [/bib_ref]. ZIKV induces apoptosis in NSCs and neurospheres. Growth areas of brain organoids from human iPSC also significantly reduced upon ZIKV exposure 57 . Fibroblast growth factor 2 (FGF2) signalling has been shown to be involved in replication of ZIKV in NSCs 58 . Li et al 59 described that Asian strain of ZIKV (SZ01) infects embryonic mouse brain. The mouse embryo was infected at day 13.5 and inspected post three and five days of infection. ZIKV efficiently replicated in the mouse brain and targeted mainly SOX-2 (stem cell marker) and Ki67 (proliferation marker) positive cells. ZIKV infection induced proliferation arrest, apoptosis and impaired differentiation of NSCs leading to cortical thinning and hence resulting in microcephaly. The Global gene expression studies of ZIKV-infected brains revealed upregulation of candidate flavivirus entry receptors and dysregulation of genes associated with immune responses, cell death and microcephaly [bib_ref] Zika virus disrupts neural progenitor development and leads to microcephaly in mice..., Li [/bib_ref].
ZIKV reduces expression of adherens junction proteins which help in anchoring radial glial cells (RGCs) and alter their properties in mouse brain. These adherens junction proteins include ZO-1 (zonula occludens-1), SMAD7 (β-catenin, SMAD family member 7) and NUMBL (NUMB like endocytic adaptor protein). ZIKV NS2A protein, when expressed in developing brain, also led to a similar reduction in adherent junction proteins [bib_ref] Zika-virus-encoded NS2A disrupts mammalian cortical neurogenesis by degrading adherens junction proteins, Yoon [/bib_ref]. ZIKV impairs brain development by inducing apoptosis in NES (neuroepithelial stem) cells and RGCs. Another study discovered induction of caspase-1 and gasdermin D (GSDMD)-mediated pryroptotic cell death in neural progenitor cells, in response to ZIKV infection 61 .
ZIKV induces mitosis defects in human NSCs including formation of multipolar spindle, chromosome laggards, smaller nuclei and death of the progeny following cell division. An increased frequency of aneuploidy, such as monosomy, trisomy and polyploidy was observed in infected cells [bib_ref] Zika virus infection induces mitosis abnormalities and apoptotic cell death of human..., Souza [/bib_ref] [bib_ref] Zika virus increases mind bomb 1 levels, causing degradation of pericentriolar material..., Wen [/bib_ref]. ZIKV infection perturbed centrosome (the hallmark of premature differentiation) culminating into depletion of stem cell niche, impaired VZ, disrupted neurogenesis and ultimately cortical thinning [bib_ref] Recent Zika virus isolates induce premature differentiation of neural progenitors in human..., Gabriel [/bib_ref]. DNA damage response has also been observed in ZIKV infected NSCs [bib_ref] Zika virus infection induces DNA damage response in human neural progenitors that..., Hammack [/bib_ref]. Global transcriptome analysis (RNA-seq) of ZIKV infected human NSCs derived from iPSCs showed a drastic change in the transcriptome of virus-infected NSCs [bib_ref] Bioinformatic analysis reveals the expression of unique transcriptomic signatures in Zika virus..., Rolfe [/bib_ref]. Gene ontology (GO) analysis of differentially expressed transcripts highlighted the involvement of downregulated transcripts in cell-cycle-related pathways and upregulated transcripts in transcription, protein transport and catabolic processes [bib_ref] Zika virus infects human cortical neural progenitors and attenuates their growth, Tang [/bib_ref]. In an independent study, transcriptome analysis of human embryonic stem cells derived brain organoids showed activation of innate immune response post-ZIKV infection [bib_ref] Zika virus depletes neural progenitors in human cerebral organoids through activation of..., Dang [/bib_ref]. Toll-like receptors-3 (TLR3) were upregulated in ZIKV infected brain organoids, and the crucial role of TLR3 in ZIKV neuropathogenesis has been predicted. The growth of ZIKV infected neurospheres and organoids could be rescued with TLR3 competitive inhibitor (thiophenecarboxamide propionate), confirming the role of TLR3 67 . Two major immune pathways, nuclear factor κB (NF-κB) and signal transducer of activators of transcription (STAT), are highlighted by transcriptomic analysis. Activation of these two pathways is responsible for the increase in inflammatory molecules post-host-virus interaction;
the study points to chronic inflammatory conditions of the brain post-ZIKV infection [bib_ref] Bioinformatic analysis reveals the expression of unique transcriptomic signatures in Zika virus..., Rolfe [/bib_ref].
Autophagy plays a pro-viral role in ZIKV replication. The virus induces autophagy in human foetal NSCs as indicated by the microtubuleassociated protein 1A/1B-light chain 3, LC3-I to LC3-II conversion [bib_ref] Zika virus infection induces mitosis abnormalities and apoptotic cell death of human..., Souza [/bib_ref]. Mouse embryonic fibroblasts (MEF) with autophagy-related gene 3 (Atg3), an essential component for autophagy, knockout showed reduced ZIKV replication as compared to wild-type MEFs suggesting that ZIKV exploits host autophagosome machinery for its replication [bib_ref] Zika virus NS4A and NS4B proteins deregulate Akt-mTOR signaling in human fetal..., Liang [/bib_ref]. ZIKV proteins NS4A and NS4B when co-overexpressed in human foetal NSCs, resulted in reduced neurosphere growth and induced autophagy in cells. NS4A and NS4B also impaired differentiation of NPCs into neurons and astrocytes in vitro [bib_ref] Zika virus NS4A and NS4B proteins deregulate Akt-mTOR signaling in human fetal..., Liang [/bib_ref]. ZIKV replication suppressed phosphorylation of Akt at Thr308 and Ser473, hence reducing mTOR phosphorylation at Ser2448. Similar response was found when NS4A and NS4B were expressed together [bib_ref] Zika virus NS4A and NS4B proteins deregulate Akt-mTOR signaling in human fetal..., Liang [/bib_ref]. Sofosbuvir, an Food and Drug Administration (FDA) approved antiviral drug, protects NES cells from ZIKV-induced cell death, perhaps because the compound also inhibits the re-localization of pTBK1 to mitochondria (pTBK1 plays a significant role in inducing innate antiviral immunity and cell proliferation) [bib_ref] Zika virus disrupts phospho-TBK1 localization and mitosis in human neuroepithelial stem cells..., Onorati [/bib_ref]. ZIKV infection also alters secreted proteome of mouse NSCs. Specific proteins crucial during repair or neuronal damage and proteins that antagonize and promote infection are found to be upregulated in the secretome of ZIKV infected NSCs. TLR3 was also upregulated in NSCs post-ZIKV infection in secretome [bib_ref] Zika virus induces astrocyte differentiation in neural stem cells, Lossia [/bib_ref].
Overexpression of ZIKV E protein in human foetal NSCs (fNSCs) reduces proliferation and leads to cellcycle arrest at the G0/G1 phase. E protein induced apoptosis in differentiating NSCs at day three and hampered the migration of newly differentiating cells from neurospheres [bib_ref] Zika virus E protein alters the properties of human fetal neural stem..., Bhagat [/bib_ref]. Similar effects were observed in the mouse brain in the SVZ region when E protein was electroporated in utero at embryonic day 13.5 and analyzed at embryonic day 15. It was observed that proliferation was attenuated as assessed by a significant decrease in Ki67 71 .
MicroRNA (miRNA) are non-coding regulatory small RNA (19-24 nucleotide long) molecules. These molecules downregulate their target genes by binding [bib_ref] Zika virus E protein alters the properties of human fetal neural stem..., Bhagat [/bib_ref]. mir-1273g-3p and mir-204-3p were highly expressed in response to E protein in NSCs, interestingly, mir-1273g-3p and mir-204-3p are molecular targets of important development-related genes paired box gene 3 (PAX3) and notch receptor 2 (NOTCH2), respectively [bib_ref] Zika virus E protein alters the properties of human fetal neural stem..., Bhagat [/bib_ref]. This study provided compelling data to suggest the crucial role of ZIKV E protein and miRNA circuitry in ZIKV neuropathology.
In human infections, ZIKV was found to trigger endoplasmic reticulum (ER) stress and unfolded protein response (UPR) in the cerebral cortex of ZIKVinfected post-mortem foetuses [bib_ref] Stress-induced unfolded protein response contributes to Zika virus-associated microcephaly, Gladwyn-Ng [/bib_ref]. A similar response was also observed in cultured human NSCs and mouse embryos [bib_ref] Stress-induced unfolded protein response contributes to Zika virus-associated microcephaly, Gladwyn-Ng [/bib_ref]. ER stress markers -calnexin, calreticulin, MX, PDI1, PERK-ATF4 and ratio of spliced/unspliced XBP1 were upregulated post-ZIKV infection in fNSCs and mouse brain [bib_ref] Stress-induced unfolded protein response contributes to Zika virus-associated microcephaly, Gladwyn-Ng [/bib_ref]. Induction of DNA methylation was observed in human cerebral organoid-derived neural precursor cells, astrocytes and neurons following ZIKV infection [bib_ref] Zika virus alters DNA methylation of neural genes in an organoid model..., Janssens [/bib_ref]. ZIKV was also found to modulate long non-coding profile of human NSCs and alternate splicing pattern [bib_ref] ZIKV infection effects changes in gene splicing, isoform composition and lncRNA expression..., Hu [/bib_ref].
ZIKV modulates the fate of NSCs in the developing brain. ZIKV infection increased glial fibrillary acidic protein (GFAP) and aldehyde dehydrogenase 1 family member A (ALDH1a)-markers of astrocytic differentiation, in mouse NSCs indicating differentiation of NSCs towards glial lineage [bib_ref] Zika virus induces astrocyte differentiation in neural stem cells, Lossia [/bib_ref]. Induction of glial differentiation could be to initiate an immune response against ZIKV [bib_ref] Zika virus induces astrocyte differentiation in neural stem cells, Lossia [/bib_ref]. ZIKV proteins NS4A and NS4B when co-overexpressed in NSCs also inhibited neural differentiation 68 . Moreover, ZIKV E protein also induced massive apoptosis in differentiating human foetal NSCs into neurons [bib_ref] Zika virus E protein alters the properties of human fetal neural stem..., Bhagat [/bib_ref]. ZIKV NS5 is localized to the nucleus and during mitosis, re-localized to centrosomes. Cells expressing NS5 took longer to complete the cell cycle, suggesting that ZIKV induced cell cycle arrest by translocation of NS5 to the nucleus and by interacting with several centrosome proteins [bib_ref] Zika virus NS5 localizes at centrosomes during cell division, Kesari [/bib_ref]. Hence, the virus has the capability to deregulate the fate of NSCs that may impact the overall pool of brain cells.
Neurons: Lack of functional neuron and impaired neurogenesis are hallmarks of ZIKV neuropathogenesis. The primary target of ZIKV is NSCs; however, ZIKV can infect mature and immature neurons in vitro as well as in vivo in mouse and human model systems though at low efficiency [bib_ref] Zika virus infects human cortical neural progenitors and attenuates their growth, Tang [/bib_ref] [bib_ref] Zika virus persistently and productively infects primary adult sensory neurons in vitro, Swartwout [/bib_ref] [bib_ref] Zika virus directly infects peripheral neurons and induces cell death, Oh [/bib_ref]. ZIKV targets newly forming special AT-rich sequencebinding protein 2 (SATB2+) neurons [bib_ref] Zika virus cell tropism in the developing human brain and inhibition by..., Retallack [/bib_ref]. Human foetal brain tissues ageing from 14 to 21 gestation weeks with ZIKV infection (Strain MR766) were found to have many neurons affluent areas positive for ZIKV. The virus infects Tbr2-expressing intermediate progenitor cells and microtubule-associated protein 2 (MAP2+) mature neurons to some extent 79 .
ZIKV, strains MR766 and IbH30656 were found to infect sensory neurons of the trigeminal and dorsal root ganglia but without inducing cell death. Primary adult murine sympathetic neurons of the superior cervical ganglia and parasympathetic neurons of the ciliary ganglia did not show signs of active ZIKV [bib_ref] Zika virus persistently and productively infects primary adult sensory neurons in vitro, Swartwout [/bib_ref]. Human iPSCs derived cortical neurons as well as the motor neurons are susceptible to ZIKV infection (by Strains: MR766 from the African lineage and PRVABC59 from Asian lineage), where Asian lineage showed an almost ten-fold replication rate [bib_ref] Replication of the Zika virus in different iPSC-derived neuronal cells and implications..., Lanko [/bib_ref]. Another study explored ZIKV infection in the peripheral nervous system (PNS), both in vivo and in vitro model systems. Peripheral neurons were permissive to the infection leading to cell death 77 .
Astrocytes: Astrocytes are associated with a plethora of roles including formation of blood-brain barrier (BBB) and its maintenance, synaptogenesis, repair of tissue damage, neurotransmission, metabolic regulation, innate and adaptive responses moderation to pathogens and supporting the synaptic transmission. Astrocytes are thought to be an early target of ZIKV during in utero infections due to their participation in BBB formation. ZIKV infects astrocytes with high propensity [bib_ref] MicroRNA and mRNA dysregulation in astrocytes infected with zika virus, Kozak [/bib_ref] [bib_ref] Zika virus dysregulates the expression of astrocytic genes involved in neurodevelopment, Shereen [/bib_ref] [bib_ref] Zika virus promotes CCN1 expression via the CaMKIIα-CREB pathway in astrocytes, Sun [/bib_ref]. Astrocytes respond to all forms of central nervous system (CNS) damage and disease resulting in changes in their gene expression, cellular structure and function. Such responses of astrocytes are referred to as astrogliosis. Astrogliosis is identified as a major contributor to CNS disorders, including viral infections. During ZIKV infection, astrogliosis and the associated inflammatory scenario may lead to massive cell death in the developing brain. Astrogliosis has been observed in ZIKV-infected human foetuses [bib_ref] Disruption of glial cell development by Zika virus contributes to severe microcephalic..., Li [/bib_ref]. GFAP is a hallmark of glial activation, robust upregulation of GFAP is observed in ZIKV infected mouse brain [bib_ref] Disruption of glial cell development by Zika virus contributes to severe microcephalic..., Li [/bib_ref]. ZIKV also inhibits activation of NLRP3-dependent inflammasome in bone marrow-derived macrophages (MDMs) and mixed glial cells from mouse brain [bib_ref] Zika virus impairs host NLRP3-mediated inflammasome activation in an NS3-dependent manner, Gim [/bib_ref].
The susceptibility of astrocytes could be due to high expression of AXL, TAM family receptor, but more pieces of evidence are needed to claim AXL as candidate entry receptor for ZIKV. Furthermore, satellite glial cells which surround and protect the soma of ganglionic neurons in peripheral sensory and autonomic ganglia can be infected with ZIKV; the virus was found to kill satellite glial cells in vitro [bib_ref] Zika virus persistently and productively infects primary adult sensory neurons in vitro, Swartwout [/bib_ref]. ZIKV susceptibility of astrocytes results in progressive cell death inducing necroptosis in infected cells. It is indicated by increased phosphorylation of receptorinteracting serine/threonine-protein kinase (RIPK)1, RIPK3, and mixed lineage kinase domain-like (MLKL) protein. It was found that ZIKV-induced necroptosis was RIPK3-dependent. Human astrocytes when pretreated with an inhibitor for RIPK3 activation, showed restricted viral replication. This study suggested that induced necroptosis could play a protective role in viral pathogenesis [bib_ref] RIPK3-Dependent Necroptosis Is Induced and Restricts Viral Replication in Human Astrocytes Infected..., Wen [/bib_ref].
As in NSCs, ZIKV modulates miRNA circuitry in astrocytes. Kozak et al 23 described the differential expression of miRNA post-ZIKV infection in SVG human astrocyte cell line. Several miRNAs were differentially expressed at 24 h post-infection; 11 miRNAs were found to be upregulated by more than 2-fold. At 48 h post-infection, three miRNAs -miR-411-3p, miR-323a-5p and miR-194-5p were upregulated by more than 2-fold, and at 72 h postinfection, miR-9-5p was upregulated over 7-fold. These observations highlighted that miRNAs might have a role to play in ZIKV neuropathology. GO analysis of targets of differentially expressed miRNAs in SVG cell line (human astrocyte) are involved in viral life cycle, viral transcription, cell cycle, immune system response and innate immune response. is a schematic representation of the role of miRNAs in ZIKV-induced alterations in NSCs and astrocytes.
Oligodendrocytes: Infection of oligodendrocytes is not pursued much till date; hence, minimal information is available regarding the role of oligodendrocytes during ZIKV infection. Since oligodendrocytes are essential for the functioning of neurons, alterations in their properties would impair neural functioning. Cumberworth et al [bib_ref] Zika virus tropism and interactions in myelinating neural cell cultures: CNS cells..., Cumberworth [/bib_ref]
## Brain microvascular endothelial cells (bmecs):
BMECs form the BBB providing a tight seal which imparts selective permeability across the BBB. BMECs, when experience insult, release cytokine and chemokines, which ultimately lead to the infiltration of monocytes into the brain. BMECs infected by ZIKV upon activation release inflammatory molecules -interferon (IFN)-β, IFN-λ, interferon induced protein with tetratricopeptide repeats (IFIT)1, IFIT2, interferon-stimulated gene (ISG)15, interleukin (IL)-1 and IL-6 [bib_ref] Zika virus persistently infects and is basolaterally released from primary human brain..., Mladinich [/bib_ref] [bib_ref] Zika virus infects, activates, and crosses brain microvascular endothelial cells, without barrier..., Papa [/bib_ref]. The viability of BMECs remains unaffected in in vitro model systems of BBB [bib_ref] Zika virus persistently infects and is basolaterally released from primary human brain..., Mladinich [/bib_ref] [bib_ref] Zika virus crosses an in vitro human blood brain barrier model, Alimonti [/bib_ref]. NS5 sumoylation has been found to be associated with persistent ZIKV infection in human BMECs. Though effects of ZIKV on BMECs remain mild, the virus creates an inflammatory scenario 90 .
Microglia: Microglial cells play a pivotal role during the brain development and significantly contribute in neuroinflammation and neuropathogenesis in viral infections. It plays additional roles in synapse formation, brain inflammatory response, neurodegenerative diseases, stroke, trauma and regeneration. ZIKV has been shown to target microglia, primary human monocytes and MDMs thereby altering their properties [bib_ref] Zika virus infects human fetal brain microglia and induces inflammation, Lum [/bib_ref]. ZIKV induced an anti-viral response by enhancing IFN-α expression and other pro-inflammatory molecules, i.e., IL-1β, IL-6, IL-8, macrophage inflammatory protein (MIP) -1α, MIP-1β, monocyte chemoattractant protein (MCP)-1, tumour necrosis factor alpha (TNF-α), IL-1Rα, IL-1α and IL-12p70 [bib_ref] Zika virus infects human fetal brain microglia and induces inflammation, Lum [/bib_ref].
Pro-inflammatory molecules released from astrocytes impair the properties of NSCs and neurons. In vitro studies suggested that conditioned medium from ZIKV-infected microglia hampered the growth of neurospheres and proliferation of NSCs. Moreover, conditioned medium from infected microglial cells reduced neurogenesis and enhanced astrogliogenesis. TNF-α and IL-6 cytokines were preeminent among others, as blocking them in conditioned medium rescued neurogenesis and inhibited enhanced astrogliogenesis 92 . Furthermore, the role of microglia is evident as infected microglial cells appear to be phagocytosing pyknotic nuclei post-ZIKV infection, suggesting chronic inflammatory conditions in ZIKV infected brains [bib_ref] Zika virus tropism and interactions in myelinating neural cell cultures: CNS cells..., Cumberworth [/bib_ref]. Investigations into the detailed molecular pathways induced by ZIKV in microglial cells are hence necessary.
## Open questions
In spite of the recent advances in the understanding of mechanisms that may underlie ZIKV neuropathogenesis, basic and clinical researchers are striving to find answers for some open questions in a hope to reduce the burden of consequences arising following ZIKV infections. Some of the open questions are as follows -What makes ZIKV infect such a wide variety of cell types? Which are the receptors/co-receptors and what are the mechanisms involved in the entry of virus into the host cells? What are the receptor-independent (if any) mechanisms of ZIKV invasion? How did the virus gain neurotropism? What all signalling pathways are involved in disrupting the proliferative, differentiating and migrating properties of NSCs? What can be the possible regimen for blocking/ treating intrauterine ZIKV infection?
# Remarks & conclusions
ZIKV being an arbovirus with the existence of various lineages, can continue to stay a global threat, especially in densely populated developing countries that lack health awareness. The virus can infect multiple cell types of brain as well as other organs of the body. Sustained efforts for prevention and clinical management of ZIKV related diseases are critical.
The virus modulates host cell machinery for its successful replication. The virus and its proteins inhibit innate antiviral response by modulating TLR3 activity; it also enhances autophagosome formation in cells for its replication. Once the virus enters brain, it deteriorates development of the brain at various levels: (i) induction of apoptosis and hampered proliferation in early progenitors, (ii) inhibition of neurogenesis and astrogliogenesis, (iii) activation of brain immune cells (astrocytes and microglia) to release many cytokines and chemokines. The virus causes massive cell death in the developing brain. The damage caused to the brain in utero is irreversible and leaves the newborn with various severe motor and cognitive disabilities. . Role of microRNAs in the zika virus-induced alterations in neural stem cells and astrocytes. In human foetal neural stem cells, zika virus E protein disrupted miRNA circuitry, as investigated through small RNA sequencing. NOTCH2 and PAX3, two significant development-related transcripts, were downregulated by miR-204-3p and miR-1273-3p 62, respectively. Altered expression of NOTCH2 and PAX3 may cause hindered brain development leading to ZIKV-associated microcephaly. Similarly, SVG-astrocytes infected with ZIKV (strain PRVABC59) resulted in altered miRNA circuitry at 24 hpi, 48 hpi and 72 hpi. Upregulated miRNAs were found to target genes involved in viral process and cell cycle.
## Conflicts of interest:
[fig] Figure 1: Genetic organization of Zika virus. [/fig]
[fig] Figure 2: Vertical transmission and congenital Zika Syndrome: [/fig]
[fig] Financial: support & sponsorship: The first author (RB) acknowledges the Council of Scientific & Industrial Research, New Delhi for providing junior and senior research fellowships. The authors also acknowledge the support of the facilities provided under the Biotechnology Information System Network (BTISNET) grant, Department of Biotechnology (DBT), India, and Distributed Information Centre at NBRC, Manesar, India. [/fig]
[table] Table: Current update and model systems used for zika virus infection in various brain cell types ++, +++ depicts the degree of susceptibility of cells towards ZIKV. CNS, central nervous system; PNS, peripheral nervous system; NSC, neural stem cell; TLR3, toll-like receptors-3; UPR, unfolded protein response; CASP-3, Caspase-3 [/table]
|
Recombinant expression of osmotin in barley improves stress resistance and food safety during adverse growing conditions
BackgroundAlthough many genetic manipulations of crops providing biofortified or safer food have been done, the acceptance of biotechnology crops still remains limited. We report on a transgenic barley expressing the multi-functional protein osmotin that improves plant defense under stress conditions.MethodsAn Agrobacterium-mediated technique was used to transform immature embryos of the spring barley cultivar Golden Promise. Transgenic barley plants of the T0 and T1 generation were evaluated by molecular methods. Transgenic barley tolerance to stress was determined by chlorophyll, total protein, malondialdehyde and ascorbate peroxidase content. Methanol extracts of i) Fusarium oxysporum infected or ii) salt-stressed plants, were characterized by their acute toxicity effect on human dermal fibroblasts (HDF), genotoxicity and affection of biodiversity interactions, which was tested through monitoring barley natural virus pathogen-host interactions-the BYDV and WDV viruses transmitted to the plants by aphids and leafhoppers.
# Results
Transgenic plants maintained the same level of chlorophyll and protein, which significantly declined in wild-type barley under the same stressful conditions. Salt stress evoked higher ascorbate peroxidase level and correspondingly less malondialdehyde. Osmotin expressing barley extracts exhibited a lower cytotoxicity effect of statistical significance than that of wild-type plants under both types of stress tested on human dermal fibroblasts. Extract of Fusarium oxysporum infected transgenic barley was not able to damage DNA in the Comet
# Introduction
Osmotin is a 26-kDa protein belonging to the 5 th class of pathogenesis-related proteins (PR 5) together with thaumatin, zeamatin and others. Osmotin was first characterized in Nicotiana tabacum plants adapted to saline conditions [bib_ref] Molecular Cloning of Osmotin and Regulation of Its Expression by ABA and..., Singh [/bib_ref] [bib_ref] Proteins Associated with Adaptation of Cultured Tobacco Cells to NaCl, Singh [/bib_ref]. Later, its homologs were found in both monocots and dicots across the whole plant kingdom. Osmotin is a multifunctional protein that plays an important role in the plant immune system during stress. Abiotic originators of stress such as drought, salt and cold induce osmotin expression. As the name indicates, osmotin plays a role as an osmoprotectant, providing enzyme protection and protein chaperone functions. Biotic stress resistance is related to the ability of osmotin to activate the fungal receptor that induces the programmed cell death of fungi. Being a promising protein, it has served as a tool for biotechnology engineering for many decades [bib_ref] Molecular Cloning of Osmotin and Regulation of Its Expression by ABA and..., Singh [/bib_ref]. The first transgenic plants containing the heterologous osmotin gene in their genomes were potato and tobacco [bib_ref] Disease responses to fungal infection in transgenic tobacco and potato plants that..., Liu [/bib_ref] and many others followed, as summarized in reviews. More recently, transgenic monocotyledonous plants bearing the osmotin gene were developed as well. Osmotin was first expressed in a monocot wheat in 2008 [6] followed by rice another three years later . The genetic manipulation of monocots is a promising and still developing field, because their agriculture is significantly endangered by climatic changes. Monocots play an important role in agriculture for animal feed and for human nutrition, as they are among the most often planted crops (e.g. wheat, barley, rice, maize and others), however, the application of their GM variants still remains limited.
Even though transgenic crops are one of the most characterized and toxicologically tested plants, many of them have never been applied (e.g. Golden Rice . The economic aspects, environmental benefits and higher yield of GM crops [9] usually do not affect public opinion significantly. Even though biofortified crops (reviewed e.g. in [10]) with improved properties and safer food [bib_ref] Genetically Modified Crops and Food Security, Qaim [/bib_ref] were reported years ago, the usage of GM crops is still restricted. No conventionally bred crop has been so thoroughly tested for toxicity, allergenicity or effect on the environment, intestinal microbiome and nutrition. However, what is often neglected is the increased toxicity of crops protecting themselves against harsh environmental conditions. As a protection against stress, plants have adapted their own immune system [bib_ref] Plant Immune System: Crosstalk Between Responses to Biotic and Abiotic Stresses the..., Nejat [/bib_ref] in order to survive. Nonetheless, many of the secondary metabolites formed this way can be toxic compounds, e.g. glycoalkaloids [bib_ref] The steroidal glycoalkaloids solamargine and solasonine in Solanum plants, Sinani [/bib_ref] , steroidal alkaloids [bib_ref] Cyclopamine: From Cyclops Lambs to Cancer Treatment, Lee [/bib_ref] , flavonoids [bib_ref] Detection of weak estrogenic flavonoids using a recombinant yeast strain and a..., Breinholt [/bib_ref] [bib_ref] Genotoxicity of the isoflavones genistein, daidzein and equol in V79 cells, Virgilio [/bib_ref] and glycosides [bib_ref] Human Breast Tumor Cells Are More Resistant to Cardiac Glycoside Toxicity Than..., Clifford [/bib_ref] [bib_ref] Inhibition of export of fibroblast growth factor-2 (FGF-2) from the prostate cancer..., Smith [/bib_ref]. Furthermore, the toxicity of plant immune system compounds is often supplemented by the toxicity of the pathogen's secondary metabolites, e.g. mycotoxins [bib_ref] Occurrence, Toxicity, and Analysis of Major Mycotoxins in Food, Alshannaq [/bib_ref].
The aim of this work is the development of a transgenic barley with improved resistance to stress originators, documentation of this higher resistance by biochemical methods, and a proof of better toxicological properties for consumers compared to commercially used crops. As an often-mentioned disadvantage of GM plants, its influence on biodiversity is described.
Here, we also focused on studying the effect of this GM barley on the spreading of viruses by aphids and leafhoppers. Two economically most important virus pathogens for barley cereals in Europe were selected-Barley yellow dwarf virus, transmitted by aphids, and Wheat dwarf virus, transmitted by leafhoppers.
# Materials and methods
## Barley transformation
The osmotin gene (OSM, GenBank: M29279.1, NCBI) was synthetized artificially (GeneArt Gene Synthesis, Thermo Fisher Scientific), therefore codon optimization for barley was performed. The osmotin gene was further cloned into the vector pDONR207 (Invitrogen) by BP clonase reaction (Gateway, Thermo Fisher Scientific). The construct was verified by digestion. The osmotin gene was inserted by LR clonase reaction (Gateway LR Clonase, Thermo Fisher Scientific) into the destination vector pBract214 (http://www.bract.org). The pBract214 vector was designed for the transformation of barley. The gene of interest is under the control of the maize Ubi promoter. The vector contains the hpt gene conferring hygromycin resistance under the CaMV 35S promoter. Correct orientation of the transgene was verified by restriction analysis and sequencing. The Vector pBract214::osm (S1 the helper plasmid pSoup were transformed into the Agrobacterium tumefaciens strain AGL1 by electroporation. The Agrobacterium-mediated transformation of the immature zygotic barley embryo genotype 'Golden Promise' was performed according to the transformation protocol by Harwood et al [bib_ref] Barley transformation using Agrobacterium-mediated techniques, Harwood [/bib_ref]. Explants were cultivated in vitro on selection-callus induction, and regeneration medium and transferred into soil. Putative transgenic T0 plants were screened by PCR analysis. The analysis was performed with genomic DNA that was isolated from leaf tissue of the regenerated plants. For PCR reaction, premix REDTaq ReadyMix PCR Reaction Mix (Sigma-Aldrich, USA,) was used. The presence of the osmotin gene was determined by amplifying a 222-bp fragment using the primers F: 5´-GCCCTGCCTTCATACGCTAT-3´and R: 5´-TACGGGCAGTTGTTCCTCA C-3´. The presence of the hpt selection gene by amplifying a 275-bp amplicon using the primers F: 5´-GATTGCTGATCCCCATGTGT-3´and R: 5´-GCTGCTCCATACAAGCCAAC-3´.
Transgene expression was verified at the mRNA level. Where not stated otherwise, all procedures were done according to the manufacturer's instructions. RNA was extracted from young leaf tissue of the transgenic plant with an RNAqueous Total RNA Isolation Kit (Thermo Fisher Scientific). The sample was treated with a Turbo DNA-free Kit (Thermo Fisher Scientific) and RNA concentration was assessed spectrophotometrically (DeNovix, DS-11 Spectrophotometer). 1 ug of total RNA was reverse transcribed using RevertAid H minus Reverse transcriptase (Thermo Fisher Scientific). To analyze the reaction efficiency, dilution series of the selected cDNA samples were prepared. The endogenous gene for elongation factor [bib_ref] Internal standards for quantitative RT-PCR studies of gene expression under drought treatment..., Rapacz [/bib_ref] was selected as an internal control. For mRNA expression verification, a SensiFAST SYBR No-ROX Kit (Bioline) was used. Three-step PCR was conducted using a MyGo Mini real-time cycler (IT-IS Life Science Ltd.). Primer sequences for osmotin transcript detection were F: 5´-TCAGGTCCAGCTTCGTGTTC-3´and R: 5´-TACGGGCAGTTG TTCCTCAC-3´and produced an amplicon of 85 bp. Initial denaturation at 95˚C for 180 s was followed by forty cycles of denaturation at 95˚C, annealing at 60˚C and elongation at 72˚C. Reaction was terminated by a final 5 min extension at 72˚C. Melting analysis and electrophoretic separation of PCR products were done to verify primer specificity. Transgenic barley plants with verified expression of osmotin gene were grown in greenhouse to maturity. Immature embryos (T1 generation) were dissected from young caryopses of T0 plants and were selected on half-MS medium containing hygromycin 75 mg/L. Subsequently, the germinating plants were transferred to pots. All the germinated plants were analyzed for the presence of the osmotin and hpt transgenes by PCR.
## Exposure of transgenic and control plants to the stress
Isolated embryos of the T1 generation of obtained GM plants and of the non-GM barley were sown in individual pots (10 x 10 cm) filled with universal gardening substrate and placed in growing chambers under controlled conditions (temperature 24/16˚C, air humidity 70%, 16/8 day/night). Until 50 days after sowing, all plants had four developed leaves and were divided into groups according to the planned stress treatment. Each group consisted of 4 transgenic and 4 non-transgenic plants. The control group was planted under the same conditions, another group was irrigated with 200 mM NaCl every second day, the last group was treated with spores of Fusarium oxysporum DBM 4199 (OD = 0.7) once per week. One fully developed leaf was cut off every 3 days and kept at -80˚C until the biochemical analysis. 16 days after the beginning of the stress conditions, all biomass was harvested and lyophilized for toxicity testing.
The BYDV virus infection was implemented using the aphids Rhopalosiphum padi carrying the BYDV-PAV species [bib_ref] Discrimination of Three BYDV Species by One-step RT-PCR-RFLP and Sequence Based Methods..., Kundu [/bib_ref]. At the stage of two unfolded leaves, approximately five viruliferous aphids were transmitted onto each of 6 transgenic and 6 control group plants. The aphids were maintained there for four days, after that, the plants were treated with insecticide (Mospilan, Nippon Soda Co.) and kept under controlled conditions (temperature 19/15˚C, 16/8 day/night).
The WDV virus infection was administered using leafhoppers Psammotettix alienus carrying the WDV-B barley strain [bib_ref] Discrimination of Three BYDV Species by One-step RT-PCR-RFLP and Sequence Based Methods..., Kundu [/bib_ref]. At the stage of the main shoot and one tiller being detectable, approximately five viruliferous leafhoppers were transmitted onto each of 6 transgenic and 6 control group plants. The leafhoppers were maintained there for fourteen days, after that the leafhoppers were removed by hand and the plants were closely checked for any missing leafhoppers. The plants were then kept under controlled conditions as before (temperature 19/15˚C, 16/8 day/night).
Both WDV and BYDV isolates, as well as infected aphids and leafhoppers, were obtained from the Virus Collection of the Crop Research Institute, Prague (Virus Collection, 2017). The aphid and the leafhopper count of insect vectors required for successful infection were estimated based on the long-term experience with testing of cereal cultivars in the Crop Reseach Institute, Prague [bib_ref] A comparative study of the Barley yellow dwarf virus species PAV and..., Jarosova [/bib_ref] [bib_ref] Discrimination and genetic diversity of Wheat dwarf virus in the Czech Republic, Kundu [/bib_ref].
## Biochemical assays
For chlorophyll content determination, 0.02 g of barley leaf tissue was used. Samples were incubated in 99% alcohol (1:100 (w/V)) in the dark for 24 h. The subsequent measurement was the same as described in the method by [bib_ref] Transgenic tobacco plants expressingBoRS1 gene from-Brassica oleracea var.acephala show enhanced tolerance to..., Tang [/bib_ref]. The preparation of samples for the determination of protein content and enzyme activities was universal in order to decrease sample weight and consumption of plant material. 0.1 g of leaf tissue was homogenized in liquid nitrogen into powder. Then, 2 ml of extraction buffer (50 mM phosphate buffer, 1 mM EDTA and 2% (w/v) polyvinylpyrrolidone of pH 7.8) was added. The homogenized sample was centrifuged at 14,000 × g at 4˚C for 30 min. Protein content was measured according to the method by Bradford [bib_ref] A rapid and sensitive method for the quantitation of microgram quantities of..., Bradford [/bib_ref]. The reaction mixture contained plant extract, extraction buffer and reagent in the ratio 1:1:8. The reagent and subsequent measurement was the same as in the original paper. The enzyme activity of ascorbate peroxidase was determined according to [bib_ref] Hydrogen Peroxide is Scavenged by Ascorbate-specific Peroxidase in Spinach Chloroplasts, Nakano [/bib_ref] without any modifications. Malondialdehyde content was determined according to [bib_ref] Improving the thiobarbituric acid-reactive-substances assay for estimating lipid peroxidation in plant tissues..., Hodges [/bib_ref] with slight modifications. Extract was prepared by the homogenization of 0.1 g of leaf tissue in liquid nitrogen and the addition of 2 ml of 80% ethanol followed by centrifugation at 14,000 × g at 4˚C for 20 min. In contrast to the original paper, a 0.4 ml aliquot was used for the preparation of the reaction mixture. After heating followed by cooling, the mixture was centrifuged at 1,000 g at 4˚C for 20 min.
## Hemolytic and cytotoxicity studies
Hemolytic activity was determined according to a previous report [bib_ref] New findings in potential applications of tobacco osmotin, Viktorova [/bib_ref]. 1 μl of plant methanol extract (100 mg.ml -1 ) was used per spot. Triton X-100 (1%, 1 μl per spot, Sigma Aldrich) was used as a known hemolytic agent. The toxicity in mammalian tissue culture was studied on HDF-human dermal fibroblasts, Sigma-Aldrich, 106-05N).
HDF cells were cultivated in DMEM (Sigma-Aldrich) enriched with 10% fetal bovine serum (Sigma-Aldrich). The cells were maintained in media without antibiotics; however, for experiments media supplemented with Antibiotic Antimycotic Solution were used (commercial mixture of penicillin, amphotericin and streptomycin, Sigma-Aldrich).
Cells were harvested from exponential-phase cultures by a standardized detachment procedure using 0.25% Trypsin-EDTA, and the cell number was counted automatically using a Roche's CASY Cell Counter and Analyzer. 100 ml of 10 5 cells.ml -1 was seeded into the wells for cytotoxicity experiments. Each concentration was tested in quadruplicate within the same experiment in the concentration range 62.5-1,000 μg.ml -1 . Viability was evaluated after 72 h by standard resazurin assayusing fluorescent measurement (560/590 nm). Viability was calculated as (sample fluorescence-fluorescence of resazurin) / (fluorescence of cells without treatment-fluorescence of resazurin).
Genotoxicity assay was determined using Hek 293T cells (Human embryonal kidney cells, Sigma-Aldrich, USA) according to the Comet assay as previously described [bib_ref] Structure, mechanical characteristics and in vitro degradation, cytotoxicity, genotoxicity and mutagenicity of..., Kubá Sek [/bib_ref]. The cells were cultivated as described for HDF cells. For the experiment, 10 5 cells.ml -1 were seeded into the 12-well plates. After 24 hours of cultivation in CO 2 incubator, the medium was changed and the plant extracts were added to the final concentration of 3 mg.ml -1 . After 24 hours of incubation, the positive control of genotoxicity was realized by addition of H 2 O 2 (4.2 mg.l -1 ) for 10 min to the cells. As a negative control served cells without any treatment. After the cells harvesting and electrophoresis, fluorescent microscopy (Olympus IX81 equipped with Texas Red filtr) was used for comets visualization and ImageJ software was used for evaluation TailDNA (%).
## Detection of bydv infection
RNA was isolated by the traditional method using TRIzol reagent (Thermo Fisher Scientific, USA). RT-qPCR assay was performed in a Roche LightCycler 480 Instrument II using Light-Cycler 480 SYBR Green I Master (Roche Applied Science, Germany), RT Enzyme Mix (Array-Script UP Reverse Transcriptase and RNase Inhibitor, Thermo Fisher Scientific) and PVinterF and YanRA primers [bib_ref] Validation of reference genes as internal control for studying viral infections in..., Jarošová [/bib_ref]. The RT-qPCR was performed at 48˚C for 30 min, 95˚C for 10 min, followed by 40 cycles at 95˚C for 15 s and 60˚C for 1 min and a melting-curve analysis (95˚C 15 s, 60˚C 1 min, 95˚C 15 s, 60˚C 15 s). The qPCR efficiency was determined as E = 105.51%, R2 = 0.9964 and the linear standard curve interval as 6.89×10 1 -6.89×10 7 copies using triplicates for the standard and tested samples. For the standard curve, a specific BYDV nucleotide sequence (294 bp, primers PVinterF+YanRA) amplified by RT-PCR was inserted into the vector pGem-T Easy (Promega) and cloned into E. coli JM-109. The selected colony with confirmed insertion sequence was cultivated and DNA extracted (Plasmid Plus Midi Kit, Qiagen, Netherlands) and then excised from the gel (Sigma-Aldrich x-tracta Gel Extraction Tool, Sigma-Aldrich, USA) and purified (GenElute Gel Extraction kit, Sigma-Aldrich). Thereafter, tenfold serial dilution of the transcripts were prepared.
For all samples, the mean detected BYDV concentration was calculated based on the tested triplicates and subsequently normalized using the RNA sample concentration detected spectrophotometrically. For these normalized BYDV titers, their log10 values were calculated and for each week, the log10 mean BYDV titers of the tested group and control group are depicted, together with the interval of plus and minus one standard error of the mean.
## Detection of wdv infection
DNA was isolated by adding 0.5 ml of extraction buffer (1 M guanidine thiocyanate, 20 mM Na2H2EDTA, 0.1 M MOPS, pH 4.6, mercaptoethanol added to 0.2% just prior to use) to 50-100 mg of sampled tissue that had been disrupted and homogenized in liquid nitrogen. The solution was incubated for 30 min in a 60˚C water bath with occasional vortexing followed by phenol-chloroform-isoamyl alcohol (25:24:1, Affymetrix, USA) extraction, chloroform extraction, isopropanol and sodium acetate precipitation and two steps of 70% ethanol purification. The qPCR assay was performed on a Roche LightCycler 480 Instrument II using LightCycler 480 Probes Master (Roche Applied Science, Germany), UniWDVfw and UniWDVrv primers [bib_ref] Real-time PCR assay for the discrimination and quantification of wheat and barley..., Gadiou [/bib_ref] and a TaqMan probe (6-FAM-AGGCGAAGAATGATTCACCCT-BHQ-1). The qPCR was performed at 95˚C for 10 min, followed by 40 cycles at 95˚C for 15 s and 60˚C for 1 min. The qPCR efficiency was determined as E = 99.49%, R 2 = 0.9989 and the linear standard curve interval as 2.9×10 3 -2.9×10 8 copies using triplicates for the standard and tested samples. For the standard curve, a specific WDV nucleotide sequence (140 bp, primers UniWDVfw+UniWDVrv) amplified by PCR was inserted into the vector pGem-T Easy (Promega) and cloned into E. coli JM-109. The selected colony with confirmed insertion sequence was cultivated and DNA extracted (Plasmid Plus Midi Kit, Qiagen, Netherlands) and then excised from the gel (Sigma-Aldrich x-tracta Gel Extraction Tool, Sigma-Aldrich, USA) and purified (GenElute Gel Extraction kit, Sigma-Aldrich). Thereafter, tenfold serial dilution of the transcripts were prepared.
For all samples, the mean detected WDV concentration was calculated based on the tested triplicates and subsequently normalized using the DNA sample concentration detected spectrophotometrically. For these normalized WDV titers, their log10 values were calculated and for each week, the log10 mean WDV titers of the tested group and control group are depicted, together with the interval of plus and minus one standard error of the mean.
## Statistical analysis of experimental data
The statistical significance of results was tested by analysis of variance followed by Duncan's test in STATISTICA 12 (data analysis software system, StatSoft. Inc. 2013). For all statistical tests, the significance level was established at p < 0.05.
# Results and discussion
## Transgenic barley expressing tobacco osmotin gene
In order to avoid potential mutations originating from the cloning process, the osmotin gene was commercially synthetized de novo. The codon optimized osmotin gene was cloned into the expression vector pBract214. The possibility of tobacco osmotin gene expression under the constitutive promoter of the cauliflower mosaic virus has been already by our group demonstrated in barley previously [bib_ref] New findings in potential applications of tobacco osmotin, Viktorova [/bib_ref]. Therefore, the same promoter was used in this case as well. Transgenic barley plants expressing the osmotin gene were prepared via Agrobacterium-mediated transformation. In total, 210 immature embryos were transformed, providing 26 regenerating plants (S2 of the T0 generation that were transferred into pots and grown to maturity in a greenhouse. The presence of the osmotin transgene was confirmed by PCR in 25 regenerated plants [fig_ref] Fig 3: Effect of salinity on lipid peroxidation demonstrated as malondialdehyde [/fig_ref] The ratio corresponds to a transformation efficiency of 12%, which is typically lower than for the transformation of dicotyledonous plants [bib_ref] Problems and possibilities of monocot transformation, Sood [/bib_ref] , but comparable with other barley transformation experiments [bib_ref] Selection of transformation-efficient barley genotypes based on TFA (transformation amenability) haplotype and..., Hisano [/bib_ref]. Notably, transformation efficiency for Agrobacterium-mediated barley immature embryos utilizing hygromycin selection is 25% in average [bib_ref] Barley transformation using Agrobacterium-mediated techniques, Harwood [/bib_ref]. Incomplete T-DNA integration into host genomic DNA might occur during Agrobacterium-mediated transformation leading to an unintended loss of the selection marker. Previously, it was reported that incomplete integration of T-DNA can reach up to 44% in monocot wheat [bib_ref] Characterisation of T-DNA loci and vector backbone sequences in transgenic wheat produced..., Wu [/bib_ref]. In our experiment, hpt selection marker was detected in all transgenic plants thus confirming complete integration of T-DNA cassette (S4 Heterologous peptide expression can be driven by tissue specific or constitutive promoters. [bib_ref] The ZmRCP-1 promoter of maize provides root tip specific expression of transgenes..., Onyango [/bib_ref] demonstrated usage of root tip specific promoter to induce resistance to biotic stress induced by nematodes. Alternatively, constitutive promoter CaMV35S [bib_ref] Overexpression of a fusion defensin genes from radish and fenugreek improves resistance..., Bala [/bib_ref] [bib_ref] Overexpression of Jatropha curcas Defensin (JcDef) Enhances Sheath Blight Disease Resistance in..., Wang [/bib_ref] can be used for heterologous protein production in plants. In our study, expression of osmotin protein was driven by the strong constitutive maize ubi promoter that provides strong stable expression in all plant tissues. Broad expression of osmotin should enhance plant response to various biotic and abiotic stresses such as salt stress or fungal infection whose symptoms are not restricted into specific plant tissues but more likely tend to affect the whole plant. The expression of the osmotin gene was confirmed by transgene-specific RT-PCR, while transgene-specific amplicons were not detected in the WT. The specificity of PCR product was additionally verified by melting analysis after performed PCR reaction.
## Higher stress resistance of transgenic barley
Both transgenic and non-transgenic tobacco plants were exposed to stress caused by salinity (200 mM NaCl) and pathogen infection (Fusarium oxysporum). As can be seen e.g. in the S5 Fig, the transgenic plants showed a significant reduction in disease symptoms. Both types of stress were able to induce a decrease in protein content in wild-type barley. Many researchers have reported that the level of total soluble protein in crop plants decreases under abiotic stress. Stress usually leads to protein damage caused by e.g. reactive oxygen species [bib_ref] Salt-tolerance in plants. I. Ions, compatible organic solutes and the stability of..., Brady [/bib_ref] or by increased activity of proteases [bib_ref] Changes of Proline Content, Activity, and Active Isoforms of Antioxidative Enzymes in..., Wang [/bib_ref]. A downregulation of photosynthesis under several stresses was reported at the proteome level [bib_ref] Vítá mvá s P, Prá šil IT. Proteomics of stress responses in..., Kosová [/bib_ref]. Salt and drought induced a decrease in the main proteins of photosystem II and in both chlorophyll a and b binding proteins as well as producing a downregulation of RuBisCO and key Calvin cycle enzymes in barley and wheat [bib_ref] Proteomics study reveals the molecular mechanisms underlying water stress tolerance induced by..., Ghabooli [/bib_ref] [bib_ref] Identification of changes in Triticum durum L. leaf proteome in response to..., Caruso [/bib_ref]. However, under stress the transgenic barley expressing the osmotin gene maintained the same protein level as the non-stressed transgenic control plants [fig_ref] Fig 1: Protein content determined under both types of stress [/fig_ref]. Similarly, a higher protein level was detected in strawberries recombinantly expressing osmotin [bib_ref] Development of transgenic strawberry (Fragaria x ananassa Duch.) plants tolerant to salt..., Husaini [/bib_ref] in comparison to non-transgenic plants during salt conditions. As was mentioned above, photosynthesis is significantly affected during stress conditions; therefore, chlorophyll content in barley was measured in the presence of both types of stressing factors. Ongoing stress was detected in wild-type barley, where both Fusarium infection and salinity exhibited an influence on chlorophyll content . However, similarly to protein content, the transgenic barley maintained the same chlorophyll level as the control nonstressed plants. In agreement with our results, it has been [bib_ref] Overexpression of osmotin gene confers tolerance to salt and drought stresses in..., Goel [/bib_ref] already reported that transgenic tomato plants expressing the osmotin gene had higher chlorophyll content during the drought and salt stress than the non-transgenic plants. Similarly, osmotin-expressing transgenic soybean, chilli pepper and strawberry exhibited higher chlorophyll content than the non-transgenic variants during salinity [bib_ref] Development of transgenic strawberry (Fragaria x ananassa Duch.) plants tolerant to salt..., Husaini [/bib_ref] [bib_ref] Overexpression of tobacco osmotin (Tbosm) in soybean conferred resistance to salinity stress..., Subramanyam [/bib_ref]. The connection between osmotin and photosynthesis has been already reported, demonstrating an osmotin affinity to brassinosteroids, plant hormones affecting photosynthesis activity [bib_ref] Affinity chromatography as the method for brassinosteroid-binding protein isolation, Kamlar [/bib_ref] [bib_ref] 24-epibrassinolide and 20-hydroxyecdysone affect photosynthesis differently in maize and spinach, Rothova [/bib_ref].
As a major part of their defense system, plants have evolved an antioxidant strategy for overcoming stress conditions. Antioxidants (both enzymatic and non-enzymatic) prevent osmotic stress, oxidative stress, molecular damage, and even cell death [bib_ref] Plant responses to stresses: Role of ascorbate peroxidase in the antioxidant protection, Caverzan [/bib_ref]. Salt stress induces the production of reactive oxygen species (ROS), which causes oxidative stress. Therefore, the amount of antioxidant plays an important role during stressful conditions. Here, attention was focused on APX (ascorbate peroxidase). When influenced by stress, the transgenic barley plants exhibited a higher level of this antioxidant [fig_ref] Fig 3: Effect of salinity on lipid peroxidation demonstrated as malondialdehyde [/fig_ref] , indicating a lower susceptibility to salinity than the non-transgenic control plants. The connection between a higher level of APX and salt stress tolerance was demonstrated in genetically modified sweet potato [bib_ref] Overexpression of CuZnSOD and APX enhance salt stress tolerance in sweet potato, Yan [/bib_ref]. Similarly to our results, transgenic chilli pepper and soybean expressing the tobacco osmotin gene exhibited a higher level of APX and improved salt stress tolerance [bib_ref] Overexpression of tobacco osmotin (Tbosm) in soybean conferred resistance to salinity stress..., Subramanyam [/bib_ref].
Lipid peroxidation is a process caused by free radicals (e.g. ROS) attacking unsaturated lipids in membranes. Malondialdehyde (MDA) is one of the end products of lipid peroxidation. As we demonstrated that an increased level of APX prevents radical formation in transgenic plants, logically, we also found a lower amount of MDA in transgenic plants during salinity [fig_ref] Fig 4: Cytotoxicity of methanol barley extracts on human dermal fibroblasts [/fig_ref]. Less MDA, indicating the effect of osmotin on cell membrane protection from damage by lipid peroxidation, has been already reported in transgenic olive plants exposed to drought [bib_ref] Olea europaea L.) plants transgenic for tobacco osmotin gene are less sensitive..., Silvestri [/bib_ref].
Both types of stress led to the induction of stress markers such as a decrease in chlorophyll and protein content in wild-type barley plants, however, osmotin-expressing barley plants did not show evidence of ongoing stress, indicating their better preparedness for coping with the
## Lower toxicity of stressed transgenic barley in comparison to wt
At the end of the exposure to stress, the aboveground biomass was extracted by methanol. The extracts were then added to the growth medium of human fibroblasts in the concentration range from 0 to 1,000 μg.ml -1 . In both types of stress, there is evidence that transgenic plant extracts are less toxic than these of the non-transgenic (wild-type) ones. The cytotoxicity experiment was done in four technical repetitions for each plant sample, and the viability of fibroblasts [fig_ref] Fig 5: Genotoxicity was evaluated after 24 h of co-cultivation of human embryonic kidney... [/fig_ref] was evaluated as the average of four biological repetitions (meaning four independent plants, both transgenic and wild-type variant). The viability of cells decreased with a higher concentration of plant extracts. However, the toxicity of wild-type barley extracts of plants exposed to both types of stress was detected at the lowest tested concentration (62.5 μg.ml -1 ). On the other hand, a toxicity of osmotin-expressing barley extracts was detected at a significantly higher concentration (500 μg.ml -1 ). This finding could confirm our hypothesis that transgenic plants are better prepared for stressful conditions by osmotin expression and therefore do not produce so many secondary metabolites, which are mostly responsible for their toxicity. This finding was confirmed by genotoxicity comparison of transgenic and non-transgenic extracts as well where the plants expose to Fusarium GM plants, which were modified to cope with environmental stress, have their internal metabolism significantly changed, preventing plant defense system over-response and the accumulation of toxicants, anti-nutrients and secondary metabolites during the ongoing stress. The changes in toxic secondary metabolite content have been already demonstrated [bib_ref] Toxic secondary metabolite production in genetically modified potatoes in response to stress, Matthews [/bib_ref] in transgenic potatoes exposed to a pathogen. The genetic manipulation of carbohydrate metabolism and pathogen resistance in these potatoes led to changes in the profile of plant defense compounds, which were mainly characterized by a reduction in the level of the main glycoalkaloids R-solanine and R-chaconine.
As well as the expression of plant secondary metabolites, the secondary metabolites formed by the pathogen could have a significant effect on the acute toxicity of crop extracts. In particular, the toxicity of mycotoxins has been reported many times. A lower amount of mycotoxins as a secondary effect of genetic manipulation was detected e.g. in a comprehensive study focused on transgenic maize [bib_ref] A review on comparative data concerning Fusarium mycotoxins in Bt maize and..., Ostry [/bib_ref].
The mycotoxins, as a secondary metabolite of fungi, could be responsible for the genotoxicity, which we detected in case of methanol extracts from the non-transgenic barleys infected by Fusarium oxysporum. However, the extracts from transgenic barley expressing the antifungal protein osmotin showed no toxicity in the same test as shown in [fig_ref] Fig 5: Genotoxicity was evaluated after 24 h of co-cultivation of human embryonic kidney... [/fig_ref] The genotoxicity was evaluated after co-cultivation of plant extracts with human embryonal kidney cells (Hek 293T) by standardized Comet assay with appropriate controls.
## Weak or no impact of transgenic barley on viral infection spread by aphids and leafhoppers
The influence of GM crops on biodiversity has been discussed and tested many times (reviewed e.g. byand [bib_ref] Effects of biotechnology on biodiversity: herbicide-tolerant and insect-resistant GM crops, Ammann [/bib_ref] , mostly demonstrating that GM crops have reduced the impacts of agriculture on biodiversity. Nevertheless, confirmation of this hypothesis is still needed. In this paper, we focused on the effect of barley expressing a multi-functional osmotin protein on virus pathogen-host interactions. For barley, the effect of aphids spreading BYDV and the effect of leafhoppers spreading WDV was studied. Both viruses cause worldwide diseases of the most important crops including barley, wheat, rice and maize [bib_ref] Barley yellow dwarf viruses: infection mechanisms and breeding strategies, Choudhury [/bib_ref]. As is shown in [fig_ref] Fig 6: BYDV and WDV titres measured in plant leaves in 42-day period following... [/fig_ref] the genetic manipulation of barley by the osmotin gene insertion had no effect on obtained virus titres through the whole tested period of first 6 weeks after inoculation. Both aphids and leafhoppers were able to attack GM barley and insert the virus into the phloem and to infect the tested plants. For phenotype observations, the difference in plant height and tiller count between wild-type infected and transgenic infected plants was tested 6 weeks after inoculation. For both the WDV and BYDV virus, neither the plant height nor the tiller count after 6 weeks from inoculation was proved to differ. The smallest t-test value for two-sample twotailed t-test was measured for tiller count for WDV infection (t = 0.086), where the infected transgenic barley plants showed smaller, but not significantly different, tiller count (2.29) than the infected wild-type plants , i.e. exhibiting slightly less severe symptoms of WDV infection, where strongly infected plants often exhibit a high tiller count at the tillering stage. Similarly no impact of potato genetic manipulation on aphids was determined neither by [bib_ref] Biodiversity analyses for risk assessment of genetically modified potato, Lazebnik [/bib_ref] or by another research group, which found Bt corn to have no effect on aphids.
# Conclusions
Many investigations have been carried out to elucidate the mechanisms of the response of the transgenic plant to abiotic and biotic stress, however, the acceptance of transgenic crops in agriculture and industry is still limited, particularly in the EU. We compared osmotin-expressing GM and non-GM barley exposed to biotic and abiotic stress in order to investigate whether their toxicity level under adverse conditions is comparable. The results clearly show that our transgenic barley has a decreased toxicity to human cells under conditions of a/biotic stress for which it is better prepared and exhibits higher stress resistance. These findings provide a new perspective that could help to evaluate the safety of products from genetically modified crops.
## Supporting information
[fig] Fig 1: Protein content determined under both types of stress: (a) plants under abiotic stress and (b) plants under biotic stress. WT: non-transgenic (wild type) barley Hordeum vulgare L. var. Golden Promise, T: transgenic barley expressing tobacco osmotin gene. The results are shown as the average value of four plants measured in four replicates. Data are presented with the standard error of the mean (SEM). The statistical significance of results was tested by analysis of variance followed by Duncan's test (p < 0.05). � the groups were compared by two-choice t-test, the p values are presented. https://doi.org/10.1371/journal.pone.0212718.g001Fig 2. Chlorophyll content determined in both types of stress: (a) plants under abiotic stress and (b) plants under biotic stress. WT: non-transgenic (wild type) barley Hordeum vulgare L. var. Golden Promise, T: transgenic barley expressing tobacco osmotin gene. The results are shown as the average value of four plants measured in four replicates. Data are presented with the standard error of the mean (SEM). The statistical significance of results was tested by analysis of variance followed by Duncan's test (p < 0.05). https://doi.org/10.1371/journal.pone.0212718.g002 [/fig]
[fig] Fig 3: Effect of salinity on lipid peroxidation demonstrated as malondialdehyde (MDA) content (a) and antioxidant activity presented as ascorbate peroxidase (APX) activity (b). WT: non-transgenic (wild type) barley Hordeum vulgare L. var. Golden Promise, T: transgenic barley expressing tobacco osmotin gene. The results are shown as the average value of four plants measured in four replicates. Data are presented with the standard error of the mean (SEM). The statistical significance of results was tested by analysis of variance followed by Duncan's test (p < 0.05). https://doi.org/10.1371/journal.pone.0212718.g003 [/fig]
[fig] Fig 4: Cytotoxicity of methanol barley extracts on human dermal fibroblasts (HDF). The toxicity was evaluated after 72 hours of cells cultivation over a range of concentrations of extracts by standardized resazurin-based viability assay. The results are shown as the average value of four plants measured in four replicates. Data are presented with the standard error of the mean (SEM). The statistical significance of results was tested by analysis of variance followed by Duncan's test (p < 0.05). https://doi.org/10.1371/journal.pone.0212718.g004 stressful conditions. Moreover, during conditions of salt stress, transgenic barley has a higher level of antioxidant and corresponding lower amount of MDA. [/fig]
[fig] Fig 5: Genotoxicity was evaluated after 24 h of co-cultivation of human embryonic kidney cells with methanol extracts of both transgenic and non-transgenic barley exposed to Fusarium oxysporum infection. Cells incubated with H 2 O 2 served as a positive control of DNA damage. Cells without any treatment served as a negative control. Data are presented as the average of 50 individual determinations with the standard error of the mean (SEM). https://doi.org/10.1371/journal.pone.0212718.g005 [/fig]
[fig] Fig 6: BYDV and WDV titres measured in plant leaves in 42-day period following inoculation (DPI-days post inoculation). BYDV and WDV titres were measured as the mean qPCR identified in titres from triplicates, normalized to (divided by) the RNA sample concentration (for BYDV) or DNA sample concentration (for WDV) and then a base-10 logarithm transformation was applied. From each tested period, the mean from all six tested plants of each group is depicted (WT: non-transgenic (wild type) barley Hordeum vulgare L. var. Golden Promise, T: transgenic barley expressing tobacco osmotin gene). Data are presented with the standard error of the mean (SEM). The statistical significance of results was tested by analysis of variance followed by Duncan's test (p < 0.05).https://doi.org/10.1371/journal.pone.0212718.g006 [/fig]
[fig] S1: Fig. Scheme of the expression vector pBRACT214 containing the osmotin gene. (TIF) S2 Fig. Regeneration of transgenic barley var. Golden Promise after transformation with osmotin gene. a) Regenerating plantlets from calluses after 6 wk on selection medium. b) Putative transgenic plants on regenerating medium. c) Transgenic plants T0 generation in greenhouse. (TIF) S3 Fig. Detection of osmotin transgene in the genomic DNA of T0 regenerants. Lane 1-11: samples; lane 12: negative control (DNA/RNA free water); lane 13: positive control (plasmid pBRACT214::osm); lane 14: negative control (genomic DNA of WT plants); lane 15: DNA standard (50 bp DNA ladder, Bioline). Size of PCR product is 222 bp. (TIF) S4 Fig. Detection of transgene encoding the marker for selection (hpt gene) in the genomic DNA of T0 regenerants. Lane 1-3: samples; lane 4: negative control (DNA/RNA free water); lane 5: negative control (genomic DNA of WT plants); lane 6: positive control (hpt positive plant), lane 7: DNA standard (2-kb DNA ladder, Bioline). Size of PCR product is 275 bp. (TIF) S5 Fig. Exposition of transgenic and control barley to fungi infection. left: Non-transgenic barley (left) versus transgenic barley bearing tobacco osmotin gene (right) after biotic stress (15 days after first spraying of Fusarium oxysporum spores). Right: symptoms recognized on non-transgenic barley leaves after stress (chlorosis, necrosis, premature leaf drops and wilt of whole plant). (JPG) 6. Noori SAS, Sokhansanj A. Wheat plants containing an osmotin gene show enhanced ability to produce roots at high NaCl concentration. Russ J Plant Physl. 2008; 55:256-8. https://doi.org/10.1134/ s1021443708020143 PubMed PMID: WOS:000254839800014. 7. Rao MVR, Parameswari C, Sripriya R, Veluthambi K. Transgene stacking and marker elimination in transgenic rice by sequential Agrobacterium-mediated co-transformation with the same selectable marker gene. Plant Cell Rep. 2011; 30:1241-52. https://doi.org/10.1007/s00299-011-1033-y PubMed PMID: WOS:000291604100009. PMID: 21327387 8. Beyer P, Al-Babili S, Ye XD, Lucca P, Schaub P, Welsch R, et al. Golden rice: Introducing the beta-carotene biosynthesis pathway into rice endosperm by genetic engineering to defeat vitamin A deficiency. J Nutr. 2002; 132:506S-510S. PubMed PMID: WOS:000174189800032. https://doi.org/10.1093/jn/132. 3.506S PMID: 11880581 9. Capellesso AJ, Cazella AA, Schmitt AL, Farley J, Martins DA. Economic and environmental impacts of production intensification in agriculture: comparing transgenic, conventional, and agroecological maize crops. Agroecol Sust Food. 2016; 40:215-36. https://doi.org/10.1080/21683565.2015.1128508 PubMed PMID: WOS:000374240100003. 10. Talsma EF, Melse-Boonstra A, Brouwer ID. Acceptance and adoption of biofortified crops in low-and middle-income countries: a systematic review. Nutr Rev. 2017; 75:798-829. https://doi.org/10.1093/ nutrit/nux037 PubMed PMID: WOS:000412812700003. PMID: 29028269 [/fig]
|
Differential oestrogen receptor binding is associated with clinical outcome in breast cancer
[formula] B%D%(1D)F) BFDCCFD%+F A' B 0EDB%)DC0B EBDE%FDC1F A' ) BEDEF0D%0C BFD(0+D+%( A' B )11D1CEDCEF E)D((+D1EB A' ) BEDE1(DC1) BFD01CD+B1 A' B B0D10+D0+) BFD%0BDE(B A' ) FBDBFFD0)C BCDC01DFF1 A' B FEDFFCD1+1 FBD)FCDFB) A' B BCD(1%D%(C )0D)C+DBF) A' B F+DFEFDBBB B0DB1(D(B( #8G.) ) FBD+C0D)CE BCD%EEDC)C #8G.) ) F(D)1CD+B0 B0D0)BDFEF #8G.) B FCDF(%DC)) F1DFC+DC(B #8G.) B FCD%E1DF(F F1DE(%DE+1 BEDBBCDC0) BBD)0CD10C )(DC%1D)B1 )BD)0+DEBB )(DEE0D+0F )BD1C(D+CB B(D+B+D+BB )%D+BBDF0( F0D(%(D(+1 FBD0(CD(%( MACS Peaks %+D1B0 +0D1%( C%D)F1 %+D0%) %1D(C1 BED%)F BFD(%( +)D0B+ +1D%EF (BD()% +%D1BF (%DF01 CED1CF %BDCFB CEDC%1 %FD1EE )1FD00F [/formula]
## Genes near core er binding events
[formula] DHRS3 KRT15 HDAC11 FOXN1 GRIK3 KRT19 IL17RB TRAF4 KIAA0040 MAP3K14 SLC6A6 RAPGEFL1 ECE1 CUEDC1 MAPKAPK3 RARA FABP3 CYB561 PALLD IGFBP4 C1orf34 NAT9 BTF3 TUBG1 PPM1J TBC1D16 PDLIM4 CACNG4 TPM3 FASN SLC22A5 SLC9A3R1 MYOG ATP8B1 MAPK13 TMEM104 LAPTM4A EPS15L1 CCR6 ATP9B CISH ZBTB7A HSPB1 CYP2B7P1 SIAH2 XAB2 ZNF282 NFIC SSR3 JUND PVT1 JMJD2B MAML3 ZNF91 FLJ10661 HPN PITX1 CYP2A7 EBAG9 PVRL2 FLOT1 ZMYND8 LAPTM4B DHX34 IER3 SDC4 SPINK4 ATP5D SPDEF DSCR3 S100G CIRBP PLEC1 TFF1 PFKFB3 C19orf24 COX6C XBP1 DOCK1 STXBP2 KLF10 A4GALT NCOA4 MAP2K7 DPM2 LRP5L MYST4 SNAPC2 FAM102A [/formula]
LGALS2
[formula] MSMB SCN1B LAMP2 RPL3 SLC29A3 CD22 SOX3 THRAP3 PPRC1 C19orf15 LDB1 SARS TncRNA PAK4 ARRB1 CELSR2 TSKU BCL3 SUV420H1 OLFML3 TPCN1 GPR77 ALG8 PMF1 FKBP4 OXT SCNN1A PRDX6 ITFG2 TGIF2 MAX QSOX1 LTBR C20orf24 SERPINA6 ELF3 CSK WISP2 ANXA2 GPR37L1 UBE2I ACO2 CA12 IL19 ABAT CDC42EP1 ANPEP IL24 MSLN MAFF NARFL RCOR3 ALDOA MGAT3 Magmas H3F3A PPP4C SMCR7L UNC119 GREB1 HSD11B2 AGPS CCR7 CENPA OSGIN1 [/formula]
C2orf18 MAPT Supplementary figure 7. Heatmap representing clustering of patient tumour samples based on ER binding events. The clustering represents correlations between individual samples and is based on the 1,791 differential ER binding events (i.e. 1,192 poor outcome and 599 good outcome ER binding regions). Interestingly, the only tumour to cluster incorrectly is P5, which is from a patient who did not relapse and die for 13.4 years, significantly longer than the other poor outcome tumours with long term survival information (see Supplementary figure 1). The three tumours that were split into two sections and mapped separately are highlighted. In all three cases, the replicates clustered together. Supplementary figure 11. Genes within the poor and good outcome predictor. Genes in black text represent positive weight genes and blue text represent negative weight genes. ACAA1 KCNK4 ACAA1 BAMBI HCK VAMP5 Supplementary figure 12. Survival analyses in independent datasets. The genes adjacent (< 20kb) to ER binding events enriched in patients with poor outcome tumours and metastases or genes near (< 20kb) ER binding events enriched in patients with good outcome tumours, were used to stratify patients and the top one third (according to 'risk index') and bottom one third (according to 'risk index') were compared. Survival analysis of good and poor outcome gene predictor identified in ER+ tumours, but tested in cohorts of ER negative tumours. As with the ER+ tumours, distant metastasis free survival (DMFS) was used as the clinical endpoint. The predictors generated from differential ER binding events in ER+ tumours only have predictive value when tested in independent ER+ tumour cohorts and not ER-cohorts. An additional ER negative dataset was included: ;9:R65.
## Good outcome genes poor outcome genes
[formula] LOXL1 PKP1 RACGAP1 SPAG9 DIXDC1 GAD2 TMEM97 PPP2R4 ZYX APOA4 HSPB1 PLXNB2 GAS7 QSOX1 NUDT4 RPL23A EPAS1 PCTK1 POLDIP2 SPINK4 DAAM2 BPNT1 DDX23 GPX2 PRKCSH SLURP1 MAPRE1 RHOH FMOD NGFB PTTG3 RAP2A TPSB2 KCNK4 EIF4EBP1 OBSCN KLF13 USF2 FADD TSPAN5 ALOX5AP EPB41 SDHA ZNF91 CRHR1 NCAM2 DNAJA1 KIAA0100 SPARCL1 RNF2 GATAD2A ENPP4 EVC SGCB FBP1 CTNND1 RHBDF1 TPSG1 RNF10 VAV3 TMEM63A ASMTL U1SNRNPBP TRIM14 FASN ASAH1 NR2F6 CDC42EP3 SPTLC1 PDK2 SNRPD3 PCDH12 COL4A3BP HTR3A RRP12 AGTR1 CST3 CDKN1A RNF14 TMBIM4 ATF3 KCNE1 ABCC5 FAM55C RCAN1 STK19 NDUFB5 ADAM2 OXT FZD10 PGAM1 NAT2 USP12 ESRRA NDUFA4 SLC16A6 JAG2 FDFT1 BCAS1 STARD3 TAF1C SMOX BCL2L13 SPTBN2 UBA1 ALDOC ALG8 ACCN1 CAMK1D NPC1L1 C6orf64 GRB10 GRM4 DYM CACNG4 SLA EPHA1 LEFTY1ICK CTSK PTPRN2 KIF13B BAZ1B ANKFY1 KIAA1467 LPP SLK PTGIS LOXL1 PKP1 RACGAP1 SPAG9 DIXDC1 GAD2 TMEM97 PPP2R4 ZYX APOA4 HSPB1 PLXNB2 GAS7 QSOX1 NUDT4 RPL23A EPAS1 PCTK1 POLDIP2 SPINK4 DAAM2 BPNT1 DDX23 GPX2 PRKCSH SLURP1 MAPRE1 RHOH FMOD NGFB PTTG3 RAP2A TPSB2 KCNK4 EIF4EBP1 OBSCN KLF13 USF2 FADD TSPAN5 ALOX5AP EPB41 SDHA ZNF91 CRHR1 NCAM2 DNAJA1 KIAA0100 SPARCL1 RNF2 GATAD2A ENPP4 EVC SGCB FBP1 CTNND1 RHBDF1 TPSG1 RNF10 VAV3 TMEM63A ASMTL U1SNRNPBP TRIM14 FASN ASAH1 NR2F6 CDC42EP3 SPTLC1 PDK2 SNRPD3 PCDH12 COL4A3BP HTR3A RRP12 AGTR1 CST3 CDKN1A RNF14 TMBIM4 ATF3 KCNE1 ABCC5 FAM55C RCAN1 STK19 NDUFB5 ADAM2 OXT FZD10 PGAM1 NAT2 USP12 ESRRA NDUFA4 SLC16A6 JAG2 FDFT1 BCAS1 STARD3 TAF1C SMOX BCL2L13 SPTBN2 UBA1 ALDOC ALG8 ACCN1 CAMK1D NPC1L1 C6orf64 GRB10 GRM4 DYM CACNG4 SLA EPHA1 LEFTY1EIF4EBP1 OBSCN KLF13 USF2 FADD TSPAN5 ALOX5AP EPB41 SDHA ZNF91 CRHR1 NCAM2 DNAJA1 KIAA0100 SPARCL1 RNF2 GATAD2A ENPP4 EVC SGCB FBP1 CTNND1 RHBDF1 TPSG1 RNF10 VAV3 TMEM63A ASMTL U1SNRNPBP TRIM14 FASN ASAH1 NR2F6 CDC42EP3 SPTLC1 PDK2 SNRPD3 PCDH12 COL4A3BP HTR3A RRP12 AGTR1 CST3 CDKN1A RNF14 TMBIM4 ATF3 KCNE1 ABCC5 FAM55C RCAN1 STK19 NDUFB5 ADAM2 OXT FZD10 PGAM1 NAT2 USP12 ESRRA NDUFA4 SLC16A6 JAG2 FDFT1 BCAS1 STARD3 TAF1C SMOX BCL2L13 SPTBN2 UBA1 ALDOC ALG8 ACCN1 CAMK1D NPC1L1 C6orf64 GRB10 GRM4 DYM CACNG4 SLA EPHA1 LEFTY1ICK CTSK PTPRN2 KIF13B BAZ1B ANKFY1 KIAA1467 LPP SLK PTGIS LOXL1 PKP1 RACGAP1 SPAG9 DIXDC1 GAD2 TMEM97 PPP2R4 ZYX APOA4 HSPB1 PLXNB2 GAS7 QSOX1 NUDT4 RPL23A EPAS1 PCTK1 POLDIP2 SPINK4 DAAM2 BPNT1 DDX23 GPX2 PRKCSH SLURP1 MAPRE1 RHOH FMOD NGFB PTTG3 RAP2A TPSB2 KCNK4 EIF4EBP1 OBSCN KLF13 USF2 FADD TSPAN5 ALOX5AP EPB41 SDHA ZNF91 CRHR1 NCAM2 DNAJA1 KIAA0100 SPARCL1 RNF2 GATAD2A ENPP4 EVC SGCB FBP1 CTNND1 RHBDF1 TPSG1 RNF10 VAV3 TMEM63A ASMTL U1SNRNPBP TRIM14 FASN ASAH1 NR2F6 CDC42EP3 SPTLC1 PDK2 SNRPD3 PCDH12 COL4A3BP HTR3A RRP12 AGTR1 CST3 CDKN1A RNF14 TMBIM4 ATF3 KCNE1 ABCC5 FAM55C RCAN1 STK19 NDUFB5 ADAM2 OXT FZD10 PGAM1 NAT2 USP12 ESRRA NDUFA4 SLC16A6 JAG2 FDFT1 BCAS1 STARD3 TAF1C SMOX BCL2L13 SPTBN2 UBA1 ALDOC ALG8 ACCN1 CAMK1D NPC1L1 C6orf64 GRB10 GRM4 DYM CACNG4 SLA EPHA1 LEFTY1 [/formula]
[fig] Supplementary figure 8: Genome region enrichment analysis tool (GREAT) identification of enriched processes/pathways based on the differential ER binding events. [/fig]
[fig] Supplementary figure 9: Charafe-Jauffret, E., Ginestier, C., Monville, F., Finetti, P., Adelaide, J., Cervera, N., Fekairi, S., Xerri, L., Jacquemier, J.,Birnbaum, D. et al. 2006.Gene expression profiling of breast cell lines identifies potential new basal markers. Oncogene 25(15): 2273-2284. Climent, J., Dimitrow, P., Fridlyand, J., Palacios, J., Siebert, R., Albertson, D.G., Gray, J.W., Pinkel, D., Lluch, A., and Martinez-Climent, J.A. 2007. Deletion of chromosome 11q predicts response to anthracycline-based chemotherapy in early breast cancer. Cancer research 67(2): 818-826. Creighton, C.J., Massarweh, S., Huang, S., Tsimelzon, A., Hilsenbeck, S.G., Osborne, C.K., Shou, J., Malorni, L., and Schiff, R. 2008. Development of resistance to targeted therapies transforms the clinically associated molecular profile subtype of breast tumor xenografts. Cancer research 68(18): 7493-7501. Nikolsky, Y., Sviridov, E., Yao, J., Dosymbekov, D., Ustyansky, V., Kaznacheev, V., Dezso, Z., Mulvey, L., Macconaill, L.E., Winckler, W. et al. 2008. Genome-wide functional synergy between amplified and mutated genes in human breast cancer. Cancer research 68(22): 9532-9540. Smid, M., Wang, Y., Zhang, Y., Sieuwerts, A.M., Yu, J., Klijn, J.G., Foekens, J.A., and Martens, J.W. 2008. Subtypes of breast cancer show preferential site of relapse. Cancer research 68(9): 3108-3114.For the genes in each of the poor and good outcome differential binding event lists, we computed a P-value of association with DMFS (Cox regression log-rank test), in each of the nine breast cancer cohorts considered. A. The histogram of these P-values. B. Plots of the statistics against the -log10(P-values). The number of P-values passing the 0.001 threshold and for each directionality are given. 10. A. For the supervised principal components algorithm (SPCA) predictors derived from the poor and good outcome differential binding analysis (DBA) gene lists, we compare the number of genes with positive (Up-mRNA) and negative (Down-mRNA) PCA weights. B. The specific weights for the 174 poor outcome genes which are up-regulated in patients with poor outcome tumours, and the weights for the 62 good outcome genes which are downregulated in patients with poor outcome tumours. [/fig]
|
The Potential Use of Metabolic Cofactors in Treatment of NAFLD
Non-alcoholic fatty liver disease (NAFLD) is caused by the imbalance between lipid deposition and lipid removal from the liver, and its global prevalence continues to increase dramatically. NAFLD encompasses a spectrum of pathological conditions including simple steatosis and non-alcoholic steatohepatitis (NASH), which can progress to cirrhosis and liver cancer. Even though there is a multi-disciplinary effort for development of a treatment strategy for NAFLD, there is not an approved effective medication available. Single or combined metabolic cofactors can be supplemented to boost the metabolic processes altered in NAFLD. Here, we review the dosage and usage of metabolic cofactors including l-carnitine, Nicotinamide riboside (NR), l-serine, and N-acetyl-l-cysteine (NAC) in human clinical studies to improve the altered biological functions associated with different human diseases. We also discuss the potential use of these substances in treatment of NAFLD and other metabolic diseases including neurodegenerative and cardiovascular diseases of which pathogenesis is linked to mitochondrial dysfunction.
# Introduction
Non-alcoholic fatty liver disease (NAFLD) is a consequence of the imbalance between deposition and removal of lipids from the liver. The global prevalence of NAFLD continues to increase dramatically and has reached 25% at the population level. NAFLD includes a spectrum of pathological conditions, ranging from simple steatosis to hepatic inflammation referred as non-alcoholic steatohepatitis (NASH), which can progress to cirrhosis and hepatocellular carcinoma (HCC). Even though hepatic steatosis and fibrosis are reversible conditions, decompensated cirrhosis is frequently associated with irreversible hepatic damage, and transplantation is considered to be the only possible treatment option.
Oxidative stress and inflammatory responses have major role in the pathogenesis of NAFLD and its progression to NASH. NAFLD is also closely linked to complex metabolic conditions including obesity, type 2 diabetes mellitus (T2DM) and cardiovascular diseases (CVD). Considering the complexity associated with the drug development that can be used in effective treatment of the patients, supplementation of natural substances that can activate the altered metabolic pathways in NAFLD can be used in treatment of the patients. Supplementation of such metabolic cofactors may also improve the metabolic parameters in NAFLD patients and stop progression of the disease to severe form of the diseases including NASH, cirrhosis and HCC. To date, a number of metabolic cofactors have been supplemented to patients with different metabolic disorders and its positive effect has been shown in placebo controlled human clinical studies .
The pathogenesis of NASH is a rather complex process consisting of lipid accumulation, hepatocellular injury, recruitment and activation of inflammatory and hepatic stellate cells, and fibrogenesis. At earlier stages, lipid accumulation is a critical phase arising from increased dietary fatty acids and de novo lipogenesis and insufficient capacity for lipid removal by fatty acid oxidation and lipoprotein secretion. Elucidation of pathophysiological responses and cellular dysfunction induced by lipid accumulation is essential to develop efficient treatment strategies for NAFLD and other complex metabolic diseases including T2DM and CVD. Such diseases have been frequently studied based on a single gene or a specific pathway. With the "omics" revolution that enable the generation of enormous amounts of data (i.e., genomics, transcriptomics, proteomics, lipidomics, metabolomics, fluxomics, epigenetics, and metagenomics), identification of key genes and essential pathways driving steatosis, steatohepatitis and fibrosis in NAFLD is possible using a holistic systems biology approach. However, analyzing multi-layer omics data to understand the pathophysiology is a challenging task.
Systems biology enables the integration and analysis of multi-layer omics data to make predictions that can be experimentally tested. This integrative approach has successfully been used in various contexts in medical research over the last decade. We have recently performed systems level analysis by combining clinical studies with stable isotopes, in-depth multi-omics profiling and personalized network analysis, and revealed the underlying mechanisms of NAFLD to developing novel strategies for its prevention and treatment. Our integrated analysis indicated that there is an augmented requirement for Nicotinamide adenine dinucleotide (NAD)+ and reduced glutathione (GSH) in patients with NAFLD. This observation was further validated using liver transcriptomics and plasma metabolomics data. We also performed a mouse supplementation study with the precursors for NAD+ (Nicotinamide riboside and GSH (l-serine [serine] and N-acetyl-l-cysteine [NAC]), and a proof of concept human supplementation study with serine to decrease the hepatic steatosis.
Our analysis indicated that a three-step strategy including (i) increasing mitochondrial fatty acid uptake, (ii) increasing mitochondrial fatty acid oxidation, and (iii) increasing the availability of GSH can be applied to decrease the amount of hepatic steatosis in NAFLD patients. Single or combined metabolic cofactors can be supplemented to boost these metabolic processes altered in NAFLD. Supplementation of (i) l-carnitine (carnitine) may stimulate the transfer of fatty acids from cytosol to mitochondria, (ii) NR may provide required amount of NAD+ which is essential for mitochondrial fatty acid oxidation, and (iii) serine and NAC may increase the level of GSH in the cells. Here, we conducted a literature review about the use of individual metabolic cofactors including carnitine, NR, serine and NAC, in human clinical studies to treat different disorders. We also provided a guideline for use of these metabolic cofactors by reviewing different human clinical studies and observed that careful screening of the patients during human clinical studies may decrease the risk associated with the combination of metabolic cofactors. . The dosages of the Nicotinamide Riboside, l-serine, and l-carnitine used in previous human trials. The data retrieved from https://clinicaltrials.gov. The dosage of the N-acetyl-l-cysteine is provided in .. The red and blue arrows indicate the upregulation and down regulation of metabolic pathways in NAFLD, respectively. Metabolic cofactors can be supplemented to treat such metabolic abnormalities for effective treatment of the patients. The three-step strategy can be applied by supplementing (1) L-carnitine to enhance the transport of fatty acids across the mitochondrial membrane, (2) the NAD + precursor nicotinamide riboside to enhance the β-oxidation of fatty acids in mitochondria, and the (3) glutathione (GSH) precursors including serine and N-acetyl-L-cysteine (NAC) to form GSH that is required to protect liver against free radical-mediated oxidative stress generated by the increased β-oxidation of fatty acids in the mitochondria. TCA: The citric acid cycle, VLDL: Very-low-density lipoprotein, ROS: Reactive oxygen species. PKLR: pyruvate kinase L/R, FASN: Fatty acid synthase.
## Nct
## Potential risks and benefits of metabolic co-factors
Chronic metabolic disorders and aging are commonly associated with reduced plasma NAD + and GSH levels. Since hepatocellular synthesis of GSH by de novo or by the salvation pathways are required; intracellular GSH levels cannot be increased by simply providing GSH supplements itself. Based on our integrative analysis, we found that the level of GSH is not enough to maintain and regulate the thiol redox status of the liver in subjects with high hepatic steatosis at fasting stage due to depletion of glycine and serine. It has been shown that serine synthesis is downregulated in NAFLD patientsand supplementation of serine improved liver tissue function and decreased the liver steatosis in a proof of concept human study. We aimed to decrease the level of liver fat content in NAFLD patients by increasing the hepatic levels of pivotal metabolic cofactors via simultaneous dietary supplementation of carnitine, NR, serine, and NAC. The study is based on a three-step strategy to increase the amount of fat oxidized in their liver. First, we included carnitine, which is transported by carnitine palmitoyltransferase (CPT) I and II and generates a long chain acetyl carnitine ester to facilitate the transport of fatty acids (1) l-carnitine to enhance the transport of fatty acids across the mitochondrial membrane, (2) the NAD + precursor nicotinamide riboside to enhance the β-oxidation of fatty acids in mitochondria, and the (3) glutathione (GSH) precursors including serine and N-acetyl-l-cysteine (NAC) to form GSH that is required to protect liver against free radical-mediated oxidative stress generated by the increased β-oxidation of fatty acids in the mitochondria. TCA: The citric acid cycle, VLDL: Very-low-density lipoprotein, ROS: Reactive oxygen species. PKLR: pyruvate kinase L/R, FASN: Fatty acid synthase.
## Potential risks and benefits of metabolic co-factors
Chronic metabolic disorders and aging are commonly associated with reduced plasma NAD + and GSH levels. Since hepatocellular synthesis of GSH by de novo or by the salvation pathways are required; intracellular GSH levels cannot be increased by simply providing GSH supplements itself. Based on our integrative analysis, we found that the level of GSH is not enough to maintain and regulate the thiol redox status of the liver in subjects with high hepatic steatosis at fasting stage due to depletion of glycine and serine. It has been shown that serine synthesis is downregulated in NAFLD patientsand supplementation of serine improved liver tissue function and decreased the liver steatosis in a proof of concept human study. We aimed to decrease the level of liver fat content in NAFLD patients by increasing the hepatic levels of pivotal metabolic cofactors via simultaneous dietary supplementation of carnitine, NR, serine, and NAC. The study is based on a three-step strategy to increase the amount of fat oxidized in their liver. First, we included carnitine, which is transported by carnitine palmitoyltransferase (CPT) I and II and generates a long chain acetyl carnitine ester to facilitate the transport of fatty acids across the mitochondrial membrane. In addition to that, carnitine is important for exchanging acetyl groups and stabilization of acetyl-CoA and coenzyme A. Second, we included NR, a precursor of NAD + , to boost the level of hepatic mitochondrial β-oxidation of fatty acids and activate mitochondria.
Decreased electron transport chain function combined with impaired rates of fatty acid β-oxidation leads to the accumulation of incomplete products of β-oxidation, which combined with increased levels of reactive oxygen species (ROS) contribute to insulin resistance. Finally, we included the two GSH precursors serine and NAC to increase GSH levels in the hepatocytes. Increased GSH levels may protect against free radical-mediated oxidative stress generated by the increased β-oxidation of fatty acids in mitochondria. Combined metabolic cofactors may concomitantly stimulate these three different metabolic pathways and activate the liver tissue metabolism in NAFLD patients.
Metabolic cofactors including carnitine, NR, serine, or NAC have been supplemented individually for treatment of different disorders . Neither of them is known to have significant side effects or toxicity. Human clinical studies related to these metabolic cofactors as well as pharmacological data on each metabolic-cofactor is summarized below.
## L-carnitine
Carnitine is a white crystalline, hygroscopic powder and readily soluble in water and hot alcohol, and is insoluble in acetone. The absolute bioavailability of carnitine is~15-16%. The mean distribution half-life is~0.6 h and the mean apparent terminal elimination half-life is 17.4 h. Total body clearance of carnitine (Dose/AUC including endogenous baseline concentrations) is a mean of 4.00 L/h. Carnitine is not bound to plasma protein or albumin when tested at any concentration.
The major metabolites are trimethylamine N-oxide, excreted primarily in urine (8% to 49% of the administered dose) and [3H]-γ-butyrobetaine, excreted primarily in feces (0.44% to 45% of the administered dose). Urinary excretion of carnitine is about 4 to 8% of the dose. Fecal excretion of total carnitine is less than 1% of the administered dose.
Carnitine is a naturally occurring substance required in mammalian energy metabolism. It is a carrier molecule that facilitates the transport of long-chain fatty acids across the inner mitochondrial membrane and delivers substrate for oxidation of fatty acids and subsequent energy production. Carnitine supplementation is recommended to the patients with inborn errors of metabolism and patients undergoing hemodialysis for kidney disease.
Carnitine deficiency is characterized with very low carnitine levels in plasma and tissues and may be either primary or secondary. Primary carnitine deficiency is an autosomal recessive disorder caused by a deficiency in the plasma membrane carnitine transporter and leads to urinary carnitine wasting. SLC22A5 mutations can also affect carnitine transport and decrease plasma carnitine levels.
In primary systemic deficiency, the clinical complications are associated with the recurrent episodes of Reye-like encephalopathy, hypoketotic hypoglycemia, and/or cardiomyopathy. Associated symptoms also include hypotonia, muscle weakness, and failure to thrive. In some patients, particularly those presenting with cardiomyopathy, carnitine supplementation may rapidly alleviate signs and symptoms. Secondary carnitine deficiency is associated with inadequate carnitine intake, decreased carnitine synthesis due to liver disorders, loss of carnitine during diarrhea, diuresis, or hemodialysis.
Carnitine is used in the treatment of primary and secondary carnitine deficiency. It may also be recommended to the patients taking certain drugs (such as valproic acid for seizures or antibiotics for tuberculosis), or during medical procedures (hemodialysis for kidney disease) that deplete the body's carnitine deficiency.
Carnitine has also been supplemented to patients with heart failure, ischemic heart diseases, peripheral arterial diseases, HIV, male infertility, anorexia, chronic fatigue syndrome, and fatigue associated with chronic diseases and chronic obstructive pulmonary disease . It is also used as a replacement supplement in strict vegetarians, dieters, and low-weight or premature infants. In athletes, carnitine has been used to improve performance, although the beneficial effects in athletes are not consisted in all studies.
## Dosage
The recommended daily dosage of carnitine is 1 to 3 g, but higher doses have been used in clinical studies (see . Numerous studies have been performed:
In patients with carnitine depletion in peripheral blood mononuclear cells, carnitine has been supplemented at a dose of 6 g/day for 2 weeks. Several studies have tested if carnitine supplementation promotes weight loss in obese subjects (4 g/L, 8 weeks). Efficacy and effectiveness of carnitine supplementation for cancer-related fatigue has been analyzed in a systematic literature review and meta-analysis. Nine studies used a dose between 2 to 6 g per day. Impact of carnitine supplementation on plasma lipoprotein(a) concentrations have been analyzed in a recent systematic review and meta-analysis of human clinical trials. Studies used 2-4 g/day. A systematic review was conducted to determine the effects of carnitine on all-cause mortality and cardiovascular morbidities in the setting of acute myocardial infarction (meta-analysis of five controlled trials, n = 3108). There were no significant differences between the effects of daily carnitine supplementation of 2 g and 6 g on heart failure, unstable angina, or myocardial reinfarction. The effect of carnitine supplementation on the regression of NASH was evaluated in 74 patients with a clinical and pathologic diagnosis of NASH. The study subjects were randomly allocated to the placebo or to the carnitine (2 g per day divided into two equal doses for 24 weeks) groups. At the end of the study, carnitine-treated patients showed significant improvements in AST, ALT, gamma-GT, total cholesterol, LDL-cholesterol, HDL-cholesterols, triglycerides, glucose, HOMA-IR, C-reactive protein, TNF-alpha, and histological scores. Thus, carnitine supplementation reduced inflammation, and improved liver function, glucose plasma level, lipid profile, HOMA-IR, and histological manifestations of NASH.
## Safety aspects
Carnitine is usually well-tolerated and the FDA granted carnitine with Generally Recognized as Safe (GRAS) status. Various mild gastrointestinal complaints have been reported during the long-term administration of oral l-or d, carnitine; including transient nausea and vomiting, abdominal cramps, and diarrhea. At doses of approximately 3 g/day, carnitine supplements can cause nausea, vomiting, abdominal cramps, diarrhea, and a "fishy" body odor. Rare side effects include muscle weakness/mild myasthenia in uremic patients and seizures in those with seizure disorders.
Gastrointestinal adverse reactions with carnitine may be avoided by a slow consumption of the solution or by a greater dilution. Decreasing the dosage often diminishes or eliminates drug-related patient body odor or gastrointestinal symptoms when present. Tolerance should be monitored very closely during the first week of administration, and after any dosage increases.
It has been reported that the intestinal bacteria may metabolize carnitine to form a substance called Trimethylamine-N-oxide (TMAO) that might increase the risk of cardiovascular disease. This effect appears to be more pronounced in people who consume meat than in vegans or vegetarians. The implications of these findings are not well understood and require more research.
There have been no reports of toxicity from carnitine over dosage. Carnitine is easily removed from plasma by dialysis. Large doses of carnitine may cause diarrhea. Pregnancy risk category of carnitine is B and it should be used during pregnancy only if it is clearly needed. Drug interaction with carnitine is very rare but it may decrease the effectiveness of the thyroid hormone and increase the effects of warfarin.
## Nicotinamide riboside
NAD is an endogenous substance that is involved in several important cell functions such as signal transduction, DNA repair, and post-translational protein modifications. NAD-consuming activities and cell division necessitate ongoing NAD synthesis, either through a de novo pathway that originates with tryptophan or via salvage pathways from three NAD + precursor vitamins, NR, nicotinamide, and nicotinic acid. In animals, NAD generation is vital, since it is linked to several redox reactions in the body. NAD plays a central role in fatty acid metabolism, energy metabolism, and oxidative phosphorylation and is a key component of many metabolic pathways for carbohydrates, lipids, and amino acids.
NR is in wide use as an NAD+ precursor vitamin. It is available as an over-the-counter dietary supplement. The therapeutic potential of NAD has been investigated in several clinical conditions (e.g., aging), improving exercise performance, assisting weight loss and treating chronic fatigue syndrome. Its effect has also been tested in improving depression, cognitive function, as well as treating patients with dyslipidemia, diabetes, Parkinson's disease, Alzheimer's disease, and dementia .
## Dosage
The daily dose of NR is 500-2000 mg in previously performed human clinical studies . Numerous studies have been performed:
Trammell et al. determined the time and dose-dependent effects of NR on blood NAD + level in humans. They reported that human blood level of NAD + can rise as much as 2.7-fold with a single oral dose of NR in a pilot study. They also demonstrated that single doses of 100, 300, and 1000 mg of NR produce dose-dependent increases in the blood NAD + metabolome in the first clinical trial of NR pharmacokinetics in humans. reported an open-label, non-randomized study of the pharmacokinetics of NR and its effects on blood NAD + levels. In eight healthy volunteers, 250 mg NR was orally administered on days 1 and 2, then uptitrated to peak dose of 1000 mg twice daily on days 7 and 8. On the morning of day 9, subjects completed a 24-hour pharmacokinetic study after receiving 1000 mg NR at t = 0. They analyzed whole-blood levels of NR, clinical blood chemistry, and NAD + levels and reported that oral NR was well tolerated with no adverse events. Significant increases comparing baseline to mean concentrations at steady state were observed for both NR (p = 0.03) and NAD + (p = 0.001); the latter increased by 100%. Absolute changes from baseline to day 9 in NR and NAD + levels correlated highly (R 2 = 0.72, p = 0.008). The authors concluded that NR increases circulating NAD + in humans and it may be used in treatment of patients with diseases associated to mitochondrial dysfunction.
## Safety aspects
In 2016, the FDA granted NR with GRAS status on the basis of existing clinical study, which showed that the "no observed adverse effect level (NOAEL) was 300 mg/kg/day." NR seems safe for most people when used appropriately and short-term, up to 12 weeks. It is also approved as a food ingredient in enhanced water products, protein shakes, nutrition bars, and chewing gum at no more than 0.027% by weight.
There was no mortality at an oral dose of 5000 mg/kg. Based on the results of a 14-day study, a 90-day study was performed comparing NR at 300, 1000, and 3000 mg/kg/day to an equimolar dose of nicotinamide at 1260 mg/kg/day as a positive control. Results from the study show that NR had a similar toxicity profile to nicotinamide at the highest dose tested. The lowest observed adverse effect level for NR was 1000 mg/kg/day, and the no observed adverse effect level was 300 mg/kg/day.
Flushing is the most sensitive side effect of nicotinic acid supplementation. After ingestion of supplemental nicotinamide, no cases of flushing or glucose intolerance have been reported and only one case of hepatitis was reported following the ingestion of greater than 3 g/day for several days. Such side-effects have not been reported for supplementation of NR. NR was not genotoxic but data on its use during pregnancy and breast-feeding is inadequate. There is no known drug interaction with the supplementation of NR.
## L-serine
Serine is a non-essential amino acid; however, under certain circumstances, vertebrates cannot synthesize it in sufficient quantities to meet necessary cellular demands. Serine is biosynthesized in the mammalian system from 3-phosphoglycerate and serves as a precursor for the synthesis of the amino acids including glycine and cysteine. Physiologically, it has a variety of roles, perhaps most importantly as a phosphorylation site in proteins. Mutations in the metabolic enzymes that synthesize serine have been implicated in various human diseases.
## Dosage
The FDA determined serine with GRAS status and it appears to be neuroprotective. The daily dose of serine is 1-30 g in previous human clinical studies , but higher doses have been used in some clinical studies . Numerous studies have been performed including:
Serine supplementation (three daily doses of 5 g of serine [i.e., 190 mg/kg]) has also been shown to be safe even in pregnancy, as shown by in pre-and postnatal treatment of 3-phosphoglycerate-dehydrogenase deficiency. Hereditary sensory and autonomic neuropathy type 1 (HSAN1) is a disorder caused by missense mutations in the enzyme serine palmitoyltransferase (SPT). Subjects received daily supplements of powdered serine (mixed in water) on a low-or high-dose schedule (200 or 400 mg/kg body weight, respectively; n = 7 per group). Results showed that an altered substrate selectivity of the mutant SPT is key to the pathophysiology of HSAN1 and raise the prospect of serine supplementation as a first treatment option for this disorder. In our previous study, we assessed the effect of dietary supplementation with serine (200 mg/kg per day) for 2 weeks on fatty liver and fasting levels of plasma markers of liver functions in six obese subjects with NAFLD. Our analysis showed that supplementation of serine improved markers of liver tissue function and significantly decreased liver fat. Fridman et al. performed a randomized, double-blind, placebo-controlled trial (n = 18) to evaluate the efficacy and safety of serine treatment for adults with hereditary sensory and autonomic neuropathy type 1 (HSAN1). The study subjects were randomized to serine (400 mg/kg/day) or placebo for one year. All participants received serine during the second year. Analysis of vital signs, physical examination findings, and clinical laboratory examinations did not reveal adverse effects of serine. Thus, long-term serine supplementation did not reveal adverse effects of serine.
## N-acetyl-l-cysteine
NAC is a white to white with light yellow cast powder, and has a pKa of 9.5 at 30 - C. NAC is stable in gastric and intestinal fluids and rapidly absorbed after oral administration. It is not affected by food intake. It reaches peak plasma concentration in 30-60 min after application. The distribution volume (Vd) is between 0.33 and 0.47 L/kg, which is evident in extracellular fluids and passes primarily to the lung, kidney and liver. After oral administration, 48% of the amount passed to the blood is determined in the lungs. The rate of binding to plasma proteins is about 50%. NAC is extensively metabolized in liver, and 22-30% is excreted in urine in the form of sulfate and taurine. NAC has a half-life of 5.6-6 h in adults.
NAC is the N-acetyl derivative of the amino acid l-cysteine and it is an essential precursor in the formation of the antioxidant glutathione within the human body. Hence, administration of NAC replenishes glutathione stores.
Glutathione may act as an endogenous neuromodulator, modulate the redox state of the N-methyl-d-aspartate receptor complex, and activate ionotropic receptors that are different from any other excitatory amino acid receptor, which may constitute glutathione receptors, potentially making it a neurotransmitter. Since NAC is a prodrug of glutathione, it may modulate some/all of the glutathione receptors. It has potential to confer antioxidant effects and may reduce free radicals. NAC also possesses some anti-inflammatory effects possibly via inhibiting NF-κB and modulating cytokine synthesis.
NAC has two approved indications: (i) treatment of paracetamol (acetaminophen) overdose associated with liver damage and (ii) removal of thick mucus in individuals with cystic fibrosis or chronic obstructive pulmonary disease. NAC has also been tested for contrast induced nephropathy, infertility, cystic fibrosis, ischemic heart diseases, HIV, hypercholesterolemia, and schizophrenia . It is beneficial effect has been reported in any kind of acute hepatic failure and treatment of liver diseases as discussed below.
## Dosage
The daily dose of NAC is 1-5 g in previous human clinical studies , but higher doses have been used in different clinical studies .
NAC is available as intravenous (IV) and oral formulations. The IV injection and inhalation preparations are, in general, prescription only, whereas the oral solution and the effervescent tablets are available over the counter in many countries including the United States. Daily dose of NAC is 200 mg capsules taken 3 times a day (morning, lunch, evening) or a single dose of 600 mg (3 capsules) in the evening. In paracetamol poisoning, the loading dose is~140 mg/kg, and the maintenance dose is 70 mg/kg every 4 h (17 doses). Numerous studies have been performed:
In early psychosis, NAC was administered at a dose of 2700 mg/day for 6 months in a double-blind placebo-controlled trial. NAC was administered to enhance performance of elite sport. A recent systematic review of the literature evaluated the effect of NAC supplementation. The typical daily dose of NAC reported was 5.8 g/day; with a range between 1.2 and 20.0 g/day. The effect of NAC supplementation on oxidative stress status and alveolar inflammation was analyzed in a double-blind, randomized clinical trial using a dose of 1800 mg/day for 4 months in people exposed to asbestos. NAC was administered in oral doses of 6000-8000 mg daily for several months in HIV-infected patients. It had a good safety profile and minimal adverse effects. NAFLD patients (n = 30) were randomly selected to receive either NAC (600 mg per 12 h) or vitamin C (1000 mg per 12 h). Liver function tests (ALT, AST and ALP) were measured as well as the grade of steatosis, the pattern of its echogenicity, the span of the liver and the spleen, and the portal vein diameter before the intervention. Patients were followed up using the same method of evaluation repeated in the first, second, and third months. NAC resulted in a significant decrease of serum ALT after three months, compared to vitamin C. This effect was independent of the grade of steatosis in the initial diagnosis. It has been reported that NAC significantly decrease the span of the spleen and it can be used to improve liver function in patients with NAFLD. The therapeutic effect of NAC in the treatment of NASH was investigated in 35 patients diagnosed with NASH based on liver biopsy. Patients were divided into two groups: the first (18 patients) was administered NAC 600 mg/day orally for 4 weeks, while the control group (17 patients) was followed up without therapy. Results did not show improved liver function in this study. It has been reported that the daily amount of glutathione synthesis in humans is 10-15 g and most of the sources of this are provided from the natural sources of the organism. Therefore, the authors hypothesized that the lower dosage of NAC (600 mg/day) might not affect glutathione synthesis to a great extent. To test whether glutathione deficiency occurs due to the diminishd synthesis and contributes to oxidative stress, eight elderly (60-75 years) and eight younger (30-40 years) subjects received stable-isotope infusions of [ 2 H( 2 )]glycine, after which red blood cell (RBC) glutathione synthesis and concentrations, plasma oxidative stress, and markers of oxidant damage were measured. Results showed that glutathione deficiency in elderly humans occurs because of a marked reduction in synthesis, and that dietary supplementation with the two glutathione precursors cysteine (as NAC) and glycine fully restores glutathione synthesis and concentrations and lowers levels of oxidative stress and oxidant damages.
## Safety aspects
Side effects occur rarely with NAC usage. Even at very high doses, serious adverse events or signs of intoxication have not been observed. Most common adverse reactions (incidence greater than 2%) are rash, urticaria/facial flushing, pruritus, nausea, and vomiting. Adverse effects for oral formulations of NAC have been reported to include nausea, vomiting, rash, and fever.
Although unclear, there was a trend of increasing side effects with increasing doses and IV usage of NAC compared with placebo. Anaphylactic reactions, i.e., pruritus, rash, angioedema, bronchospasm, tachycardia, and hypotension have been previously reported to occur within 30 min after IV loading dose of NAC in~3-6% of people. Hypersensitivity reactions, including generalized urticaria, have been observed in patients receiving oral NAC for acetaminophen overdose, but anaphylaxis has rarely been reported with oral administration.
Occasionally, severe and persistent vomiting occurs as a symptom of acute acetaminophen overdose. Treatment with NAC may aggravate the vomiting and increase the risk of upper gastrointestinal hemorrhage in at risk patients (e.g., those with esophageal varices, peptic ulcers, etc.).
Large doses in a mouse model showed that NAC could potentially cause damage to the heart and lungs. It has been shown that NAC was metabolized to S-nitroso-N-acetylcysteine, which increased blood pressure in the lungs and right ventricle of the heart (pulmonary artery hypertension) in mice treated with NAC. The effect was similar to that observed following a three-week exposure to an oxygen-deprived environment (chronic hypoxia). The authors also found that S-nitroso-N-acetylcysteine induced a hypoxia-like response in the expression of several important genes both in vitro and in vivo.
No specific antidote is currently available; supportive and symptomatic treatments are performed. NAC is contraindicated in patients with previous allergic/anaphylactoid reaction to NAC. Patients with acute asthma attacks can also not use NAC. Pregnancy risk category of NAC is B. There is insufficient data on the usage in pregnancy and lactation and it should only be used if it is necessary. Drug interaction with NAC is very rare. Minor interactions have been reported with parenteral nitroglycerine and oral forms of nitrates. NAC has no effect on the use of vehicles and machinery.
# Conclusions
Here, we reviewed the dosage and known safety of carnitine, NR, serine, and NAC in different human clinical trials with a special emphasis on human safety. Three of the compounds are found naturally as constituents in food, except for NAC, which is a drug, despite its primary metabolite being cysteine. These metabolic cofactors can be tested in future and clinical implications in single or combined form for treatment of different diseases.
The concept that substrate deficiencies for metabolic enzymes underlie the progression of and may even be the causes of human diseases, and it has rarely been mechanistically explored for different diseases. Based on our integrative systems level analysis, we suggested the use of carnitine, NR, serine, and NAC for effective treatment of NAFLD. Of note, following the failure of recent clinical trials targeting single molecule or pathway; current NASH studies show that the trend towards targeting multiple pathways employing combination therapeutics. This is in line with changing our understanding of NASH pathogenesis from double-hit to multiple-hit hypothesis. Therefore, a combination of supplements targeting several NASH pathways would probably be more appropriate.
From a toxicological point of view, using nutritional substances rarely causes safety issues in humans. The "safe doses" for most nutritional/food substances are already known in humans and, unlike most pharmaceuticals, there is often a very large margin to toxicity using nutritional compounds. As an example, the dose at which vitamin B 12 becomes toxic for humans is over 100-fold higher than the recommended daily intake. The reason for this low toxicity of nutritional substances is because they are generally directly or indirectly metabolized via cellular metabolic pathways. The products are easily dealt with either by shuffling substrate-overload to other metabolic pathways, or used as building blocks of other compounds (such as glycogen, fat, proteins), or excreted from the cell or body.
Despite this, many calls to national poison centers occur every year due to ingestion of high doses of vitamins and minerals, especially for the fat-soluble vitamins A, D, and iron. Regarding pharmaceuticals, with the exception of some antibiotics, very few drugs are given in gram (g) doses.
Most drugs are metabolized in the liver in order to aid excretion from the body. Almost all medical drugs are foreign to the body, and their individual structures and use are largely directed to alteration of an enzyme or receptor activity, before being excreted from the body. Hence, during the supplementation of these metabolic cofactors in human clinical trials, patients should be screened carefully to avoid potential side effects on liver, heart and kidney functions.
We demonstrated that simultaneous supplementation of combined metabolic cofactors may improve the efficacy of the intervention in patients with NAFLD and decreases liver fat. Considering that NAFLD, obesity, type 2 diabetes, neurodegenerative diseases, and cardiovascular diseases are common conditions that regularly co-exist and act synergistically. Furthermore, it has a similar pathogenesis with alcoholic fatty liver disease; such metabolic cofactors can also be used in the treatment of the subjects with such disorders after testing its effect in placebo-controlled human clinical studies.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6643/11/7/1578/s1, : The dosages of the Nicotinamide Riboside, L-serine, L-carnitine and N-acetyl-L-cysteine used in previous human trials. The data retrieved from https://clinicaltrials.gov.
Funding: This research was funded by Knut and Alice Wallenberg Foundation, grant number 2017.0303 and The APC was also funded by Knut and Alice Wallenberg Foundation. |
Demographic, Clinical and Management Characteristics of Newly Diagnosed COPD Patients in Turkey: A Real-Life Study
Purpose: In order to determine the clinical and sociodemographic characteristics of newly diagnosed treatment-naïve asthma and COPD patients in Turkey, a multicenter study in 2012 was initiated . We aimed to investigate the characteristics and therapies of COPD patients in the original study in more detail. Patients and Methods: This nation-wide, multicentric, non-interventional, prospective, reallife observational cohort study was conducted in 122 centers. The newly diagnosed patients were not receiving any treatment before the recruitment. Their general characteristics, the combined GOLD 2011 COPD categories and exacerbation histories were noted. The patients were followed up with 3 voluntary visits for 1 year. Their adherence to the inhaled treatment according to GOLD 2011 was evaluated during follow-up visits. Results: The study included 776 COPD patients. Their mean age was 59.4±9.1 years, and 11.9% of the patients were female. 35.1% of the patients were in the GOLD 2011 C and D category. 12.6% are frequent exacerbators, and 52.8% had at least one comorbid condition. 71.8% overtreatment rate was detected. Their attendance rates for three follow-up visits became 55.9%, 32.9% and 18.7%, respectively. The adherence rate to the treatment was measured as 81.9%. Conclusion: Although these patients were diagnosed for the first time, the GOLD C and D categories and frequent exacerbator phenotype were found at a high rate. They were usually prescribed an overtreatment regimen. We think that newly diagnosed COPD patients should be evaluated carefully, and best effort should be made to treat these patients in accordance with the recommendations of the major COPD guidelines.
# Introduction
Chronic obstructive pulmonary disease (COPD) is one of the leading causes of mortality and morbidity worldwide. [bib_ref] Worldwide burden of COPD in high-and low-income countries. Part I. The Burden..., Buist [/bib_ref] The population-based surveys of chronic obstructive pulmonary disease (COPD) have confirmed that COPD underdiagnosis is universally high and significant airflow obstruction is often present at the first diagnosis. [bib_ref] Standards for the diagnosis and treatment of patients with COPD: a summary..., Celli [/bib_ref] [bib_ref] Determinants of underdiagnosis of COPD in national and international surveys, Lamprecht [/bib_ref] However, the newly diagnosed patients are expected to be in the early stages. The studies for understanding of different characteristics of newly diagnosed COPD patients are almost lacking. The newly diagnosed COPD patients are presumably to be different from usual COPD patients due to the perception of symptoms and adherence to treatment.
Turkey is among the countries with the highest prevalence of COPD. In the BOLD-Adana Turkey study, the prevalence of COPD was found [bib_ref] Is exposure to biomass smoke the biggest risk factor for COPD globally?, Salvi [/bib_ref].1% (28.5% in men and 10.3% in women) with a fixed ratio (FEV1/FVC <70%) in adults over 40 years of age. [bib_ref] Prevalence of COPD in Adana, Turkey (BOLD-Turkey Study), Kocabas [/bib_ref] We published the clinical, therapeutic features and treatment adherence of newly diagnosed treatment-naïve Turkish asthma and COPD patients in an observational multi-centric cohort study. [bib_ref] Real life profile of asthma and chronic obstructive pulmonary disease patients in..., Misirligil [/bib_ref] In this study, since the clinical and demographic characteristics of asthma and COPD patients were presented by comparison, the data of COPD patients could not be presented and discussed in detail. In this study, we aimed to analyze the clinical and therapeutic characteristics of COPD patients in this cohort in detail.
## Materials and methods study design
This study is a nation-wide, multi-centric, prospective and non-interventional study focusing on the general characteristics, diagnosis and treatment approaches for newly diagnosed patients with COPD under real-life conditions for 1 year. Written informed consent was obtained from all subjects. The protocol, which was in accordance with the Declaration of Helsinki, was approved by the ethics committee of Erciyes University, Faculty of Medicine (February 10, 2012). The study was conducted between June 2012 and March 2014 in 122 centers (university hospital, public hospital and private hospital) in Turkey. All participants in the study were chest disease specialists.
COPD patients were enrolled on the basis of the GOLD (Global Initiative for COPD) 2011.A standard web-based questionnaire including items related to demographics, clinical, laboratory and treatment parameters was applied to the patients. All procedures were administered by the investigators (pulmonary physicians) at 122 centers and supervised by the authorized Executive Board consisting of 11 academics for this research organization.
## Study population
To select the study patients and to avoid bias, the physicians screened all patients consecutively and all eligible patients were informed about the study. Outpatient COPD patients meeting the inclusion criteria below were included in the study:
- Age of diagnosis: ≥40 years to ≤80 years. - Diagnosis of COPD within the previous 3 months or after the initiation visit. - Diagnosis of COPD was based on a post-bronchodilator forced expiratory volume in 1 s (FEV 1 )/forced vital capacity (FVC) ratio of <0.7, in the absence of a primary diagnosis of bronchiectasis, asthma or any other significant respiratory disease. The patients also had a smoking history of at least 10 packs/year, or occupational exposure to irritant or toxic gases or biomass exposure. - Patients (or their legal representatives) who accepted to sign the informed consent form.
## Procedures and measures
All procedures were supervised by the authorized Executive Board of 11 academics for this research organization. A web-based "Case Report Form" was prepared by the Executive Board and contract research organization. The Case Report Form included previous and current medical history, demographics (the region of life spent longest, the region of life spent in last 3 months, height, body weight, body mass index and education status), risk factors (smoking, occupation, and biomass using), spirometry, disease severity (combined GOLD 2011 criteria for COPD), and the treatment prescribed. At the first visit, the patients were taken medical history and examined. Spirometry tests were performed, chest radiographs were examined, treatments were arranged and the Case Report Form were completed. After the initial visit, all patients were followed-up for 12 months. Intended timing for the visits was 0-1 month for visit-1, 1-3 months for visit-2 and 3-12 months for visit-3; however, the frequency of the visits was not determined strictly, and the physicians were completely free to plan the follow-up procedures. Data collected during the follow-up visits included the information on exacerbations and assessment of adherence to the treatment. Two or more exacerbations within 1 year were considered to be "frequent exacerbation."
The physician evaluated patients' adherence to the medication at each visit using a scale. According to this scale, "high compliance" defines the patients who use medications regularly, "medium compliance" and "low compliance" terms define those who use medications with partial or significant interruptions, respectively. And finally, "noncompliance" defines the patients who did not use medication at all. The assessment for patients' adherence to the medications was completely based upon their statement.
## Data management
Physicians were blind to patients of other centers. All lower/upper limits regarding the laboratory test results, decimals, birth dates, etc. were identified in the software and entries out of limits were not allowed. System retrieved and weekly updated the package information (name, dosage, pharmaceutical form) of all the medications from the list at the website of Turkish Drug and Medical Device Institution (www.iegm.gov.tr) and allowed physicians to choose the correct and valid name of medication.
To improve data quality, the audit visits were performed at 20 centres (15% of all) which included 450 patients (22% of all) and data were compared with the source documents. Data were protected at high capacity servers located at data center of Türk Telekom (communication and substructure provider company). System equipments were designed binary so as to provide back-up in case of any breakdown. Security of the system was provided with software like firewall, antivirus, antispam, IPS (Intrusion Prevention System) and VPN (Virtual Private Network). SSL/VPN (Secure Socket Layer/Virtual Private Network) technology was used for encryption of database. To ensure the secure data transfer, a "https://" (Hyper Text Transfer Protocol Secure) link was used.
# Statistical analysis
Patients' demographics and disease characteristics were presented as mean ± SD for continuous variables and described as percentages for categorical variables.
# Results
A total of 776 newly diagnosed treatment-naïve COPD patients from 122 centers were included. 11.9% of the patients were female. Their mean age was 59.4±9.1 years, and more than 50% of the patients were below 60 years. Current smoker rate was 56.3%. Education level of the study group was low, and approximately 80% of the patients had less than high school education. 23.4% of the patients were obese (BMI≥30). 60% of the whole COPD group was living in the cities. [fig_ref] Table 1: Demographic and General Characteristics of the Patients [/fig_ref] demonstrates detailed data about the demographic and different characteristics of the patients included in the analysis.
Disease severity assessed during the initial visit using the combined GOLD 2011 criteria revealed the rates from A to D categories 25.7%, 39.2%, 28.5% and 6.6%, respectively. More than one-third of the patients (35.1%) were in the GOLD 2011 C and D categories despite their being newly diagnosed and treatment-naïve COPD patients. The detailed past history pointed out that 12.6% of the patients were frequent exacerbators (≥2 exacerbations in the preceding year), and 4.3% had hospitalization due to an exacerbation. More than half of the patients (52.8%) had at least one comorbid condition. The most frequent comorbidities were hypertension (21.1%), coronary artery disease (13.6%) and diabetes mellitus (10.1%) [fig_ref] Figure 1: Frequency of comorbidities in newly diagnosed COPD patients [/fig_ref].
Distribution of the preferred inhaled drug regimens according to the GOLD categories is shown in [fig_ref] Table 2: Distribution of the Preferred Inhaled Drug Regimens According to the GOLD COPD... [/fig_ref]. Considering the study group as a whole, the most frequently prescribed drug regimen was ICS+LABA+LAMA combination (54.9%). The adherence of the physicians to GOLD 2011 medication recommendations is shown in [fig_ref] Table 3: Adherence to GOLD 2011 Treatment Recommendations of Physicians [/fig_ref]. Considering the GOLD 2011 treatment strategy criteria, the overall overtreatment rate was determined as 71.8%. Overtreatment rates were 66.1%, 78.7%, 81.6% and 10.6% according to the GOLD 2011 COPD categories A, B, C and D, respectively.
The rates for visits 1, 2 and 3 in the 1-year follow-up were 55.9%, 32.9% and 18.7%, respectively. According to the patients' statements, 81.9% of the patients had high adherence to the treatment prescribed. The compliance rates were partially stable between visits in the GOLD groups.
# Discussion
This multicentre, non-interventional, prospective, real-life observational study demonstrated the demographical, clinical and management characteristics of the newly diagnosed COPD patients in our country. To our knowledge, this is the first prospective study conducted on newly diagnosed and treatment-naïve COPD patients. Since COPD is closely related to age and is very rare among subjects below 40 years of age and elderly patients have difficulty performing spirometry correctly, 8-10 patients 40 years and over and 80 years and under were included in the study. The important findings of the study were that approximately one-third of newly diagnosed COPD patients had the GOLD stage C or D disease severity, and more than 70% of the patients were over treated. The detailed past history pointed out that 12.6% of our patients were frequent exacerbators before COPD diagnosis. This high exacerbation rate in newly diagnosed COPD patient group is a very important finding because COPD exacerbations are a major cause of morbidity and mortality. Generally mild to moderate COPD patients are underdiagnosed and the diagnosis of COPD is usually made in moderate or severe stages. [bib_ref] Underdiagnosis of chronic obstructive pulmonary disease: a qualitative study in primary care, Walters [/bib_ref] Delay in diagnosis of COPD is a global and multifactorial problem. [bib_ref] Determinants of underdiagnosis of COPD in national and international surveys, Lamprecht [/bib_ref] In a Spain study, COPD prevalence was found 10% and only 27% of these patients had COPD diagnosis before. [bib_ref] Prevalence of COPD in Spain: impact of underdiagnosis COPD on quality of..., Miravitlles [/bib_ref] In a Norway study, only 20% of male patients and 30% of female patients were diagnosed COPD by a physician. [bib_ref] Increased prevalence of chronic obstructive pulmonary disease in a general population, Waatevik [/bib_ref] The rate of underdiagnosed COPD patients in Turkey is much higher than European countries. In the BOLD-Adana study, 8.4% patients had COPD diagnosis. [bib_ref] Prevalence of COPD in Adana, Turkey (BOLD-Turkey Study), Kocabas [/bib_ref] The most important reason of delay in COPD diagnosis is the slowly progressive nature of COPD. 1 Therefore, the disease usually remains undetected, and a large number of patients are first determined during the exacerbation period. Another important reasons for the late diagnosis are the lack of awareness about COPD in health care providers, 13 underuse of spirometry in health care centers [bib_ref] Screening for and early detection of chronic obstructive pulmonary disease, Soriano [/bib_ref] [bib_ref] Underuse of spirometry by general practitioners for the diagnosis of, Caramori [/bib_ref] and low awareness of COPD among the general population. [bib_ref] Chronic respiratory symptoms, spirometry and knowledge of COPD among general population, Miravitlles [/bib_ref] Only 5.6% of the patients were nonsmokers. 7.2% of patients had biomass fuels use in their history. In lowincome countries, exposure to biomass fuels for cooking and heating is an important cause of COPD in rural areas of developing and the least developed countries. [bib_ref] Indoor air pollution in developing countries: a major environmental and public health..., Bruce [/bib_ref] [bib_ref] Is exposure to biomass smoke the biggest risk factor for COPD globally?, Salvi [/bib_ref] Biomass smoke is an important contributing factor to the development of chronic airway diseases in non-smoker females in the Central, East, and South Eastern Anatolian Regions of Turkey. [bib_ref] Prevalence of COPD: first epidemiological study of a large region in Turkey, Günen [/bib_ref] [bib_ref] Obstructive airway diseases in women exposed to biomass smoke, Ekici [/bib_ref] Underdiagnosis of COPD was frequent in women with exposure biomass fuels due to a lack of knowledge of hazards of biomass smoke, the low level of socioeconomic status and the ignorance of the health care providers. [bib_ref] The prevalence of chronic obstructive pulmonary disease and the determinants of underdiagnosis..., Mahishale [/bib_ref] In the same period with this study, another study called the ALPHABET with a similar methodology was conducted on usual COPD outpatients (n:1610).In both studies, the demographic characteristics of the patients were similar. However, the mean age of our study groups was found approximately 3 years younger than the usual COPD outpatients. Our findings have shown that 35.1% of the patients were labeled as GOLD C and D category. Interestingly, the ALPHABET study also yielded that 38.2% of the usual outpatients were identified as in the GOLD C and D categories.These figures from newly diagnosed treatment-naïve outpatients and usual outpatients were quite similar.
COPD patients have a large number of comorbidities, and these comorbidities can also be seen in newly diagnosed patients. [bib_ref] Patterns of comorbidities in newly diagnosed COPD and asthma in primary care, Soriano [/bib_ref] The comorbidities of COPD may increase the risk of morbidity and mortality. [bib_ref] Mortality in COPD: role of comorbidities, Sin [/bib_ref] We found that at least one comorbidity was present in 52.8% of the patients. Although the comorbidity rate in our study was found comparably less than the previous studies (65-97%), their rank order was almost the same. [bib_ref] Five comorbidities reflected the health status in patients with chronic obstructive pulmonary..., Frei [/bib_ref] [bib_ref] Comorbidity, systemic inflammation and outcomes in the ECLIPSE cohort, Miller [/bib_ref] Our analysis demonstrated that the majority of COPD patients with the GOLD stage A, B and C disease severity are overtreated. Although the patients in the present study were newly diagnosed and treatment-naïve ones, the overtreatment rate was found as 71.8%. The most frequently prescribed medication was triple regimen (ICS+LABA +LAMA) with an overall frequency of 54.9%. The current studies show that very high over-treatment rates in COPD are a major global problem. [bib_ref] Guideline adherence in management of stable chronic obstructive pulmonary disease, Sharif [/bib_ref] In a study conducted in 2010 in Turkey, overtreatment was determined in 100% of the GOLD stage 1 and 91.1% of the GOLD stage 2 contrary to GOLD recommendations.In the ALPHABET study, over-treatment was also noted in >70% of GOLD 2013 A, B, and C patients.Compliance with COPD guidelines has been shown to be globally low in several studies, and adherence to the GOLD treatment strategy is not sufficient. [bib_ref] General practitioner's adherence to the COPD GOLD guidelines: baseline data of the..., Jochmann [/bib_ref] [bib_ref] Health care utilization history, GOLD guidelines, and respiratory medication prescriptions in patients..., Seaman [/bib_ref] [bib_ref] Variation in adherence with Global Initiative for Chronic Obstructive Lung Disease (GOLD)..., Fitch [/bib_ref] Other reasons for nonadherence in treatment may be the inability of physicians to perceive the severity of illness and the maintenance of ICS use habits.
The patients' adherence to the visits during 1-year of follow-up is poor in our population. We planned 3 voluntary follow-up visits for our patients during 1-year after the initial visit. Considering the 1-year follow-up visits, the attendance rates of our patient groups were moderate, 55.9%, 32.9% and 18.7%, respectively. Based upon the patients' statements, the adherence rate to the prescribed inhaled regimens was reportedly very high (81.9%). Adherence rates in COPD are known to vary from 22% to 78%. [bib_ref] Adherence to inhaled therapies, health outcomes and costs in patients with asthma..., Makela [/bib_ref] [bib_ref] Refill adherence for patients with asthma and COPD: comparison of a pharmacy..., Krigsman [/bib_ref] [bib_ref] Medication adherence issues in patients treated for COPD, Restrepo [/bib_ref] Good adherence is associated with reduced exacerbation rates in patients with COPD. [bib_ref] Impact of patients' satisfaction with their inhalers on treatment compliance and health..., Chrystyn [/bib_ref] A subgroup analysis of the TORCH (Towards a Revolution in COPD Health) study showed that good compliance to study medication in patients with COPD was associated with lower mortality rates compared with poor compliance (11.3% vs 26.4%). [bib_ref] Adherence to inhaled therapy, mortality and hospital admission in COPD, Vestbo [/bib_ref] Our study showed that adherence rates to the treatment decrease as the category becomes more severe from the GOLD stage A to D. This general trend is reasonable since less severe patients get more benefit from the prescribed therapy, and thus become comparably more adherent to the treatment.
In this study, 56% of COPD patients were active smokers. In a study to identify subgroups of COPD using cluster analysis in ECLIPSE (Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints), the rate of active smokers was between 25% and 42% (mean 36%). [bib_ref] Identification of five chronic obstructive pulmonary disease subgroups with different prognoses in..., Rennard [/bib_ref] In a previous study by Gunen et al, the smoking rate in COPD was found to be unexpectedly high in the Malatya region of Turkey. [bib_ref] Prevalence of COPD: first epidemiological study of a large region in Turkey, Günen [/bib_ref] In this study, 25.5% of the women and 57.2% of the men in the general population were current smokers. The prevalence of COPD was 18.1% in current smokers over 40 years of age, and 77.5% of the participants with COPD were current smokers.
This study has several limitations. The major limitation of the study is that the follow-up of patients was not regular due to the study design. Second, the determination of treatment adherence during a 1-year follow-up has some limitations. Selection bias could be a problem because surprisingly high adherence rates were obtained from the patients who attended the follow-up visits. Additionally, since the determination of treatment adherence was completely based upon the patients' statements, some misleading statements might have increased our adherence rates significantly. On the other hand, we believe that the general trends of treatment adherence revealed and the distributions according to the GOLD categories were less affected by this relatively subjective approach.
# Conclusion
To our knowledge, this is the first prospective real-life study on newly diagnosed and treatment-naïve COPD patients. Although these patients were diagnosed for the first time, the GOLD C and D categories and frequent exacerbator phenotype were found at a high rate. Our data have shown that although this subgroup of patients is younger than the usual COPD patients with fully established disease, their comorbidity rates are high and 35% of these patients have a high risk for exacerbations (the GOLD categories C and D). They were usually prescribed an overtreatment regimen. In conclusion, every tool should be used to increase the awareness of COPD and the effective early detection program for COPD in our community. The newly diagnosed, treatment-naïve COPD patients should be evaluated carefully, and best effort should be made to treat these patients in accordance with the recommendations of the major COPD guidelines.
# Disclosure
The study is financially supported by "Abdi Ibrahim Ilac San. ve Tic. AS, Turkey." All authors are members of the advisory board of the study and received payment from "Abdi Ibrahim Ilac San. ve Tic. AS, Turkey" for this. Professor Hakan Gunen reports non-financial support from Abdiİbrahim Pharmaceutical Company, during the conduct of the study; Abdiİbrahim Turkey Pharmaceutical Company paid the publication processing fee, outside the submitted work. Dr Tevfik Ozlu reports personal fees from Abdiibrahim, during the conduct of the study; Prof. Dr. Oznur Abadoglu reports grants from Abdi İbrahim, during the conduct of the study. Dr Bilun Gemicioglu reports Clinical research grants from Sanofi, AstraZeneca, Novartis; Consultancy or lecture grants from Abdi Ibrahim, Deva, Novartis, during the conduct of the study. Prof. Dr. Zeynep Misirligil report grants, personal fees from Abdi İbrahim İlaç San. ve Tic. AS, during the conduct of the study. The authors report no other conflicts of interest in this work.
## International journal of chronic obstructive pulmonary disease
## Dovepress
## Publish your work in this journal
The International Journal of COPD is an international, peer-reviewed journal of therapeutics and pharmacology focusing on concise rapid reporting of clinical studies and reviews in COPD. Special focus is given to the pathophysiological processes underlying the disease, intervention programs, patient focused education, and self management protocols. This journal is indexed on PubMed Central, MedLine and CAS. The manuscript management system is completely online and includes a very quick and fair peer-review system, which is all easy to use. Visit http://www.dovepress.com/testimonials.php to read real quotes from published authors.
Submit your manuscript here: https://www.dovepress.com/international-journal-of-chronic-obstructive-pulmonary-disease-journal
[fig] Figure 1: Frequency of comorbidities in newly diagnosed COPD patients. [/fig]
[table] Table 1: Demographic and General Characteristics of the Patients [/table]
[table] Table 2: Distribution of the Preferred Inhaled Drug Regimens According to the GOLD COPD Categories in Newly Diagnosed COPD Patients [/table]
[table] Table 3: Adherence to GOLD 2011 Treatment Recommendations of Physicians [/table]
|
Large-scale synthesis of graphene and other 2D materials towards industrialization
The effective application of graphene and other 2D materials is strongly dependent on the industrial-scale manufacturing of films and powders of appropriate morphology and quality. Here, we discuss three state-of-the-art mass production techniques, their limitations, and opportunities for future improvement.Two-dimensional (2D) van der Waals (vdW) layered materials including graphene, transition metal dichalcogenides (TMDs), hexagonal boron nitride (hBN), and MXenes have attracted much attention in recent years. This is due to their distinctive physical and chemical properties, such as their quantum Hall effects and quantum valley Hall effects, indirect-to-direct bandgap transition, and strong spin-orbit coupling 1 , which have not been accessible with conventional 3D bulk materials. In addition, vertical vdW heterostructures constructed by layer-by-layer stacking enable interesting applications for atomically thin quantum wells, p-n junctions, Coulomb drag transistors, and twistronic devices 1-3 . However, applications based on such structures are limited by the fact that most vdW materials are currently only available with a lateral size of up to a few tens of micrometers. Techniques for the large-scale synthesis of 2D materials will therefore be required for industrialization. Moreover, since specific applications of these materials are strongly dependent on characteristics such as their morphology and quality, mass production techniques should also be developed that can accommodate such requirements(Fig. 1). In general, most applications rely on either films or powders of vdW materials. Films require high crystal quality, and can be used in the context of electronics, spintronics, optoelectronics, twistronics, or solar cells, whereas powders exhibit large surface areas and are employed in the construction of batteries, sensors, and catalysts. At present, only large-area graphene films and graphite oxide powders are currently available in the commercial market 4 . In this Comment, we briefly examine research trends in synthesis techniques and their associated challenges for the industrialization of 2D layered materials.Three representative synthesis techniquesThere are currently three representative synthesis techniques available for the large-scale synthesis of 2D materials. The first is chemical vapor deposition (CVD); although a variety of thin-film deposition techniques have been investigated for growing large-area 2D films, https://doi.
including pulsed laser deposition (PLD) [bib_ref] Cooling rate and energy dependence of pulsed laser fabricated graphene on nickel..., Koh [/bib_ref] , atomic layer deposition (ALD) [bib_ref] Atomic layer deposition of a MoS 2 film, Tan [/bib_ref] , and molecular beam epitaxy (MBE) [bib_ref] Epitaxial graphene growth by carbon molecular beam epitaxy (CMBE), Park [/bib_ref] , CVD is most feasible for industrialization when one takes into account the uniformity and crystallinity of 2D films as well as requirements of high throughput, cost effectiveness, and scalability. The other two techniques being investigated for mass production are top-down liquid exfoliation of 2D materials and bottom-up wet chemical synthesis.
CVD for growing large-scale 2D thin films There are multiple examples of CVD synthesis of thin films at wafer scale [fig_ref] Figure 2: Historical landmarks in the development of three representative synthesis techniques [/fig_ref]. For example, wafer-scale polycrystalline monolayer and multilayer graphene films have been successfully synthesized by CVD on polycrystalline Cu and Ni foils since 2009 [bib_ref] Review of chemical vapor deposition of graphene and related applications, Zhang [/bib_ref] [bib_ref] Large-area synthesis of high-quality and uniform graphene films on copper foils, Li [/bib_ref] [bib_ref] Synthesis of large-area graphene layers on poly-nickel substrate by chemical vapor deposition:..., Chae [/bib_ref] [bib_ref] Poly(ethylene co-vinyl acetate)-assisted one-step transfer of ultra-large graphene, Han [/bib_ref] , and wafer-scale single-crystal monolayer graphene has been synthesized by using single-crystal substrates such as H-Ge (110) and Cu (111) [bib_ref] Seamless stitching of graphene domains on polished copper (111) foil, Nguyen [/bib_ref] [bib_ref] Towards wafer-scale monocrystalline graphene growth and characterization, Nguyen [/bib_ref]. Single-crystal multilayer graphene films have been also grown on Si-Cu alloys at wafer scale [bib_ref] Layer-controlled single-crystalline graphene film with stacking order via Cu-Si alloy formation, Nguyen [/bib_ref]. In 2012, centimeter-scale polycrystalline monolayers of hBN and TMDs were grown on polycrystalline Cu foils and SiO 2 /Si substrates, respectively [bib_ref] High-mobility three-atom-thick semiconducting films with wafer-scale homogeneity, Kang [/bib_ref] [bib_ref] Chemical vapor deposition growth and applications of two-dimensional materials and their heterostructures, Cai [/bib_ref]. And more recently, single-crystal hBN and TMD films were successfully synthesized on liquid Au, highindex single-crystal Cu surfaces, and atomic sawtooth Au surfaces [bib_ref] Wafer-scale single-crystal hexagonal boron nitride film via selfcollimated grain formation, Lee [/bib_ref] [bib_ref] Epitaxial single-crystal growth of transition metal dichalcogenide monolayers via the atomic sawtooth..., Choi [/bib_ref].
CVD produces relatively high-quality 2D films under atmospheric or low pressure, and the size of the film can easily be scaled up by increasing the chamber size. However, high temperature reactions (above 500°C) are required, which could be a drawback for industrialization. The growth of a vast range of 2D materials, including graphene, hBN, and TMDs, is still limited by the absence of appropriate precursors. Perhaps the most important technical challenge presented by this method is the poor control over the number of synthesized layers, because the absence of dangling bonds on the surface of 2D vdW materials makes epitaxial growth difficult.
## Liquid exfoliation
Liquid exfoliation is a powerful process for the mass production of pristine 2D bulk materials by dispersing them into individual sheets. Bulk materials have been synthesized by chemical vapor transport (CVT) [fig_ref] Figure 2: Historical landmarks in the development of three representative synthesis techniques [/fig_ref] since the late 1960s, but most 2D bulk materials are currently only available in small quantities.
Nanodispersion into monolayers is often required to manifest the unique 2D nature, but the strong vdW energy of micron-scale materials hinders facile exfoliation. Thus, two additional steps should be considered for liquid exfoliation processes: (i) weakening the layer-layer interaction by expanding the interlayer distance, and (ii) physical agitation for dispersion [bib_ref] Preparation and characterization of graphite nanosheets from ultrasonic powdering technique, Chen [/bib_ref] [bib_ref] Synthesis and exfoliation of isocyanate-treated graphene oxide nanoplatelets, Stankovich [/bib_ref]. In 1958, it was demonstrated that the interlayer distance can be increased from 3.4 to 7.0 Å by the oxidation of graphite, known as "graphite oxide", and such an expansion of the interlayer distance made it possible to disperse the individual graphite oxide layers by sonication. Graphite oxide layers can subsequently be reduced to graphene nanosheets through chemical treatment with reducing agents and thermal annealing treatment [bib_ref] Synthesis of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide, Stankovich [/bib_ref] [bib_ref] Efficient reduction of graphite oxide by sodium borohydride and its effect on..., Shin [/bib_ref] [bib_ref] Facile physical route to highly crystalline graphene, Jin [/bib_ref].
The lattice of graphene nanosheets is often severely damaged during oxidation and reduction processes. To prevent this, the interlayer distance can be increased by intercalating ions and molecules between layers. Electrochemistry enables effective intercalation of both cations and anions in an electrolyte solution by applying negative and positive bias, respectively [bib_ref] Recent new additives for electric vehicle lead-acid batteries for extending the cycle..., Kozawa [/bib_ref]. Alkali metals, organic solvents, and surfactants with similar surface energies to those of the 2D materials can also be directly intercalated in liquid or vapor phase [bib_ref] Tuning two-dimensional nanomaterials by intercalation: Materials, properties, and applications, Wan [/bib_ref] [bib_ref] A chemical route to carbon nanoscrolls, Viculis [/bib_ref]. After intercalation, agitation methods such as sonication, homogenization, and microwave treatment can be employed to exfoliate materials into individual 2D layers [bib_ref] Two-dimensional nanosheets produced by liquid exfoliation of layered materials, Coleman [/bib_ref]. Liquid exfoliation enables mass production of 2D nanosheets under atmospheric pressure at room temperature. However, this approach also leads to presumably unavoidable damage and non-uniform nanosheet thickness.
## Wet chemical synthesis
Hydrothermal and solvothermal syntheses are representative wet chemical synthesis methods, in which materials are respectively solubilized in aqueous solution and organic solvent under high vapor pressure at elevated temperatures (~300°C). A variety of nanomaterials have been synthesized in this fashion since the first report of hydrothermal synthesis of microscopic quartz crystals in 1845 [fig_ref] Figure 2: Historical landmarks in the development of three representative synthesis techniques [/fig_ref]. The wet chemical synthesis of pure 2D materials such as graphene and MoS 2 surged in the early 21st century [bib_ref] Solvothermal synthesis of nanocrystalline MoS 2 from MoO 3 and elemental sulfur, Zhan [/bib_ref] [bib_ref] Mass production of twodimensional materials beyond graphene and their applications, Yang [/bib_ref] , and more recently, doped 2D materials, nanocomposites, and their heterostructures have been synthesized in this fashion by adding various precursors and dopants in solvent to enhance the material properties for specific applications [bib_ref] Toward N-doped graphene via solvothermal synthesis, Deng [/bib_ref] hydrogen evolution reaction in graphene oxide was dramatically enhanced by introducing boron dopants [bib_ref] Metal-free B-doped graphene with efficient electrocatalytic activity for hydrogen evolution reaction, Sathe [/bib_ref]. The strengths of wet chemical synthesis include the controllability of surface morphology, crystallite size, and dopants in 2D materials for catalyst, energy storage, and chemical/biological sensor applications. Reaction temperatures, precursors, and additives have been optimized for various types of 2D materials and their composites, enabling essentially unlimited mass production. The direct synthesis of 2D materials on a desired substrate is also possible, although such synthesis takes a relatively longer time-up to a few days. Growth temperature is often limited to below 300°C due to the limited durability of equipment under harsh conditions including high pressure and exposure to corrosive chemicals. It is worth noting that bottom-up synthesis tends to yield low-quality 2D materials with defects, but it is still possible to employ these for catalytic applications.
## Perspectives and challenges toward industrialization
The abovementioned techniques enable mass production of 2D materials, but considerable further advances will be required for some specific applications. Single-crystal graphene films have been successfully synthesized at wafer scale with layer control, but the synthesis of other 2D materials such as hBN and TMDs are limited exclusively to single-crystal monolayer films. Thickness control of such materials is essential for tunneling barrier and high-performance electronics. The combination of tunable bandgap semiconductors, metals, and insulators in 2D systems can generate versatile heterostructures with remarkable physical properties. Several planar and vertical heterostructures have been generated to date, but these remain limited to micron scale [bib_ref] Waals layered materials: Opportunities and challenges, Duong [/bib_ref] [bib_ref] Heterostructures based on twodimensional layered materials and their potential applications, Li [/bib_ref]. More generally, the growth of various heterostructures at wafer scale is still challenging . Atomic sawtooth surfaces could be ideal as a growth platform for single-crystal 2D materials including graphene, hBN, TMDs, and their heterostructures, but surface facet control remains elusive. The formation of wrinkles in 2D films after high temperature growth is another important issue, originating from the thermal expansion coefficient mismatch between 2D materials and growth substrate. Recently, the growth of fold-free single-crystal graphene films at 750°C has been reported 36 , but further investigation will be required to see if this method is applicable for other 2D materials, and lowertemperature growth methods should be established. High temperature processes (above 400°C) are not compatible with current Si technology, and 2D thin films grown by CVD at high temperature must therefore be transferred onto a target substrate. A conventional transfer process can give rise to serious problems such as folding and cracking of 2D films, ultimately degrading film homogeneity and device performance. Furthermore, the polymer contaminants that are commonly introduced as a supporting layer for the transfer process can give rise to unintentional doping and high contact resistance in heterostructure interfaces and devices. Therefore, methods for the direct growth of large-area 2D films by CVD or advanced roll-toroll transfer technique would be highly desirable. For industrialization, the manufacturing process including scalable techniques (roll-to-roll, batch-to-batch, etc.), production capacity/ cost, reproducibility, and large-area uniformity are further considered [bib_ref] Synthesis challenges for graphene industry, Lin [/bib_ref].
Wet chemical processes including liquid exfoliation and wet chemical synthesis also face several challenges for the mass production of 2D materials. Liquid exfoliation employs pristine 2D bulk materials synthesized by CVT or flux methods for the mass production of 2D nanosheets. Those synthetic methods typically take at least one week, lowering the throughput of production, and companies need the capacity to provide these bulk materials at a larger scale. Additionally, the production yield of liquid exfoliation generally remains poor, and although some materials show relatively high yield, most 2D materials like hBN and telluride are not effectively exfoliated with current techniques. In addition, it is difficult to obtain 2D nanosheets of uniform size and thickness with this method . In order to remedy this, improved techniques for sorting the synthesized nanosheets in terms of size and thickness (e.g., density gradient ultracentrifugation) are needed.
Bottom-up chemical synthesis typically produces 2D materials with low crystal quality. The defect sites (i.e., edges) often serve as active sites for 2D catalyst, but also give rise to low durability and instability issues. Moreover, 2D materials generated by chemical synthesis are not uniformly distributed in terms of size and thickness, requiring special care during synthesis. In addition, the byproducts frequently generated during chemical synthesis can inhibit catalytic activity. To resolve these material quality and byproduct issues, post-treatments such as thermal annealing and purification have been suggested, but a simple process without post-treatment would greatly improve productivity. Another important issue is the environmental pollution caused by the large amount of hazardous chemical wastes used in synthesis , and the use of supercritical fluid regions could be considered as a shortcut to minimize chemical use [bib_ref] Supercritical hydrothermal synthesis of MoS 2 nanosheets with controllable layer number and..., Takahashi [/bib_ref].
In addition, rapid and reliable non-destructive characterization tools are highly required to evaluate the wafer-scale 2D materials in terms of sample quality and uniformity. The current state-ofthe-art terahertz image, phase-shift interferometry, and wide-field Raman imaging could be adopted to analyze the electrical and optical properties of 2D films with short acquisition time of a few seconds per mm 2 and high spatial resolution of an order of micrometers [bib_ref] Facile morphological qualification of transferred graphene by phase-shifting interferometry, Lee [/bib_ref] [bib_ref] High-throughput graphene imaging on arbitrary substrates with widefield Raman spectroscopy, Havener [/bib_ref] [bib_ref] Systematic THz study of the substrate effect in limiting the mobility of..., Scarfe [/bib_ref]. It still requires a prolonged period to thoroughly inspect 12-inch wafer-scale sample, and therefore, the development of advanced characterization tools is further desired.
From a materials point of view, there is plenty of room for unexplored novel 2D materials and their vdW heterostructures. Since it is almost impossible to explore all such materials experimentally, artificial intelligence-based material design could prove useful for the industrialization and large-scale manufacture of such newly-developed 2D materials [bib_ref] Machine learning-driven new material discovery, Cai [/bib_ref] Current challenges in the mass production of 2D materials. a Single-crystal homo/heteroepitaxial growth, wrinkle formation by thermal expansion coefficient mismatch between 2D materials and growth substrates, and cracking/contamination during the transfer process are all issues presented by the CVD technique. b Inhomogeneous size and thickness of 2D nanosheets and poor production yield are problems associated with liquid exfoliation. c Low durability and instability of 2D materials by defects and environmental pollution remain challenges for wet chemical synthesis.
[fig] Figure 2: Historical landmarks in the development of three representative synthesis techniques. Lefthand panels show timelines of milestones for a chemical vapor deposition (CVD), b liquid exfoliation, and c wet chemical synthesis methods. The abbreviations correspond to: metal-organic CVD (MOCVD), graphene (Gr), graphite oxide (GO), reduced GO (rGO), and molybdenum disulfide (MoS 2 ). Righthand panels show the corresponding strengths and weaknesses of these methods in terms of mass production (MP), thickness controllability (THK), temperature variation (TEMP), uniformity (UNI), material diversity (MAT), crystal quality (QLTY), morphology (MORPH [/fig]
[table] .: Received: 3 November 2021; Accepted: 24 February 2022;Single-crystal homo/heteroepitaxy [/table]
|
LC-MSsim – a simulation software for liquid chromatography mass spectrometry data
Several of the tools that we tested depend a lot on suitable parameters. Nevertheless, most parameters are not well documented. On the other hand, some tools work "outof-the-box" and do not offer any parameters that could be adopted.
We chose the parameter settings for each tool as described below. Three tools (msInspect, Decon2LS and Mzmine) still produced large numbers of false positives, as stated below. We therefore removed for each of these three tools all peptide features with an intensity below the first quantile. The first quantile is the value such that 25% of all feature intensities are below it and the rest above. msInspect: This tools does not have any parameters besides m/z and rt ranges searched. Each isotope pattern is scored using the Kullback-Leibler (KL) distance between an averagine model and the true peak intensities. A KL distance of zero indicates a perfect match between averagine model and signal and thus a high-quality peptide feature. To find a suitable cutoff for this score, we executed msInspect on a complex LC-MS run, a digest of Halobacterium NRC-1 proteins recorded on a API QSTAR Pulsar I instrument (downloaded from the PeptideAtlas database, ref:
Pae000245, data set # 25, http://www.peptideatlas.org/). We manually annotated this data set and choose 20 intense and well-resolved peptide signals to determine a suitable cutoff. We choose a cutoff for the KL distance of 0.8 such that 80% of the annotated features were detected. Applying this filtering threshold to the whole data set, the number of detected features was reduced from 3366 to 2608 features.
Decon2LS: This tool offers plenty of parameters, some are documented and some are not. According to the authors (personal communication), the fit intensity and the fit score threshold should have significant influence on the result. However, we made the experience that the fit intensity threshold has not much influence on the result. Consequently, we optimized the fit score threshold (a distance measure between 0 and 1, where 0 is best) in the same way as the KL distance described above.
We choose a cutoff for the fit score of 0.2 which resulted in 85% annotated features recovered. The overall number of features was reduced from 19603 to 14592.
MZmine: This software offers different peak detection strategies. We used the "Recursive threshold peak detecter" algorithm. Furthermore, the software has several parameters that influence the feature detection process. This process consists of two parts: a peak detection and a de-isotoping step. Peak detection is influenced by parameters such as bin sizes, min intensity etc. The de-isotoping step groups peaks into isotopic pattern, estimates a charge and removes incomplete isotope pattern and single peaks. We contacted the developers of Mzmine and asked them which parameters would have significant influence. MZmine does not compute a score and s/n threshold that could be used as single filter criterion. According to the recommendations of the MZmine developers, we adopted the "Chromatographic threshold level" and the "m/z bin size" as well as the noise threshold and minimum peak height which are both given in absolute intensity units.
For the high-resolution data (FWHM 0.02), we choose a set of parameters that recovered 100% of the annotated features from the QSTAR mass spectrometer. With decreasing mass resolution, we relaxed the bin width and the chromatographic threshold level but without satisfying results. In each case, we estimated the noise level as the median intensity on a small m/z interval in empty region (i.e. without peptide signals) of the LC-MS map. We set the minimum peak intensity to the same value.
## Specarray:
No parameter changes possible, no manual tuning.
OpenMS + Superhirn : several parameter offered, moderate optimization to achieve trade-off between false positives and false negatives.
## Avaibility of software and simulation parameters:
All peptide feature detection algorithms were tested in the version that was available online in January 2008.
In the case of OpenMS, we used a slightly modified version of the 1.0 Release. It can be downloaded from sourceforget using: svn co https://open-ms.svn.sourceforge.net/svnroot/open-ms/FF10
The installations instructions are the same as for the most recent OpenMS release version (1.2) and can be obtained from www.openms.de (=> follow link "installation").
The parameter settings and peptide feature lists for each feature detection algorithm, as far as they could be stored, are available in the supplement.
The parameter sets are available for download from http://lcms-sim.sourceforge.net/ The simulation parameter are contained in the svn repository of LC-MSsim: svn co https://lcms-sim.svn.sourceforge.net/svnroot/lcms-sim lcms-sim (check subdirectory "ini_files").
The simulated data sets are several GBs large and can be downloaded from the PRIDE database (http://www.ebi.ac.uk/pride/, Accession numbers 8161-8168 incl.).
Ole Schulz-Trieglaff, [email protected] |
Exercise Addiction and Its Relationship with Health Outcomes in Indoor Cycling Practitioners in Fitness Centers
We studied the prevalence and possible association between exercise addiction and health in indoor cycling practitioners. In 1014 (492 women) adult indoor cyclists and 926 (597 women) controls with low levels of physical activity according to the short form of the International Physical Activity Questionnaire, we examined the risk of exercise addiction according to the Exercise Addiction Inventory and several health outcomes through a web-based experiment. The prevalence of a high risk of exercise addiction in cyclists was 13.3%, and it was higher in men than in women (16.5% vs. 10.0%, p = 0.002). Women cyclists with a high risk of exercise addiction had higher levels of physical activity (p < 0.001; effect size = −0.62, 95% CI: (−0.91, −0.32)) and anxiety symptom severity (p = 0.001; Effect Size (ES) = −0.59 (−0.89, −0.30)) than those with a low risk. For both sexes, cyclists with a low risk of exercise addiction had better social function, emotional role, and anxiety symptom severity compared with the controls (all p < 0.002; ES ranged from 0.25 to 0.47). Higher anxiety symptom severity and cardiorespiratory fitness were the main determinants of exercise addiction in cyclists (both p < 0.001). Our data suggest the importance of considering exercise addiction in indoor cyclists.
# Introduction
Exercise addiction, defined as "an abnormal reliance on exercise behavior to cope with chronic stress or the hassles and challenges of the everyday life and featuring the core components of addiction found in more traditional addictions", is a popular but relatively new research topic. Many researchers claim that exercise addiction should be accepted as a mental disorder in diagnostic manuals; however, the accumulated literature is still scarce and limited by a lack of conceptual and methodological consistency.
Two key factors to consider in improving the understanding of exercise addiction are its prevalence and health consequences. The sports discipline and the screening scale employed affect the results regarding prevalence. Along these lines, a recent systematic reviewsuggested that endurance sports (e.g., cycling) seem to be associated with a higher risk of exercise addiction than power (e.g., weightlifting), mixed (e.g., soccer), and health and fitness disciplines among fitness center users. Additionally, one studyrecommended the use of the Exercise Addiction Inventory because it is a brief, internationally validated, widely used, and highly sensitive screening instrument to detect individuals at risk of exercise addiction. Researchers should use adequate tools to assess the prevalence of exercise addiction in those who engage in specific sports disciplines.
Most studies on exercise addiction have been conducted with fitness center users (e.g.,. While helpful, these studies include heterogeneous groups of exercisers that practice different disciplines and provide heterogeneous results. Therefore, a key factor for improving our understanding of the impact of exercise addiction is to develop scientific studies in homogeneous samples that are specific for a single exercise discipline. To know which exercise disciplines may be potentially associated with an increased risk of exercise addiction could help to develop better exercise practices to prevent and/or treat exercise addiction. To the best of our knowledge, the study by Lichtenstein and Jensenwas the only prior study that used a sample of gym users who practiced a specific fitness discipline such as CrossFit. Other of the most popular specific disciplines in fitness centers with thousands affiliated that may benefit from these kinds of research advances is indoor cycling. This study aimed to explore the prevalence of exercise addiction and its possible association with health outcomes in indoor cycling practitioners. As with other addictions, exercise addiction may co-occur with other behavioral dysfunctions related to diet, smoking, or alcohol consumption and may have a negative impact on physical and mental health outcomes, such as quality of life and sleep, anxiety, depression, fitness, and body mass index. Based on a recent meta-analysis, we hypothesized that a high risk of exercise addiction could be associated with worse mental health in indoor cycling practitioners.
# Materials and methods
## Participants
This cross-sectional study was part of a research project focused on exploring health and its determinants among amateur endurance outdoor cyclists, indoor cycling practitioners, and controlsthrough a web-based experiment on a freely available webpage. This study included respondents who were 18-65 years old and who engaged in a minimum of 2 h/week and 6 continuous months of indoor cycling practice in fitness centers (indoor cycling practitioners) or who were classified as having low levels of physical activity according to the Spanish short-form version of the International Physical Activity Questionnaire(controls).
## Procedures
An invitation to participate in the study was sent via e-mail to the representatives of the 4627 fitness centers officially registered in Spain. The study was also made known through media coverage, including TV and social media such as Twitter, Facebook, and LinkedIn. The invitation included a brief introduction to the study, an explanation of the anonymous and voluntary nature of participation in the study, and the link to the experiment. The experiment included questions on health behaviors and health status and took an average of 40 min to complete. At the end of the experiment, individuals needed to provide informed consent for the scientific use of the data and provide an email address to receive the results. Data were collected in the last week of May from 2016 to 2018 with the same procedure throughout the study period. To avoid duplicate samples, cases that contained the same email address were eliminated. In 2017 and 2018 we included a warning message before starting the experiment indicating that those who had already completed it should not do it again. Data analysis was conducted in February 2019. The protocol was approved by the Committee on Biomedical Ethics of the Aragón Government (PI17/0252). There was no compensation for participation.
## Measures
## Exercise addiction
The risk of exercise addiction was measured with the Spanish versionof the Exercise Addiction Inventory, which has satisfactory psychometric properties in the Spanish population (Cronbach's α = 0.70 and intraclass correlation coefficient = 0.92). The inventory, which can be found in the aforementioned free full-text articles, is composed of six items with 5-point Likert-scale responses (1 = strongly disagree, 5 = strongly agree), with scores ≥ 24 indicating a high risk of addiction.
## Health behaviors
The indoor cycling training variables were volume (hours/week) and frequency (days/week) in the last month and experience (years). Physical activity was measured with the Spanish short-form version of the International Physical Activity Questionnaire(freely available on the IPAQ website) by totaling the walking, moderate-and vigorous-intensity activities and expressing the result in metabolic equivalent of task-min/week. The Spanish short-form version of the International Physical Activity Questionnaire has been validated in the Spanish population showing a correlation coefficient of 0.51 (95% CI: 0.30-0.70) for total amount of physical activity against accelerometry measures. Adherence to the Mediterranean diet was assessed with the Spanish version of the 14-Item Mediterranean Diet Tool, with higher scores indicating higher adherence. This Spanish version of the 14-Item Mediterranean Diet Tool has been validated in the Spanish population showing a correlation coefficient of 0.52 and an intraclass correlation coefficient of 0.51 (both p < 0.001) for total score against the 137-item Food Frequency Questionnaire. Smoking dependence was evaluated with the Spanish version (α value = 0.66)of the revised version of the Fagerström Test for Nicotine Dependence, with lower scores indicating lower levels of dependence. Alcohol use was calculated by transforming the volumes of beer, wine, and spirits consumed in the last week into standard alcohol units, with lower units indicating lower levels of alcohol consumption.
## Health status
Quality of life was measured with the Spanish versionof the 12-Item Health Survey 2.0, which examines eight domains constituting the physical and mental component summary scores, with higher scores indicating better quality of life. The 12-Item Health Survey showed a high criterion validity because it explained more than 90% of the variability in the physical and mental component summary scores on the 36-Item Health Survey. Quality of sleep was assessed with the Spanish versionof the Pittsburgh Sleep Quality Index, which examines seven sleep components that yield a global score, with lower scores indicating better quality of sleep. The Spanish version of the Pittsburgh Sleep Quality Index has satisfactory psychometric properties in the Spanish population (Cronbach's α values ranged from 0.67 to 0.69 and test-retest correlation coefficient for global score = 0.51; p < 0.001). Anxiety and depression symptom severity were separately evaluated with the Spanish versionof the Hospital Anxiety and Depression Scale, with lower scores indicating lesser symptom severity. The Spanish version of the Hospital Anxiety and Depression Scale showed satisfactory psychometric properties in the Spanish population including a good internal consistency (Cronbach's α = 0.85 for the anxiety subscale and 0.84 for the depression subscale). Body mass index was calculated using weight and height. The Spanish version of the International Fitness Scalewas used to measure the physical fitness level (overall and specific components: cardiorespiratory fitness, muscular strength, speed-agility, and flexibility). The individuals were asked how they perceived their own level compared with their friends' physical fitness using a 5-point Likert-scale (very poor = 1, poor = 2, average = 3, good = 4, and very good = 5). The Spanish version of the International Fitness Scale showed satisfactory psychometric properties in the Spanish adults including test-retest reliability (averaged weighted Kappa of 0.59) and validity against objective measures.
## Sociodemographic data
Sex, age, educational status, occupational status, marital status, and number of children were determined.
# Statistical analysis
The characteristics of indoor cycling practitioners with low and high levels of the risk of exercise addiction and controls stratified by sex were compared with chi-square test, unpaired t-test, and one-way ANOVA followed by a Bonferroni post hoc test, according to the nature of the variable. Possible interactions with sex were tested using two-way ANOVA. To minimize the risk of type I statistical error, analyses were corrected for multiple comparisons with the Bonferroni method, in which the threshold p-Value is obtained by dividing 0.05 by the number of comparisons; in this case, 0.050 × 32 −1 = 0.002. Pearson correlation coefficients were calculated between the risk of exercise addiction and the characteristics of all indoor cycling practitioners. Multiple stepwise regression model analysis was applied to examine the possible determinants of the risk of exercise addiction (dependent variable) and significant correlates from the univariate tests (independent variables). Residuals were tested for normality, linearity, and homoscedasticity. The variance inflation factor never exceeded three, indicating that multicollinearity was not a concern. Data were analyzed using SPSS Statistics for Windows, Version 22.0 (IBM Corp, Armonk, NY, USA), with statistical significance set at α = 0.050. Additionally, the standardized mean difference was quantified with the unbiased version of Cohen's d (d unb ), also known as Hedges' g, and associated 95% confidence intervals using an online effect size calculator (http://www.cem.org/effect-size-calculator). The d unb was calculated through formula 4 provided by Lakens.
# Results
Of the 8257 respondents, 1014 individuals met the inclusion criteria for indoor cycling practitioners, and 926 were controls. The prevalence of a high risk of exercise addiction in indoor cycling practitioners was 13.3% (135/1014), and it was higher in men (16.5%; 86/522) than in women (10.0%; 49/492) (p = 0.002). The characteristics of men and women cycling practitioners and controls are presented in Tables 1 and 2, respectively. Values are the mean ± standard deviation or n (%); CI, confidence intervals; d unb , unbiased version of Cohen's d; MET, metabolic equivalent of task; REA, risk of exercise addiction; SAU, standard alcohol units'; chi-square test was used to examine differences in the proportion of comparing educational, occupational, and marital status and number of children; unpaired t-test was used to examine differences of training volume and frequency and experience in indoor cycling; one-way ANOVA with Bonferroni post hoc test was used to examine differences of resting variables; significant p-Values are in bold; the threshold p-Value was set at 0.002 (=0.050/32). Values are the mean ± standard deviation or n (%); CI, confidence intervals; d unb , unbiased version of Cohen's d; MET, metabolic equivalent of task; REA, risk of exercise addiction; SAU, standard alcohol units; chi-square test was used to examine differences in the proportion of comparing educational, occupational, and marital status and number of children. Unpaired t-test was used to examine differences of training volume and frequency and experience in indoor cycling. One-way ANOVA with Bonferroni post hoc test was used to examine differences of resting variables; significant p-Values are in bold; the threshold p-Value was set at 0.002 (=0.050/32).
Women cycling practitioners with a high risk of exercise addiction had significantly higher levels of weekly training hours (p = 0.001; d unb = −0.53 95% CI: (−0.82, −0.23)) and days (p = 0.001; d unb = −0.56 (−0.85, −0.26)), physical activity (p < 0.001; d unb = −0.62 (−0.91, −0.32)), and anxiety symptom severity (p = 0.001; d unb = −0.59 (−0.89, −0.30)) than those with a low risk of exercise addiction. Independent of sex and risk of exercise addiction, indoor cycling practitioners reported better values for most of the health outcomes than the controls; however, only indoor cycling practitioners with a low risk of exercise addiction showed significant results for certain outcomes. Specifically, in men, physical role (p < 0.001; d unb = 0.36 (0.20, 0.52)), social function (p < 0.001; d unb = 0.37 (0.21, 0.53)), emotional role (p < 0.001; d unb = 0.30 (0.14, 0.46)), and anxiety symptom severity 16)). The two-way ANOVA showed an interaction of sex with physical activity (p < 0.001) and muscular strength (p < 0.002). shows the descriptive data (mean and standard deviation or percentage) and the simple correlation models for the associations of the risk of exercise addiction and the characteristics of all indoor cycling practitioners.presents the significant final model that emerged for indoor cycling practitioners. Sex, age, physical activity, anxiety, cardiorespiratory fitness, and muscular strength were identified as determinants, explaining 13.9% of the variance in the risk of exercise addiction. . Descriptive data (mean and standard deviation or percentage) and associations with risk of exercise addiction in all indoor cycling practitioners (n = 1014). ; significant when p-Value < 0.050.
## Variables
# Discussion
Two major findings of the present study can be highlighted. First, 13.3% of indoor cycling practitioners in fitness centers reported a high risk of exercise addiction. Second, male sex, lower age, higher level of physical activity, muscular strength, and, especially, higher level of anxiety symptom severity and cardiorespiratory fitness were associated with a higher risk of exercise addiction.
The prevalence of exercise addiction for the overall sample of indoor cycling practitioners is similar to the mean reported in systematic reviews of studies that used the Exercise Addiction Inventory in habitual exercisers (13.2%)and in endurance athletes (14.2%), including one study in amateur endurance outdoor cyclists (17%), but it was slightly higher than that reported in studies in gym users (8.2%), which ranged from 3.6%to 42.5%. Although the aforementioned studies in gym users might be more useful for contextualizing our results, all included heterogeneous samples that were not specific to a particular discipline. The study of Lichtenstein and Jensenshowed that the prevalence of exercise addiction in people engaging in a specific fitness discipline such as CrossFit according to the Exercise Addiction Inventory was 5%, suggesting that the problem is less prevalent than among those who engage in indoor cycling. More studies in homogeneous samples that are specific for a single fitness discipline are needed to understand which disciplines are potentially associated with an increased risk of exercise addiction.
The higher prevalence of a high risk of exercise addiction in men than in women is consistent with the findings of two studies in gym usersbut seems to contradict the study by Lichtenstein et al.that showed no sex differences in gym users. According to studies of regular exercisers in sports clubs and leisure centersand long-term fitness club members, we suggest that men could be more motivated than women by mastery and competition, which seem to be associated with exercise addiction. Nevertheless, research on whether motivation to exercise actually explains sex differences in the prevalence rate of exercise addiction in indoor cycling practitioners is needed.
Our finding that women indoor cycling practitioners with a high risk of exercise addiction had more severe anxiety symptoms than those with a low risk is consistent with previous studies in college students who habitually engaged in exercise, outdoor cyclists, and gym users. However, the finding of higher levels of fitness in the group with a high risk of exercise addiction than in the group with a low risk of exercise addiction is inconsistent with the finding of the abovementioned study in outdoor cyclists, which was the only study prior to ours that explored the possible association between exercise addiction and fitness. Again, possible differences in attitudes towards exercise among participants in the current study may explain these discrepancies. More research that compares the possible association of exercise addiction and fitness between "more" and "less" competitive indoor cycling practitioners is needed. Of interest, e-racing, which is a recent development in indoor cycling that involves social media, could be a possible future enabler of exercise addiction due to the competitive nature and use of interactive rewards, unlike traditional methods of indoor cycling.
The positive association between exercise addiction and age in indoor cycling practitioners is consistent with the findings of studies in habitual exercisersand people who practice CrossFit. Finally, our finding that cardiorespiratory fitness was, along with anxiety, the main determinant of exercise addiction in indoor cycling practitioners suggests the importance of considering fitness as a means of identifying indoor cycling practitioners who are potentially addicted to exercise. This result could be partially due to the higher volume and frequency of training and level of physical activity observed in the group of women with a high risk of exercise addiction compared with those with a low risk of exercise addiction. However, the use of a self-report-based cardiorespiratory fitness measurement limits the accuracy of our findings and suggests caution with this claim. Additionally, is important to consider possible inconsistent interpretations of the Exercise Addiction Inventory related to the nature of the studied sample as explained by Szabo et al.. It may be possible that some of our sample of indoor cyclists may have confounded exercise addiction (associated with adverse effects on health)with exercise harmonious passion (associated with beneficial effects on health).
## Interpretation and practical application of current research findings
The current study could help raise awareness among health professionals of the importance of considering exercise addiction as a health risk factor in indoor cycling practitioners and may encourage the scientific community to research the prevalence and effects of exercise addiction in homogeneous samples that are specific for a single exercise discipline. This study identified determinants of exercise addiction that should be considered. However, our model explained only approximately 14% of the variance in the risk of exercise addiction, suggesting that other outcomes not included in this study, such as attitudes towards exercise, passion, narcissism, self-esteem, eating disorders, emotional distress, body appearance anxiety, and use of fitness supplements, could explain some of the remaining variance. Better indoor cycling practices may be needed to prevent and/or treat exercise addiction. Multidisciplinary teams including qualified exercise professionals should play an important role in addressing exercise addiction.
# Limitations and strengths
The two major limitations were that the study was cross-sectional in design and based on self-reported data, which preclude causal inferencesand involve bias, respectively. Other aspects that may also limit generalizationwere the following: (1) we do not know whether the representatives of fitness centers sent the information to all users, the number of people who declined to participate, the reasons for refusal to participate, or the intensity of the indoor cycling training sessions; (2) indoor cycling practitioners may undertake other physical activities; (3) among the small number of participants with a high risk of exercise addiction, few were women.
Despite these limitations, this study has strengths. The Internet design of the experiment, media coverage, and anonymous participation prevented problems of missing data, increased access to more potential participants, and may have encouraged individuals to respond to very personal questions. Furthermore, well-established, validated, and reliable measures that apply norm-based scoring methodology were used (for details, see Section 2).
## Highlights
## -
A total of 13.3% of indoor cycling practitioners reported a high risk of exercise addiction.
## -
Exercise addiction was more prevalent in men and associated with worse health outcomes. - Cardiorespiratory fitness and anxiety were the main determinants of exercise addiction.
# Conclusions
In total, 13.3% of indoor cycling practitioners exhibited a high risk of exercise addiction; the proportion was higher in men and was associated with more severe anxiety symptoms in women. Independent of sex and the risk of exercise addiction, indoor cycling practitioners reported better values for most of the health outcomes than the controls. However, only indoor cycling practitioners with a low risk of exercise addiction showed significant results for certain outcomes, such as social function, emotional role, and anxiety symptom severity. In addition, anxiety and cardiorespiratory fitness were identified as major determinants of exercise addiction. Finally, the findings on anxiety seems to confirm our principal hypothesis that a high risk of exercise addiction could be associated with worse mental health in indoor cycling practitioners. numbers PUI/2018-337 (JBA) and PUI/2018-336 (MAOC)). JRM was supported by Universidad de Lleida, Cátedra ASISA (grant number X18010). DMI was supported by the Biomedical Research Networking Center on Frailty and Healthy Aging (CIBERFES) and FEDER funds from the European Union (reference number CB16/10/00477). The funders had no role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication. There was no external financial support. |
A Research Agenda for Malaria Eradication: Cross-Cutting Issues for Eradication
Discipline-specific Malaria Eradication ResearchAgenda (malERA) Consultative Groups have recognized several cross-cutting issues that must be addressed to prevent repetition of some of the mistakes of past malaria elimination campaigns in future programs. Integrated research is required to develop a decision-making framework for the switch from malaria control to elimination. Similarly, a strong economic case is needed for the very long-term financial support that is essential for elimination. Another cross-cutting priority is the development of improved measures of intensity of transmission, especially at low and nonrandom levels. Because sustained malaria elimination is dependent on a functioning health system, a further key cross-cutting research question is to determine how inputs for malaria can strengthen health systems, information systems, and overall health outcomes. Implementation of elimination programs must also be accompanied by capacity building and training to allow the assessment of the impact of new combinations of interventions, new roles for different individuals, and the operational research that is needed to facilitate program expansion. Finally, because community engagement, knowledge management, communication, political, and multisectoral support are critical but poorly understood success factors for malaria elimination, integrated research into these issues is vital.Research for Readiness to Attempt Regional or National EliminationThe GMAP has identified the need to continue and scale up control of malaria in highly endemic areas for maximal reduction of Review articles synthesize in narrative form the best available evidence on a topic.
# Introduction
During their deliberations, scientists in the various Consultative Groups contributing to the Malaria Eradication Research Agenda (malERA) concentrated on research questions relevant to their thematic areas. But, in addition, they also briefly noted many issues of relevance beyond their own domains. Some of these issues are likely to be critically important in malaria elimination/eradication programs. Consequently, they received special attention from the malERA Consultative Group on Integration Strategies. In this paper, we focus on the research and development needs of these important cross-cutting issues, especially in the context of historical reports of reasons for the failure of past campaigns. Consideration of these cross-cutting issues, we argue, is essential for regional elimination and, ultimately, global eradication of malaria, but is also relevant for scaled-up and improved control of disease.
## The historical context
The Consultative Group identified many cross-cutting topics of special significance by examining reports of the failures and successes of earlier approaches to regional elimination of malaria. History reveals that political, social and human factors are likely to be just as important as, if not more important than, biological and technological factors, and that a multidisciplinary approach to elimination/eradication is essential. Accordingly, special attention was given during the malERA consultations to finding synergies and strategies to prevent the ''silo effects'' that can occur when specialist groups work in isolation. It is important to identify critical partnerships between malaria elimination/eradication programs and programs in health or education, such as integrated management of childhood illness. Similarly, it is important to recognise the need to address social determinants of health for successful malaria eradication campaigns. Finally, ongoing critical analysis of the success or failure of current elimination efforts constitutes a research agenda in its own right, as exemplified in numerous campaigns against other diseases [bib_ref] The role of research in viral disease eradication and elimination programs: Lessons..., Breman [/bib_ref] [bib_ref] Some lessons for the future from the Global Malaria Eradication Programme (1955-1969), Nájera [/bib_ref].
## The global malaria action plan and research for eradication
The Global Malaria Action Plan (GMAP)is focused predominantly on control, but nevertheless includes eradication as an ultimate goal. The malERA process, with its paradigm shift from control to elimination, has produced significant additions to GMAP by defining a research agenda that will assist in interruption of transmission. The malERA process emphasises the importance of clearly defining the essential research and development needed to achieve specific goals. That is, it focuses on the minimal essentials-what we ''need to know''-rather than what would be maximally possible to know or even ''nice to know.'' morbidity and mortality, and recognises this as a priority for the foreseeable future. As the malaria map shrinks and malaria incidence falls, some countries may consider attacking remaining foci with an elimination agenda. Many pre-elimination considerations are related directly to the competence and readiness of the health system, and are discussed in the malERA paper on health systems and operational research. Decision-makers must also balance the consequences of diverting resources from urgent clinical needs to a problem that by definition is causing little morbidity.
Importantly, decision makers at national and regional levels may need to be reminded that successful elimination for a few years will inevitably lead to loss of the naturally acquired immunity that is a good defence against malaria. Attempts to eliminate malaria that are not sustained can therefore provide the grounds for serious epidemics in people of all ages, with rapid loss of the gains accrued during an elimination program if the program fails or is stopped prematurely [bib_ref] Some lessons for the future from the Global Malaria Eradication Programme (1955-1969), Nájera [/bib_ref] [bib_ref] DDT house spraying and reemerging malaria, Roberts [/bib_ref].
The Malaria Elimination Group has recently highlighted the immediate needs of governments that are currently facing important decisions about malaria elimination/eradication [bib_ref] Shrinking the malaria map: a guide on malaria elimination for policy makers...., Feachem [/bib_ref]. Political commitment is essential; local research agendas for drug and insecticide resistance must be completed, health system readiness assessed, and cost-effectiveness analyses undertaken before deciding to make the long-term investment in elimination.
From its discussions, the malERA Consultative Group on Integration concluded that the cross-cutting research and development agenda in the context of the paradigm shift from control to eradication must take into account the research developments of the last few decades. Since the end of the Global Malaria Eradication Program (GMEP), innovations such as rapid diagnostic tests, insecticide-treated bednets, and improved information systems and communication systems have been developed, and a partially effective vaccine should be available in the foreseeable future. Thus, an algorithm needs to be defined and a tool developed for deciding the readiness of the system for elimination, or even for introduction of one of these innovations.
## The case for long-term investment for eradication
Cross-cutting research is needed to make the case for long-term investment in eradication for the global public good and to ensure that financial support is available for the ''last mile'' before elimination [bib_ref] Malaria eradication: The economic, financial and institutional challenge, Mills [/bib_ref]. This case should align with, and complement, important and related development themes such as global security, migration, food security, and climate change. If research findings suggest that the case is strong, malaria eradication could be included in global policies for health that follow on from the Millennium Development Goals beyond 2015. Importantly, a development agenda consistent with the Paris Declaration on Aid Effectiveness and the Accra Agenda for Actionshould be accompanied by strong harmonization with the GMAP and the goals of the Roll Back Malaria Partnership.
## Cross-cutting research for a good measure of transmission in the later stages of elimination
Malaria elimination has a very different endpoint from malaria control and this change of paradigm demands the development of specific measures of progress. New infections are a direct measure of ongoing transmission but require labor-intensive, active surveillance studies, particularly during the elimination phase in regions previously experiencing high transmission where immune individuals are unlikely to experience symptomatic disease. After some years, as immunity declines, infection is more likely to be symptomatic and may then be a good surrogate marker for the detection of continued or resumed transmission during surveillance. Thus, at the end of the process, some years after elimination has been achieved and the population has lost clinical immunity, surveillance of clinical cases can become a guide to transmission. However, there are many years between the time when transmission can be measured in endemic areas (albeit with difficulty and high cost) and the time when active surveillance of occasional cases becomes a useful measure (see also.
Accordingly, elimination programs need rapid, sensitive, standardised, and reproducible transmission measurement methods to monitor progress towards the desired goal [bib_ref] Measuring malaria endemicity from intense to interrupted transmission, Hay [/bib_ref] , particularly when transmission continues at low and nonrandom levels. Research into and development of new measures that are simpler than surveillance for incident infections is a high priority in the cross-cutting research and development agenda. Such measures could potentially be based on serological or other biomarkers and used as indicators of readiness for elimination, progress towards that goal, and as markers of residual foci or reintroduced infection.
In particular, the new and improved measures of transmission could be used for measurement and certification of the absence of transmission. Such measures are essential to ensure that the decision to stop expensive entomological studies or indoor spraying that inconvenience communities is made at the appropriate time. Sustained funding is, of course, required to detect ongoing transmission or reintroduction of disease.
## Integration with strengthened health systems
Many past efforts at malaria elimination have failed because the health system failed during the implementation of stand-alone programs [bib_ref] Some lessons for the future from the Global Malaria Eradication Programme (1955-1969), Nájera [/bib_ref]. This failure, through neglect or at least underresourcing during implementation of vertical programs, resulted in the pessimistic view that malaria can only be eliminated in regions where economic progress and stable governance are in place that Summary Points N Several important cross-cutting issues must be addressed as the international community or an individual program moves from malaria control to malaria elimination/eradication: an integrated decision-making framework must be constructed for this paradigm shift N Methods to measure transmission rapidly and costeffectively in populations, particularly in low transmission settings, must be developed; very sensitive indicators of transmission are particularly important late in the elimination phase N Elimination programs must be integrated for mutual benefit with strengthened health systems; better training and capacity building, better information systems, and modeling must also be developed N New or improved tools alone will not be enough; community engagement and good communication between everyone involved in malaria elimination/ eradication is essential N A research and development agenda for cross-cutting issues is presented that should facilitate progress as programs aiming at malaria elimination/eradication supersede malaria control programs support well-functioning health systems. Even if a region initially opts for a purely vertical approach, when transmission declines, patient needs for appropriate diagnosis and treatment in the general health system become part of the surveillance system and need to be integrated with existing health system structures for local responses and central monitoring [bib_ref] Some lessons for the future from the Global Malaria Eradication Programme (1955-1969), Nájera [/bib_ref]. Moreover, diagnosis and appropriate treatment can contribute to reduction in transmission, and good health facilities are essential for management of other febrile illnesses. For these reasons, a malaria elimination program simply cannot succeed in the absence of an effective health system.
The importance of health systems thinking, the need for settingspecific and phase-specific integration, and the need for new approaches to replace the old separation into ''horizontal'' or ''vertical'' programs have been discussed by most of the other malERA Consultative Groups but particularly by the group that focused on health systems. The consultative groups also highlighted relevant cross-cutting research and development agenda topics such as the need to measure synergies between malaria-specific programs and health systems strengthening, and the extent to which inputs for malaria elimination can be used to strengthen population health. Our group concluded that tailoring an approach to each setting is required, maximising synergy with the health system for mutual benefit, while maintaining the integrity of categorical program objectives, and the important activities of the health system.
## Training
All of the consultative groups recognized the need for training and capacity building in the context of elimination, from discovery research in the laboratory, through social sciences research in communities, and on to operational research in the context of health systems thinking. Master's level research training that introduces the principles of a scientific approach, epidemiology, and evidence-based decision making would benefit anyone involved in deciding about resource allocation, timing, and refinement of the elimination approach before, during, and after any elimination/eradication program. Training for the eradication research agenda also needs to be accompanied by training of public health leaders and managers with substantial knowledge of malaria.
In addition, communities of health systems experts require research training to help them measure the impacts of an integrated approach to malaria elimination. ''Elimination science'' would assess the implementation of changed diagnostic or surveillance methods, or expanded roles of community health workers or reporters engaged in active surveillance (''learning in action''). The information gleaned through such assessments could be used for operational research or social science research relevant to community participation and engagement. It could also be used by a new cohort of experts in database development, management, or information technology.
For basic research, which has a longer time frame, academic expertise needs to be developed and sustained in fields relevant to technological development such as bioinformatics, genetics, drug and vaccine discovery, systems thinking, and mathematical modeling. It also needs to be developed in fields relevant to health promotion and communication and the enhancement of these fields by new technology.
Together, these training requirements, particularly those that focus on the needs of disease-endemic countries, are substantial and should be the subject of a later specific review.
## Information systems and modeling for assessing combinations of intervention strategies
All the consultative groups acknowledged the importance of strong information systems that are reliable and responsive to local needs for rapid intervention, and that provide inputs to national and regional databases. The requirements for information systems will change over time with changes in transmission but an important attribute of these systems should be harmonization and the avoidance of unnecessary duplication to meet, for example, special or frequent requests from funding agencies. Importantly, additional sources of information have to be integrated into existing information systems to allow modeling of future interventions, to facilitate the analysis of system-wide effects for costing and implementation, and to provide a resource for researchers who are modeling transmission, as discussed in other malERA articles (also see.
In common with surveillance systems, information systems need to be envisaged as tools for intervention (with a target product profile and standards to be developed and monitored), rather than as ends in themselves. The consideration of information systems as interventions (just as surveillance was defined as an intervention by the WHO Global Malaria Eradication Program), provides a useful perspective for the definition of the malERA research and development agenda and is well discussed elsewhere in this series.
Finally, because the costs and benefits, potential synergies, and operational assessments of combination strategies are likely to be different in different environments, modeling emerged as one of the key cross-cutting themes during the malERA consultation process. In particular, the use of modeling to assist discussions and decisions on intervention mixes in time and space emerged as a high priority cross-cutting theme that is discussed further in the relevant article in this Supplement.
## Community engagement
Successful public health programs are characterized by community engagement and good communication, but how to achieve these critical success factors is not well understood. Community case management and treatments such as piloted in Tigray, can be effective, but support from all sectors of society is critical, particularly where there is a requirement for behavioural change. Strategies are required to explain why efforts against malaria need to be maintained, even when malaria cases are extremely rare. Conversely, governments also have to choose the correct time, and explain the rationale for stopping certain interventions. We need to understand how public perception affects such decisions and provide guidance for countries on when certain interventions will no longer be cost-effective, and we have to communicate this information effectively.
Good communication is essential among malaria researchers. It is also essential that malaria researchers communicate well with people involved in health systems, malaria control specialists, health care workers, funders, stakeholders from public and private nongovernment sectors, communities, the general population, and the international community. Research should be undertaken on the range of factors that influence connectivity, from cultural aspects to technology, which could be revolutionised by the advent and availability of new means of communication.
# Conclusions
An important part of the malERA process was to identify crosscutting issues that could facilitate the achievement of the goal of elimination, particularly in the light of past failures, and build on the GMAP that already includes eradication as a long-term goal.
As recognized by the whole malaria community, integration is a prerequisite for success.
Tools alone are not enough, but need to be accompanied by excellent and ongoing coordination, operational research, information systems, and monitoring and evaluation supplemented by active surveillance. Integration with the health system and a multidisciplinary approach are also essential, providing new tools and approaches for modeling and for systems thinking about the concepts and strategy needed to achieve the ultimate goal. In addition, communication and research into its improvement and local adaptation are critical; without excellent communication and community and political engagement, elimination/eradication programs will not succeed. Moreover the community and the health system need to be ready with appropriate tools and trained personnel in place to take on new or specific tasks that need to be integrated into ongoing activities.
Before attempting elimination, a realistic feasibility assessment is required to determine readiness for this challenge. Some countries fall far short of readiness, having tools that are inadequate to complete the task where force of infection is very high, having health systems that are weak, or suffering from socio-political and civil disturbances that make public health practice nearly impossible. Other countries may simply lack one major prerequisite such as political will, or a drug to overcome resistance to available antimalarial therapy. Unrealistic promises about malaria elimination will inevitably lead to disappointment and disillusion with public health approaches and should be avoided.
We cannot provide estimates of the cost of the research and development agenda for cross-cutting issues that we present in Box 1, and recognise that further work will be required to delineate fully all the regulatory and ethical implications of new tools that have been envisaged or described here. Technology that may provide solutions may currently be beyond our imagination, but could be available within a short few years. Importantly, however, we recognize that very long-term investments will be needed for the research and development agenda that we have outlined. We also recognize that we need to build on public/private partnerships and connections with industry to facilitate new advances. Nevertheless, we emphasize that, even if elimination programs are decades away for some countries with very high transmission, now is the time to start work on the broad and integrated portfolio of long-term research that is essential if the goal of malaria eradication is to be achieved. Box 1. Summary of the Research and Development Agenda for Cross-Cutting Issues N Develop and validate a framework of essential information required for making the decision to progress from scaled-up control to elimination that includes political, economic, and financial factors. The framework should recognise variability in epidemiology, the need for political will to prioritise and/or finance and support such a long-term project, and the need for locally effective tools powerful enough to finish the task.
N Develop a long-term investment case for elimination that should align with important development themes such as global security, migration, food security, and climate change.
N Document current and past efforts towards elimination. N Develop methods and approaches to measure and monitor transmission in a rapid and cost-effective way at a population level, especially in very low transmission settings. These methods and approaches should be used as metrics for the very sensitive indicators of progress needed for active surveillance systems required in the last phases of elimination.
N Define and develop the tools required for a communication and knowledge management strategy that encourages community engagement, local health system involvement, and the participation of national, and international stakeholders.
PLoS Medicine | www.plosmedicine.org
January 2011 | Volume 8 | Issue 1 | e1000404
AcknowledgmentsThe Malaria Eradication Research Agenda (malERA) Consultative Group on Integration Strategies was chaired by Fred Binka, INDEPTH -Author ContributionsGB is the corresponding author on behalf of the malERA consultative group on Cross-Cutting Issues. ICMJE criteria for authorship read and met: PLA, GB, MT. Participated in one or more discussion meetings resulting in the writing of this article: The malERA Consultative Group on Cross-Cutting Issues. |
Spatial early warning signals of social and epidemiological tipping points in a coupled behaviour-disease network
the resurgence of infectious diseases due to vaccine refusal has highlighted the role of interactions between disease dynamics and the spread of vaccine opinion on social networks. Shifts between disease elimination and outbreak regimes often occur through tipping points. it is known that tipping points can be predicted by early warning signals (eWS) based on characteristic dynamics near the critical transition, but the study of eWS in coupled behaviour-disease networks has received little attention. Here, we test several eWS indicators measuring spatial coherence and autocorrelation for their ability to predict a critical transition corresponding to disease outbreaks and vaccine refusal in a multiplex network model. the model couples paediatric infectious disease spread through a contact network to binary opinion dynamics of vaccine opinion on a social network. through change point detection, we find that mutual information and join count indicators provided the best EWS. We also show the paediatric infectious disease natural history generates a discrepancy between population-level vaccine opinions and vaccine immunity status, such that transitions in the social network may occur before epidemiological transitions. these results suggest that monitoring social media for eWS of paediatric infectious disease outbreaks using these spatial indicators could be successful.
Resurgences of vaccine-preventable diseases severely stress public health systems, interrupt tourism and public services, and disrupt economies through the huge costs of large-scale interventions. These impacts motivate the study of factors that support vaccine uptake. Undervaccination may be attributed to vaccine refusal 2 , the spread of anti-vaccine opinion facilitated by media coverage and its sensationalisation of true adverse vaccine effects, the expectation of adverse effects 4 , misstatement of the cause of illnesses, the spread of other rumours and false information, and the effect of social norms.
These phenomena show how the social diffusion of information is heavily responsible for the trajectory of disease spread through its ability to alter individual behaviour. Much work has modelled opinion dynamics for different applications through the use of voter modelsand majority opinion models, among other frameworks, and their combination with network structure has revealed much about the occurrence of opinion cascades 10 and forecasting. For instance, models coupling behavioural dynamics and spreading processes change the predicted dynamics of influenza transmissionand climate changealike.
Opinion propagation can be represented by information diffusion through social networks. Similarly, infection spread is often conceptualized as spreading through a physical contact network. Many models explore the dynamics of n-layer multiplex networks, where each layer represents a different aspect of the dynamics of a single coupled system. In these cases, the theory of phase transitions in spatially structured systems is important. For instance, epidemic regimes have previously been modelled as the outcome of phase transitions in physical systems and are featured widely in the epidemiology literature. Generally, phase transitions occur when a physical system moves from one state to another after going through some critical point. Second-order transitions occur when a macroscopic variable varies continuouslyand are called critical transitions.
Systems approaching these critical transitions sometimes display characteristic spatial or temporal behaviours called early warning signals (EWS) that can predict coming epidemic disease outbreaks and other events. EWS are perhaps better defined as statistically significant, recognisable and characteristic behaviours known to precede critical transition in dynamical systems. For instance, critical slowing down can precede both first-and second-order transitionsand is accompanied by the divergence of correlation length in a physical system Many statistics have been used to study EWS in spatially extended systems; temporaland spatial correlationhave been found to precede transitions in spreading processes. Other measurements have been applied to spin systems, where each site in a lattice may be in one of two possible states, possibly partially dependent on the state of neighbouring sites. The spin model has also been applied to opinion dynamics; a simple voter model with binary opinion dynamics is analogous to a physical spin system, where particles represent agents and spins represent different opinions. Consensus formation can be seen as a second-order phase transition to an ordered state (where all spins are aligned). In this regime, knowledge of the opinion of a single agent predicts the opinion of all other agents in the system. Since the transition in finite networks is smooth, the distance across the network over which the opinions of connected agents are strongly correlated increases smoothly; this is analogous to divergence of the correlation length of a physical system.
Above some critical temperature, disordered systems take the form of a spin glass. In a spatial opinion model, this describes a state where opinions between neighbours are generally uncorrelated. On a static network, this state induces a larger number of edges between dissimilar neighbours as compared to that of consensus regimes. This is related to join count statistics, where the numbers of edges between like neighbours are compared to the number between dislike neighbours as a test of geographical distribution. This is arguably the most natural and well-defined measure for graphs presenting binary data and is used for spatial analysis.
The necessity of disease surveillance and early warning signals for outbreaks has been discussed in multiple contexts, from epidemic mitigation to bioterrorism prevention. Potential mitigation of unnecessary expense motivates us to find reliable EWS that remain easily computable on large high-resolution data sets. Furthermore, the study of EWS in coupled disease-behaviour multiplex networks has received relatively little attention, suggesting a significant gap in the literature. Our objective is to evaluate and compare the relative merits of the mutual information, Moran's I, Geary's C and join count statistics as EWS of the occurrence of epidemics and changes in aggregate opinion on a coupled disease-behaviour network model. We use three differently parametrised models (V1, V2 and V3) coupling a binary vaccination opinion dynamic to an SIRV epidemic process. The resulting trends in the EWS for model V2 will be explored in the Results and Discussion sections, with V1 and V3 presented in Supplementary Information S5.
The outline of this paper is as follows: the Methods section will present the EWS and their derivations and give the details of the model used. The Results section will analyse the trends in the warning signals and the Discussion section will present a review of the study and any shortcomings of our approach, with further results pertinent to the study presented in the Supplementary Information.
# Methods
We assume an acute, self-limiting infection that confers lifelong natural immunity upon recovery, and for which a vaccine is readily available. Similar premises have been used to represent the natural history of many paediatric infectious diseases such as measles. In particular, we assume an SIRV p natural history consisting of four mutually exclusive disease states. Agents are initially susceptible to infection (S). Upon infection the agent enters the infected state ( → S I), which we treat as a combination of both the latent, ill and infectious periods. Upon clearing the infection, agents enter the recovered state of lifelong immunity ( → I R); additionally, susceptible agents may be vaccinated and so enter the vaccinated state ( → S V p ). We also include injunctive social norms (i.e. peer pressure) as well as a perceived cost of vaccination that captures both economic costs and the fear of perceived adverse vaccine effects. As in some models, we include a noise parameter ξ to account for environmental and demographic stochasticity 51 with the simplifying assumption of perfect vaccination 52 (reversion from the recovered state to the susceptible only through agent death). During simulation, each time step represents a single week. the model. We model a multilayer network where each layer is given an identical undirected Erdös-Rényi random graph with size N and mean node degree Q n . Each agent n can be described by a pair of states; for instance, each agent is assigned the joint state V V ( , ) s p at the start of the simulation with probability α (they are a pro-vaccine vaccinated agent), else they are initialised with joint state N S ( , ) (an anti-vaccine susceptible agent) with probability α − 1 . The social process follows an NV s dynamic, representing pro-(V s ) and anti-vaccine (N ) opinion for each agent n. ξ represents the probability of any agent switching opinion randomly in each week and → N V ( ) n s represents the probability of switching from anti-vaccine opinion to pro-vaccine opinion ( → N V s ) upon interaction with a disagreeing neighbour. We introduce an imitation dynamic by having each agent n compare its opinion with a single randomly chosen social contact (a neighbouring agent on the social layer) each week; n then changes its vaccination opinion only if there is disagreement (the agent and the neighbour have different vaccine opinions). This change of opinion depends on the perceived risk of vaccine adverse effects κ ("vaccine risk") and I n (the number of infected physical neighbours of n) according to the rules Here, σ represents the strength of an injunctive social norm to maintain the current opinion (referred to as "social norm"), while ⁎ Q n represents the number of neighbours of n with opinion ⁎ and Q n represents the total number of social contacts of n.
[formula] → = + −∆ → = + −∆ → → N V U V N U ( ) 1 1 exp( ) , ( ) 1 1 exp( ) ,(1)σ κ σ κ ∆ =− − − − ∆ =− − + − . → → U Q Q Q I U Q Q Q I ( ),( ) (2) [/formula]
The epidemiological dynamics (which we term 'physical dynamics') follow an SIRV p process, in which an agent n can progress through each of four disease compartments: S (susceptible), I (infected), R (recovered) and V p (vaccinated). Every week (i.e. in each time step), each susceptible agent n interacts with all its physical neighbours; each effective interaction carries the probability p of infection ( → S I), so that each susceptible agent faces the total probability − − p 1 (1 ) I n of infection in a single week. The duration of the illness is weeks (with no impact on mortality), after which n gains lifelong natural immunity ( → I R). Alternately, if a susceptible agent n adopts a pro-vaccine opinion, they are immediately vaccinated → ( ) S V p and gain lifelong vaccine-derived immunity. We also assume that only susceptible agents are vaccinated. Thus, individual agents may change their opinion about vaccination multiple times in their life ( → → N V N s ), but once they are vaccinated they may not become unvaccinated ( → S V p ). This in turn creates an asymmetry between disease dynamics and social dynamics that will have implications for the model predictions. We will discuss this in the Results section.
Each agent n has probability µ of dying each week, upon which they are replaced by a new susceptible individual who is a pro-vaccine vaccinated agent V V ( , ) s p with probability α, or an anti-vaccine susceptible agent N S ( , ) with probability α − 1 , keeping the same physical and social contacts as the agent they replaced (that is, the network is static). Case importation is accounted for by infecting a randomly selected proportion ι of susceptible agents at the start of each week, and noise is introduced to the model by changing the vaccine opinions of a randomly selected proportion ξ of the entire population weekly.
At the start of the simulation, some susceptible agent is randomly selected as an index patient and infected; subsequent disease spread is governed solely by environment and inter-agent interaction. Models V1, V2 and V3 all use these model dynamics, and a flowchart and detailed written description of the model transitions are given in Supplementary Information S2. A complete list of the variables used is given in early warning signals. Mutual information M is defined as
[formula] → N V ( ) n s and → V N ( ) n s [/formula]
respectively upon interaction with a dissenting neighbour. α gives the probability of being birthed with provaccine opinion V s .
[formula] M X Y X Y ∑ ∑ = ⋅ ⋅ ∈ ∈ x y x y x y ( , ) (, ) log ( , ) ( ) ( ) ,(3) [/formula]
x y 2 where X and Y are discrete random variables; x takes value on the set X x x x { , , , } = … and y on set y y y { , , , } 1 2 3 Y = …, with a joint probability mass function of X and Y 53 . Mutual information is an entropy-based quantification of the "shared information" of two random variables quantifying how knowledge of one decreases the uncertainty of the other and vice versa. Mutual information peaks at the critical temperature of spin systems during second-order transitions and has been widely used in detecting phase transitions; an advantage of this statistic is its ability to quantify non-linear dependence, unlike Moran's I and covariance which only account for linear dependence.
Join counts quantify the degree of clustering by giving the number of adjacencies between agents of different types. We divide the population into two attributive classes, with V s the compartment of pro-vaccine agents and N the compartment of anti-vaccine agents. Let [Ψ, Ω] be the number of social interactions between agents with vaccine opinions Ψ and Ω; then N N represents the number of nearest-neighbour interactions between anti-vaccine agents, V V
[ , ] s s the number of interactions between pro-vaccine agents and N V [ , ] s the number of interactions between pro-and anti-vaccine agents. These can be written as
[formula] ∑ ∑ ∑ ω ω ω = − = − − = N V x x N N x x V V x x [ , ] 1 2 ( ) , [ , ] 1 2 (1 )(1 ), [ , ] 1 2 ,(4)s j k jk j k j k jk j k s s j k jk j k , 2 , , where ω = 1 j k , [/formula]
if agents j and k are social neighbours, with ω = 0 j k , otherwise (ω is the adjacency matrix of the social network); x n represents the opinion score of agent n, defined as
[formula] { x k V x 1 ( has a pro vaccine opinion) 0 else ( 5) n s n = ∈ . - [/formula]
In an opinion model, clustering manifests as agents consistently having a higher number of like-minded neighbours than expected based on the global prevalence of the opinion; join counts are then used to test the null hypothesis of positive correlation. Join counts are used in many fields as a categorical test of spatial autocorrelation, including ecology 58 and geographical information systems. In all parameter realisations here, the number of joins are counted naïvely rather than calculated. Joins between like-minded agents (e.g. N N
[ , ] and V V [ , ] s s joins) will be called similar joins, and edges between disagreeing neighbours (e.g. N V
[ , ] s ) will be called dissimilar joins.
The Moran's I coefficient J quantifies spatial correlation and is defined
[formula] J ω = ⋅ ∑ − − ∑ − N W x x x x x x ( )( ) ( ) ,(6)j k jk j k j j , 2 where ω = ∑ W j k jk , and = ∑ x x N j j 1 [/formula]
represents the mean opinion score of the population. Used as a global statistic, Moran's I gives the degree of correlation between the values of neighbouring patches (agents and their social contacts); here, the numerical value of the vaccine opinion is the same as described in Eq.. Algebraic manipulation of Eq. (6) using Eq. (5) gives
[formula] J N W V V x V V N V W x x V N x 2 [ , ] 2 (2 [ , ] [ , ]) (1 2 ) [ ] ,(7) [/formula]
s s s s s
[formula] s 2 2 = ⋅ ⋅ − ⋅ + + ⋅ − ⋅ + ⋅ [/formula]
(full derivation given in Supplementary Information S4); we can then consider Moran's I as a measure derived from the linear combination of join counts. Positive values signify spatial correlation, with negative values signifying anticorrelation. The Geary's C coefficient C is yet another measure of spatial correlation based on the cross-product (like Moran's I), but unlike Moran's I it accounts for the difference in opinion between two neighbours 60 . It is given as
[formula] ω = − ∑ − ∑ − . C N W x x x x 1 ( ) ( ) (8) j k jk j k j j , 2 2 [/formula]
Lower values show spatial correlation, and large values represent anticorrelation. Like Moran's I (Eq. (7)), Geary's C can also be expressed as a linear combination of join counts www.nature.com/scientificreports www.nature.com/scientificreports/ proportion ratio ι = . × − 2 5 10 5 was added to provide periodic impulses of infection as a test of resilience in endemic disease regimes. Here, an ensemble of 100 simulations using parameters κ = 0, σ = 0 and α = .
[formula] C N W N V x V N x 1 2 [ , ] (1 2 ) [ ] ;(9) [/formula]
0 05 returned the values < ., where Ψ represents the mean number of agents with (social or physical) state Ψ, averaged over all realisations of that combination of parameter values.
[formula] ⟨ ⟩ ⟨ ⟩ S R , 005 at equilibrium (defined in [/formula]
The infectivity = .
p 0 2 was chosen to reflect the reproduction rate of a measles infection commonly estimated from empirical data 61 ; effective contacts occur in the simulation once per week during the period of infection. The probability of randomly switching opinion ξ = × − 1 10 1 4 was included as a source of noise. We found that the parameter ranges for vaccine risk κ ∈ −and social norm σ ∈ [0, 3] were sufficiently broad to capture transitions in both social and physical dynamics, as well as the corresponding trends in the mutual informationand dissimilar join count.
The contours in each panel ofshow the obvious correspondence between transitions in the social and physical dynamics of the model, and substantial changes in M and N N , ; here, the dissimilar join count N V , sand mutual information Mincrease while the vaccine risk κ and social norm σ parameters increase towards their respective (pre-transition) threshold values. These trends are generally asymmetric about both transitions; this can be seen in, where post-transition trends do not exhibit similarly detectable warnings (if any). This Parametrisation applies to model V2. The corresponding parametrisations and contour plots of models V1 and V3, as well as their post-transition trends are presented through the Supplementary Information.
# Results
Due to the low initial vaccine coverage α = .
0 05, all realisations demonstrated an initial epidemic spread (defined in Supplementary Information S3) over the first 7 weeks. After this period, the dynamics settled down to a quasi-equilibrium state characterized by fluctuations around a mean value that is the focus of our study -the following subsections are grouped by major findings of the model. The term model variables refer to the outputs S , I , R , V p , N and V s . Population vaccine immunity status can differ from aggregate vaccine opinion. Because only susceptible individuals are vaccinated and individuals cannot become 'unvaccinated' (but may change their opinion about the vaccine over their lifetime), the population-averaged vaccine opinion is not equal to the population-averaged vaccine immunity status, even at the quasi-equilibrium state. With no social pressure (σ = 0), a small increase in vaccine risk κ → . 0 03125 pushes the system to endemic infection and anti-vaccine consensusdespite a high vaccination rate. Towards an explanation, if an agent n is newly birthed into this regime, the probability of having an infected neighbour vanishes 〈 〉 → under the assumptions that the average agent with anti-vaccine opinion is almost certain to interact with a disagreeing contact. A similar calculation explains the phenomenon of high vaccination ratedespite mixed consensuswhen vaccine risk becomes neutral (κ → 0). A vaccine then perceived as beneficial (κ → − . 0 03125) intuitively results in a high vaccination rateand pro-vaccine consensus; in both these regimes, the disease survives only through case importation. In the absence of social norms and vaccine risk, the population's aggregate vaccine opinion may not be a good indicator of its vaccine immunity profile (and vice versa); Eq.shows that the probability of changing opinion depends only on I n when social norm σ = 0. In this region, the pattern of disease spread will be determined by the initial conditions of the physical dynamics; slight changes in vaccine risk κ will push the network towards either of the consensuses, with minimal www.nature.com/scientificreports www.nature.com/scientificreports/ effect on the high vaccination rate. This phenomenon is shared by models V1 and V3, as shown in Supplementary Information S5.1.
[formula] I ( 0 ) n , so that → = → ≈ N V V N ( ) ( ) 1 2 ,(10) [/formula]
## Ews trends identify approaching transitions in both social and physical layers. the trends in
both model dynamics and the proposed EWS are shown in, with social norm σ = 0 (column A) and σ = .
0 25 (column B). The first vertical black line in all panels ofrepresents the transition in the social dynamics K s , defined as the smallest κ value at which V N s 〈 〉 ≈ 〈 〉 (the mean number of pro-vaccine agents equals the number of anti-vaccine agents); the second vertical black line represents the transition in the physical dynamics K p , similarly defined as the earliest κ where R V p 〈 〉 ≈ 〈 〉. Multiple physical and social transitions were found for some parameter combinations; these trends and the attendant behaviour of the EWS for models V1 and V3 can be seen in Supplementary Information S5.3. We also note that the equilibrium values of EWS and model variables were averaged over − realisations of all parameter combinations. All EWS show recognisable trends preceding both transitions for both social norm values σ = 0 (column A) and σ = .
0 25 (column B); for instance N N ,, N V , s, Jand Cincrease sharply preceding K s with all but N V , sapproaching some maximum value preceding K p , while V V , s sand Msharply decrease and approach some minimum value before K s and K p respectively. Mutual information Min particular shows clear changes in trend well before the social transition K s occurs. Though N V , s shows a similar rising-falling pattern for both σ = . 0, 0 25, its maximum value with σ = .
0 25is much lower than that for σ = 0. For σ = . 0 25, the mean of the Geary's C Cshows almost no change, though its envelope broadens post-transition; we see this as a failure of the EWS (no forewarning given). Similar observations hold for model V1 (Supplementary Information S5.2), with the failure of the Geary's C coefficient C shown clearly in Supplementary Figs. S6 and S7. As stated in the Methods section, the pre-and post-transition trends of the EWS do not generally resemble each other; asymmetry of the EWS about K s can be seenand Supplementary Figs. S6 and S7, showing that (in general) less of a warning is given (if any) when the κ-series is reversed. This is explicitly demonstrated in Supplementary Figs. S15 and S16, where skewness γ 1 is used to quantify asymmetry of the trend of each EWS.
We can then say that all proposed EWS other than C give appreciable signals approaching K s and K p when σ = .
0 25. K s precedes K p, showing that a shift in consensus will always precede a crisis in vaccination coverage in this model. Also shown is a marked decrease in − K K p s (the gap between the two transitions K s and K p , which we call the intertransition distance) as the social norm strengthens (for example, σ → . 0 25 in column (B) of. The generalisation of these trends to all tested values of σ is confirmed in, where − K K p s is everywhere positive, though the distance between K s and K p vanishes with increasing σ ; the inset ofshows the location of K s (blue) and K p (red) with respect to σ, so that − K K p s (purple) gives the width of the area between the two curves in the inset graph at each σ. Other disparate models of the disease display largely similar concave decreases in the intertransition distance, suggesting that this behaviour arises generally from the model dynamics rather than in some specific subspace of the parameter space (see.
## Stronger social norms result in decreased lead time for all ews.
The findings of the preceding subsection are intuitive, as vaccination depends more heavily on individual vaccination opinion than the number of infected neighbouring agents in Eq., so that the opinion dynamics exert more influence than any feedback effect occurring in the physical dynamics. However, the vanishing intertransition distance − K K p s presents a problem if we depend on predictions of K s to enact interventions avoiding the collapse of the system to a non-vaccinated regime (i.e. avoiding K p ). For social norms of increasing strength, we can therefore look at the K s , where Π Ψ represents some κ value at which we can assert that a signal occurs in some sequence Ψ of κ values; since we've established that K s precedes K p everywhere, then necessarily any warning of a social transition also warns of the following physical transition, so the quantity − Π Ψ K p is not discussed here.
[formula] b) V [ ] s , κ = 0. (c) V [ ] s , κ = − . 0 03125. (d) V [ ] p , κ = . 0 03125. (e) V [ ] p , κ = 0. (f) V [ ] p , κ = − . [/formula]
One way for us to quantify this lead time is to use a change point detection tool to find κ values at which the two classes EWS (subscript WS) and model variables (subscript S/P) give signals (i.e. display statistically significant changes in trend/behaviour). Here, the Lanzante test 62 from the trend, the positivity of some curves shows that some of the proposed EWS do indeed give early warnings of coming transitions (largely for σ ≤ .
1 875). Failures of N V , s and M occur in the range σ .
≤ ≤ . 1 875 2 125, while all other tests give valid warning signals everywhere 2 5 σ < . ; model variables R and 〈 〉 V p fail in the range σ . ≤ ≤ . 12 125. The failure of all the tests after 2 5 σ = . likely results from insufficient length of the EWS' κ-series; the inset ofshows that → − K 1 s as σ increases. 2 5 (as in. In line with our focus on social dynamics as a predictor, V s appears to be the best performing signal of all the model variables; as was reasonably expected, I appears to perform badly, since its role as a transitory compartment in the disease dynamics means that it never "gathers" sufficient agents over the course of each realisation to give a true indication of the state of the system (other than indicating the presence or absence of endemic infection).
Since the perceived cost of vaccination κ ≥ −1 in this study, our method of detecting the change point will not accurately predict a change point Π Ψ close to −1. Since no one warning signal gives the highest lead time for any large contiguous range of σ values, there is unfortunately no single objective way to choose a "strongest" signal; they are all suitable tools to predict coming crises in aggregate opinion and vaccination dynamics. However, it is worth noting that mutual information M and the dissimilar join count N V , s perform better than the other indicators;shows that mutual information M gave the largest lead time measured for 45% of tested social norm σ values, and V V ,
## Ews can provide better forewarning than trends in model variables. one final question is whether
the proposed EWS (mutual information, Moran's I, Geary's C, join counts) give earlier warnings than simply monitoring trends in model variables (such as the number of infections or pro-vaccine agents, using a change point test for prediction in both cases). There are many ways to quantify this, including maximin comparison (finding the larger of minimum values of classes WS and S/P) and maximax comparison (finding the larger of the maxima of each class) at each value of the social norm σ.
To compare the minima of the EWS and model variable lead times, we define χ min Lan as 36 4% of tested σ values. The second part of the comparison, red curve) is between the maxima of the lead times; as above, we define the comparison variable χ min Lan and tolerance ε max Lan as, there is no apparent pattern to the positivity/negativity of χ min Lan and χ max Lan . These comparisons are given for other tests and models in Supplementary Information S4.
[formula] χ = − B B min( ) min( ),(12)χ ε χ χ = − = − . B B max( ) max( ), max( ) min( ) 100(14) [/formula]
# Discussion
Here we studied a range of early warning signals for critical transitions in a two-layer coupled behaviour-disease model for paediatric infectious diseases. We compared the indicators to one another and the approach of simply monitoring trends in model variables. We found that the performance of the indicators was variable depending on model parameters, but the mutual information statistic and the dissimilar join count showed consistently high pre-transition lead times over various strengths of the social norm, many times giving the highest lead times of all the EWS. Through maximin and maximax comparisons, we found that using EWS provide more advance warning than simply monitoring trends in model variables in a clear majority of cases.
We note that join counts have the additional advantage of easy computability, since they require only counting pairs of a given type. This contrasts with other more computationally intensive indicators such as autocorrelation which require making decisions about whether to study lag-1 or higher order lags, as well as choosing parameter values governing computation of residuals. Moran's I was also shown to predict the approach of transitions, although perhaps this finding is trivial considering that it is a linear combination of similar and dissimilar join counts. Its predictive power was not as strong as many of the other indicators such as join count and mutual information, hence the added complexity of its calculation may not justify its use. Potential downfalls of the mutual information statistic include its computational complexity and the availability of a suitable data set pairing the personal health of each agent with their social activity.
We also showed that a population may have relatively high vaccine coverage despite a low pro-vaccine opinion. This discrepancy between social and physical dynamics is due to the paediatric infectious disease natural history we assumed in the model. Unlike influenza, where revaccination must occur seasonally, an individual who receives a sufficient number of measles or chickenpox vaccine doses generally has lifelong immunity and therefore the opinion towards the vaccine can decline well before the level of vaccine immunity does. (Individuals can change their opinion but never become 'unvaccinated' .) The implication of this asymmetry is that monitoring social media networks for changes in opinion using early warning signals like mutual information might provide advance warning of outbreak hot-spots.
The distance between the change point in the EWS indicators and the critical transition in the social dynamics decreases as the social norm grows stronger, as does the distance between the transitions in social and physical dynamics of the model. Given the relative scale of the social norm and vaccine risk parameter values used, stronger social norms decrease the time interval between birth and vaccination decision (the vaccination rate converges to its equilibrium value in fewer time steps than in other regimes); feedback between this and the disease incidence in the network (which affects the number of infected neighbours in each agent's neighbourhood) alters the probability function controlling the vaccination decisions, effecting faster alignment of majority opinion and vaccination coverage.
This study only lays the foundation for the investigation of spatial EWS for such coupled-behaviour systems. There is much work to be done before they can be meaningfully applied to empirical data. For instance, we assumed that the network was static. This simplifying assumption could be relaxed in future work by using an evolving social dynamic in which agents are allowed to form or break links with new agents based on their node degree 64 or vaccine opinion and associated social pressures. The rate of interaction between agents was also assumed fixed in our model (relative to the speed of other dynamics, such as the birth/death interval and the length of illness). A valid extension of the model would be a variable rate of communication between agents, since the rate of communication has been shown to influence the rate and efficiency of opinion formation. A further avenue of research would explore interventions to turn populations away from critical transitions. This could answer research questions such as: How far in advance must we act to prevent a social or physical critical transition, and does this change our interpretation of the EWS? If we assume that any of the EWS can be used for monitoring, how would this change in trend alter the reliability of the EWS?
Many researchers and public health bodies are drawing attention to global resurgences of vaccine-preventable illness, and speak to the vast efforts and multiple approaches taken to mitigating outbreaks. A few of these approaches have focused on human behaviour and opinion dynamics, either by directly tracking aggregate vaccine opinion, or monitoring alerts and media reports. Our work demonstrates the potential uses of early warning systems of critical transitions in preventative epidemiology. In particular, our work provides proof-of-concept for the idea of monitoring social networks for early warning signals of both social and epidemiological shifts, and also suggests several EWS indicators that might work well for this purpose.
## Data availability
The datasets generated during and analysed during the current study are available from the corresponding author upon reasonable request. |
Biomarkers of collagen synthesis predict progression in the PROFILE idiopathic pulmonary fibrosis cohort
Idiopathic pulmonary fibrosis (IPF) is characterised by excessive extracellular matrix (ECM) deposition and remodelling. Measuring this activity provides an opportunity to develop tools capable of identifying individuals at-risk of progression. Longitudinal change in markers of ECM synthesis was assessed in 145 newly-diagnosed individuals with IPF.
## Take home message
This study demonstrates that synthesis neoepitopes for collagen type 3 and 6, but not the N-terminal propeptide of collagen type 1, distinguish individuals with IPF who have progressive disease from those with more stable disease.
# Background
Idiopathic pulmonary fibrosis (IPF) is a fatal condition with a dismal untreated prognosis and despite the approval of two anti-fibrotic therapies remains a major clinical challenge. At present there are no available measures for stratifying individuals with IPF to specific therapies nor, importantly, to determine response to therapy. Furthermore, although forced vital capacity (FVC) is accepted as a regulatory end-point in clinical trials, the relative insensitivity of physiological measures to change in fibrosis necessitates studies with a minimum duration of 52 weeks [bib_ref] Idiopathic pulmonary fibrosis: clinically meaningful primary endpoints in phase 3 clinical trials, Rose [/bib_ref]. Therefore, there is an urgent need to identify biomarkers for various aspects of disease behaviour in IPF; particularly those that detect individuals at greater risk of progression and which stratify individuals into better defined endotypes (as a path to enabling effective targeting of therapies to individuals most likely to benefit).
Excessive extracellular matrix (ECM) deposition is central to the pathogenesis of IPF. The production, deposition and remodelling of collagen, a major component of ECM, is a dynamic process [bib_ref] The extracellular matrix: not just pretty fibrils, Hynes [/bib_ref]. In healthy lung, there is a balance between degradation and synthesis of collagen and other ECM proteins [bib_ref] The extracellular matrix: not just pretty fibrils, Hynes [/bib_ref] , which is disturbed in the lungs of individuals with fibrosis [bib_ref] Concentration, biosynthesis and degradation of collagen in idiopathic pulmonary fibrosis, Selman [/bib_ref] [bib_ref] Serological investigation of the collagen degradation profile of patients with chronic obstructive..., Leeming [/bib_ref]. During collagen turnover, distinct peptides are generated as by-products of both collagen breakdown and synthesis. The pro-peptide of pro-collagen is cleaved before collagen is incorporated into the matrix. This results in the release of unique fragments for each individual form of collagen [bib_ref] Always cleave up your mess: targeting collagen degradation to treat tissue fibrosis, Mckleroy [/bib_ref]. Similarly, during degradation, specific metalloproteases (MMPs) cleave collagen fibres revealing distinct neo-epitopes. Both synthesis peptides and cleavage neo-epitopes are released into the circulation and can be detected in the blood.
We have previously demonstrated, in subjects recruited to the PROFILE (Prospective Observation of Fibrosis in the Lung Clinical Endpoints) study, that serum levels of neoepitopes are higher in individuals with IPF when compared with healthy age-matched controls. Furthermore, we demonstrated that longitudinal changes in serum concentrations of several neo-epitopes track with progression of fibrosis and, by three months, can predict mortality in individuals with IPF [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref]. In the current study, we assessed collagen formation markers for their prognostic utility in IPF in a separate cohort of individuals recruited into the PROFILE study. We also compared these synthesis markers to their collagen-equivalent degradation neo-epitopes thus enabling assessment of the relationship between synthesis and degradation during IPF progression.
# Methods
## Study design and participants
As previously described, the PROFILE study is a prospective, multicentre, observational cohort study that recruited individuals with IPF and fibrotic NSIP from two co-ordinating UK centres; Nottingham and Royal Brompton [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref] [bib_ref] An epithelial biomarker signature for idiopathic pulmonary fibrosis: an analysis from the..., Maher [/bib_ref]. Participants were enrolled if they had an incident diagnosis confirmed at multidisciplinary discussion according to international consensus guidelines [bib_ref] An official ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence-based guidelines for diagnosis and..., Raghu [/bib_ref]. The cohort of individuals used in the current study represents a sub-group of 145 participants from PROFILE; none of whom were included in the previous PRO-FILE study on neo-epitopes [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref] and all of whom had at least 1 year of follow up at the time neoepitope analyses were performed. A total of 50 physician-verified control participants without history of respiratory disease were recruited from primary care clinics. Samples from 20 of these subjects were used in our earlier study [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref]. All participants provided written informed consent. The PROFILE study is registered on Clini-calTrials.gov, numbers NCT01134822 and NCT01110694. For survival analyses, a censoring date of 1st March 2016 was used.
## Procedures
## Patient samples
Serum samples collected from study participants at baseline, 1, 3, and 6 months, and a single serum sample from controls were stored at − 80°C until assayed. Studyspecific operating procedures for collection of the serum were used to ensure standardisation across both centres. Briefly, blood was collected with anticoagulant free, serum separation tubes, with coated silica as the clot activator (Becton Dickinson, Winnersh, UK). All samples were processed within 2 h of collection. Samples were allowed to clot at room temperature for 30 min and serum was then isolated through centrifugation, and aliquoted before freezing.
## Quantification of extracellular matrix related biochemical markers
Serum samples from patients were analysed using competitive ELISAs for the formation markers of type 1 collagen (P1NP) [bib_ref] Enzyme-linked immunosorbent serum assays (ELISAs) for rat and human N-terminal pro-peptide of..., Leeming [/bib_ref] , type 3 collagen (PRO-C3) [bib_ref] The neo-epitope specific PRO-C3 ELISA measures true formation of type III collagen..., Nielsen [/bib_ref] and type 6 collagen (PRO-C6) [bib_ref] Type VI collagen turnover-related peptides-novel serological biomarkers of muscle mass and anabolic..., Nedergaard [/bib_ref] as previously described. Additionally, assessment of matrix metalloproteinase (MMP) degraded biglycan (BGM), − type 1 collagen (C1M) [bib_ref] A novel marker for assessment of liver matrix remodeling: an enzyme-linked immunosorbent..., Leeming [/bib_ref] , −type 3 collagen (C3M) [bib_ref] A novel assay for extracellular matrix remodeling associated with liver fibrosis: an..., Barascuk [/bib_ref] , −type 6 collagen (C6M) [bib_ref] MMP mediated degradation of type VI collagen is highly associated with liver..., Veidal [/bib_ref] and C-reactive protein -(CRPM) [bib_ref] Investigation of two novel biochemical markers of inflammation, matrix metalloproteinase and cathepsin..., Skjøt-Arkil [/bib_ref] ; was undertaken as previously described [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref]. A single aliquot for each subject at each time-point was used for all assays to avoid repeated freeze/thaw of samples. Samples were analysed blinded to the associated clinical data. Performance characteristics of the assays and nonstandard abbreviations are described in the Additional file 1: [fig_ref] Table 1: Baseline clinical characteristics in participants with idiopathic pulmonary fibrosis [/fig_ref].
## Follow up
Lung function measurements for forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLCO) were performed on participants at baseline, 6 months and 12 months and annually thereafter according to current ATS/ERS guidelines. Percent-predicted values are based on the European Coal and Steel Board reference equations using age and height at the baseline visit. Date of death data was captured from centralised UK National Health Service records.
# Statistical analysis
As previously described, 'progressive disease' was defined as ≥10% decline in FVC and/or death within 12 months. Missing lung function data were not imputed. Where no lung function data beyond baseline was available and the subject had not died within the relevant period, progression status was adjudicated by the responsible physician, after case note review and blinded to biomarker data at the time of adjudication. A total of 9 subjects were adjudicated for progression: 1 subject was adjudicated to have progressed and 8 were adjudicated as stable.
In cases where biomarker data were below the lower limit of detection, values were imputed to be half the lower limit of detection. Values above the upper limit were conservatively imputed as the upper limit of detection. Imputation rates ranged between 0.9% to below 10% for all markers except C3M that had approximately 19% imputation. Before statistical analysis all biomarkers were transformed (log 10 transformed where the biomarker is treated as response and log 2 where the biomarker is treated as a covariate). A linear regression model adjusted for gender and age was used to evaluate the association between patient group (healthy controls and IPF subjects) and biomarker values at baseline. Adjusted estimates of group means, 95% CIs and p-values of the estimated mean difference are reported. For serial biomarker data, change from baseline for each participant is represented by a gradient on a linear model where the single explanatory variable is the number of months between the baseline visit and subsequent visits using the recorded visit date. We used a linear mixed-effects model (Mixed Model Repeated Measures (MMRM)) to investigate associations between repeated continuous biomarker response and progression status (progression vs stable), visit, visit by progression status (controls are not included in the model). This analysis was adjusted for age, site, gender and smoking status. Unstructured covariance was assumed for all repeated biomarker response values within a participant. Adjusted estimates of biomarker means (95% CIs) by progression status at each visit and p-values of the test of mean differences between progression vs. stable at each visit are reported. Apart from biomarker data above or below the limits of quantification, all other missing biomarker data was not imputed and was assumed Missing at Random (MAR) implying that after adjusting for important covariates in the model the missing data was independent of the observed data.
Univariate analysis was conducted to examine the effect of baseline biomarker measures on overall survival. Two separate procedures were used; the Kaplan-Meier method for survival curves and a proportional hazards model to derive Hazard Ratios. A proportional hazards model was used to evaluate association between time to event (overall survival) as response and continuous biomarker value as a covariate. Similarly, in a separate model, baseline % predicted FVC and separately % predicted DLco were included as covariates. A proportional hazards model was used to evaluate association between time to event (overall survival) as response and biomarker slope as a factor (defined as 'slope to month 3': HIGH (rising concentrations, slope > 0) versus LOW (stable or falling concentrations, slope < =0). Hazard ratios represent either risk of mortality for a 2-fold increase in the biomarker value or risk of mortality in participants with rising neoepitope concentrations relative to those with stable or falling concentrations.
Associations between explanatory variables and endpoints were deemed to be significant at the 5% level. All statistical analyses were conducted with SAS (version 9.3) and were independently quality controlled by a second statistician.
# Results
145 IPF participants recruited in the PROFILE study were assessed in the current analysis. As expected, participants were predominantly male (81%) with a mean age of 71.7 years (SD 7.7) and moderate lung function impairment; FVC 79.8 (SD 20.4) %predicted and DLco 48.2 (SD 17.9) %predicted. Subject demographics, divided in to those with progressive and stable disease, are presented in [fig_ref] Table 1: Baseline clinical characteristics in participants with idiopathic pulmonary fibrosis [/fig_ref]. The 50 control subjects were 80% male with a mean age of 64.6 (SD 7.27).
At baseline, levels of PRO-C3 and PRO-C6 but not P1NP, were significantly elevated in individuals with IPF compared with healthy controls [fig_ref] Figure 1: Baseline comparison of collagen synthesis neoepitope concentrations for collagen-type 1 [/fig_ref]. To enable paired assessment of collagen synthesis and degradation, levels of neoepitopes previously linked with disease progression and survival in the PROFILE study [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref] were measured. These included the three degradation neoepitopes which pair with the synthesis markers assessed; C1M (pairs with P1NP), C3M (PRO-C3) and C6M (PRO-C6) and two additional neoepitopes from our previous study, CRPM and BGM. In keeping with our previously presented data, C1M, C3M, C6M and CRPM neoepitopes were significantly elevated in the serum of subjects with IPF compared with controls at baseline (Additional file 1: [fig_ref] Figure 1: Baseline comparison of collagen synthesis neoepitope concentrations for collagen-type 1 [/fig_ref]. In subjects with IPF there was mild to moderate correlation between levels of different synthesis neoepitopes (Additional file 1:. This correlation was strongest for PRO-C3 and PRO-C6 (r = 0.51). 73 subjects had progressive disease whilst 71 had stable disease. Levels of both PRO-C3 and PRO-C6 were higher in individuals with progressive disease compared with those with stable disease at baseline and they remained consistently and significantly higher in individuals with progressive disease at all visits (1,3 and 6 months) . When serial measurements were adjusted for baseline FVC and DLco the numerical differences between progressive and stable subjects remained albeit with a reduction in the significance level at each timepoint (Additional file 1: . There were no significant differences between stable and progressive disease at any visit for P1NP .
Consistent with our previously reported findings, C1M, C3M, C6M and CRPM were also higher in those with progressive disease compared with individuals with stable disease across all time-points beyond baseline (Additional file 1: . When values were adjusted for baseline lung function variables this trend remained albeit with reduced statistical significance (see Figs. S3 and S4 in the Additional file 1).
When comparing ratios of synthesis and degradation markers P1NP:C1M was significantly lower in participants with progressive disease from 1 month onwards compared with stable disease, whilst PRO-C3:C3M was significantly lower in participants with stable disease at each time point compared with progressive disease . There were no significant differences between stable and progressive disease at any timepoint for the ratio of PRO-C6:C6M .
In were predictive of survival ( Change over time of the three synthesis markers failed to associate with overall survival [fig_ref] Figure 4: Effect of 3 month change in neoepitope concentrations on overall survival [/fig_ref] [fig_ref] Figure 4: Effect of 3 month change in neoepitope concentrations on overall survival [/fig_ref]. Rising levels of the ratio of PRO-C6:C6M were associated with lower mortality (HR 0.55 [0.32-0.95], p = 0.032), however, no relationship was observed with levels of P1NP:C1M or PRO-C3:C3M [fig_ref] Figure 4: Effect of 3 month change in neoepitope concentrations on overall survival [/fig_ref]. Following correction for baseline lung function the association between change in C1M and PRO-C6: Comparisons of neoepitope concentrations for collagen synthesis markers in healthy control and participants with stable and progressive IPF, measured at baseline, 1, 3 and 6 months after baseline for collagen-type 1 (P1NP), −type 3 (PRO-C3) and -type 6 (PRO-C6). Plots represent mean and 95% CI (error bars) adjusted for age, sex, site and smoking status. Disease progression was defined as all-cause mortality or ≧10% decline in forced vital capacity at 12 months. The number of evaluable samples available for analysis at each time point are provided in the graph. P values are provided where significant (p < 0.05) differences were observed between stable and progressive disease at a particular time point. Triangles represent which groups were compared at each time point on the graph, based on colour Comparisons of absolute values in the ratio of neoepitope collagen synthesis to degradation markers in healthy controls and participants with stable and progressive IPF, measured at baseline, 1, 3 and 6 months for: collagen-type 1 (P1NP:C1M), −type 3 (PRO-C3:C3M) and -type 6 (PRO-C6:C6M). Plots represent mean and 95% CI (error bars) adjusted for age, sex, site and smoking status. Disease progression was defined as allcause mortality or ≧10% decline in forced vital capacity at 12 months. The number of evaluable samples available for analysis at each time point are provided in the graph. P values are provided where significant (p < 0.05) differences were observed between stable and progressive disease at a particular time point. Triangles represent which groups were compared at each time point on the graph, based on colour C6M no longer remained significant (Additional file 1: . There was moderate correlation between changes in different neoepitopes (Additional file 1: .
# Discussion
Remodelling of extracellular matrix is pivotal to the development and progression of IPF and results in distinct alterations in collagen turnover. In the current study, we measured synthesis markers of three collagens important in the development of IPF; namely collagen type-1 (P1NP), − 3 (PRO-C3) and − 6 (PRO-C6) [bib_ref] Serum type III procollagen peptide concentration in cryptogenic fibrosing alveolitis and its..., Kirk [/bib_ref] [bib_ref] Increased expression of type VI collagen in lung fibrosis, Specks [/bib_ref]. Given that fibrogenesis in IPF likely reflects an imbalance between synthesis and degradation of collagen, we also measured paired markers of degradation for the same collagens: type-1 (C1M), − 3 (C3M) and − 6 (C6M). For comparison with our prior observations [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref] , we also measured degradation of the matrix molecule biglycan (BGM) and C reactive protein (CRPM). Our data demonstrate that synthesis markers for type-3 (PRO-C3) and-6 (PRO-C6) collagen are elevated in subjects with IPF compared with healthy controls and distinguish individuals with progressive disease, suggesting that these biomarkers may reflect underlying pathophysiology driving disease activity in IPF.
Collagen type-3, a fibrillar collagen, forms a major component of the lung interstitial extracellular matrix and is synthesised by activated fibroblasts [bib_ref] The extracellular matrix at a glance, Frantz [/bib_ref]. Our findings support earlier studies demonstrating that synthesis of collagen type-3 is increased in IPF, with high levels of type-3 pro-collagen peptide reported in serum [bib_ref] Serum type III procollagen peptide concentration in cryptogenic fibrosing alveolitis and its..., Kirk [/bib_ref] and lavage fluid [bib_ref] Serum and bronchoalveolar lavage n-terminal type III procollagen peptides in idiopathic pulmonary..., Low [/bib_ref] [bib_ref] Hyaluronan and type III procollagen peptide concentrations in bronchoalveolar lavage fluid in..., Bjermer [/bib_ref]. In the case of collagen type-3, the synthesis neoepitope, PRO-C3, appears to be a better biomarker of progressive disease than the paired degradation biomarker, C3M. Interestingly, in subjects with more stable disease, the ratio of PRO-C3 to C3M is significantly lower; supporting the idea that synthesis, and increased turnover in general, may be reflective of active remodelling and disease progression.
Alongside collagen type-3, our data suggest that changes in collagen type-6 are also informative of disease behaviour in IPF and may be an important driver of fibrosis. Collagen type-6 filaments are an integral component found at the interface between the basement membrane and interstitial matrix of the lung, and act as an important network for attachment of basement membranes, collagen fibres and cells. Our results support the importance of this collagen in the development of fibrosis. Previous research suggests that collagen type-6 fragments have signalling properties during wound healing and fibrogenesis and may drive poorer outcomes in fibrotic diseases [bib_ref] The good and the bad collagens of fibrosis-their role in signaling and..., Karsdal [/bib_ref]. Compared with other collagens, type-6 collagen has been shown to be a potent inducer of the differentiation and activation of fibroblasts to highly synthetic myofibroblasts; a key pathway leading to the production of collagen and ECM proteins [bib_ref] Type VI collagen induces cardiac myofibroblast differentiation: implications for postinfarction remodeling, Naugle [/bib_ref]. In addition, increased levels of serum collagen type-6 markers have been demonstrated to act as an early predictor of fibrosis in the liver [bib_ref] MMP mediated degradation of type VI collagen is highly associated with liver..., Veidal [/bib_ref]. This evidence, coupled with data from the current study, suggest that collagen type-6 may in itself be a driver of the fibrotic process, which in turn serves as a useful marker of IPF progression.
Intriguingly, whilst increases in collagen type-1 production are associated with TGFβ-activation of IPF fibroblasts [bib_ref] Fibroblasts from idiopathic pulmonary fibrosis and normal lungs differ in growth rate,..., Ramos [/bib_ref] , and have been observed in IPF lung [bib_ref] Extracellular matrix in normal and fibrotic human lungs, Raghu [/bib_ref] , we did not find elevated levels of collagen type 1 synthesis neoeptiopes in the serum of IPF subjects. However, collagen type-1 is also the predominant collagen found in bone and it has been estimated that 90% of theUnadjusted and adjusted (for baseline FVC and DLco) univariate analysis of overall survival in relation to baseline measures of collagen synthesis neoepitopes. Data shown for collagen-type 1 (P1NP), −type 3 (PRO-C3) and -type 6 (PRO-C6), the ratio of collagen synthesis: degradation ratios for collagen-type 1 (P1NP:C1M), −type 3 (PRO-C3: C3M), and -type 6 (PRO-C6: C6M) and lung function measurements (% predicted FVC and % predicted DLCO) in IPF subjects. The Hazard Ratio is based on a two-fold increase in the variable. Data are expressed as mean hazard ratio (HR) with 95% CIs for all thresholds and represent the associated change in mortality risk based on a 2-fold increase in the explanatory value synthesis of this collagen measurable in serum is bone derived [bib_ref] An update on biomarkers of bone turnover and their utility in biomedical..., Leeming [/bib_ref]. This may explain why there was no difference observed in P1NP, the measured marker of collagen type-1 synthesis, between IPF subjects and health controls and why differences were not seen between stable and progressive disease. In contrast to P1NP, the serum degradation marker of collagen type-1, C1M, does not measure bone-derived degradation as the epitope is denatured in bone by osteoclast-released cathepsin K [bib_ref] A novel marker for assessment of liver matrix remodeling: an enzyme-linked immunosorbent..., Leeming [/bib_ref]. This disparity in the pools of collagen type-1 being measured may account for the differences observed between this pair of synthesis and degradation neoepitopes. Consistent with our previous study, baseline levels of C1M, C3M, C6M and CRPM were elevated in subjects with IPF compared with healthy controls. Furthermore, consistent with our prior observations, elevated CRPM distinguished progressive from stable disease from as early as one month after recruitment. In our previous study BGM, C1M, C3M and C6M were also different between individuals with stable and progressive IPF, but these differences only became significant at 6 months. Interestingly, in the current study, the levels of C1M and C6M were significantly higher from 1 month onwards.
Rising levels over 3 months of matrix synthesis markers were not predictive of mortality. This is in distinct contrast with changing levels of the degradation markers CRPM, C1M, C3M and C6M; for each of these markers rising levels over 3 months identified subjects with an increased risk of death. The predictive value of these markers, as shown by the hazard ratios, was consistent with previous results [bib_ref] Longitudinal change in collagen degradation biomarkers in idiopathic pulmonary fibrosis: an analysis..., Jenkins [/bib_ref]. A changing ratio over 3 months of synthesis and degradation markers for type 6 collagen, PRO-C6:C6M, was also predictive of mortality albeit less strongly than C6M alone especially after adjustment for baseline lung function. It is possible that the different characteristics of the synthesis and degradation markers may reflect the specific location of individual collagens within the fibrotic lung and their susceptibility to degradation as fibrosis progresses. Alternatively, as is likely to be the case for P1NP, other anatomical pools of collagen turnover, including bone [bib_ref] An update on biomarkers of bone turnover and their utility in biomedical..., Leeming [/bib_ref] and liver [bib_ref] MMP mediated degradation of type VI collagen is highly associated with liver..., Veidal [/bib_ref] , may dilute the serum signal for individual neoepitopes thus limiting their suitability as markers of IPF severity or disease progression.
The absence of reliable and validated short-term measurements that predict disease progression, survival and response to treatment represent a major obstacle in the management of IPF. Physiological measurements, particularly FVC, remain the primary method for monitoring IPF disease progression in both registration clinical trials and clinical practice [bib_ref] Prognostic factors for idiopathic pulmonary fibrosis: clinical, physiologic, pathologic, and molecular aspects, Lee [/bib_ref] [bib_ref] Idiopathic pulmonary fibrosis: lung function is a clinically meaningful endpoint for phase..., Bois [/bib_ref]. Whilst changes in lung function are likely to remain a gold standard for assessing IPF progression, there are several short-comings of these measurements. Firstly, they are relatively insensitive to subtle changes in disease severity necessitating measurement over longer timeframes (52 weeks in clinical trials) before disease progression can be reliably determined. Although a recent study has demonstrated that daily home spirometry is predictive of disease progression within 3 months, there are some potential limitations with this method, particularly long-term compliance [bib_ref] Daily home spirometry: an effective tool for detecting progression in idiopathic pulmonary..., Russell [/bib_ref]. Secondly, current treatments only slow decline in FVC making determination of treatment response in individuals almost impossible. Serum biomarkers, such as those measured in the current study, have the potential to overcome some of these challenges, by providing a minimally invasive, short-term readout that could supplement FVC measurements to give a more reliable assessment of disease progression and response to anti-fibrotic therapy.
We have shown that neoepitopes reflecting collagen synthesis and degradation in IPF are informative of prognosis, are able to distinguish progressive from non-progressive disease and provide a 3-month read out of change in disease severity. The utility of these observations is currently being tested in a randomised placebo control trial of nintedanib in individuals with IPF and preserved lung function (NCT02788474) [bib_ref] Investigating the effects of nintedanib on biomarkers of extracellular matrix turnover in..., Maher [/bib_ref]. If neo-epitopes can be shown to change with therapy, then these markers may be of use in shortening future clinical trials and for determining response to therapy in the clinic.
A major strength of our study is the prospective longitudinal design, which enabled blood sampling to be standardised and enabled collection of robust outcome data. Furthermore, the PROFILE cohort is representative of real-world patients with IPF and includes many subjects who would not have been enrolled in clinical trials [bib_ref] PROFILEing idiopathic pulmonary fibrosis: rethinking biomarker discovery, Maher [/bib_ref]. As such our findings are more translatable to clinical practice than insights derived from clinical trial cohorts. Furthermore, recruitment of the PROFILE cohort commenced before effective therapy for IPF became available. As such, none of the cohort described in this manuscript were on antifibrotic therapy during the first 12 months of follow up.
The major weakness of this current study is that we undertook analysis of synthesis neoepitopes in a single cohort and did not adhere to a discovery-validation design. However, as PROFILE is the only longitudinal, realworld study that includes treatment-naive patients with multiple biological samples, and without the limitations posed by clinical trial cohorts, that we are aware of, external replication of this study is not currently possible. We feel that this limitation is partially offset by our demonstration that degradation neo-epitopes behave consistently with our previous observations, which were performed in discovery, and validation sample sets. As such we believe that this confirmation of the usefulness of collagen degradation biomarkers provides indirect validation of our findings and supports our conclusions regarding the potential utility of collagen synthesis biomarkers. However, we hope that others will, in future be able to validate the relationships which we have observed between synthesis biomarkers and disease activity and confirm the relevance of ratios of paired collagen synthesis and degradation markers.
# Conclusion
This study demonstrates that collagen type-3 and -6 synthesis biomarkers are elevated in individuals with IPF compared with controls and differ in levels between stable and progressive disease. In addition, this study re-confirms the value of a panel of ECM degradation biomarkers that are predictive of disease progression and survival in IPF. This panel of biomarkers provides informative insights into matrix turnover and underlying pathophysiology driving disease progression, which may assist in the development of precision approaches to IPF treatment. Furthermore, the predictive prognostic value of change in these biomarkers is discernible within 3 months, providing much needed short-term measurements to assist the identification of at-risk IPF patients, which could be used to supplement physiological assessment to allow for better disease management through targeted and earlier therapeutic intervention strategies. Further validation of these findings is required in additional real-world cohorts of individuals with IPF and other fibrosing lung diseases.
## Additional file
Additional file 1: [fig_ref] Table 1: Baseline clinical characteristics in participants with idiopathic pulmonary fibrosis [/fig_ref]. Target and assay performance characteristics for the ELISA assays specific to each neoepitope.. Spearman correlations for baseline serum concentrations of synthesis neoepitopes in control and IPF subjects. . Unadjusted and adjusted (for baseline FVC and DLco) univariate analysis of overall survival in relation to 3-month change in neoepitope concentrations. . Spearman correlations for 3-month change in serum concentrations of synthesis and degradation neoepitopes in IPF subjects. [fig_ref] Figure 1: Baseline comparison of collagen synthesis neoepitope concentrations for collagen-type 1 [/fig_ref]. Baseline comparison of collagen degradation neoepitope (BGM, C1M, C3M, C6M and CRPM) concentrations in healthy controls (n = 50) and participants with idiopathic pulmonary fibrosis (n = 145). Plots represent mean and 95% CI (error bars) adjusted for gender. . Comparison of neoepitope concentrations in healthy controls (green) and participants with stable (blue) and progressive (red) idiopathic pulmonary fibrosis at baseline and subsequently at 1-, 3-and 6months. . Comparison of synthesis neoepitope concentrations adjusted for baseline FVC and DLco %predicted in healthy controls (green) and participants with stable (blue) and progressive (red) idiopathic pulmonary fibrosis at baseline and subsequently at 1-, 3-and 6-months. [fig_ref] Figure 4: Effect of 3 month change in neoepitope concentrations on overall survival [/fig_ref]. Comparison of degradation neoepitope concentrations adjusted for baseline FVC and DLco %predicted in healthy controls (green) and participants with stable (blue) and progressive (red) idiopathic pulmonary fibrosis at baseline and subsequently at 1-, 3-and 6-months. . Comparison of neoepitope concentrations adjusted for baseline FVC and DLco %predicted in healthy controls (green) and participants with stable (blue) and progressive (red) idiopathic pulmonary fibrosis at baseline and subsequently at 1-, 3-and 6-months. (DOCX 1380 kb)
## Availability of data and materials
Data is available on reasonable written request to the PROFILE study steering committee.
## Ethics approval and consent to participate
The PROFILE study is registered on clinicaltrials.gov (NCT01134822 and NCT01110694). The study was approved by the Royal Free Hospital Research Ethics Committee Ref 10/H0720/12.
## Consent for publication
All subjects provided signed informed consent including for publication of all results arising from the PROFILE study.
Competing interests WAF, JKS, EO, CBN, ARK are employees and shareholders of GSK. ST, BN, YK and AMD are working on behalf of GSK. TMM has, via his institution, received industry-academic funding from GlaxoSmithKline R&D and UCB and has received consultancy or speaker fees from Apellis, Astra Zeneca, Bayer, Biogen Idec, Boehringer Ingelheim, Galapagos, GlaxoSmithKline R&D, Indalo, Pliant, ProMetic, Roche, Samumed and UCB. RGJ has received industryacademic funding from GlaxoSmithKline R&D, Novartis and Biogen Idec and has received consultancy or speaker fees from InterMune, MedImmune, Pulmatrix, Boehringer Ingelheim, and GlaxoSmithKline R&D. MAK and DJL are employees and shareholders of Nordic Biosciences and hold patents relating to neoepitopes.
[fig] Figure 1: Baseline comparison of collagen synthesis neoepitope concentrations for collagen-type 1 (P1NP), −type 3 (PRO-C3) and -type 6 (PRO-C6) in healthy controls and participants with idiopathic pulmonary fibrosis. Plots represent mean and 95% CI (error bars) adjusted for age and gender [/fig]
[fig] Figure 4: Effect of 3 month change in neoepitope concentrations on overall survival. Subjects with IPF were dichotomised into HIGH or LOW groups, to determine the effect of 3-month change in neoepitope concentration on overall survival, where: HIGH = rising (0 > ng/mL per month) neoepitope concentrations from baseline to month 3, >or LOW = stable or falling (≤0 ng/mL per month) neoepitope concentration from baseline to month 3. For ratio values, the slope values were derived following the creation of the ratio variables and subjects were dichotomised into HIGH or LOW slope groups. Hazard ratio represents the mortality risk in participants with rising neoepitope concentrations relative to those with stable or falling concentrations.Data shown in: top panel -collagen synthesis markers P1NP, PRO-C3, PRO-C6; mid panel -ratio of collagen synthesis/degradation markers: P1NP:C1M, PRO-C3:C3M and PRO-C6:C6M; bottom panel -BGM, C1M, C3M, C6M, CRPM). The number of deaths in each group is shown for each biomarker. P values presented relate to the comparison of the Kaplan-Meir survival curves for each dichotomised measure [/fig]
[fig] Fibrosis: Discovery Performance Unit, GlaxoSmithKline R&D, GlaxoSmithKline Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, UK. [/fig]
[table] Table 1: Baseline clinical characteristics in participants with idiopathic pulmonary fibrosis. Data are shown for all subjects with IPF and then for the population separated in to those with progressive or stable disease at 12 months. Data are mean (SD) or n (%), unless otherwise stated. BMI = body-mass index. FVC = forced vital capacity. DLCO = diffusion capacity for carbon monoxide. CPI = composite physiological index, *Number of subjects analysed due to missing data [/table]
|
Safety assessment of genetically modified plants with deliberately altered composition
# Introduction
The food and feed risk assessment strategy for genetically modified (GM) crops in Europe, as applied by the European Food Safety Authority (EFSA) GMO Panel, compares GM plants and their derived food and feed with a conventional non-GM comparator, the comparator being a plant with a history of safe use as food (the principle of substantial equivalence). The development of GM crop varieties that have received substantial, deliberate modifications to their composition, metabolism or physiology raises the question of whether the comparative approach as applied by EFSA would be fully applicable in some cases. Indeed, current EFSA guidance on the risk assessment of food and feed from GM plants states that:
"Where no comparator can be identified, a comparative risk assessment cannot be made and a comprehensive safety and nutritional assessment of the GM plant and derived food and feed itself should be carried out. This would, for instance, be the case where the food and/or feed derived from a GM plant is not closely related to a food and/or a feed with a history of safe use, or where a specific trait or specific traits are introduced with the intention of changing significantly the composition of the plant" .
Genetically modified varieties with deliberately altered composition are sometimes called 'second generation', or described as having 'output' traits, distinguishing them from 'first-generation' varieties with 'input' traits such as herbicide tolerance or insect resistance. The 'first' and 'second generation' terms are not particularly useful, partly because the first GM varieties to be launched back in 1994 were tomatoes with improved shelf life (an output trait) and partly because GM plants may have output traits combined with well-established input traits. For the purpose of this article, therefore, the term 'novel traits' is used. This term also encapsulates the fact that regulatory authorities are much more familiar with GM plants in which the composition of the crop product is indistinguishable from that of its non-GM comparator, such as those carrying herbicide tolerance or insect resistance traits.
## Examples of gm crops with novel traits
Slow-ripening GM tomatoes (reviewed by could be seen as providing a useful precedent for the safety assessment of other GM crop varieties with altered composition. However, they had been developed, launched, marketed and eventually withdrawn years before the current European regulatory system, set out in European Union (EU) directive, GM Food and Feed Regulation (EC) No. 1829/2003, was in place. Calgene's 'Flavr Savr' variety, for example, which was marketed in the USA in 1994, was not a commercial success and was soon withdrawn. Tomato paste made from slow-ripening tomatoes was marketed in Europe by Syngenta between 1996 and 1999, but it too was withdrawn at the end of the 1990s.
A more recent example of a GM crop with altered composition and one that has been assessed under the current EFSA risk assessment process is the EH92-527-1 'Amflora' potato variety. 'Amflora' was developed by BASF and contained starch composed almost entirely of amylopectin as a result of reduced activity of a granule-bound starch synthase [bib_ref] Inhibition of the expression of the gene for granule-bound starch synthase in..., Visser [/bib_ref]. The aim was to produce starch as an industrial raw material; however, the application for permission to market the variety took 13 years to pass through the EU approval process. It was finally cleared for cultivation in 2010, but BASF announced in 2012 that it was withdrawing from plant biotechnology in Europe altogether, and although 'Amflora' is licensed for cultivation, it is currently unavailable.
This outcome highlights the need for robust, workable processes for the safety assessment of GM crops with novel traits. There is some urgency to this because other varieties have either been approved for use elsewhere in the world already, or are believed to be in the final stages of development. One of these is the 'Mavera' maize variety produced by Renessen, a joint venture between Cargill and Monsanto, which has been engineered to synthesize more lysine by introducing a gene from a bacterium, Corynebacterium glutamicum, encoding a lysineinsensitive dihydrodipicolinate synthase, . This variety is aimed at the US bioethanol production market, for which high-protein animal feed is an important co-product. It also contains a 'triple stack' of three input traits (two insect resistance and one herbicide tolerance), an example of a novel trait being launched combined with established input traits. Syngenta is also targeting the huge US bioethanol production market by developing a GM maize variety that it claims gives a better yield of ethanol . The variety contains a Thermococcales spp. gene, AMY797E, encoding a highly thermostable a-amylase and was deregulated by the US authorities in early 2011.
There are also examples of plant oils being modified by GM. Plant oils contain a range of fatty acids with different chain lengths and degrees of saturation, including some that are polyunsaturated. Polyunsaturated fatty acids (PUFAs) may be considered healthier than saturated fats, but they are prone to oxidation during cooking, giving rise to a rancid, 'off' flavour and odour. Food processors can avoid PUFA oxidation by chemical hydrogenation of the double bonds, but this can cause the formation of trans fatty acids, which have a similar effect on serum cholesterol levels as saturated fats. Soybean oil is normally rich in linoleic acid, an 18-carbon chain PUFA, and PBI, a subsidiary of DuPont, has produced a genetically modified variety, 'Plenish', that accumulates high levels of oleic acid instead of linoleic acid. Hydrogenation, with its risk of trans fatty acid formation, is therefore not required for the GM oil.
Other GM plants with more ambitious changes to their oil profile are in development. The longest PUFAs made by higher plants have an 18-carbon chain (C18), but considerably longer fatty acids are present in fish oils, including nutritionally important omega-3 LC-PUFAs such as eicosapentaenoic acid (EPA) (C20) and docosahexaenoic acid (DHA) (C22). EPA and DHA are in fact made by marine algae, not fish, but they accumulate through the marine food chain and are acquired in the human diet from marine fish or farmed fish that have been fed marine fishmeal. The development of GM plants producing EPA and DHA could provide an alternative source of these PUFAs, and the biosynthetic pathway has been successfully engineered into the seed of Camelina sativa to produce oil with up to 12% EPA and 14% DHA, levels very close to those in fish oil.
Increased vitamin content is another obvious nutritional enhancement, with elevated levels of vitamin A in particular being a target in several subsistence crops. Indeed, Golden Rice, a GM line of rice synthesizing the vitamin A precursor, b-carotene, in the grain (unlike its non-GM comparator) was developed in the late 1990s . Similar technology is being used in cassava and banana, both of which are staple crops in Africa.
While these crops are designed to alleviate deficiencies in diets in developing countries, governments may apply for consent to market in Europe in case of inadvertent mixing of the GM varieties with produce destined for export.
Genetic modification is also being used to reduce the level of undesirable substances in food. Asparagine synthetase gene expression in potato, for example, has been reduced by RNA interference to decrease the risk of formation of acrylamide, a probably carcinogenic contaminant that forms from a reaction between free asparagine and reducing sugars during hightemperature cooking and processing . Other strategies to reduce acrylamide formation target enzymes involved in starch breakdown and in 2013 Simplot Plant Sciences applied to the US authorities for deregulation of low acrylamide GM potato genotypes incorporating one or both of these strategies combined with decreased polyphenol oxidase activity to reduce bruising.
Compositional changes could also result from GM approaches to improve crop tolerance of abiotic stresses, such as drought and salt stress. Drought and salt both impose an osmotic stress and plants use changes in metabolic processes as part of their response, for example by interconverting starch or fructan with simple sugars and synthesizing compatible solutes such as proline and glycine betaine. Strategies aimed at manipulating these responses could result in compositional changes. Lastly, there is biopharming, in which GM plants are used to produce pharmaceuticals, vaccines, antibodies, enzymes or other high-value, nonfood products [reviewed in a previous ].
## Adapting the eu's risk assessment approaches to cover novel gm traits
The need to adapt the guidance for the risk assessment of GM food and feed to deal with applications for the approval of novel GM traits has been recognized by EFSA, and a report was commissioned in 2013 [bib_ref] Review of the strategies for the comprehensive food and feed safety and..., Tompkins [/bib_ref]. The report was based on a review of the scientific literature on the topic and the approaches currently being taken by international agencies.
The intention at the start of the review process was to include 'comprehensive' risk assessment strategies that did not rely on the comparative approach alone to assess safety. However, it became clear that comparative approaches based on the concept of substantial equivalence (as defined by the OECD in 1993) were the basis of all current risk assessment strategies across the world. In other words, a comprehensive assessment of GM plants and derived food and feed without the use of a comparator is not an approach that is used by international risk assessment bodies or proposed by the scientific literature. Indeed, novel GM plants that have already been approved in the USA and other countries outside the EU have been risk assessed using the same framework as 'first-generation' input traits, with no specific modifications to risk assessment criteria, although there are some differences in the way that groups define substantial equivalence in the literature.
A key area of the EFSA risk assessment process that may require revision to make it applicable to novel GM plants is the first step of hazard identification. This is largely because of the difficulty of comparing plants which may have major changes in composition with a conventional comparator. For example, EFSA's guidelines for field trials currently require that a GM variety and its conventional comparator be grown alongside each other to provide material for analysis. This may not be appropriate if, for example, the GM variety has improved stress tolerance because the conventional comparator may not be able to survive in some of the environments in which the GM variety could be expected to be cultivated. In this case, a GM plant may need to be compared with the range of natural variation of control plants grown under 'normal' conditions, but with the GM plant grown under the same conditions and in a more extreme environment. A second approach may employ a field experiment to test the response of the GM crop and its comparator to increasing degrees of stress, in the background of a 'normal' (i.e. nonstress) environment, in a fully factorial design. The report recommended that further guidance on field trials be developed, including the specific circumstances in which field trial design may need to be amended.
New approaches will also need to be developed to assess novel GM crops containing constituents that are not normally found in that species, with GM oilseeds containing LC-PUFAs that are otherwise found only in fish oils being an obvious example. The concept of 'history of safe use', which is already a valuable tool in risk assessment, would be particularly useful here: if an applicant could demonstrate that the novel constituent in the GM plant had a 'history of safe use' in food or feed, even if it had previously come from a different source, this would give confidence in its safety. However, while this concept is used by many international risk assessment bodies, it is seldom given a specific definition; in particular, there is no standard requirement for time period or level of consumption. In the report, the definition used by Health Canada was supported, but even this uses vague terms:
"the safety of the expressed trait in food or feed is confirmed from experience of use and continued use in the normal diet of a large part of the population of a country or in farmed/domestic animals for a number of generations, consumed at levels foreseen to be similar to its use in the GM plant".
It may also be more useful to compare a food product derived from plants carrying the novel trait to another food product that is more similar to it than its conventional counterpart. In the case of GM oils containing LC-PUFAs, this could be fish oil. However, GM plants that have been engineered to synthesize LC-PUFAs do not make true fish oils; they make an oil with a unique combination of some of the fatty acids that would normally be present in that plant species and some additional LC-PUFAs. So, although informative, such a comparison could only be part of the process.
Another consideration would be whether the novel GM crop would change consumption patterns and therefore exposure. This could be difficult to predict because it would be affected by consumer preferences, which could change over time, and the trend by food manufacturers to include functional ingredients in multiple products. It would be particularly important for food constituents for which there is an upper limit to the recommended dietary intake. It would also be important that consumers were given enough information about the difference between the GM and non-GM products to make informed choices over consumption. The results of exposure assessments could be validated by postmarket monitoring (PMM). However, PMM is an area that has received little discussion in the literature, and guidance would have to be developed on the sorts of scenarios where PMM should be performed, what should be measured, the methodologies for carrying it out accurately enough to provide useful data and how it would be financed.
## Concluding remarks
Any system put in place for risk assessing and licensing novel GM traits must work efficiently. This is important not only for the development of these traits (some of which have clear consumer benefits) in Europe, but also because, as the world's most lucrative market for food commodities, Europe's regulatory processes affect the development of crop biotechnology well beyond its borders. The system must also have the confidence of consumers, and the report recommended that public consultation be viewed as an important element of the risk assessment process. In particular, that dossiers of nonconfidential information be published and comments invited before consideration by the GMO Panel and that comments be sought from the public on the panel's initial opinion. This would be similar to the process adopted by the UK's Advisory Committee on Novel Foods and Processes (ACNFP) and in line with the transparency favoured by the United States Department of Agriculture Plant Health Inspection Service. The EU's own Novel Foods Regulations (258/97) are currently being revised, and it will be important to ensure that processes for assessing novel GM traits are aligned with those for assessing other novel foods. |
Obsessive-compulsive disorder in children and adolescents: epidemiology, diagnosis and management
Obsessive-compulsive disorder (OCD) can be found in about 4% of the general population and is characterized by various compulsions and obsessions that interfere with the person's quality of life from a mild to severe degree. The following discussion reflects on current concepts in this condition, including its epidemiology and etiologic underpinnings (behavioral, neurological, immunological, gastroenterological, as well as genetic). The interplay of PANS and PANDAS are included in this review. In addition, the core concepts of OCD diagnosis, differential diagnosis, and co-morbidities are considered. It is stressed that the quality of life for persons with pediatric OCD as well as for family members can be quite limited and challenged. Thus, principles of management are presented as a guide to improve the quality of life for these persons as much as possible.
# Introduction
The presence of obsessions and compulsions with or without the obsessive-compulsive disorder (OCD) has been observed and suggested in human beings for eons of time with considerable impact on human history as well as world religions ("religious scrupulosity") [bib_ref] Obsessions from the past: a study of the chapter on "blasphemous thoughts"..., Avgoustidis [/bib_ref] [bib_ref] The doubting disease: religious scrupulosity and obsessive-compulsive disorder in historical context, Cefalu [/bib_ref] [bib_ref] Did Evagrius Ponticus (AD 346-99) have obsessivecompulsive disorder, Hill [/bib_ref] [bib_ref] Scrupulosity disorder: an overview and introductory analysis, Miller [/bib_ref]. Scrupulosity focusing on morality was noted in the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM) IV-TR that identified it as a symptom of obsessive-compulsive personality disorder [bib_ref] Scrupulosity in patients with obsessive-compulsive disorder: relationship to clinical and cognitive phenomena, Nelson [/bib_ref]. Medical attention on obsessive-compulsive traits and disorder in children and adolescents-apart from that found in adultsgradually began in the 20 th century and has continued into the 21 st century as well [bib_ref] Problems in school attendance; school phobia and related conditions, Glaser [/bib_ref] [bib_ref] School phobia in older children and adolescents: diagnostic implications and prognosis, Milman [/bib_ref] [bib_ref] Childhood-onset obsessive compulsive disorder, Swedo [/bib_ref]. This discussion considers the current understanding of pediatric OCD with special focus on its epidemiology, etiology, diagnosis, differential diagnosis, co-morbidities, and management [bib_ref] Anxiety disorders, Nazeer [/bib_ref] [bib_ref] Anxiety disorders in children and adolescents, Patel [/bib_ref].
Overview OCD is currently identified by the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, 5 th edition (DSM-5) as a condition with obsessions (repeating urges or ideas that are irrational, intrusive) combined with compulsions (behaviors) interfering with the person's quality of life and often in association with other neurodevelopmental or psychiatric
## Review article
Obsessive-compulsive disorder in children and adolescents: epidemiology, diagnosis and management conditions [bib_ref] Obsessive-Compulsive Disorder, Drubach [/bib_ref] [bib_ref] Phenomenology of OCD: Lessons from a large multicenter study and implications for..., Shavitt [/bib_ref]. As research on OCD continues, the question can be raised whether the core aspect is the person's compulsions that lead to various obsessions, or do the obsessions stimulate the compulsions in a way to minimize anxiety caused by irrational fears [bib_ref] Which Is the Driver, the Obsessions or the Compulsions, in OCD?, Gillan [/bib_ref]. Other symptomatic features which may arise include impulsivity, rage episodes, aggressive behavior, sexual obsessions, religious rituals (scrupulosity), and eating obsessions in eating disorders [bib_ref] Sexual obsessions in pediatric obsessive-compulsive disorder: clinical characteristics and treatment outcomes, De La Cruz [/bib_ref] [bib_ref] The cost and impact of compulsivity: A research perspective, Hollander [/bib_ref] [bib_ref] Obsessivecompulsive symptomatology in female adolescent inpatients with restrictive compared with binge-purge eating..., Lewis [/bib_ref] [bib_ref] Distinguishing Fear Versus Distress Symptomatology in Pediatric OCD, Rozenman [/bib_ref] [bib_ref] Comparing OCDaffected youth with and without religious symptoms: Clinical profiles and treatment..., Wu [/bib_ref] [bib_ref] Rage outbursts in adult obsessivecompulsive disorder, Wu [/bib_ref] [bib_ref] Persistence of impulsivity in pediatric and adolescent patients with obsessive-compulsive disorder, Yamamuro [/bib_ref]. Various fears may arise about hygiene, contamination, cleanliness, the need for continuous checking and establishment of constant symmetry in their lives [bib_ref] School phobia in older children and adolescents: diagnostic implications and prognosis, Milman [/bib_ref] [bib_ref] Childhood-onset obsessive compulsive disorder, Swedo [/bib_ref] [bib_ref] Anxiety disorders, Nazeer [/bib_ref] [bib_ref] Anxiety disorders in children and adolescents, Patel [/bib_ref].
OCD is a chronic condition that impacts humans from early childhood through adulthood with a continuum between obsessive-compulsive traits to overt OCD [bib_ref] Obsessivecompulsive symptoms are associated with psychiatric comorbidities, behavioral and clinical problems: a..., Alvarenga [/bib_ref] [bib_ref] Characteristics of Young Children with Obsessive-Compulsive Disorder: Baseline Features from the POTS..., Skriner [/bib_ref]. OCD is often under-diagnosed and is often found with minimal insight in those who have OCD as well as considerable family accommodation that they enjoy from their families [bib_ref] The importance of insight, avoidance behavior, not-just-right perception and personality traits in..., Nissen [/bib_ref] [bib_ref] Symptom Insight in Pediatric Obsessive-Compulsive Disorder: Outcomes of an International Aggregated Cross-Sectional..., Selles [/bib_ref]. Older parental age increases risks for some neurodevelopmental disorders like OCD and autism spectrum disorder [bib_ref] Parental Age and Differential Estimates of Risk for Neuropsychiatric Disorders: Findings From..., Janecka [/bib_ref].
## Epidemiology
Studies over the past decades note that OCD is variably found in 1% to 4% of persons (children, adolescents, adults) throughout the world often with a seriously negative impact on their lives; approximately 4 in 10 with OCD develop it as a chronic condition, and many seek to conceal their OCD from others [bib_ref] Pediatric obsessive-compulsive disorder: a guide to assessment and treatment, Keeley [/bib_ref] [bib_ref] Epidemiology of obsessive-compulsive disorder, Kiejna [/bib_ref] [bib_ref] The prevalence of diagnosed obsessive compulsive disorder and associated comorbidities: A population-based..., Osland [/bib_ref] [bib_ref] Pharmacotherapy of Obsessive Compulsive Disorder, Sasson [/bib_ref] [bib_ref] Challenges in the diagnosis and treatment of pediatric obsessive-compulsive disorder, Westwell-Roper [/bib_ref] [bib_ref] Obsessive-compulsive disorder, Veale [/bib_ref]. The World Health Organization (WHO) places OCD in the top ten of the most handicapping disorders of humans [bib_ref] Challenges in the diagnosis and treatment of pediatric obsessive-compulsive disorder, Westwell-Roper [/bib_ref]. OCD can begin in childhood, and approximately 8 in 10 of those developing OCD initiate it by 18 years of age [bib_ref] Obsessive-compulsive disorder, Veale [/bib_ref]. The research seeks to find various OCD subtypes based on concepts of etiology; there is, for example, OCD in children called Early Onset OCD that reflects a neurodevelopmental perspective [bib_ref] Diagnostic Issues in Early-Onset Obsessive-Compulsive Disorder and their Treatment Implications, Burchi [/bib_ref].
## Etiology
Research reveals a variety of underlying etiologic factors in the development of OCD in humans of various ages. These include inter-connecting issues based on behavioral, neurological, infectious, and immunological underpinnings [bib_ref] Childhood-onset obsessive compulsive disorder, Swedo [/bib_ref] [bib_ref] Anxiety disorders, Nazeer [/bib_ref] [bib_ref] Anxiety disorders in children and adolescents, Patel [/bib_ref] [bib_ref] Obsessive-Compulsive Disorder, Drubach [/bib_ref].
## Behavioral etiology
Various psychological and neuropsychological models of etiology describe a complex process in which the OCD person learns or is driven to determine what are perceived as errors in their thinking patterns in order to cope with deep-seated anxiety [bib_ref] Error-processing abnormalities in pediatric anxiety and obsessive compulsive disorders, Fitzgerald [/bib_ref]. The role of over-protective parenting styles with rigid rules, a person's perspective that their actions are never acceptable to others and various other adverse childhood experiences continue to be debated, especially in light of 20 th and 21 st century methods of studying the central nervous system [bib_ref] A Preliminary Investigation of Pathways to Inflated Responsibility Beliefs in Children with..., Collins [/bib_ref] [bib_ref] Brain white matter integrity and association with age at onset in pediatric..., Rosso [/bib_ref].
## Neurological etiology
Changes leading to neurological dysfunctions are noted in various studies of the central nervous system leading to OCD and its identification as a complex, heterogeneous condition. These brain studies included brain scans, magnetic resonance imaging (MRI), functional magnetic resonance imaging (fMRI), whole-brain voxel-based morphometry (VBM), and diffusion tensor imaging (DTI) [bib_ref] Obsessivecompulsive disorder is a heterogeneous disorder: evidence from diffusion tensor imaging and..., Glahn [/bib_ref]. These changes are noted in the corticostriato-thalamocortical circuits of the brain and brain white matter [bib_ref] Changes in Gray Matter Volume and White Matter Microstructure in Adolescents with..., Zarei [/bib_ref]. This frontal-striatal-thalamic model of neurological dysfunction has been a center point of brain imaging OCD research for the past several decades showing diffuse changes including frontoparietal-limbic dysfunction, orbitofrontal dysfunction, brain volume reduction, basal ganglia dysfunction, and error-related brain activity [bib_ref] Global Resting-State Functional Magnetic Resonance Imaging Analysis Identifies Frontal Cortex, Striatal, and..., Anticevic [/bib_ref] [bib_ref] Use of Computer Vision Tools to Identify Behavioral Markers of Pediatric Obsessive-Compulsive..., Bernstein [/bib_ref] [bib_ref] Meta-analytic investigations of structural grey matter, executive domain-related functional activations, and white..., Eng [/bib_ref] [bib_ref] Development of Posterior Medial Frontal Cortex Function in Pediatric Obsessive-Compulsive Disorder, Fitzgerald [/bib_ref] [bib_ref] Frontoparietal areas link impairments of large-scale intrinsic brain networks with aberrant fronto-striatal..., Gürsel [/bib_ref] [bib_ref] Withdrawn/Depressed Behaviors and Error-Related Brain Activity in Youth With Obsessive-Compulsive Disorder, Hanna [/bib_ref] [bib_ref] Alterations in restingstate local functional connectivity in obsessive-compulsive disorder, Hao [/bib_ref] [bib_ref] Altered functional network architecture in orbitofronto-striato-thalamic circuit of unmedicated patients with obsessive-compulsive..., Jung [/bib_ref] [bib_ref] White matter structural alterations in pediatric obsessive-compulsive disorder: Relation to symptom dimensions, Lázaro [/bib_ref] [bib_ref] Brain Imaging in Pediatric Obsessive-Compulsive Disorder, Macmaster [/bib_ref] [bib_ref] Cerebral white matter volume changes in patients with obsessive-compulsive disorder: Voxel-based morphometry, Park [/bib_ref] [bib_ref] Widespread structural brain changes in OCD: A systematic review of voxel-based morphometry..., Piras [/bib_ref] [bib_ref] Cortico-striato-thalamocortical circuit abnormalities in obsessive-compulsive disorder: A voxel-based morphometric and fMRI study..., Tang [/bib_ref] [bib_ref] Intrinsic brain abnormalities in drug-naive patients with obsessive-compulsive disorder: A resting-state functional..., Yang [/bib_ref] [bib_ref] Alterations of gray and white matter morphology in obsessive compulsive disorder, Gonçalves [/bib_ref].
Studies also reveal alterations in the hippocampus (60), and serotonergic-dopaminergic pathways as well as the glutamatergic system [bib_ref] Obsessive-Compulsive Disorder, Goodman [/bib_ref] [bib_ref] Obsessive compulsive disorder and the glutamatergic system, Kariuki-Nyuthe [/bib_ref]. Brain changes (i.e., asymmetry or thickness of subcortical structures, parietal cortex thickness, others) can be influenced by various factors such as genetics, age, anti-OCD medication effects and others [bib_ref] Working Group E-O. Neuroimaging of Cortical Brain Alterations in Adult and Pediatric..., Boedhoe [/bib_ref] [bib_ref] Mapping Cortical and Subcortical Asymmetry in Obsessive-Compulsive Disorder: Findings From the ENIGMA..., Kong [/bib_ref]. Obsessive-compulsive behavior has been anecdotally reported in some taking the anti-epileptic drug, levetiracetam, possibly because of effects on the glutamatergic system [bib_ref] Obsessive-Compulsive Behavior Induced by Levetiracetam, Fujikawa [/bib_ref].
## Immunological etiology
Immunological mechanisms are postulated as being involved in pediatric OCD as noted by research linking this condition with pro-inflammatory cytokines such as TNFalpha and IL-12 [bib_ref] Cytokine, chemokine and BDNF levels in medication-free pediatric patients with obsessive-compulsive disorder, Sivri [/bib_ref]. Inflammatory damage to monocytes can be noted in persons with OCD and modified by effects of OCD-medication as seen with measurements of proinflammatory cytokines [bib_ref] Prevalence of Acute-Onset Subtypes in Pediatric Obsessive-Compulsive Disorder, Jaspers-Fayer [/bib_ref]. The role of the immune system and infection continues to be studied as seen with a report of a pediatric person developing OCD after developing acute disseminated encephalomyelitis (ADEM) which targeted the cerebral white matter [bib_ref] A Case Report of Obsessive-Compulsive Disorder Following Acute Disseminated Encephalomyelitis, Muir [/bib_ref].
## Pandas/pans
Further intriguing links of pediatric OCD with immunoinfectious underpinnings is found in the association of OCD symptomatology (1-10%) with the proposed [1998] PANDAS (pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections) with positive antistreptolysin O as well as anti-DNase B antibody titers and PANS (pediatric acute-onset neuropsychiatric syndrome) [bib_ref] Prevalence of Acute-Onset Subtypes in Pediatric Obsessive-Compulsive Disorder, Jaspers-Fayer [/bib_ref] [bib_ref] Antineuronal Antibodies in a Heterogeneous Group of Youth and Young Adults with..., Cox [/bib_ref] [bib_ref] Inflammation in Tic Disorders and Obsessive-Compulsive Disorder: Are PANS and PANDAS a..., Gilbert [/bib_ref] [bib_ref] Establishing a Pediatric Acute-Onset Neuropsychiatric Syndrome Clinic: Baseline Clinical Features of the..., Gromark [/bib_ref] [bib_ref] Individualized Immunological Data for Precise Classification of OCD Patients, Lamothe [/bib_ref] [bib_ref] Clinical-Serological Characterization and Treatment Outcome of a Large Cohort of Italian Children..., Lepri [/bib_ref] [bib_ref] Immune system and obsessive-compulsive disorder, Marazziti [/bib_ref] [bib_ref] Evaluation of autoimmune phenomena in patients with pediatric autoimmune neuropsychiatric disorders associated..., Stagi [/bib_ref] [bib_ref] PANDAS/PANS in childhood: Controversies and evidence, Wilbur [/bib_ref]. Though PANDAS remained a proposed syndrome and was not listed as a distinct entity in the 2013 DSM-V, proposed links with OCD remain an ongoing research efforts [bib_ref] An expert opinion on PANDAS/PANS: highlights and controversies, Chiarello [/bib_ref].
## Gastrointestinal (gi) microbiome etiology
Another intriguing area of research in the etio-pathogenesis of psychiatric disease is the microbiome-gut-brain axis in which gastrointestinal (GI) microbes influence the brain to affect human behavior via GI inflammation and immunological effects [bib_ref] An expert opinion on PANDAS/PANS: highlights and controversies, Chiarello [/bib_ref]. The presence of various GI microbes, as influenced by various internal and external factors (stressors), can induce or influence such conditions as depression, OCD, PANS, PANDAS and others [bib_ref] An expert opinion on PANDAS/PANS: highlights and controversies, Chiarello [/bib_ref] [bib_ref] Obsessive-compulsive disorder and gut microbiota dysregulation, Rees [/bib_ref] [bib_ref] A gut (microbiome) feeling about the brain, Sherwin [/bib_ref]. Bacteria under study include group A beta-hemolytic streptococcus, and Bacteroidetes (i.e., Bacteroides, Odoribacter, and Oscillospira) [bib_ref] Gut Microbiota Profiling and Gut-Brain Crosstalk in Children Affected by Pediatric Acute-Onset..., Quagliariello [/bib_ref] [bib_ref] What's bugging the gut in ocd?" a review of the gut microbiome..., Turna [/bib_ref] [bib_ref] The gut microbiome in psychiatry: A primer for clinicians, Van Ameringen [/bib_ref].
## Genetic etiology
If a person has OCD, this condition is increased two-times in first-degree relatives; this is increased ten-times if the OCD has a childhood-onset. The concordance rate noted for monozygotic twins is 0.57 versus 0.22 described in dizygotic twins. The genetic contribution to OCD is part of the overall research looking at genetic correlations with a wide variety of psychiatric and immunological disorders [bib_ref] Genetic correlations among psychiatric and immune-related phenotypes based on genome-wide association data, Tylee [/bib_ref]. Various genes are under study for OCD and one example is the SLC6A4 gene (serotonin transporter gene) that has been noted to influence OCD transmission via both genetic and epigenetic effects; epigenetic influences include DNA methylation (84). Other serotonin system genes under study for their role in OCD, in addition to SLC6A4, include HTR2A, HTR1B, and HTR2C [bib_ref] A review of the role of serotonin system genes in obsessive-compulsive disorder, Sinopoli [/bib_ref].
Links with OCD have been found, for example, with serotonin transporter polymorphism 5-HTTLPR and HTR2A polymorphism rs6311 (or rs6313) (85). Serotonin system gene variants have been linked with changes in brain volumes in pediatric persons with OCD [bib_ref] Serotonin system gene variants and regional brain volume differences in pediatric OCD, Sinopoli [/bib_ref]. Research is occurring on SLC1A1 that encodes the glutamate transporter (epithelial and neuronal) and potential association with OCD [bib_ref] A screen of SLC1A1 for OCD-related alleles, Wang [/bib_ref]. Research on genetic links in OCD also looks at finding copy number variations (CNVs), suggesting that CNVs can be part of the underlying etiologic mechanism for OCD [bib_ref] Uncovering obsessive-compulsive disorder risk genes in a pediatric cohort by high-resolution analysis..., Gazzellone [/bib_ref].
The complexities of genetic contributions to OCD and other neuropsychiatric/ neurodevelopmental disorders continue to be unraveled by research. One report noted deletions at two chromosomes (18q22.1 and 13q12.3-q13.1) in a person with OCD along with Tourette syndrome and dysmorphism [bib_ref] Tourette syndrome, obsessive compulsive behavior, and dysmorphic features in a patient with..., Bhatti [/bib_ref]. Given the enormous entanglements of genetic influence on OCD and the difficulties of OCD itself, genetic counseling is recommended for families (i.e., parents, others) to help them deal with OCD in a family member (i.e., a child or others) [bib_ref] Obsessive-Compulsive Disorder: The Process of Parental Adaptation and Implications for Genetic Counseling, Andrighetti [/bib_ref].
## Diagnosis
A diagnosis of OCD is based on a thorough medical history (including family history) along with a physical examination and diagnostic interviews with psychological assessments; these include such measures as the Children's Yale-Brown Obsessive-Compulsive Scale (CY-BOCS), the Children's Yale-Brown Obsessive-Compulsive Scale, Second Edition (CY-BOCS-II), the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS), the Y-BOCS self-report instrument (Y-BOCS-SR), the Yale-Brown Obsessive-Compulsive Scale Check List (Y-BOCS CL), the Anxiety Disorders Interview Schedule for DSM-IV, the Family Accommodation Scale for Obsessive-Compulsive Disorder-Patient Version (FAS-PV), Short Mood and Feelings Questionnaire, the Screen for Child Anxiety-Related Emotional Disorders, the Child Behavior Checklist-Obsessive-Compulsive Subscale (CBCL-OCS), the Child Obsessive-Compulsive Externalizing/Internalizing Scale (COCEIS) and other screening measures [bib_ref] Latent class analysis of obsessive-compulsive symptoms in a clinical sample, Atli [/bib_ref] [bib_ref] The Yale-Brown Obsessive Compulsive Scale: factor analysis, construct validity, and suggestions for..., Deacon [/bib_ref] [bib_ref] A 9-Year-Old Girl With Persistent Obsessive and Compulsive Behaviors in a Primary..., Gist [/bib_ref] [bib_ref] Assessment of obsessive-compulsive disorder: A review, Grabill [/bib_ref] [bib_ref] Parents' Perceptions of Internalizing and Externalizing Features in Childhood OCD, Guzick [/bib_ref] [bib_ref] Obsessive-compulsive disorder: tools for recognizing its many expressions, Merlo [/bib_ref] [bib_ref] Assessment of Pediatric Obsessive-Compulsive Disorder: A Critical Review of Current Methodology, Merlo [/bib_ref] [bib_ref] Obsessivecompulsive disorder in children and adolescents, Rotberg [/bib_ref] [bib_ref] The Child Behavior Checklist-Obsessive-Compulsive Subscale Detects Severe Psychopathology and Behavioral Problems Among..., Saad [/bib_ref] [bib_ref] Development and Psychometric Evaluation of the Children's Yale-Brown Obsessive-Compulsive Scale Second Edition, Storch [/bib_ref] [bib_ref] Symptom correspondence between clinicians and patients on the Yale-Brown Obsessive Compulsive Scale, Storch [/bib_ref] [bib_ref] Further psychometric properties of the Yale-Brown Obsessive Compulsive Scale -Second Edition, Wu [/bib_ref].
Clinicians should inquire about such symptoms as rituals or behaviors that are repeated over and over, things or situations (locations) that are carefully evaded, and acknowledgment of invasive thoughts or ideas [bib_ref] The Yale-Brown Obsessive Compulsive Scale: factor analysis, construct validity, and suggestions for..., Deacon [/bib_ref]. The use of transcranial sonography as a diagnostic modality revealing subcortical echogenicity in OCD and other psychiatric disorders remains in research [bib_ref] Transcranial sonography in psychiatry as a potential tool in diagnosis and research, Drepper [/bib_ref]. Studies on eye-tracking measurement have not proven diagnostic in OCD [bib_ref] Eye movement tracking in pediatric obsessive compulsive disorder, Ray [/bib_ref].
Testing for the neurocognitive deficits can be useful for the identification of trait markers in those with OCD [bib_ref] Neurocognitive deficits in obsessivecompulsive disorder: A selective review, Suhas [/bib_ref]. Also, potentially valuable are the use of computer vision tools to find behavioral markers in those with pediatric OCD (46). Research is not definitive but continues in identifying potential subtypes of pediatric OCD based on executive functioning, aggressive traits, social reciprocity, and other features [bib_ref] Cool and Hot Aspects of Executive Function in Childhood Obsessive-Compulsive Disorder, Hybel [/bib_ref] [bib_ref] Are the components of social reciprocity transdiagnostic across pediatric neurodevelopmental disorders? Evidence..., Sturm [/bib_ref] [bib_ref] A Case Report of Harm-Related Obsessions in Pediatric Obsessive-Compulsive Disorder, Wu [/bib_ref] [bib_ref] Children with Obsessive-Compulsive Symptomology in the General Population, Zijlmans [/bib_ref].
## Differential diagnosis
An essential aspect of the OCD patient assessment is to separate the OCD diagnosis from other disorders that may be the cause of the patient's symptomatology instead of OCD. A careful assessment including, thorough medical and psychological evaluations, is critical in this regard. For example, this person may have other anxiety disorders including generalized anxiety disorder, panic disorder, and different phobias. Other disorders to consider, depending on the presentation and history include depressive disorder, anorexia nervosa, bulimia nervosa, bipolar disorder, psychosis, tic disorders, substance use disorder, hoarding disorder, body dysmorphic disorder, trichotillomania, paraphilia, gambling disorder, and personality disorder (i.e., obsessive-compulsive personality disorder).
## Comorbidity
Another aspect of the evaluation is to know that OCD is well-known to co-exist with a wide variety of other disorders and an understanding of the full patient picture is critical to developing effective management strategies. Research notes that those with a pediatric-onset OCD have increased co-morbidities and, the comorbidities are similar, to some extent, with the differential diagnosis list (99).
As indicated above, a number of psychiatric disorders can co-exist with OCD including Tourette's disorder, eating disorders (i.e., anorexia nervosa, bulimia nervosa, others), major depressive disorder (lifetime co-existence in 75% of OCD persons), bipolar disorder, social phobia, panic disorder, impulse control disorders, personality disorders (i.e., obsessive-compulsive personality disorder [OCPD]), oppositional defiant disorder (ODD) and others [bib_ref] Comorbid psychopathology and clinical symptomatology in children and adolescents with obsessive-compulsive disorder, Anagnostopoulos [/bib_ref] [bib_ref] Clinical aspects of obsessive-compulsive disorder, Baup [/bib_ref] [bib_ref] Diagnosis and treatment of obsessive-compulsive disorder and related disorders, Dell'osso [/bib_ref] [bib_ref] Categorical and dimensional aspects of co-morbidity in obsessivecompulsive disorder (OCD), Ivarsson [/bib_ref] [bib_ref] Obsessive-Compulsive Disorder in Children and Adolescents: A Review of the Past 10..., March [/bib_ref] [bib_ref] Obsessive compulsive disorder: comorbid conditions, Pigott [/bib_ref] [bib_ref] Impact of Gender, Age at Onset, and Lifetime Tic Disorders on the..., Tanidir [/bib_ref] [bib_ref] Comorbid Psychopathology and the Clinical Profile of Family Accommodation in Pediatric OCD, Wu [/bib_ref]. Suicidality can be a worrisome co-morbidity from the viewpoint of the patient, family and healthcare givers [bib_ref] Suicidal and death ideation among adults with obsessive-compulsive disorder presenting for intensive..., Storch [/bib_ref].
OCD is commonly co-morbid with autism spectrum disorder (ASD) and, attention-deficit/hyperactivity disorder (ADHD) [bib_ref] Comorbidity Between Attention Deficit/Hyperactivity Disorder and Obsessive-Compulsive Disorder Across the Lifespan, Abramovitch [/bib_ref] [bib_ref] Subclinical autism spectrum symptoms in pediatric obsessive-compulsive disorder, Arildskov [/bib_ref] [bib_ref] Adapted cognitive behavior therapy for obsessive-compulsive disorder with co-occurring autism spectrum disorder:..., Flygare [/bib_ref] [bib_ref] The link between ADHD-like inattention and obsessions and compulsions during treatment of..., Guzick [/bib_ref] [bib_ref] Intra-individual variability across cognitive task in drug-naïve pediatric patients with obsessive compulsive..., Okazaki [/bib_ref] [bib_ref] Anxiety Disorders and Obsessive-Compulsive Disorder in Individuals with, Postorino [/bib_ref] [bib_ref] Elevated Autism Spectrum Disorder Traits in Young Children with OCD, Stewart [/bib_ref]. Clinicians should also be cognizant of and observant for potential OCD co-morbidity with enuresis (non-monosymptomatic primary nocturnal enuresis), self-induced dermatoses, immunoglobulin A dysgammaglobulinemia, and sleep dysfunction [insomnia, disorders of sleep initiation and maintenance (DIMS), delayed sleep phase disorder (DSPD)] [bib_ref] Comorbidity Between Attention Deficit/Hyperactivity Disorder and Obsessive-Compulsive Disorder Across the Lifespan, Abramovitch [/bib_ref] [bib_ref] Subclinical autism spectrum symptoms in pediatric obsessive-compulsive disorder, Arildskov [/bib_ref] [bib_ref] Adapted cognitive behavior therapy for obsessive-compulsive disorder with co-occurring autism spectrum disorder:..., Flygare [/bib_ref] [bib_ref] The link between ADHD-like inattention and obsessions and compulsions during treatment of..., Guzick [/bib_ref] [bib_ref] Intra-individual variability across cognitive task in drug-naïve pediatric patients with obsessive compulsive..., Okazaki [/bib_ref] [bib_ref] Anxiety Disorders and Obsessive-Compulsive Disorder in Individuals with, Postorino [/bib_ref] [bib_ref] Elevated Autism Spectrum Disorder Traits in Young Children with OCD, Stewart [/bib_ref] [bib_ref] Self-induced dermatoses: A great imitator, Gupta [/bib_ref] [bib_ref] A casecontrol study of sleep disturbances in pediatric obsessivecompulsive disorder, Jaspers-Fayer [/bib_ref] [bib_ref] Insomnia in pediatric obsessive-compulsive disorder: prevalence and association with multimodal treatment outcomes..., Sevilla-Cermeño [/bib_ref]. Also complicating management plans can be other complex psychiatric comorbid disorders as hoarding, psychosis, and obsessive-compulsive personality disorder (OCPD). Issues of PANS and PANDAS have been considered before (vida supra).
## Quality of life in ocd
T h e q u a l i t y o f l i f e f o r t h o s e w i t h O C D c a n b e compromised to a considerable extent depending on various factors including the age of onset, severity of the OCD symptomatology, complications of family accommodation, failure for the OCD diagnosis (often under-diagnosed), co-morbidities, and others [bib_ref] A Multisite Study of Family Functioning Impairment in Pediatric Obsessive-Compulsive Disorder, Stewart [/bib_ref] [bib_ref] Quality of life in children and adolescents with obsessive-compulsive disorder: The Pediatric..., Zaboski [/bib_ref]. In light of such factors, effective management plans for pediatric OCD is considered at this time.
## Management
This section reviews the current evidence for the treatments of OCD in children and adolescents, including psychotherapy and medications, with a focus on randomized controlled trials in the pediatric population. As discussed below substantial evidence supports the use of Serotonin Reuptake Inhibitors (SSRIs) and Cognitive-Behavioral Therapy (CBT) in the treatment of OCD in children and adolescents and are often used together in clinical practice.
## Psychotherapeutic management
Cognitive behavioral therapy (CBT) is widely regarded as the first-line treatment of OCD for mild-moderate cases in children and adolescents. CBT has shown comparable and, at times, superior results to medication management for treatment of OCD [bib_ref] Group Cognitive-Behavioral Therapy Versus Sertraline for the Treatment of Children and Adolescents..., Asbahr [/bib_ref] [bib_ref] Behavior Therapy Versus Clomipramine for the Treatment of Obsessive-Compulsive Disorder in Children..., De Haan [/bib_ref] [bib_ref] Sertraline in Children and Adolescents With Obsessive-Compulsive Disorder, March [/bib_ref] [bib_ref] Effectiveness of cognitive behavior treatment for pediatric obsessivecompulsive disorder: Acute outcomes from..., Torp [/bib_ref]. These studies are summarized in [fig_ref] Table 1: Evidence for psychotherapy treatments for pediatric OCD [/fig_ref]. There is minimal evidence for the use of other forms of psychotherapies for pediatric OCD. Studies demonstrating positive benefit from CBT for OCD in children typically include components of psychoeducation, cognitive retraining, and exposure response prevention. Other specific components for successful CBT for use in pediatric OCD have also been evaluated. A program as short as five weeks in duration has shown to lead to significant improvements in the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS) score and functional improvement [bib_ref] Randomized controlled trial of full and brief cognitive-behaviour therapy and wait-list for..., Bolton [/bib_ref]. Weekly CBT sessions were found to be as effective as daily sessions [bib_ref] Family-Based Cognitive-Behavioral Therapy for Pediatric Obsessive-Compulsive Disorder, Storch [/bib_ref]. Parental involvement is often heavily incorporated into the CBT model. The Positive Family Interaction Therapy model involves both individual CBT with the child or adolescent patient as well as additional family sessions demonstrated a 70% response rate compared to the more traditional model of individual CBT with parent checkins and psychoeducation [bib_ref] Optimizing Treatment for Complex Cases of Childhood Obsessive Compulsive Disorder: A Preliminary..., Peris [/bib_ref]. Given the concerns about accessibility of high-quality CBT in the community, research into the telephone-and web-based programs have yielded positive results [bib_ref] Preliminary investigation of web-camera delivered cognitive-behavioral therapy for youth with obsessive-compulsive disorder, Storch [/bib_ref] [bib_ref] Telephone Cognitive-Behavioral Therapy for Adolescents With Obsessive-Compulsive Disorder: A Randomized Controlled Non-inferiority..., Turner [/bib_ref]. The modalities have found to be superior to placebo and equivalent to traditional office-based CBT. CBT can also be administered in group format without compromising its effectiveness [bib_ref] Group Cognitive-Behavioral Therapy Versus Sertraline for the Treatment of Children and Adolescents..., Asbahr [/bib_ref] [bib_ref] Cognitive-Behavioral Family Treatment of Childhood Obsessive-Compulsive Disorder: A Controlled Trial, Barrett [/bib_ref]. In the Pediatric Obsessive-Compulsive Disorder Treatment Study for Young Children [bib_ref] The pediatric obsessive compulsive disorder treatment study for young children (POTS Jr):..., Freeman [/bib_ref] , investigators reviewed the relative efficacy of family-based CBT, involving exposure and response prevention (ERP), to family-based relaxation treatment in children 5 to 8 years old and demonstrated improvement in clinical symptoms and functional impairment.
## Pharmacotherapy of ocd
Pharmacological treatment of OCD in children and adolescents has grown over the past thirty years. Clomipramine was the first medication to be studied that showed clinical efficacy in treating pediatric OCD (148), followed shortly by fluoxetine [bib_ref] Double-Blind, Crossover Trial of Fluoxetine and Placebo in Children and Adolescents with..., Riddle [/bib_ref] , but subsequent trials have also demonstrated the efficacy of fluvoxamine [bib_ref] The Effect of Fluvoxamine and Behavior Therapy on Children and Adolescents with..., Neziroglu [/bib_ref] [bib_ref] Fluvoxamine for Children and Adolescents With Obsessive-Compulsive Disorder: A Randomized, Controlled, Multicenter..., Riddle [/bib_ref] , sertraline [bib_ref] Sertraline in Children and Adolescents With Obsessive-Compulsive Disorder, March [/bib_ref] , and paroxetine [bib_ref] Paroxetine Treatment in Children and Adolescents With Obsessive-Compulsive Disorder: A Randomized, Multicenter,..., Geller [/bib_ref]. Serotonergic medications are generally well-tolerated often with higher rates of mild side effects, most commonly nausea, but rarely serious adverse effects. Response rates, however, vary between 40-60% leaving a substantial number of non-responders. Study design often excludes children and adolescents with comorbid psychiatric conditions, including tic disorders, other primary mood disorders, and autism spectrum disorders. The studies are summarized in [fig_ref] Table 2: Evidence for pharmacological treatments for pediatric OCD [/fig_ref].
## Combination treatment
Published in 2004, the Pediatric OCD Treatment Study (POTS) was a randomized placebo-controlled trial of 112 patients, ages 7-17 years, and was designed to evaluate the efficacy of an SSRI (sertraline), CBT, and the combination across 12 weeks. Outcomes were based on the change in CY-BOCS score over the 12 weeks with remission defined as a CY-BOCS score of <10 [bib_ref] Cognitive-Behavior Therapy, Sertraline, and Their Combination for Children and Adolescents With Obsessive-Compulsive..., Stockman [/bib_ref]. Significant improvements were noticed in all three groups in the CY-BOCS score compared to the placebo group, with the greatest improvement in the combination group (53.6% remission compared to 39.3% for CBT alone, 21.4% for sertraline alone, and 3.6% for placebo). Outcomes with CBT alone did not differ significant from SSRI alone (P=0.24) but did separate from placebo (P=0.002). Sertraline alone did not differ significantly from placebo (P=0.10). No serious adverse effects were noted, but minor side effects, including nausea and diarrhea, were significantly higher in treatment groups.
POTS II, a follow-up study completed in 2011 was a randomized controlled trial of 124 pediatric patients, ages 7-17, with OCD as their primary diagnosis [bib_ref] Cognitive Behavior Therapy Augmentation of Pharmacotherapy in Pediatric Obsessive-Compulsive Disorder, Franklin [/bib_ref]. Participants were randomized to 12 weeks of medication (SSRI) only, medication and CBT "instruction", and medication management with CBT. "Instruction" in CBT was described as extended sessions with the psychiatrist providing medication management during which CBT principles including psychoeducation, reviewing stimulus hierarchy and identifying ERP targets were introduced, and subjects were instructed to implement these skills between sessions with assigned homework. CBT instruction did not include the therapist-assisted exposure and imaginal exposure. Didactic parent sessions were included in the third treatment group. This group had 14 visits, each being 1-hour long, over 12 weeks with a participating psychologist in addition to the medication provider. Outcomes were based on improvement in CY-BOCS score with response defined >30% change from baseline to week 12. Medication management with CBT yielded a significantly greater proportion of participants who responded to treatment (68.6% for medications + CBT, 34.0% for medications + instruction in CBT, and 30.0% for medication alone).
## Augmentation strategies
Despite adequate treatment trials, 40% of patients with OCD continue to have impairing symptoms. Augmentation strategies should be considered for patients who fail to respond to two or more trials of SSRIs or clomipramine of adequate dose and duration, in addition to failure to respond to CBT.
## Atypical antipsychotics
Augmentation of serotonergic agents with antipsychotics is the most commonly used strategy in clinical practice. Antipsychotic augmentation can effectively treat approximately 30% of treatment-refractory adults without significant differences in efficacy between typical or atypical medications [bib_ref] Obsessive-Compulsive Disorder, Hirschtritt [/bib_ref]. While efficacy has been demonstrated in adults [bib_ref] Antipsychotic Augmentation of Serotonin Reuptake Inhibitors in Treatment-Resistant Obsessive-Compulsive Disorder: An Update..., Dold [/bib_ref] [bib_ref] Antipsychotic augmentation of serotonergic antidepressants in treatment-resistant obsessive-compulsive disorder: A meta-analysis of..., Skapinakis [/bib_ref] , this has not been evaluated systematically in the pediatric population aside from case series and open trials.
## Other augmentation approaches
Other augmentation strategies, for example, novel medications and neuro-modulatory approaches, including Repetitive Transcranial Magnetic Stimulation (rTMS) and more invasive Deep Brain Stimulation (DBS) have also been studied. A body of evidence shows that the efficacy of DBS exists in adults [bib_ref] Deep Brain Stimulation for Obsessive-Compulsive Disorder: A Meta-Analysis of Treatment Outcome and..., Alonso [/bib_ref]. Non-surgical techniques e.g., rTMS has a limited body of evidence and questionable efficacy for OCD in adults [bib_ref] Transcranial Magnetic Stimulation for Obsessive-Compulsive Disorder, Trevizol [/bib_ref]. No randomized controlled trials exist for these approaches in children and adolescents. Novel medications that have been studied in children and adolescents are discussed below. None have shown a statistically significant benefit.
## D-cycloserine (dcs)
DCS is a partial NMDA receptor agonist, which has been used in adult trials to enhance the efficacy of exposure therapy with an anxiety disorder. In one study of 30 children and adolescents ages , participants were randomized to either CBT and placebo or CBT and DCS (weighted doses of either 25 or 50 mg). CBT consisted of seven ERP sessions. DCS or the placebo was administered one hour before each session. While no adverse effects were documented, there was no significant difference between the placebo and the DCS group. In 2016, a larger RCT of 142 children and adolescents ages 7-17 with a primary diagnosis of OCD were randomized to receive CBT and DCS or CBT and placebo; however, there was once again no significant additional benefit to the addition of DCS [bib_ref] Efficacy of Augmentation of Cognitive Behavior Therapy With Weight-Adjustedd-Cycloserine vs Placebo in..., Storch [/bib_ref].
## Riluzole
The glutamate pathway has also been considered as a target for modulation in the treatment of OCD. Riluzole is an anti-glutamatergic agent that has been studied. In an openlabel trial of riluzole [bib_ref] An Open-Label Trial of Riluzole, a Glutamate Antagonist, in Children with Treatment-Resistant..., Grant [/bib_ref] in six children ages 8-16 years old who had failed previous pharmacotherapy for OCD, four children had a significant reduction (of greater than 46%) of their baseline CY-BOCS score and a CGI-I score of "much improved" or "very much improved." Riluzole was well-tolerated. This study was followed up by a 12-week placebo-controlled trial of 60 treatment-resistant children and adolescents with moderate to severe OCD [bib_ref] 12-Week, Placebo-Controlled Trial of Add-on Riluzole in the Treatment of Childhood-Onset Obsessive-Compulsive..., Grant [/bib_ref]. Seventeen of these participants have a diagnosis of autism spectrum disorder. Subjects were randomized to receive riluzole (titrated to 100 mg/day) or placebo in addition to their existing treatment regimen. This study failed to demonstrate a significant difference between placebo and riluzole and also resulted in more drop-outs related to adverse side effects (one case of pancreatitis, five patients with elevated transaminases).
## N-acetylcysteine (nac)
NAC has also been considered as an adjuvant treatment to SRIs for OCD in children and adolescents. In a 10-week placebo-controlled randomized double-blind trial, 34 patients were randomized, to receive either citalopram and NAC or citalopram and placebo [bib_ref] Efficacy of N-Acetylcysteine Augmentation on Obsessive Compulsive Disorder: A Multicenter Randomized Double..., Ghanizadeh [/bib_ref]. The CYBOCS and Pediatric Quality of Life Inventory were used as primary outcomes. The YBOCS of NAC group significantly decreased from 21.0 to 11.3 during the study, but no statistically significant decrease of YBOCS was found in the placebo group. In the Quality of Life inventory, a similar pattern was found. There were no serious adverse effects in either group.
# Conclusions
Pediatric OCD is a chronic remitting relapsing condition that often persists into adulthood. The field has advanced significantly during the last decade and we have a better understanding of neurological and genetic underpinnings of this disease. Despite the fact that medications have a significant role in the treatment, current literature continues to converge in support of CBT as a first line treatment, which has long-term positive outcomes for pediatric OCD.
# Acknowledgments
Funding: None.
## Footnote
## Conflicts of interest:
The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the noncommercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-ncnd/4.0/.
[table] Table 1: Evidence for psychotherapy treatments for pediatric OCD [/table]
[table] Table 2: Evidence for pharmacological treatments for pediatric OCD [/table]
|
The Th1:Th2 Dichotomy of Pregnancy and Preterm Labour
Pregnancy is a unique immunological state in which a balance of immune tolerance and suppression is needed to protect the fetus without compromising the mother. It has long been established that a bias from the T helper 1 cytokine profile towards the T helper 2 profile contributes towards successful pregnancy maintenance. The majority of publications that report on aberrant Th1:Th2 balance focus on early pregnancy loss and preeclampsia. Over the last few decades, there has been an increased awareness of the role of infection and inflammation in preterm labour, and the search for new biomarkers to predict preterm labour continues. In this paper, we explore the evidence for an aberrant Th1:Th2 profile associated with preterm labour. We also consider the potential for its use in screening women at high risk of preterm labour and for prophylactic therapeutic measures for the prevention of preterm labour and associated neonatal adverse outcomes.
# Introduction
Preterm labour occurs in some 10% of pregnancies [bib_ref] Tolerance of the fetus by the maternal immune system: role of inflammatory..., Kanellopoulos-Langevin [/bib_ref]. In many developed countries, the rates are rising. Birth before 37 weeks of gestation is thought to account for up to 70% of neonatal deaths, and the extremely high neonatal intensive care costs required to support those who do survive make preterm birth both a social and economic burden. It is now widely acknowledged that the aetiology of preterm labour is multifactorial, and, as such, the underlying cause of preterm labour is often unknown. There is a strong association between preterm labour and infection and inflammation, and research in this field has dramatically increased over the last few decades [bib_ref] Targeting immune activation in the prevention of preterm labour, Sykes [/bib_ref]. However, we still have made little significant progress in the prevention of preterm labour. Evidence of the detrimental direct impact of maternal infection/inflammation on neonatal outcome is emerging, yet we do not fully understand if anti-inflammatory therapeutic agents would provide benefit or harm to the neonate born under conditions of infection/inflammation-induced preterm labour.
The immunology of pregnancy is complex, in that the mother must tolerate the "foreign" fetus, and thus requires a degree of immunosuppression whilst on the other hand needs to maintain immune function to fight off infection. One mechanism which is involved in successful pregnancy maintenance is the proposed switch from the T helper 1 (Th1) cytokine profile to the T helper 2 (Th2) profile. This paper explores the evidence for an imbalance in the Th1:Th2 profile in women at risk of and who are in established preterm labour.
## The immunology of pregnancy
The fetus can be described as a semiallogeneic graft, being a product of two histoincompatible individuals [bib_ref] The immunology of successful pregnancy, Van Nieuwenhoven [/bib_ref] [bib_ref] Genetic conflicts in human pregnancy, Haig [/bib_ref]. This poses a challenge to the mother, to both tolerate and accommodate the fetus, which will express paternal antigens, and maintain an ability to reject in case of overwhelming infection [bib_ref] Pregnancy and the immune system: between tolerance and rejection, Thellin [/bib_ref]. This challenge is undertaken in part by the immune system. The immune system has two main defence systems: the innate and the adaptive. The innate immune response is a nonspecific reaction towards foreign antigens, whereas the adaptive response forms a very specific reaction towards antigens [bib_ref] How immune mechanisms are affected by pregnancy, Luppi [/bib_ref]. Although different immune components are involved in these systems, much overlap and : Summary of the adaptive and innate immune system in pregnancy. Mediators of the adaptive and innate immune system work in parallel to facilitate a balance between immune tolerance of the fetus whilst maintaining the ability to mount a response against invading pathogens. PWBC: peripheral white blood cells.
cross-talk exist between the two. summarises the key elements of these systems during pregnancy.
The immune cells that make up the adaptive immune response include B and T lymphocytes. Activation by antigen presenting cells and cytokines leads to cytokine release by T cells in a cell-mediated response, or antibody release by B cells in a humoral response [bib_ref] Immunology of normal pregnancy, Aagaard-Tillery [/bib_ref]. Although Medawar originally hypothesised that pregnancy represents a time of immune suppression [bib_ref] Some immunological and endocrinological problems raised by the evolution of viviparity in..., Medawar [/bib_ref] , a more complex picture has recently emerged where a change in the ratio and functionrather than a complete suppression-of the maternal leukocytes occurs during pregnancy. For example, there is an increase in the total peripheral white cell count from the early stages of pregnancy with no change in the CD4 and CD8 counts [bib_ref] Changes in lymphocyte subsets during normal pregnancy, Kuhnert [/bib_ref]. Within the CD4 positive population, an increase in T regulatory cells is seen in pregnancy [bib_ref] Regulatory T cells mediate maternal tolerance to the fetus, Aluvihare [/bib_ref]. The function of the T cells adapts in pregnancy to favour the T helper 2 cytokine profile, which is more pronounced at the maternal fetal interface [bib_ref] Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy, Saito [/bib_ref]. Nonimmune cells, for example, placental trophoblasts also contribute to the Th2 cytokine predominance in pregnancy [bib_ref] Regulation of T-cell activities at the fetoplacental interface-by placenta?, Chaouat [/bib_ref]. The innate immune system provides a less specific response nevertheless is critical for the prevention of microbial invasion. Cellular components include neutrophils, monocytes, and macrophages, which protect against pathogens by phagocytosis. The Toll-like receptors (TLRs) TLR2 and TLR4 are pattern recognition receptors stimulated by Gram-positive and Gram-negative bacteria, respectively [bib_ref] Tolerance of the fetus by the maternal immune system: role of inflammatory..., Kanellopoulos-Langevin [/bib_ref]. TLRs are expressed on nonimmune cells in the placenta and fetal membranes, which mediate part of the innate immune system at the maternal fetal interface [bib_ref] Toll-like receptors at the maternal-fetal interface in normal pregnancy and pregnancy disorders, Koga [/bib_ref]. TLR2 and 4 mutations are associated with an increased risk of preterm birth [bib_ref] Relevance of mutations in the TLR4 receptor in patients with gram-negative septic..., Lorenz [/bib_ref] [bib_ref] Toll-like receptor 2 polymorphism is associated with preterm birth, Krediet [/bib_ref]. During pregnancy, there is tight regulation and considerable cross-talk between the adaptive and the innate adaptive immune system that is responsible for preventing or activating rejection of the conceptus.
## Th1:th2 cytokines
T helper 1 and 2 cell subsets originate from undifferentiated Th0 cells under the influence of interferon-gamma (IFN-γ) and interleukin-4 (IL-4), respectively. Pregnancy hormones such as progesterone [bib_ref] Progesterone favors the development of human T helper cells producing Th2-type cytokines..., Piccinni [/bib_ref] , leukaemic inhibitory factor [bib_ref] Mother's little helpers: mechanisms of maternal-fetal tolerance, Trowsdale [/bib_ref] , estradiol [bib_ref] Estradiol prevents and testosterone promotes Fas-dependent apoptosis in CD4+ Th2 cells by..., Huber [/bib_ref] , and prostaglandin D 2 (PGD 2 ) [bib_ref] Prostaglandin D2 causes preferential induction of proinflammatory Th2 cytokine production through an..., Xue [/bib_ref] promote the T helper 2 cell profile and are likely to be in part responsible for the Th2 bias associated with pregnancy.
Type 1 CD4 + T cells (Th1) produce an array of inflammatory cytokines including IFN-γ [bib_ref] Molecular aspects of T-cell differentiation, Rao [/bib_ref] , IL-2 [bib_ref] TH1 and TH2 cells: different patterns of lymphokine secretion lead to different..., Mosmann [/bib_ref] , and Tumor necrosis factor-alpha (TNF-α) [bib_ref] Th1/Th2 cytokines balance-yin and yang of reproductive immunology, Wilczynski [/bib_ref] and are the major effectors of phagocyte-mediated host defence, protective against intracellular pathogens [bib_ref] TH1 and TH2 cells: different patterns of lymphokine secretion lead to different..., Mosmann [/bib_ref] [bib_ref] T cells in normal pregnancy and recurrent pregnancy loss, Piccinni [/bib_ref] [bib_ref] The molecular basis of T helper 1 and T helper 2 cell..., O'garra [/bib_ref]. Type 2 CD4 + T cells (Th2) produce IL-4, IL-5, IL-13, IL-10 [bib_ref] Molecular aspects of T-cell differentiation, Rao [/bib_ref] , and IL-6. Whilst IL-4 and IL-10 are considered to be antiinflammatory cytokines [bib_ref] Regulation by anti-inflammatory cytokines (IL-4, IL-10, IL-13, TGFβ) of interleukin-8 production by..., Marie [/bib_ref] , IL-6 has proinflammatory properties [bib_ref] Maternal serum interleukin-6 during pregnancy and during term and preterm labor, Greig [/bib_ref]. Although IL-10 and IL-6 are frequently referred to as Th2 cytokines [bib_ref] Progesterone and the immunology of pregnancy, Druckmann [/bib_ref] [bib_ref] Accumulation of CRTH2-positive T-helper 2 and T-cytotoxic 2 cells at implantation sites..., Michimata [/bib_ref] [bib_ref] Two types of mouse T helper cell. IV. Th2 clones secrete a..., Fiorentino [/bib_ref] [bib_ref] A randomized, controlled trial of IL-10 in humans: inhibition of inflammatory cytokine..., Chernoff [/bib_ref] [bib_ref] Role of interferons and other cytokines in the regulation of the immune..., Belardelli [/bib_ref] [bib_ref] Th1-type immunity is incompatible with successful pregnancy, Raghupathy [/bib_ref] , they are both produced by other cell types including Th1 cells, macrophages, and B cells for IL-10 [bib_ref] TH1 cells control themselves by producing interleukin-10, O'garra [/bib_ref] [bib_ref] Biology of interleukin-10, Sabat [/bib_ref] , and macrophages, fibroblasts, and B Mediators of Inflammation 3 cells for IL-6 [bib_ref] Interleukin-6: from basic science to medicine-40 Years in immunology, Kishimoto [/bib_ref]. The T helper 2 cytokines are commonly associated with strong antibody responses [bib_ref] The expanding universe of T-cell subsets: Th1, Th2 and more, Mosmann [/bib_ref] , for example, IL-4 stimulates IgE and IgG 1 antibody production [bib_ref] Role of immune cells in pregnancy, Piccinni [/bib_ref]. However, the Th2 cytokines also serve other functions, for example; IL-5 promotes the growth and differentiation of eosinophils, whereas IL-13 and IL-10 inhibit the activity of macrophages [bib_ref] Role of immune cells in pregnancy, Piccinni [/bib_ref]. T helper 2 cell responses are also associated with protection against parasites, since IL-4 mediates IgE production, and IL-5 mediates an eosinophilia, both of which are hallmarks of parasitic infection [bib_ref] Regulation of immunity to parasites by T cells and T cell-derived cytokines, Sher [/bib_ref]. It is important to note that, although the Th1 and Th2 responses can be seen as discrete responses, there is considerable crosstalk and overlap between the functions of the T helper cells. For example, the Th1 cytokines can promote the production of complement-fixing antibodies involved in antibodydependent cell cytotoxicity [bib_ref] Role of hormone-controlled Th1-and Th2-type cytokines in successful pregnancy, Piccinni [/bib_ref] , and thus the dichotomy described may be an oversimplified representation of the complex immune system. Transcriptional regulation of the predominant Th2 cytokine IL-4 is by STAT-6, c-maf, GATA-3, and NFAT [bib_ref] Genes that regulate interleukin-4 expression in T cells, Szabo [/bib_ref] , whereas Th1 cell cytokine production is transcriptionally regulated by T-bet and STAT-4 [bib_ref] Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy, Saito [/bib_ref].
## Th1:th2 and pregnancy maintenance
The hypothesis of Th2 predominance and downregulation of the Th1 response originated from Wegmann and colleagues [bib_ref] Bidirectional cytokine interactions in the maternal-fetal relationship: is successful pregnancy a TH2..., Wegmann [/bib_ref] and was reinforced by evidence from both murine studies and the clinical course of Th2 and Th1 based conditions in pregnancy. IL-2, IFN-γ, and TNF-α induce miscarriage in mice, which can be reversed by inhibitors of the Th1 cytokines or by administering the anti-inflammatory Th2 cytokine IL-10 [bib_ref] IL-10 prevents naturally occuring fetal loss in the CBA x DBA/2 mating..., Chaouat [/bib_ref] [bib_ref] Cytokine-dependent abortion in CBA x DBA/2 mice is mediated by the procoagulant..., Clark [/bib_ref]. Autoimmune conditions where Th1 is involved in the pathophysiology generally improve in pregnancy (e.g., rheumatoid arthritis [bib_ref] Immunology of pregnancypregnancy as a remission inducing agent in rheumatoid arthritis, Østensen [/bib_ref] , whereas the Th2 autoimmune spectrum tends to worsen (e.g., systemic lupus erythematosus [bib_ref] Rheumatic diseases and pregnancy, Märker-Hermann [/bib_ref]. With a Th2:Th1 bias, the diminished cell-mediated immunity may be responsible for the increased susceptibility in pregnancy of conditions caused by intracellular pathogenesis (e.g., influenza, leprosy, and Listeria monocytogenes [bib_ref] Immunology of pregnancy: implications for the mother, Poole [/bib_ref].
## Peripheral blood.
Several techniques are available to establish the function of Th1 and Th2 cells in pregnancy; enzyme-linked immunosorbent assay (ELISA) can be used to measure maternal serum interleukins; peripheral T cells can be isolated and stimulated with a mitogen such as phorbol myristate acetate (PMA) or phytohaemagglutinin (PHA) to measure the cytokine production either by ELISA or flow cytometry during pregnancy compared with nonpregnant controls.
Marzi and colleagues isolated PBMCs, stimulated them with PHA, and measured interleukin secretion by ELISA showing a reduction in IFN-γ and IL-2 and an increase in IL-4 and IL-10 in pregnancy compared with nonpregnant controls [bib_ref] Characterization of type 1 and type 2 cytokine production profile in physiologic..., Marzi [/bib_ref]. In support of this study, Reinhard et al. stimulated cells with PMA and demonstrated by flow cytometry a reduction in intracellular IFN-γ and IL-2, and an increase in intracellular IL-4 production in pregnancy compared with nonpregnant controls [bib_ref] Shifts in the TH1/TH2 balance during human pregnancy correlate with apoptotic changes, Reinhard [/bib_ref]. In vivo confirmation of this bias has since been demonstrated by polymerase chain reaction (PCR) reflecting decreasing messenger ribonucleic acid (mRNA) of IFN-γ through pregnancy and a concurrent increase in IL-4 mRNA which peaks in the 7th month compared with nonpregnant controls [bib_ref] Modulations of cytokine expression in pregnant women, Tranchot-Diallo [/bib_ref]. However, not all studies support the Th1 to Th2 bias. Shimaoka et al. reported a reduction in PMA-stimulated IL-4 during pregnancy [bib_ref] Changes in cytokine production during and after normal pregnancy, Shimaoka [/bib_ref] , while Matthiesen and colleagues presented data suggesting an increase in both IL-4 and IFN-γ secreting cells in pregnancy compared with nonpregnant controls [bib_ref] Increased numbers of circulating interferon-γ-and interleukin-4-secreting cells during normal pregnancy, Matthiesen [/bib_ref] [bib_ref] In-situ detection of both inflammatory and anti-inflammatory cytokines in resting peripheral blood..., Matthiesen [/bib_ref]. Such discrepancies may be due to characterisation of cytokine profiles in either isolated cell populations or whole blood, the latter arguably being a more biologically relevant system.
## Maternal fetal interface and nonimmune cells.
While much research has been dedicated toward circulating cytokines in pregnancy, local cytokine production at the maternal interface may be of greater significance than measurements obtained in the peripheral blood [bib_ref] T cells in normal pregnancy and recurrent pregnancy loss, Piccinni [/bib_ref]. IL-4, IL-10, and macrophage colony-stimulating factor (m-CSF) production by T cells at the maternal fetal interface is associated with successful pregnancy [bib_ref] T cells in normal pregnancy and recurrent pregnancy loss, Piccinni [/bib_ref]. Trophoblast, decidua, and amnion all contribute to the Th2 cytokine environment by production of IL-13 [bib_ref] expression and localization of the Th2-type cytokine interleukin-13 and its receptor in..., Dealtry [/bib_ref] , IL-10 [bib_ref] Human placental cytotrophoblasts produce the immunosuppressive cytokine interleukin 10, Roth [/bib_ref] , IL-4 and IL-6 [bib_ref] Type-1 and type-2 cytokines in human late-gestation decidual tissue, Jones [/bib_ref] [bib_ref] Cytokine expression by models of human trophoblast as assessed by a semiquantitative..., Bennett [/bib_ref]. Coculture of trophoblasts and T cells results in an increase in the transcription factors GATA-3 and STAT-6 (which regulate Th2 cytokine production), and a reduction in the Th1 transcription factor STAT-4 and subsequently decreased production of IFN-γ and TNF-α [bib_ref] Placental trophoblasts shifted Th1/Th2 balance toward Th2 and inhibited Th17 immunity at..., Liu [/bib_ref]. The placenta also synthesises PGD 2 , which may act as a chemoattractant of Th2 cells to the maternal fetal interface via the classic Th2 receptor CRTH2 (chemoattractant receptor-homologous molecule expressed on Th2 cells) [bib_ref] Accumulation of CRTH2-positive T-helper 2 and T-cytotoxic 2 cells at implantation sites..., Michimata [/bib_ref]. CRTH2 + cells are reduced at the maternal fetal interface of women suffering from recurrent loss compared with women undergoing elective termination [bib_ref] Decrease of Thelper 2 and T-cytotoxic 2 cells at implantation sites occurs..., Michimata [/bib_ref].
Local production of IL-4 and IL-10 inhibits the function of both Th1 cells and macrophages, which serves to prevent fetal allograft rejection [bib_ref] T cell tolerance towards the fetal allograft, Piccinni [/bib_ref]. Other anti-inflammatory effects of these interleukins result in inhibition of the Th1 cytokine TNF-α [bib_ref] Interleukin-10 and transforming growth factor-β inhibit amniochorion tumor necrosis factor-α production by..., Fortunato [/bib_ref] , and TNF-α-induced cyclo-oxygenase-2 (COX-2), and/or PGE 2 synthesis in amnion-derived wish cells. Similar effects are observed in decidual and placental cells in vitro [bib_ref] Effects of interleukin-4 on the expression and activity of prostaglandin endoperoxide H..., Gilmour [/bib_ref] [bib_ref] Comparative studies on the effects of interleukin-4 and interleukin-13 on cytokine and..., Keelan [/bib_ref] [bib_ref] Interleukin-4 and transforming growth factor-β1 modulate the production of interleukin-1 receptor antagonist..., Bry [/bib_ref] [bib_ref] Anti-inflammatory effects of interleukin-4, interleukin-10, and transforming growth factor-β on human placental..., Goodwin [/bib_ref] , which is thought to inhibit the onset of labour. Consistent with such a role, decidual CD4 positive cells from women undergoing unexplained recurrent pregnancy loss typically exhibit reduced IL-4 and IL-10 production [bib_ref] Defective production of both leukemia inhibitory factor and type 2 T-helper cytokines..., Piccinni [/bib_ref].
## Th1:th2 cytokines in labour
5.1. Peripheral Blood. The Th2 cytokine predominance which exists during pregnancy has been shown to return to nonpregnant Th1:Th2 ratios by 4 weeks postpartum [bib_ref] Quantitative analysis of peripheral blood Th0, Th1, Th2 and the Th1:Th2 cell..., Saito [/bib_ref]. Labour is often seen as a proinflammatory state marking the end of the pregnancy, and thus it is plausible that labour is associated with a reversal in the bias back towards Th1 rather than Th2. Rather, Kuwajima and colleagues have shown that the Th2:Th1 ratio remained constant in favour of Th2 through pregnancy and labour, with a reversal back to nonpregnant parameters at 7 days postpartum [bib_ref] Changes in maternal peripheral T helper 1-type and T helper 2-type immunity..., Kuwajima [/bib_ref]. However, this finding is somewhat contrary to an earlier report indicating that serum IL-4 levels measured by ELISA in women through pregnancy and at different stages of labour were reduced in the later part of labour and by day 1 postpartum in both normotensive and preeclamptic women [bib_ref] Differential levels of T helper cytokines in preeclampsia: pregnancy, labor and puerperium, Omu [/bib_ref]. In this study, the Th1 proinflammatory cytokine TNFα peaked in early labour consistent with labour being a proinflammatory state. Consistent with this, an increase in IFN-γ and IL-1β in women in active labour has also been reported [bib_ref] Maternal and neonatal plasma cytokine levels in relation to mode of delivery, Buonocore [/bib_ref].
## Maternal fetal interface and nonimmune cells.
There is substantial evidence that the Th1 cytokines play a role in the initiation of labour at term [bib_ref] Th1/Th2 cytokines balance-yin and yang of reproductive immunology, Wilczynski [/bib_ref]. The importance of local rather than peripheral production of the cytokines is highlighted by their direct input into the biochemical pathways involved in parturition. Fetal membranes [bib_ref] IL-1β and IL-8 in human fetal membranes: changes with gestational age, labor,..., Elliott [/bib_ref] [bib_ref] Cytokine abundance in placental tissues: evidence of inflammatory activation in gestational membranes..., Keelan [/bib_ref] and myometriumproduce IL-1β at term, a potent inducer of NF-κB [bib_ref] 15-Deoxy-δ12,14-prostaglandin J2 inhibits interleukin-1β-induced nuclear factor-κB in human amnion and myometrial cells:..., Lindström [/bib_ref]. This transcription factor regulates the expression of numerous labour-associated genes including COX-2, the oxytocin receptor, IL-8, and matrix metalloproteinase-9 (MMP-9) [bib_ref] The role of nuclear factor kappa B in human labour, Lindström [/bib_ref]. TNF-α and IL-1β are both increased in amnion, amniotic fluid, and decidua at term [bib_ref] Cytokines, prostaglandins and parturition-a review, Keelan [/bib_ref] and can induce PGE 2 production in amniocytes and decidual cells in vitro [bib_ref] Interleukin-4 differentially regulates prostaglandin production in amnionderived WISH cells stimulated with pro-inflammatory..., Keelan [/bib_ref] [bib_ref] Mechanism of cytokine stimulation of prostaglandin biosynthesis in human decidua, Pollard [/bib_ref]. Despite the proinflammatory nature of the Th1 cytokines they are required for successful pregnancy contributing to the physiology of term labour.
## Th1:th2 cytokines in infection
Activation of the Th1 cytokines occurs as a specific response to infection caused by intracellular bacteria, parasites, and viruses [bib_ref] Th1/Th2 balance in infection, Infante-Duarte [/bib_ref]. The necessary proinflammatory type 1 response elicited by infection, along with the action of the activated T cells, drives local and systemic cytokine production that, if left unchecked, can be harmful to the host [bib_ref] Th1/Th2 balance in infection, Infante-Duarte [/bib_ref]. In some situations, the Th1 response is balanced by the production of Th2 cytokines, particularly IL-4 and IL-10 [bib_ref] Interferonγ differentially regulates interleukin-12 and interleukin-10 production in leprosy, Libraty [/bib_ref] [bib_ref] Absence of a prominent Th2 cytokine response in human tuberculosis, Lin [/bib_ref] [bib_ref] Interleukin 12 and tumor necrosis factor α are costimulators of interferon γ..., Tripp [/bib_ref] [bib_ref] Th1/Th2 profiles in tuberculosis, based on the proliferation and cytokine response of..., Surcel [/bib_ref]. In the early stages of infection, IL-12 is produced by macrophages and dendritic cells [bib_ref] Dendritic cells produce IL-12 and direct the development of Th1 cells from..., Macatonia [/bib_ref] , which lead to polarisation from Th0 to Th1 type cells [bib_ref] The molecular basis of T helper 1 and T helper 2 cell..., O'garra [/bib_ref]. IFN-γ enhances Th1 development by upregulating the IL-12 receptor and inhibiting the growth of Th2 cells [bib_ref] Cytokines induce the development of functionally heterogeneous T helper cell subsets, O'garra [/bib_ref]. IFN-γ also primes macrophages to begin phagocytosis and to stimulate the release of interleukin-1 [bib_ref] T helper cell cytokine profiles in preterm labor, Hollier [/bib_ref].
While the Th1 cytokine response may be suppressed by both the maternal and fetal immune system during pregnancy [bib_ref] Pregnancyspecific down-regulation of NF-κB expression in T cells in Mediators of Inflammation..., Mccracken [/bib_ref] , it still maintains the capacity to mount a defensive response in the context of infection. For example, cord blood mononuclear cells cultured with lipopolysaccharide (LPS) in vitro show an increased production of IFNγ concurrent with reduced IL-4 secretion [bib_ref] Correlation of Th1-type cytokine expression and induced proliferation to lipopolysaccharide, Goldberg [/bib_ref]. Similarly, neonates exposed to intrauterine infection have an increased percentage of IFN-γ-producing cells, with some neonates also showing an increase in IL-4-producing cells [bib_ref] Increase of cord blood cytokineproducing T cells in intrauterine infection, Matsuoka [/bib_ref].
In response to LPS amnion, chorion, deciduas, and placenta also release proinflammatory cytokines [bib_ref] Anti-inflammatory effects of interleukin-4, interleukin-10, and transforming growth factor-β on human placental..., Goodwin [/bib_ref] [bib_ref] Cytokine secretion by human fetal membranes, decidua and placenta at term, Denison [/bib_ref] [bib_ref] Production of pro-and anti-inflammatory cytokines of human placental trophoblasts in response to..., Griesinger [/bib_ref].
## Th1:th2 cytokines in preterm labour
Approximately, 30% of preterm births are associated with infection [bib_ref] Mechanism of parturition and preterm labor, Challis [/bib_ref] , with a higher rate of 80-85% in early preterm birth (<28 weeks) [bib_ref] Choriodecidual infection and preterm birth, Goldenberg [/bib_ref]. Immune and nonimmune cells contribute to a cytokine-rich environment in the presence of infection and inflammation. Proinflammatory cytokines such as TNF-α and IL-1β ultimately result in the production of prostaglandins and MMPs [bib_ref] T helper cell cytokine profiles in preterm labor, Hollier [/bib_ref] , via NF-κB. This triggers a cascade of prolabour events including uterine contractility and fetal membrane rupture, and if this cascade is activated early in pregnancy, preterm labour can ensue.
## Peripheral blood.
As discussed above, the peripheral response may not be as potent as the local Th1:Th2 response and may instead reflect a more significant inflammatory response at the fetal placental compartment. A large case control study of 101,042 Danish women showed that an elevated mid pregnancy IFN-γ plasma level was associated with moderate and late spontaneous preterm delivery, whereas no increased risk was seen with elevated TNF-α or IL-2 [bib_ref] Mid-pregnancy maternal plasma levels of interleukin 2, 6, and 12, tumor necrosis..., Curry [/bib_ref]. However, a study comparing women in active preterm labour and no labour looked at mitogen-stimulated production of IFN-γ and the Th2 cytokines IL-4, IL-10, and IL-13 and showed no difference in median cytokine production in the supernatant in vitro [bib_ref] T helper cell cytokine profiles in preterm labor, Hollier [/bib_ref]. The differing results between these studies could be explained by the fact that the in vitro cells lack the presence of other cells of the immune system and thus lack the ability to reflect the complexity of the immune system as a whole. This same study did however show a higher IL-12 and lower IL-4 in cervical secretions of women in preterm labour, reflecting the localised Th1:Th2 dichotomy. Bahar and colleagues did not demonstrate any difference in serum TNF-α or IFN-γ in women with preterm labour compared to term labour or matched controls not in labour [bib_ref] Maternal serum interleukin-6, interleukin-8, tumor necrosis factor-α and interferon-γ, in preterm labor, Bahar [/bib_ref]. However, those women in the preterm labour group received indomethacin, an anti-inflammatory COX-2 inhibitor, which could have dampened a typical proinflammatory response. Serum taken from women with preterm prelabour rupture of membranes (pPROM) compared to women who delivered at term exhibit a higher concentration of IFN-γ. Levels of IL-4 and IL-5 were undetectable in both groups [bib_ref] Cytokine patterns in maternal blood after premature rupture of membranes, Raghupathy [/bib_ref]. In a study of 30 women in preterm labour, mitogen-and antigenstimulated PBMCs showed a higher production of the proinflammatory cytokines IFN-γ and IL-2, along with an altered Th1:Th2 ratio favouring a Th1 response compared with controls who delivered at term [bib_ref] Pro-inflammatory maternal cytokine profile in preterm delivery, Makhseed [/bib_ref]. Taken together, these results suggest that, rather than a decrease in the Th2 response, preterm labour most likely represents an activation of the Th1 response. Thus, future development of therapeutic targets would likely be more effective if directed towards the modulation of the Th1 cytokines.
The Th1:Th2 dichotomy likely represents an oversimplification of the complexity of the cross-talk between the Th1 and Th2 cytokines. The ratio of Th1:Th2 is likely to be of more physiological importance than the actual concentrations produced. In support of such a notion, women in threatened preterm labour with high serum levels of IL-12 (which induces a Th1 cytokine response) and no change in serum IL-18 (which can induce both Th1 and Th2 response) do not show significant associations with preterm labour. However, women with high IL-12 levels and low IL-18 and thus a high IL-12 : IL-18 ratio increasing the Th1 predominance are associated with a twofold risk of preterm labour when presenting with threatened preterm labour [bib_ref] Interleukin-18 and interleukin-12 in maternal serum and spontaneous preterm delivery, Ekelund [/bib_ref].
## Maternal fetal interface and nonimmune cells.
The inflammatory response at the maternal fetal interface more likely reflects the true importance of the Th1:Th2 dichotomy and the aberrant profile in preterm labour. A recent metaanalysis concluded that proinflammatory cytokines at the maternal fetal interface play a role in the events leading to spontaneous preterm labour, while systemic inflammation does not appear to be present in asymptomatic women early on in pregnancy who then go on to deliver preterm [bib_ref] Inflammatory cytokines and spontaneous preterm birth in asymptomatic women: a systematic review, Wei [/bib_ref]. This is consistent with a more local intrauterine inflammatory response syndrome, where no organisms are identified. Understanding the pathophysiology at the maternal interface is essential for developing new therapies for the prevention of inflammation-induced preterm labour, although using such local changes for the prediction is challenging because of lack of access to the maternal fetal interface.
Placentas from women with pPROM and preterm delivery have higher Th1, inducing cytokines [bib_ref] Increased expression of pro-inflammatory cytokines in placentas of women undergoing spontaneous preterm..., El-Shazly [/bib_ref] , and placentas from women following preterm delivery compared with term delivery show a bias towards the Th1 profile with significantly higher levels of IFN-γ and IL-2 as well as the Th1-inducing cytokine IL-12 [bib_ref] Increased expression of pro-inflammatory cytokines in placentas of women undergoing spontaneous preterm..., El-Shazly [/bib_ref]. Moreover, term placentas exhibit comparatively higher levels of the Th2 cytokines, IL-4, and IL-10, compared with the preterm placentas.
TNF-α is increased in choriodecidual tissues [bib_ref] Cytokine abundance in placental tissues: evidence of inflammatory activation in gestational membranes..., Keelan [/bib_ref] and amniotic fluid [bib_ref] Prediction of preterm labour in multiple pregnancies, Ni Chuileannain [/bib_ref] in preterm labour. TNF-α is known to stimulate PG production through the TNF receptor 2, leading to uterine contractions likely via activation of NF-κB, but is also likely to contribute to MMP-9 production leading to PROM via activation of its receptor TNF Receptor 1 (TNFR1) [bib_ref] Role of tumor necrosis factor-α in the premature rupture of membranes and..., Fortunato [/bib_ref]. Interestingly, samples of myometrium collected women in preterm labour and samples collected preterm before labour express comparable mRNA levels of TNF-α. However, mRNA levels of the receptors, TNF R1 A and B, are increased in preterm labour and term labour compared with nonlabour controls [bib_ref] Cachectin/tumor necrosis factor-α formation in human decidua. Potential role of cytokines in..., Casey [/bib_ref] suggesting a receptormediated increase in sensitivity to TNF-α.
Although placental, amnion, and choriodecidual cells secrete proinflammatory cytokines, cytokine levels in tissues from preterm deliveries (with and without intrauterine infection) correlate with the extent of leukocyte infiltration in fetal membranes [bib_ref] Cytokines, prostaglandins and parturition-a review, Keelan [/bib_ref]. In the presence of infection, the primary cellular source of cytokine production in fetal membranes is likely to be infiltrating leukocytes rather than amniocytes or choriodecidual cells. [bib_ref] Cytokines, prostaglandins and parturition-a review, Keelan [/bib_ref].
## Polymorphisms of the th1 and th2
Cytokines. Studying genetic polymorphisms of the Th1 and Th2 cytokines could provide a novel screening method for determining women at high risk of preterm labour. Polymorphisms giving rise to functional alterations can also provide information on the importance of the interleukins in preterm labour. There has yet to be any promising genetic polymorphisms identified in the Th1:Th2 cytokines for the prediction of preterm labour, the work conducted warrants consideration (see [fig_ref] Table 1: Cytokine polymorphism associations with preterm labour [/fig_ref].
## Non-th1:th2 interleukins
8.1. IL-8. Interleukin 8 is a chemokine produced by many immune cells but primarily macrophages and monocytes [bib_ref] Interleukin-8, Remick [/bib_ref]. Its production is stimulated by LPS, TNF, and IL-1 [bib_ref] Interleukin-8 gene expression by a pulmonary epithelial cell line. A model for..., Standiford [/bib_ref] and, in the context of pregnancy, is thought to attract leukocytes to the gestational tissues and the cervix at the onset of term and preterm labour. IL-8 mRNA expression has been reported to be increased more than 50-fold in preterm labour and more than 1000-fold in preterm labour with evidence of chorioamnionitis in amnion and choriodecidua [bib_ref] Use of cDNA array to generate differential expression profiles for inflammatory genes..., Marvin [/bib_ref]. A number of studies have also identified increases of IL-8 in the myometrium and cervix with the onset of labour [bib_ref] Term labor is associated with a core inflammatory response in human fetal..., Bollopragada [/bib_ref] [bib_ref] Human myometrial genes are differentially expressed in labor: a suppression subtractive hybridization..., Chan [/bib_ref]. Placental IL-8 is also higher in preterm deliveries compared with term deliveries [bib_ref] Cytokine abundance in placental tissues: evidence of inflammatory activation in gestational membranes..., Keelan [/bib_ref].
## Il-6.
Although IL-6 is produced by Th2 cells, it is a proinflammatory cytokine and a major mediator of host response to inflammation and infection [bib_ref] Maternal serum interleukin-6, c-reactive protein, and matrix metalloproteinase-9 concentrations as risk factors..., Sorokin [/bib_ref]. IL-6 levels are moderately increased in placenta, significantly increased in amnion and choriodecidua in women with preterm delivery compared with term delivery [bib_ref] Cytokine abundance in placental tissues: evidence of inflammatory activation in gestational membranes..., Keelan [/bib_ref]. IL-6 appears to be among the most sensitive and specific indicators of infection-associated preterm labour [bib_ref] The preterm labor syndrome, Romero [/bib_ref] [bib_ref] Macrophage inflammatory protein-1α in term and preterm parturition: effect of microbial invasion..., Romero [/bib_ref]. The presence of an increase in IL-6 in amniotic fluid and cervicovaginal fluid is an independent risk factor for preterm labour and neonatal morbidity [bib_ref] A comparative study of the diagnostic performance of amniotic fluid glucose, white..., Romero [/bib_ref] including cerebral palsy [bib_ref] Fetal exposure to an intra-amniotic inflammation and the development of cerebral palsy..., Yoon [/bib_ref] and bronchopulmonary dysplasia [bib_ref] Increased serum levels of interleukin 6 are associated with severe intraventricular haemorrhage..., Heep [/bib_ref].
## Therapeutic modulation of th1 and th2 profile
Various therapeutic strategies have been proposed to prevent preterm labour, with the primary objectives of (1) delaying delivery to increase gestation at delivery and (2) to improve neonatal condition at birth [bib_ref] The management of preterm labor, Goldenberg [/bib_ref]. Currently, many of the strategies adopted for the prevention of preterm labour involve targeting the proposed pathways and events that result in uterine contractions and cervical shortening and dilation rather than targeting immune activation. As described here, an aberrant proinflammatory profile exists in both term and preterm labour, which is associated with neonatal morbidity. The limitation of tocolytics is the inability to counteract the exposure of the fetus to proinflammatory cytokines, which lead to the fetal inflammatory response syndrome. This may in fact worsen neonatal outcome by prolonging the exposure of the fetus to a hostile Th2
Classic Th2 cytokine. The IL-4 590 C/C genotype is associated with preterm birth but unclear. IL-4-590 SNP has been associated with both low and high IL-4 expression. Link also exists between IL-4 promoter polymorphisms and preterm birth in multiple pregnancies; however, polymorphism actually associated with increased IL-4 [bib_ref] Interleukins-1, -4, -6, -10, tumor necrosis factor, transforming growth factor-β, FAS, and..., Annells [/bib_ref] [bib_ref] Interleukin-4 and -10 gene polymorphisms and spontaneous preterm birth in multifetal gestations, Kalish [/bib_ref] IL-10
[formula] −1082G>A −819C>T −592C>A Th2 [/formula]
Anti-inflammatory Th2 cytokine inhibits production of cytokines, chemokines, and prostaglandins in LPS stimulated amnion, choriodecidual, and placental explants [bib_ref] Critical paracrine interactions between TNF-α and IL-10 regulate lipopolysaccharide-stimulated human choriodecidual cytokine..., Sato [/bib_ref] [bib_ref] The regulation of prostaglandin output from term intact fetal membranes by anti-inflammatory..., Brown [/bib_ref] [bib_ref] Evidence for interleukin-10-mediated inhibition of cyclo-oxygenase-2 expression and prostaglandin production in preterm..., Hanna [/bib_ref]. However, no clear association between IL-10 polymorphisms and PTL or adverse neonatal outcome [bib_ref] Preterm delivery and cytokine gene polymorphisms, Mattar [/bib_ref] [bib_ref] Inflammatory cytokine gene polymorphisms and spontaneous preterm birth, Moura [/bib_ref] [bib_ref] Genetic variations in fetal and maternal tumor necrosis factor-α and interleukin 10:..., Nuk [/bib_ref] [bib_ref] Interleukin-10 -1082 G/A polymorphism and risk of death or bronchopulmonary dysplasia in..., Yanamandra [/bib_ref].
environment. There is mounting evidence that periventricular leukomalacia and cerebral palsy are associated with fetal exposure to intra-amniotic inflammation and the development of fetal inflammatory response syndrome [bib_ref] Intrauterine infection and the development of cerebral palsy, Yoon [/bib_ref]. Thus, a strategy for targeting immune activation through the modulation of the Th1:Th2 bias may be beneficial for both the prevention of preterm labour as well as the reduction of neurological insult to the fetus.
## Progesterone.
There have been several studies indicating a positive response to progesterone treatment for the prevention of preterm labour in specific patient populations [bib_ref] Prevention of recurrent preterm delivery by 17 α-hydroxyprogesterone caproate, Meis [/bib_ref] [bib_ref] Prophylactic administration of progesterone by vaginal suppository to reduce the incidence of..., Da Fonseca [/bib_ref] [bib_ref] Progesterone and the risk of preterm birth among women with a short..., Fonseca [/bib_ref]. The strongest evidence for improvement in neonatal outcomes comes from the most recent multicentre randomised controlled trial which showed a 45% reduction in preterm labour (<33 weeks) and a 60% reduction in respiratory distress syndrome at <33 weeks using 90 mg of vaginal progesterone in women with a short cervix of 10-20 mm [bib_ref] Vaginal progesterone reduces the rate of preterm birth in women with a..., Hassan [/bib_ref]. The mechanism by which progesterone contributes to pregnancy maintenance has traditionally been attributed to maintenance of uterine quiescence by increasing cyclic AMP (cAMP) and a reduction in intracellular calcium thus reducing contractility [bib_ref] Effects of progesterone treatment on expression of genes involved in uterine quiescence, Soloff [/bib_ref]. Moreover, progesterone appears to inhibit the phosphorylation of myosin, a critical step in the activation of the myometrial contractile machinery required for labour onset [bib_ref] Biochemistry of myometrial contractibly, Egarter [/bib_ref] [bib_ref] Myometrial activation-coordination, connectivity and contractility, Macintyre [/bib_ref]. Progesterone also has immunomodulatory effects on the Th1:Th2 bias. Progesterone is able to suppress Th1 differentiation and enhance Th2 differentiation in peripheral blood mononuclear cells in vitro [bib_ref] Direct and indirect inhibition of Th1 development by progesterone and glucocorticoids, Miyaura [/bib_ref]. A more potent and orally bioavailable progestogen, dydrogesterone (6-dehydro-9β,10α-progesterone) upregulates IL-4 and downregulates IFN-γ in PHA-stimulated PBMCs more significantly than progesterone in vitro [bib_ref] Redirection of cytokine production by lymphocytes from women with pre-term delivery by..., Raghupathy [/bib_ref]. There is also in vivo evidence of an anti-inflammatory effect of prolonged administration of vaginal progesterone. In a study of pregnant women receiving either progesterone or placebo from 24 to 34 weeks, peripheral blood leukocytes were collected before and after treatment [bib_ref] Effect of prolonged in vivo administration of progesterone in pregnancy on myometrial..., Norman [/bib_ref]. mRNAs of the proinflammatory cytokines IL-1β and IL-8 were reduced with progesterone treatment, whereas the anti-inflammatory IL-10 was increased. A multicentre placebo controlled trial (OPPTI-MUM, https://www.opptimum.org.uk/: ISRCTN 14568373) powered on neonatal outcome will provide us with evidence of any potential beneficial effect of vaginal progesterone on neonates born preterm.
## Nf-κb inhibitors.
Inhibition of NF-κB activation is another attractive strategy to prevent preterm labour as NF-κB activation is central to the activation of labour-associated genes in labour [bib_ref] The role of nuclear factor kappa B in human labour, Lindström [/bib_ref]. NF-κB activation also leads to a proinflammatory response in various cytokines including IFN-γ [bib_ref] Transcriptional control of rapid recall by memory CD4 T cells, Lai [/bib_ref] , IL-1β [bib_ref] The role of nuclear factor kappa B in human labour, Lindström [/bib_ref] , TNF-α, and IL-8 [bib_ref] Molecular and biochemical mechanisms of preterm labour, Mohan [/bib_ref]. Ex vivo studies with the anti-inflammatory sulfasalazine suppress LPS-induced IL-6 and TNF-α production in fetal membranes via inhibition of translocation of p65 to the nucleus [bib_ref] Prevention of inflammatory activation of human gestational membranes in an ex vivo..., Keelan [/bib_ref]. The reported clinical safety profile of sulfasalazine has been variable [bib_ref] Treating inflammatory bowel disease during pregnancy. Risks and safety of drug therapy, Connell [/bib_ref] [bib_ref] Tolerability of aminosalicylates in inflammatory bowel disease, Ishaq [/bib_ref] [bib_ref] Therapeutic drug use in women with crohn's Mediators of Inflammation disease and..., Norgard [/bib_ref] , however, if used in pregnancy is often supplemented with folate. The anti-inflammatory characteristics of the cyclopentenone PG, 15-deoxy-Δ 12,14prostaglandin J 2 (15dPGJ 2 ) appears to be derived from its ability to inhibit NF-κB activation in human amnion and myometrial cell culture [bib_ref] 15-Deoxy-δ12,14-prostaglandin J2 inhibits interleukin-1β-induced nuclear factor-κB in human amnion and myometrial cells:..., Lindström [/bib_ref]. We have also shown that 15dPGJ 2 inhibits activation of NF-κB in human peripheral blood mononuclear cells and reduces the percentage of cells producing the proinflammatory cytokines, IFN-γ and TNF-α, [bib_ref] Changes in the Th1:Th2 cytokine bias in pregnancy and the effects of..., Sykes [/bib_ref]. Work conducted in our laboratory has also shown that 15dPGJ 2 is able to delay labour and provide neuroprotection by reducing pup mortality from 75% to 5% in a murine model of inflammation induced preterm labour [bib_ref] The cyclopentenone 15-deoxy-delta 12,14-prostaglandin J(2) delays lipopolysaccharide-induced preterm delivery and reduces mortality..., Pirianov [/bib_ref].
# Conclusion
There has been extensive interest in the Th1:Th2 dichotomy for the maintenance of successful pregnancy. A trend towards the Th2 cytokine profile and a suppression of the Th1 cytokine profile appears to exist both in the peripheral blood but more significantly at the maternal fetal interface. Activation of the proinflammatory Th1 profile-rather than suppression of the Th2 profile-is apparent in preterm labour and thus should be considered as the logical target for immunomodulating therapies for the prevention of preterm labour and improving neonatal outcome.
[table] Table 1: Cytokine polymorphism associations with preterm labour (PTL). [/table]
|
Double mutation (R124H, N544S) of TGFBI in two sisters with combined expression of Avellino and lattice corneal dystrophies
Purpose:The R124H mutation of the keratoepithelin gene (TGFBI) causes Avellino corneal dystrophy whereas the N544S mutation of this same gene gives rise to lattice corneal dystrophy. We now report two cases with both R124H and N544S mutations of TGFBI. Methods: Genomic DNA and cDNA were isolated from the proband and family members and were subjected to polymerase chain reaction-mediated amplification of exons 1-17 of TGFBI. The amplification products were directly sequenced. Allele-specific cloning and sequencing were applied to evaluate the compound heterozygous mutation. Results: Molecular genetic analysis revealed that the proband and one sister harbored both a heterozygous CGC→CAC (Arg→His) mutation at codon 124 and a heterozygous AAT→AGT (Asn→Ser) mutation at codon 544 of TGFBI. Slitlamp examination revealed multiple granular regions of opacity and lattice lines in the corneal stroma of the proband and her sister with the double mutation. Allele-specific cloning and sequencing revealed that the R124H and N544S mutations are on different chromosomes. Conclusions: As far as we are aware, this is the first report of a patient with a double mutation (R124H, N544S) of TGFBI causing an autosomal dominant form of corneal dystrophy. The clinical manifestations of the two cases with both R124H and N544S mutations appeared to be a summation of Avellino and lattice corneal dystrophies. Molecular Vision 2009; 15:974-979 <http://www.molvis.org/molvis/v15/a102>
Mutations of the keratoepithelin gene (TGFBI) are responsible for most corneal dystrophies. TGFBI was first identified as a transforming growth factor-β1 (TGF-β1)inducible gene in a human lung adenocarcinoma cell line [bib_ref] cDNA cloning and sequence analysis of beta ig-h3, a novel gene induced..., Skonier [/bib_ref]. The point mutations R124C, R124H, R555W, and R555Q of TGFBI were initially found to give rise to lattice corneal dystrophy (LCD), Avellino corneal dystrophy (ACD), Groenouw type I corneal dystrophy, and Reis-Bücklers corneal dystrophy, respectively [bib_ref] Kerato-epithelin mutations in four 5q31-linked corneal dystrophies, Munier [/bib_ref]. Many additional mutations of TGFBI were subsequently found to be responsible for autosomal dominant corneal dystrophies [bib_ref] BIGH3 mutation spectrum in corneal dystrophies, Munier [/bib_ref] [bib_ref] TGFBI gene mutations in corneal dystrophies, Kannabiran [/bib_ref]. ACD is characterized by the presence of granular and linear opacities in the corneal stroma. The deposits in the corneal stroma of patients with ACD are of a hyaline and amyloid nature. The only identified mutation associated with ACD is R124H of TGFBI [bib_ref] Kerato-epithelin mutations in four 5q31-linked corneal dystrophies, Munier [/bib_ref]. LCD is an inherited form of amyloidosis that is characterized by the development of lattice lines and opacities in the cornea. Several distinct mutations of TGFBI including R124C [bib_ref] Kerato-epithelin mutations in four 5q31-linked corneal dystrophies, Munier [/bib_ref] , L518P [bib_ref] Leu518Pro mutation of the beta ig-h3 gene causes lattice corneal dystrophy type..., Endo [/bib_ref] , P501T [bib_ref] A kerato-epithelin (betaig-h3) mutation in lattice corneal dystrophy type IIIA, Yamamoto [/bib_ref] , L527R [bib_ref] A new L527R mutation of the betaIGH3 gene in patients with lattice..., Fujiki [/bib_ref] , N544S [bib_ref] Association of autosomal dominantly inherited corneal dystrophies with BIGH3 gene mutations in..., Mashima [/bib_ref] , A546T [bib_ref] A new mutation (A546T) of the betaig-h3 gene responsible for a French..., Dighiero [/bib_ref] , and N622K (T1913G or T1913A) [bib_ref] BIGH3 mutation spectrum in corneal dystrophies, Munier [/bib_ref] have been associated with LCD. LCD is classified clinically [bib_ref] BIGH3 mutation spectrum in corneal dystrophies, Munier [/bib_ref] [bib_ref] TGFBI gene mutations in corneal dystrophies, Kannabiran [/bib_ref] , but standardized definitions of each subtype have not been achieved to date. The subtype of LCD caused by the N544S mutation of TGFBI is characterized by tiny nodular deposits with thin lattice lines in the middle portion of the corneal stroma.
Several case reports have suggested that corneal dystrophies caused by homozygous point mutations of TGFBI are characterized by an earlier onset, more severe symptoms, and a higher frequency of recurrence after keratoplasty compared with those attributable to the corresponding heterozygous mutations [bib_ref] Severe form of juvenile corneal stromal dystrophy with homozygous R124H mutation in..., Mashima [/bib_ref] [bib_ref] Granular corneal dystrophy with homozygous mutations in the kerato-epithelin gene, Okada [/bib_ref] [bib_ref] Severe corneal dystrophy phenotype caused by homozygous R124H keratoepithelin mutations, Okada [/bib_ref] [bib_ref] Homozygotic patient with betaig-h3 gene mutation in granular dystrophy, Fujiki [/bib_ref] [bib_ref] Different recurrence patterns after phototherapeutic keratectomy in the corneal dystrophy resulting from..., Inoue [/bib_ref]. A few case reports have also described individuals with corneal dystrophy who harbor two distinct mutations in TGFBI, the membrane component, chromosome 1, surface maker 1 (M1S1), or both [bib_ref] Q118X mutation of M1S1 gene caused gelatinous drop-like corneal dystrophy: the P501T..., Ha [/bib_ref] [bib_ref] Compound heterozygous mutations of M1S1 gene in gelatinous droplike corneal dystrophy, Tian [/bib_ref] [bib_ref] A novel missense mutation in a Japanese patient with gelatinous droplike corneal..., Taniguchi [/bib_ref] [bib_ref] Two mutations in the TGFBI (BIGH3) gene associated with lattice corneal dystrophy..., Klintworth [/bib_ref] [bib_ref] Lattice corneal dystrophy associated with the Ala546Asp and Pro551Gln missense changes in..., Aldave [/bib_ref] [bib_ref] A novel variant of granular corneal dystrophy caused by association of 2..., Dighiero [/bib_ref]. It has remained unclear, however, how the phenotype of patients with such a double mutation differs from that of those with the corresponding single mutations. We now describe the first cases of corneal dystrophy associated with both R124H and N544S mutations of TGFBI.
# Methods
This study was approved by the ethical review committee for gene analysis research of Yamaguchi University School of Medicine and Yamaguchi University Hospital. After obtaining informed written consent, we extracted genomic DNA from white blood cells of peripheral blood collected from patients in the presence of an anticoagulant. Total RNA was also extracted from the white blood cells with the use of a QIAmp RNA Blood mini kit (Qiagen, Valencia, CA) and was then subjected to reverse transcription with the use of TaqMan Reverse Transcription Reagents (Applied Biosystems, Foster City, CA). The resulting cDNA as well as genomic DNA were subjected to polymerase chain reaction (PCR) with primers that amplify exons 1, 4, 11, 12, 13, 14, 2-9, or 9-17 of TGFBI [fig_ref] TABLE 1: PCR PRIMERS USED FOR SEQUENCING EXONS OF TGFBI [/fig_ref] Taq; Takara, Tokyo, Japan). The reaction mixture was overlaid with 10 μl of mineral oil, and amplification was performed with a Gene Amp PCR System PC808 (ASTEC, Tokyo, Japan) with an initial denaturation at 95 °C for 2 min followed by 30 cycles of denaturation at 94 °C for 30 s, annealing at 58 °C, 60 °C, or 62 °C [fig_ref] TABLE 1: PCR PRIMERS USED FOR SEQUENCING EXONS OF TGFBI [/fig_ref] for 20 s, and extension at 72 °C for 30 s. The PCR products were separated by electrophoresis on a 2% agarose gel and stained with ethidium bromide. For sequencing, 2.5 µl of the PCR products were incubated with 1 μl of ExoSAP-IT (Amersham Bioscience, Tokyo, Japan) first for 20 min at 37 °C and then for another 20 min at 80 °C. Sequencing reactions were then performed with the use of a BigDye Terminator Cycle Sequencing FS Ready Reaction Kit (Applied Biosystems). After purification with ethanol, the reaction products were applied to an ABI 3100-Avant Genetic Analyzer (Applied Biosystems).
An allele-specific cloning and sequencing approach was applied to characterize the compound heterozygous mutation of R124H and N544S. In brief, cDNA of the proband was subjected to PCR with KOD FX DNA polymerase (Toyobo, Tokyo, Japan) and with the primers, 5′-TGT CCA GCA GCC CTA CCA CTC-3′ (forward) and 5′-AGG ATA TCC CCT CTT TCC TGA GGT C-3′ (reverse; containing an EcoRV restriction site at its 5′ end), to obtain products that included both mutation sites. The PCR products were purified by electrophoresis and digested with EcoRV and BamHI (site in exon 4), and the released fragments were ligated into the multiple cloning site of a sequencing vector (pcDNA3.1[+]; Promega, Madison, WI). The resulting plasmids were expanded in competent Escherichia coli JM109 cells (Invitrogen, Carlsbad, CA), and the inserts were then sequenced as described above.
# Results
The proband, a 67-year-old Japanese woman (II-1), visited our corneal clinic in January 2000 with a main complaint of gradual impairment of vision. We diagnosed her condition as ACD on the basis of slit-lamp examination. [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] , [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] , and [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] [fig_ref] Figure 3: Lattice lines in patient's cornea [/fig_ref]. The lattice lines are easily seen in A and C (black arrows), but not in B.
brothers and two sisters. Her father's brother (I-2) and her sisters (II-2, II-3) were also diagnosed at our clinic with ACD by slit-lamp examination. Her reporting suggested that her father (I-1) had corneal dystrophy. We also performed phototherapeutic keratectomy on the left eye of II-2 in March 2000 and on the right eye of II-2 in May 2000.
Slit-lamp examination subsequently revealed multiple granular regions of opacity in the surface-to-middle portion of the corneal stroma in both eyes of II-1, II-2, and II-3. Lattice lines were also observed in II-1 [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] and II-3 [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] but not in II-2 [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref]. These lattice lines can be seen better in the higher magnifications of [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] , [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] and [fig_ref] Figure 2: Slit-lamp photographs of the proband and her two sisters [/fig_ref] [fig_ref] Figure 3: Lattice lines in patient's cornea [/fig_ref]. Both II-1 and II-3 were found to harbor both a heterozygous CGC→CAC (Arg→His) mutation at codon 124 and a heterozygous AAT→AGT (Asn→Ser) mutation at codon 544 of TGFBI whereas II-2 harbored only the heterozygous CGC→CAC (Arg→His) mutation at codon 124 . The mutations were identified at both the genomic and cDNA levels.
To investigate whether the two TGFBI mutations are on the same or different chromosomes of the proband, we adopted an allele-specific cloning and sequencing approach. PCR products containing both mutation sites were subcloned and sequenced. Of three independent clones analyzed, one contained only the R124H mutation and the other two contained only the N544S mutation, indicating that the two mutations are on different chromosomes.
# Discussion
As far as we are aware, this is the first report of a patient with a double mutation of TGFBI causing an autosomal dominant form of corneal dystrophy. The clinical manifestations of the two cases with both R124H and N544S mutations appeared to be a summation of those of Avellino and lattice corneal dystrophies. We observed lattice lines in the corneas of II-1 and II-3, both of whom have the N544S mutation of TGFBI, but not in II-2, who harbors only the R124H mutation.
We were not able to perform genetic analysis on I-1 and I-2 because they were no longer alive at the time of this analysis. However, allele-specific cloning and sequencing revealed that the R124H and N544S mutations are on different chromosomes, consistent with our clinical findings. Slit-lamp examination of I-2 did not reveal the presence of lattice lines, suggesting that the R124H mutation was transmitted to the proband and her two sisters from I-1. Although slit-lamp examination was not performed on the mother of the three sisters because of her being confined to bed, it is likely that she harbors the N544S mutation of TGFBI. Given that the clinical manifestation of the N544S mutation has a late onset and that the mutation does not have a pronounced effect on visual acuity, the mother may not experience a visual disturbance.
Several cases of double mutations associated with corneal dystrophies other than macular corneal dystrophy have been described previously . However, no case of a double mutation of TGFBI causing an autosomal dominant form of corneal dystrophy has previously been reported. The presence of a homozygous Q118X mutation of M1S1 and a heterozygous P501T mutation of TGFBI in the same individual was described [bib_ref] Q118X mutation of M1S1 gene caused gelatinous drop-like corneal dystrophy: the P501T..., Ha [/bib_ref]. The Q118X mutation of M1S1 causes gelatinous drop-like corneal dystrophy (GDLD) with an autosomal recessive mode of inheritance. The P501T mutation of TGFBI causes LCD type IIIA [bib_ref] A kerato-epithelin (betaig-h3) mutation in lattice corneal dystrophy type IIIA, Yamamoto [/bib_ref]. The clinical manifestation in this patient resembled that of GDLD but not that of LCD type IIIA. A patient with a clinical diagnosis of GDLD and heterozygous Q118X and Y184C mutations of M1S1 has also been described [bib_ref] Compound heterozygous mutations of M1S1 gene in gelatinous droplike corneal dystrophy, Tian [/bib_ref]. No other case of the Y184C mutation in M1S1 has been presented, so it is not clear whether this mutation in the homozygous state can cause GDLD. A patient with a clinical diagnosis of GDLD was found to be heterozygous for both Q118X and L186P . Genetic analysis of TGFBI in the proband and her two sisters. Direct sequencing of genomic amplification products corresponding to exon 4 (upper panels) or exon 12 (lower panels) of TGFBI was performed for II-1, II-2, and II-3. A heterozygous CGC→CAC mutation was detected at codon 124 in II-1, II-2, and II-3. A heterozygous AAT→AGT mutation was detected at codon 544 in II-1 and II-3. mutations of M1S1 [bib_ref] A novel missense mutation in a Japanese patient with gelatinous droplike corneal..., Taniguchi [/bib_ref]. Patients with a clinical diagnosis of atypical LCD were found to be heterozygous for both A546D and P551Q mutations of TGFBI [bib_ref] Two mutations in the TGFBI (BIGH3) gene associated with lattice corneal dystrophy..., Klintworth [/bib_ref] [bib_ref] Lattice corneal dystrophy associated with the Ala546Asp and Pro551Gln missense changes in..., Aldave [/bib_ref]. The A546D mutation of TGFBI causes polymorphic corneal amyloidosis [bib_ref] Polymorphic corneal amyloidosis: a disorder due to a novel mutation in the..., Eifrig [/bib_ref] or atypical LCD [bib_ref] The TGFBI A546D mutation causes an atypical type of lattice corneal dystrophy, Correa-Gomez [/bib_ref] with an autosomal dominant mode of inheritance. There have been no other reports of the P551Q mutation of TGFBI, so it is not clear whether a heterozygous P551Q mutation causes corneal dystrophy. Finally, a patient with a clinical diagnosis of granular corneal dystrophy was found to be heterozygous for both R124L and ΔT125-ΔE126 mutations of TGFBI [bib_ref] A novel variant of granular corneal dystrophy caused by association of 2..., Dighiero [/bib_ref]. There have been no other reports of the ΔT125-ΔE126 mutation of TGFBI.
A few studies have addressed the penetrance of inherited corneal dystrophy. LCD type IIIA caused by the P501T mutation of TGFBI [bib_ref] Q118X mutation of M1S1 gene caused gelatinous drop-like corneal dystrophy: the P501T..., Ha [/bib_ref] and atypical granular corneal dystrophy caused by the D123H mutation of TGFBI [bib_ref] A novel mutation of the TGFBI gene found in a Vietnamese family..., Ha [/bib_ref] are thought to have a low penetrance. Non-penetrance of ACD has also been described [bib_ref] Phenotypic non-penetrance in granular corneal dystrophy type II, Kim [/bib_ref]. The penetrance of corneal dystrophies caused by the R124H or N544S mutations of TGFBI remains unclear.
In all previously reported cases of double mutations, the clinical phenotype resembled that of one but not both of the associated corneal dystrophies. In the cases described in the present study, the phenotype associated with the double mutation is the summation of both corneal dystrophies. These cases thus indicate that R124H and N544S mutations of TGFBI independently determine clinical manifestation.
[fig] Figure 1: Each PCR reaction was performed in a total volume of 10 μl containing template DNA (80 ng/μl), 10 pmol of each primer, 200 μM of each deoxynucleoside triphosphate, 20 mM MgCl2, 20 mM Tris-HCl (pH 8.0), 100 mM KCl, and 1 U of Taq polymerase (Ex Pedigree of the proband. Black symbols indicate individuals with a diagnosis of corneal dystrophy by genetic analysis. Gray symbols indicate individuals suspected of having been affected by corneal dystrophy but not subjected to genetic analysis. The arrow indicates the proband. [/fig]
[fig] Figure 2: Slit-lamp photographs of the proband and her two sisters. Slit-lamp photographs of the right eye of II-1 (A-C), the left eye of II-2 (D-F), and the left eye of II-3 (G-I) are shown. Granular deposits (gray arrowheads) and star-shaped deposits (black arrowheads) were observed in all three patients (C,F,I) whereas thin lattice lines (black arrows) were observed only in II-1 (C) and II-3 (I). Nodular deposits were apparent mostly in the superficial-to-middle portion of the corneal stroma in all three patients (B,E,H). [/fig]
[fig] Figure 3: Lattice lines in patient's cornea. The lattice lines referred to inFigure 2are better visualized in the higher magnifications of [/fig]
[table] TABLE 1: PCR PRIMERS USED FOR SEQUENCING EXONS OF TGFBI. [/table]
|
Circulating tumor DNA in hepatocellular carcinoma: trends and challenges
Molecular characterization of individual patients' tumor cells is becoming increasingly important in offering effective treatment for patients in clinical practice. Recent advances in the field have indicated that circulating tumor DNA (ctDNA) has huge potential to serve as a biomarker for early detection and precision treatment as well as prognosis of hepatocellular carcinoma (HCC). As ctDNA in HCC patients harbors the molecular characteristics of HCC tumor cells, ctDNA analysis in the blood may be sufficient for convenient, non-invasive and accurate detection, providing information for HCC diagnosis, treatment and prognosis. In this review, we will summarize and discuss current trends and challenges of ctDNA application in HCC.
# Background
Liver cancer, with 782,500 new cases and 745,500 deaths occurring worldwide in 2012, is the second leading cause of global cancer death, with China alone accounting for about 50 % of the total number of cases and deaths [bib_ref] Global cancer statistics, Torre [/bib_ref]. It was estimated that 466,100 new cases and 422,100 deaths would occur in China in 2015, accounting for about 15 % of all cancer deaths in China [bib_ref] Cancer statistics in China, Chen [/bib_ref]. Primary liver cancer includes three histologic subtypes, hepatocellular carcinoma (HCC), cholangiocarcinoma and combined hepatocellular and cholangiocarcinoma. HCC, which is the major histological subtype of primary liver cancers, accounts for between 85 and 90 % of all cases worldwide [bib_ref] Hepatocellular carcinoma: epidemiology and molecular carcinogenesis, El-Serag [/bib_ref]. A key to effective prevention and treatment of HCC is early diagnosis of HCC.
The diagnosis of HCC is increasingly made with the use of noninvasive imaging tests such as ultrasonography, computed tomography (CT) and magnetic resonance tomography (MRI), along with use of an alpha-fetoprotein (AFP) level, a predictive biomarker for HCC. Imaging tests can only determine HCC with confidence to some degree when nodules, benign or malignant, grow to at least 1 cm in size. Invasive biopsy is considered for diagnosis of HCC when imaging tests are less assuring. In either case, patients may by then have malignant tumors in an advanced stage, with limited treatment options and poor prognosis. Meanwhile, it is recognized that not all HCC can produce a higher level of AFP [bib_ref] Guide for diagnosis and treatment of hepatocellular carcinoma, Attwa [/bib_ref]. In fact, if early-stage HCC, currently difficult to diagnose and characterize, can be detected, it can be effectively treated by surgical resection with an 5-year survival rate of 90 % [bib_ref] Hepatocellular carcinoma: epidemiology and molecular carcinogenesis, El-Serag [/bib_ref]. In addition to surgical resection, several options, including liver transplantation, transarterial chemoembolization (TACE), radiofrequency ablation, high-intensity focused ultrasound and targeted molecular therapy (e.g. sorafenib treatment), are currently used in the clinic to treat HCC. The effectiveness of these treatments can be significantly improved with early detection and convenient monitoring for possible HCC relapse following treatment. Thus clinician and scientists in the field have been actively developing sensitive, reliable and convenient methods for surveillance and early detection of HCC and post-treatment monitoring of HCC relapse. Hepatocellular tumorigenesis is a slow, progressive and complex process due to the accumulation of genetic and epigenetic alterations in hepatocytes whose activities intimately interact with surrounding microenvironment
## Open access
Cell & Bioscience [bib_ref] Genetic and epigenetic aspects of initiation and progression of hepatocellular carcinoma, Kanda [/bib_ref]. Despite serious challenges in detecting these genetic and epigenetic alterations, one method that holds great promise is the detection of circulating tumor DNA (ctDNA) in the peripheral blood of HCC patients. This method, along with circulating tumor cells, is termed "liquid biopsy". By decoding the information of nucleic acids from patients' serum or plasma, not only can clinicians make accurate diagnosis and proper treatment on HCC patients, but the scientists can also use this "liquid biopsy" technology to better understand the biology of HCC and help clinicians to design cancer therapy.
## Ctdna of hcc
Among varieties of circulating cell-free DNA (cfDNA), ctDNA is released into circulation specifically from tumor cells that undergo metabolic secretion, apoptosis or necrosis [fig_ref] Figure 1: ctDNA release and extraction in HCC patients [/fig_ref]. cfDNA is defined as extracellular DNA present in plasma or serum samples. It can be detected not only in patients suffering from cancer or other diseases but also in healthy individuals. Unlike normal cfDNA, ctDNA carries tumor-specific genetic or epigenetic alterations, such as point mutations, copy number variations, chromosomal rearrangements, DNA methylation patterns, etc. Thus, a minimally invasive examination of ctDNA with a small amount of peripheral blood from patients could reveal genetic and epigenetic alterations related to specific cancer and its metastatic state, offering a unique opportunity for serially monitoring tumor genomes in a non-invasive, convenient and accurate manner.
Since 2006, cancer genome sequencing has delivered robust and comprehensive tumor genome information, which provides a fertile ground for the development of ctDNA testing. Along with the advancement of sequencing technology, huge effort and investment are being taken worldwide in putting this "liquid biopsy" into clinical practice. Potential applications of ctDNA testing in HCC patients may include: (a) early detection of cancer; (b) monitoring of tumor heterogeneity and metastasis; (c) identification of therapeutic targets; (d) real-time evaluation of treatment response and tumor relapse; and (e) real-time assessment of evolution of drug resistance [fig_ref] Figure 2: Monitoring response and relapse of HCC patients with targeted therapies by ctDNA... [/fig_ref]. These applications require detection of genetic and epigenetic alterations in ctDNA specifically associated with different stages of HCC (e.g. hepatocellular dysplasia, early HCC, progressed HCC and metastatic HCC) and with different treatment options or stages of treatment. However, progress is limited in exploring the clinical utility of ctDNA in cancer diagnosis, treatment and prognosis, particularly in HCC, due to serious technological obstacles in detection and analysis of ctDNA.
## Detection and analysis of ctdna in hcc patients
The pioneer work by Mandel and Metais welcomed the discovery of cfDNA in 1948 [bib_ref] Les acides nucléiques du plasma sanguin chez l'homme, Mandel [/bib_ref]. But during the following two decades, we witnessed the futility in this field. Until 1977, by comparing the blood samples between 173 cancer patients and 53 healthy individuals using a radioimmunoassay that can detect DNA in nanogram, Leon et al. found the cancer patients had a relative higher level of cfDNA than healthy controls and an increased level after radiation therapy [bib_ref] Free DNA in the serum of cancer patients and the effect of..., Leon [/bib_ref]. This finding exhibited the potential of cfDNA as a cancer biomarker, which can also be applied to patients with HCC. In 1989, ctDNA was noted to be a fraction of cfDNA in the blood [bib_ref] Neoplastic characteristics of the DNA found in the plasma of cancer patients, Stroun [/bib_ref]. Two groups recently found that the cfDNA levels in patients with hepatitis C virus (HCV)-related HCC were correlated with the overall survival and extrahepatic recurrence in distant organs after curative hepatectomy [bib_ref] Relation between serum levels of cell-free DNA and inflammation status in hepatitis..., Iida [/bib_ref] [bib_ref] Circulating cell-free DNA as a predictive marker for distant metastasis of hepatitis..., Tokuhisa [/bib_ref]. In addition, several studies have examined global alterations of cfDNA, such as DNA methylation alterations, microsatellite alterations, point mutations, chromosomal rearrangements and viral DNA integration [fig_ref] Table 1: Detection of cfDNA and its alterations in HCC patients* Includes all articles... [/fig_ref] , some of which have genetic and epigenetic alteration patterns similar to those detected in primary tumors. Among a pool of cfDNA, ctDNA likely harbors these alterations, demonstrating the potential utility of ctDNA in clinical applications.
## Methylation alterations
Methylation alterations occur on many genes associated with initiation and progression of HCC. Several studies have revealed the alterations of DNA methylation in HCC patients' tumor tissues including the aberrant methylation of the promoter region of Glutathione S-transferase P1 (GSTP1) [bib_ref] GSTP1 CpG island DNA hypermethylation in hepatocellular carcinomas, Tchou [/bib_ref] [bib_ref] Silencing of GSTP1 gene by CpG island DNA hypermethylation in HBV-associated hepatocellular..., Zhong [/bib_ref] and the cyclin-dependent kinase inhibitor p15 [bib_ref] Frequent p15 promoter methylation in tumor and peripheral blood from hepatocellular carcinoma..., Wong [/bib_ref] and p16 [bib_ref] p16(INK4) is inactivated by extensive CpG methylation in human hepatocellular carcinoma, Matsuda [/bib_ref]. Efforts have been made to detect such methylation alterations of cfDNA in HCC patients' blood. The successful detection of hypermethylated GSTP1 [bib_ref] Detection of aberrant promoter methylation of GSTP1 in the tumor and serum..., Wang [/bib_ref] , p15 [bib_ref] Frequent p15 promoter methylation in tumor and peripheral blood from hepatocellular carcinoma..., Wong [/bib_ref] and p16 [bib_ref] Detection of aberrant p16 methylation in the plasma and serum of liver..., Wong [/bib_ref] in cfDNA from HCC patients may allow the development of a bloodbased assay for HCC diagnosis. Methylation alterations of RAS association domain family 1A (RASSF1A) were also detected in cfDNA of HCC patients and account for 40 % cases of matched plasma [bib_ref] High frequency of promoter hypermethylation of RASSF1A in tumor and plasma of..., Yeo [/bib_ref]. Furthermore, hepatitis B virus (HBV) carriers undergoing surveillance and subsequently developing HCC had significantly higher levels of RASSF1A from the time of enrollment to cancer diagnosis [bib_ref] Quantitative analysis of circulating methylated DNA as a biomarker for hepatocellular carcinoma, Chan [/bib_ref]. Another gene with methylation abnormality that has been detected in HCC patients is long interspersed nucleotide elements (LINE-1) [bib_ref] Serum LINE-1 hypomethylation as a potential prognostic marker for hepatocellular carcinoma, Tangkijvanich [/bib_ref]. The progression and invasiveness of HCC are highly associated with elevated hypomethylated LINE-1. These lines of evidence indicate that a combined assessment of circulating methylated DNA may yield a promising tool in HCC diagnosis and management.
Methylation-based assays on ctDNA may offer the best hope for early detection, as some of methylation changes are mechanistically early events in tumor progression. Unlike genetic alterations such as mutations and deletions, epigenetic changes are also potentially reversible [bib_ref] PCR-based methods for detecting singlelocus DNA methylation biomarkers in cancer diagnostics, prognostics,..., Kristensen [/bib_ref] and can therefore be targeted for prevention of tumor initiation and progression. Compared with the detection of cfDNA with microsatellite instability as described below, the methylation-based approach is more sensitive and generates fewer false negatives [bib_ref] Detection of aberrant p16 methylation in the plasma and serum of liver..., Wong [/bib_ref]. In contrast, mutations occur in a large number and in varied frequencies, and many of highly frequent mutations are associated with different cancers. This complicates the use of mutations for HCC diagnosis by ctDNA analysis. It appears that the methylation approach is better suited at present for HCC screening using plasma or serum samples.
## Microsatellite alterations
Highly polymorphic DNA repeat regions, termed microsatellites, are commonly present in eukaryotic genomes. Loss and length alteration in microsatellites are frequent in a variety of cancers, providing a set of markers suitable for diagnostic detection. Since the discovery of tumor-derived microsatellite alterations in cfDNA [bib_ref] Microsatellite alterations in serum DNA of head and neck cancer patients, Nawroz [/bib_ref] [bib_ref] Microsatellite alterations in plasma DNA of small cell lung cancer patients, Chen [/bib_ref] , interest in uncovering these alterations of cfDNA is growing. The comparative genomic hybridization (CGH) technique has enabled scientists to define some microsatellite alterations in HCC such as chromosome 8p, 17p and 19p deletions [bib_ref] Allelic imbalance regions on chromosomes 8p, 17p and 19p related to metastasis..., Zhang [/bib_ref] [bib_ref] The prognostic value of circulating plasma DNA level and its allelic imbalance..., Ren [/bib_ref] , and the loss on 8p and 19p might contribute to HCC metastasis [bib_ref] Allelic imbalance regions on chromosomes 8p, 17p and 19p related to metastasis..., Zhang [/bib_ref] [bib_ref] The association of chromosome 8p deletion and tumor metastasis in human hepatocellular..., Qin [/bib_ref].
In the early 20th century, a microsatellite marker screen was performed in primary tumor and serum sample in 21 cases of HCC patients, 76.2 percent of which harbor at least one allelic imbalance [bib_ref] Molecular diagnosis of primary liver cancer by microsatellite DNA analysis in the..., Chang [/bib_ref] , providing early evidence for clinical utility of this approach. A combined set of microsatellite loci gives a higher probability of detecting a risk population. Two microsatellite markers on [bib_ref] Allelic imbalance regions on chromosomes 8p, 17p and 19p related to metastasis..., Zhang [/bib_ref]. Interestingly, only an allelic imbalance at D8S258 was found in the cfDNA of HCC patients, and combination of both the allelic imbalance and a higher level of cfDNA is well correlated with the decrease in disease-free and overall survival rates [bib_ref] The prognostic value of circulating plasma DNA level and its allelic imbalance..., Ren [/bib_ref].
## Point mutations
Tumor progression involves the accumulation of both inactivation of tumor suppressor genes and activation of proto-oncogenes, for instance, TP53 and CTNNB1 in HCC. Ser249 of TP53 is the most reported mutation hotspot in HCC patients, and mutation of this site leads to the deficiency in its specific DNA binding ability [bib_ref] Selective G to T mutations of p53 gene in hepatocellular carcinoma from..., Bressac [/bib_ref] [bib_ref] p53 mutations in human cancers, Hollstein [/bib_ref] [bib_ref] Hepatocellular carcinoma: from gene to public health, Montesano [/bib_ref] [bib_ref] Specific p53 mutations detected in plasma and tumors of hepatocellular carcinoma patients..., Jackson [/bib_ref] [bib_ref] Ser-249 p53 mutations in plasma DNA of patients with hepatocellular carcinoma from..., Kirk [/bib_ref] [bib_ref] Ser-249TP53 mutation in tumour and plasma DNA of hepatocellular carcinoma patients from..., Szymanska [/bib_ref]. Recently, TP53 Ser249 mutant in plasma has been reported to be highly associated with cirrhosis and HCC in China and Africa [bib_ref] The Gambia liver cancer study: infection with hepatitis B and C and..., Kirk [/bib_ref] [bib_ref] 249(ser) TP53 mutation in plasma DNA, hepatitis B viral infection, and risk..., Kirk [/bib_ref] [bib_ref] Ser-249 TP53 and CTNNB1 mutations in circulating free DNA of Egyptian patients..., Hosny [/bib_ref] , a region with high HBV prevalence and high Aflatoxin B1 exposure. Interestingly, this mutation was also detected in noncancerous hepatic tissues of HCC [bib_ref] Geographic variation of p53 mutational profile in nonmalignant human liver, Aguilar [/bib_ref] , in plasma DNA of a few healthy individuals and in patients with relatively more severe cirrhosis [bib_ref] 249(ser) TP53 mutation in plasma DNA, hepatitis B viral infection, and risk..., Kirk [/bib_ref] , indicating this mutation might be involved in early development of HCC and accumulate during HCC progression. However, since this mutation, like many other mutations, occurs in other types of cancers, it cannot be excluded that ctDNA harboring this mutation is released from other tissues.
## Chromosomal rearrangements
Genomic sequencing has revealed many recurrent chromosomal rearrangements including deletions, insertions, amplifications, translocations and more complex rearrangements in HCC. Detection of such chromosomal rearrangements in circulating tumor cell requires highly sensitive PCR. To date, rearrangement detection assay in plasma has just succeeded in a small population of patients, mostly in hematological malignancies. Nevertheless, whole genome sequencing of ctDNA provides the opportunity to identify chromosomal rearrangements or copy number changes in HCC patients, and ultimately offer a reliable and robust method for HCC detection. In addition, some non-coding DNA, such as LINE-1 which distributes throughout the genome, significantly increases in the serum of patients with HCC, in a hypomethylated form [bib_ref] Serum LINE-1 hypomethylation as a potential prognostic marker for hepatocellular carcinoma, Tangkijvanich [/bib_ref] , providing a target for ctDNA detection. Several other DNA abnormalities have also been investigated. For instance, the presence of longer relative telomere length (RTL) predicts an elevated risk for non-cirrhotic HCC patients with HBV [bib_ref] Relative telomere length: a novel non-invasive biomarker for the risk of non-cirrhotic..., Fu [/bib_ref] while telomere shortening occurs in cirrhosis samples. This difference could be used as a biomarker in blood to distinguish stages of severe liver diseases.
## Viral dna
Virus infection is the major contribution to severe liver diseases. HBV and HCV are important etiological factors for HCC, and the combined fraction of cases attributable to virus infection is estimated to 75 % of all HCC cases. The specific virus DNA level may potentially be used as molecular biomarkers of disease activity. High level of serum HBV DNA has a strong association with the incidence of HCC. Interestingly, the HBV DNA-based prediction is independent of the level of hepatitis B surface antigen or alanine aminotransferase level, and the presence of cirrhosis [bib_ref] Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B..., Chen [/bib_ref] , making it a sensitive and reliable tool for monitoring the disease. The detection of circulating HBV DNA is also applied for patients who underwent transcatheter arterial embolization (TAE). TAE is an important palliative treatment for HCC patients who are poor candidates for surgery or percutaneous ablative therapy. A recent study showed that the elevated plasma HBV DNA persistently correlates with lipiodol retention, but not with age or tumor size, making it an early indicator to assess the success or failure of TAE [bib_ref] Quantitative analysis of plasma HBV DNA for early evaluation of the response..., Su [/bib_ref]. In addition, some conservative mutations in HBV provide new targets in plasma detection, such as a double 1762T/1764A mutation in HBV genome in HCC tumors from Qidong area, China [bib_ref] Specific mutations of hepatitis B virus in plasma predict liver cancer development, Kuang [/bib_ref].
## Challenges in clinical utility of ctdna in hcc
It has been demonstrated that the concentration of ctDNA in the plasma increased along with stages of several human malignancies [bib_ref] Detection of circulating tumor DNA in early-and late-stage human malignancies, Bettegowda [/bib_ref] , providing potential utility of ctDNA in diagnosing advanced stages of cancer and monitoring cancer relapse following cancer treatment, especially for breast cancer and lung cancer patients [bib_ref] Analysis of circulating tumor DNA to monitor metastatic breast cancer, Dawson [/bib_ref] [bib_ref] An ultrasensitive method for quantitating circulating tumor DNA with broad patient coverage, Newman [/bib_ref]. In a recent study, ctDNA-based detection preceded clinical detection of metastasis for 86 % patients with an average lead time of 11 months following primary surgery for primary breast cancer patients [bib_ref] Serial monitoring of circulating tumor DNA in patients with primary breast cancer..., Olsson [/bib_ref]. However, it remains a challenge in applying the ctDNA technology in early detection of cancer including HCC. Even for advanced stage of HCC, this "liquid biopsy" methodology is yet to be established for clinical applications largely due to low level of ctDNA, poorly characterized genetic and epigenetic alterations in HCC patients and high level of tumor heterogeneity of HCC.
## Low levels of ctdna
Over past decades, many methodologies, such as Sanger sequencing (dideoxy-terminator sequencing), pyrosequencing, next-generation sequencing, real-time PCR, and amplification refractory mutation system (ARMS), have been employed to detect genetic alterations in tumor tissues. However, these methodologies are not sensitive or accurate enough to quantitatively assess cfDNA in cancer patients for clinical purposes due to several limitations. First, cfDNA is only in trace amount in the serum and plasma of cancer patients and is too low to be efficiently isolated. Secondly, ctDNA represents only a very small fraction of cfDNA, making it extremely difficult to detect. In addition, the increase in cell death caused by tumor-promoting inflammation and/or tissue repair processes dilutes tumor-specific genetic and epigenetic alterations by adding more small, fragmented "background" DNA, as well as proteins and other molecules, into blood, making it impossible for consistent detection and accurate assessment of ctDNA using the above existing technologies. As a result, the clinical utility of ctDNA in diagnosing and monitoring patients with cancers including HCC has been limited. In order to reliably enrich, detect and analyze ctDNA, new methodologies, such as PCR-based digital assaysand DNA sequencing-based assays, have been developed. PCR-based digital assays include droplet digital PCR [bib_ref] Digital PCR analysis of circulating nucleic acids, Hudecova [/bib_ref] and BEAMing on the basis of four components (beads, emulsion, amplification, and magnetics) [bib_ref] BEAMing: singlemolecule PCR on microparticles in water-in-oil emulsions, Diehl [/bib_ref] , whereas DNA sequencing-based assays include pyrophosphorolysis-activated polymerization (PAP) [bib_ref] Pyrophosphorolysis-activated polymerization detects circulating tumor DNA in metastatic uveal melanoma, Madic [/bib_ref] , tagged-amplicon deep sequencing (TAM-Seq) [bib_ref] Noninvasive identification and monitoring of cancer mutations by targeted deep sequencing of..., Forshew [/bib_ref] , safesequencing system (Safe-SeqS) [bib_ref] Detection and quantification of rare mutations with massively parallel sequencing, Kinde [/bib_ref] , cancer personalized profiling by deep sequencing (CAPP-Seq) [bib_ref] An ultrasensitive method for quantitating circulating tumor DNA with broad patient coverage, Newman [/bib_ref] and personalized analysis of rearranged ends (PARE) [bib_ref] Development of personalized tumor biomarkers using massively parallel sequencing, Leary [/bib_ref] [bib_ref] Detection of chromosomal alterations in the circulation of cancer patients with whole-genome..., Leary [/bib_ref] [fig_ref] Table 2: Potential methods for HCC ctDNA detection [/fig_ref]. BEAMing and digital droplet PCR have been developed for diagnosing patients with breast cancer [bib_ref] Analysis of circulating tumor DNA to monitor metastatic breast cancer, Dawson [/bib_ref] , colon cancer [bib_ref] Circulating mutant DNA to assess tumor dynamics, Diehl [/bib_ref] [bib_ref] Multiplex picodroplet digital PCR to detect KRAS mutations in circulating DNA from..., Taly [/bib_ref] and gastric cancer [bib_ref] Monitoring gastric cancer progression with circulating tumour DNA, Hamakawa [/bib_ref]. CAPP-Seq has successfully identified 85 % of non-smallcell lung cancer (NSCLC) patients with stage II-IV and 50 % of patients with stage I NSCLC [bib_ref] An ultrasensitive method for quantitating circulating tumor DNA with broad patient coverage, Newman [/bib_ref]. A later study compared digital sequencing of plasma-derived cfDNA to tissue-based sequencing on 165 consecutive matched samples in solid tumor cancers, proved the clinical sensitivity was 85.0 and 80.7 %, respectively [bib_ref] Analytical and clinical validation of a digital sequencing panel for quantitative, highly..., Lanman [/bib_ref]. These new advances have expanded our ability to detect tumor-specific genetic and epigenetic alterations including DNA methylations, point mutations, amplifications, chromosomal rearrangements, and aneuploidy in ctDNA. Recently, the approach termed shotgun massively parallel sequencing was applied to establish correlation between the fractional concentrations of ctDNA and the tumor size and surgical treatment [bib_ref] Cancer genome scanning in plasma: detection of tumor-associated copy number aberrations, single-nucleotide..., Chan [/bib_ref]. Notably, this approach has the ability to scan genome-wide landscape ranging from genomic aberrations to point mutations. Despite these technological advances, the low level of ctDNA remains a major factor that limits the utility of ctDNA in HCC diagnosis. The field urgently needs a method to efficiently enrich ctDNA from a pool of cfDNA.
## Poorly characterized dna alterations of hcc
Cancer somatic alterations form the basis for ctDNA detection and analysis. DNA alterations harbored in ctDNA reflect those occurring in tumors. However, unlike breast cancer or lung cancer, which has multiple welldefined genetic aberrations that dictate tumor behavior, the profiles of genetic and epigenetic alterations in HCC are poorly characterized. To date, qualitative analysis of abnormal concentrations of ctDNA or single-gene methylation alterations alone is not recommended for HCC diagnosis base on a meta-analysis, while combining with AFP improves the diagnostic performance [bib_ref] Value of quantitative and qualitative analyses of circulating cell-free DNA as diagnostic..., Liao [/bib_ref]. During the recent 5 years, the comprehensive genome-wide deep sequencing of HCC tumor samples led to identification of many HCC-specific driver mutations, including aberrations in TERT promoter, TP53, CTNNB1, ARID1A/ARID1B, Axin1, APC (adenomatous polyposis coli), TSC2 and many others (Lin and Cai, manuscript in preparation) [bib_ref] Whole-genome sequencing of liver cancers identifies etiological influences on mutation patterns and..., Fujimoto [/bib_ref] [bib_ref] Integrated analysis of somatic mutations and focal copy-number changes identifies key genes..., Guichard [/bib_ref] [bib_ref] Exome sequencing of liver fluke-associated cholangiocarcinoma, Ong [/bib_ref] [bib_ref] Advances in targeted therapies for hepatocellular carcinoma in the genomic era, Llovet [/bib_ref] [bib_ref] Exome sequencing of hepatocellular carcinomas identifies new mutational signatures and potential therapeutic..., Schulze [/bib_ref]. Mapping somatic changes in HCC tumors, in combination with other HCC-specific chromosomal rearrangements and epigenetic alterations, may pave the way for development of ctDNA detection and analysis technologies for HCC patients.
## Hcc heterogeneity
Genetic and epigenetic profiles vary in different populations of tumor cells within the same primary tumor, as well as their metastases from the same patient. This phenomenon was termed "tumor heterogeneity", which poses serious clinical barriers to targeted therapy. The development of tumor heterogeneity is attributed to clonal evolution associated with acquisition of differential genetic and epigenetic alterations. In HCC, different genetic and epigenetic alterations in individual tumor cells, together with selection pressure upon them, may cause populations of tumor cells within a tumor to undergo molecularly heterogeneous transformation, even with the seemingly identical histopathological traits [bib_ref] Evolution of the cancer genome, Yates [/bib_ref]. This evolution can start at varying times points during initiation and progression of a tumor and/or at varying sites within a tumor, resulting in the spatial and temporal heterogeneity of HCC. As a result, a single-site biopsy is certain to miss clinically important mutations from a heterogeneous HCC tumor [bib_ref] Intratumor heterogeneity and branched evolution revealed by multiregion sequencing, Gerlinger [/bib_ref] [bib_ref] The translational potential of circulating tumour DNA in oncology, Patel [/bib_ref]. In contrast, ctDNA analyzed by a personalized ctDNA detection approach is a pool of DNA fragments released from nearly every part of a tumor and every tumor in the patient and maintains tumor heterogeneity in term of genetic and epigenetic alterations. The capability of dissecting tumor heterogeneity, along with convenience and less invasiveness, makes the ctDNA approach much more desirable in diagnosis of cancers, including HCC. A proof-of-concept study has been taken recently that analysis of ctDNA can monitor somatic genetic alterations during the tumor progression, covering the whole geographical region of the tumor. In this study performed on a 66-year-old woman, 16 mutation signatures in liver metastasis and nine in primary tumor were identified, all of which were detected in ctDNA [bib_ref] Capturing intra-tumor genetic heterogeneity by de novo mutation profiling of circulating cell-free..., Mattos-Arruda [/bib_ref]. Although carried out in one single patient, this attempt presents a promising approach for overcoming the clinical challenges of heterogeneity and improving therapeutic effectiveness.
# Conclusions
Although several protein-based HCC biomarkers have been reported, very few of them demonstrate solid diagnostic performance [bib_ref] Biomarkers for the early diagnosis of hepatocellular carcinoma, Tsuchiya [/bib_ref]. In future medical management, more patients will allow their physicians to make therapeutic decisions guided by genetic analysis of ctDNA. A considerable amount of studies make it clear that clinicians are entering the age in which ctDNA analysis will be a key part of tumor management, as this convenient, non-invasive and accurate diagnostic approach will reduce the anxiety of patients, as well as clinicians, and help prevent cancer progression or even cure cancer. But prior to the clinical applications, a mechanistic understanding of the biology of ctDNA is urgently needed. Previous efforts revealed that HCC patients mostly contain genomic DNA in plasma [bib_ref] Circulating free DNA in plasma or serum as biomarker of carcinogenesis: practical..., Gormally [/bib_ref] , but recent work uncovered that the tumor-associated aberrations preferentially distribute in short DNA molecules [bib_ref] Lengthening and shortening of plasma DNA in hepatocellular carcinoma patients, Jiang [/bib_ref]. Such evidence helps improve our understanding of the nature of ctDNA and might provide guideline for technology improvement and clinical practice. In addition, it is unclear to what extent the spectra of genetic and epigenetic alterations of ctDNA resemble those in tumors of a patient. Our understanding in this direction will help determine the scope of future applications of ctDNA.
Technological advances in ctDNA enrichment and analysis help expand the potential of the ctDNA-based liquid biopsy in cancer diagnosis and therapy. However, current technologies for ctDNA enrichment and analysis are yet to work in early detection of cancer due to fast decay of ctDNA and extremely low amount of ctDNA from early stage of HCC. Identification of the increasing number of alterations in ctDNA of early HCC makes it possible to improve ctDNA enrichments and analysis for early detection of HCC, which will benefit patients the most when the tumor will be most amenable to cure. Further, due to needle biopsy sampling bias and the limited availability of "research biopsies" in advanced cancer patients, the investigation of HCC metastasis and drug resistance has been challenging. It is believed that ctDNA carries different genetic or epigenetic signatures when tumor cells in a HCC patient become metastatic or when a patient develops drug resistance. It is therefore desirable to take repetitive monitoring of these signatures (or events) in ctDNA using blood samples and have sufficient information to devise treatment options. Together, our efforts in these cDNA-related areas would provide many healthcare advances and improve the life of cancer patients.
[fig] Figure 1: ctDNA release and extraction in HCC patients. ctDNA is released from HCC cells undergoing apoptosis or necrosis and can be extracted from a blood sample. Genetic and epigenetic aberrations in ctDNA can be detected and quantified. These genetic alterations include mutations, rearrangements, methylation, microsatellite alteration and integrated viral DNA. The detection of these alterations in the background of "normal" cfDNA molecules in principle offers a higher diagnostic specificity in comparison with only quantitative measurement of total cfDNA alone [/fig]
[fig] Figure 2: Monitoring response and relapse of HCC patients with targeted therapies by ctDNA detection. From early stage lesion to late phase of carcinogenesis, an excess of apoptotic cell death, as occurs in large and rapidly proliferating HCC tumors (solid line), can lead to an increase of ctDNA levels in plasma (dash line). Moreover, the levels of ctDNA correlate well with HCC progression as well as various therapy including surgical resection, local ablation, TACE and targeted molecular therapy [/fig]
[table] Table 1: Detection of cfDNA and its alterations in HCC patients* Includes all articles published to date that assess cfDNA detection and alterations of cfDNA in HCC patients, which were summarized and clarified the significance [/table]
[table] Table 2: Potential methods for HCC ctDNA detection [/table]
|
Subsets and Splits