text
stringlengths 66
377k
|
|---|
Effects of Extreme-Duration Heavy Load Carriage on Neuromuscular Function and Locomotion: A Military-Based Study
Trekking and military missions generally consist of carrying heavy loads for extreme durations. These factors have been separately shown to be sources of neuromuscular (NM) fatigue and locomotor alterations. However, the question of their combined effects remains unresolved, and addressing this issue required a representative context.Purpose:The aim was to investigate the effects of extreme-duration heavy load carriage on NM function and walking characteristics.Methods: Ten experienced infantrymen performed a 21-h simulated military mission (SMM) in a middle-mountain environment with equipment weighing ,27 kg during battles and ,43 kg during marches. NM function was evaluated for knee extensors (KE) and plantar flexors (PF) pre-and immediately post-SMM using isometric maximal voluntary contraction (MVC) measurement, neural electrical stimulation and surface EMG. The twitch-interpolation method was used to assess central fatigue. Peripheral changes were examined by stimulating the muscle in the relaxed state. The energy cost, mechanical work and spatio-temporal pattern of walking were also evaluated pre2/post-SMM on an instrumented treadmill in three equipment conditions: Sportswear, Battle and March.Results: After the SMM, MVC declined by 210.263.6% for KE (P,0.01) and 210.7616.1% for PF (P = 0.06). The origin of fatigue was essentially peripheral for both muscle groups. A trend toward low-frequency fatigue was detected for KE (5.5%, P = 0.08). These moderate NM alterations were concomitant with a large increase in perceived fatigue from pre-(rating of 8.362.2) to post-SMM (15.962.1, P,0.01). The SMM-related fatigue did not alter walking energetics or mechanics, and the different equipment carried on the treadmill did not interact with this fatigue either.Conclusion: this study reports the first data on physiological and biomechanical consequences of extreme-duration heavy load carriage. Unexpectedly, NM function alterations due to the 21-h SMM were moderate and did not alter walking characteristics.Clinical Trial Registration: Name: Effect of prolonged military exercises with high load carriage on neuromuscular fatigue and physiological/biomechanical responses. Number: NCT01127191.
# Introduction
Outdoor activities requiring self-sufficiency, specific gear and/or involving bivouac (e.g. trekking and military missions) are characterized by the carriage of considerable equipment and supplies. This specific feature implies high-to-severe conditions of locomotion [bib_ref] Effects of heavy load carriage during constant-speed, simulated, road marching, Beekley [/bib_ref] [bib_ref] Load carriage using packs: a review of physiological, biomechanical and medical aspects, Knapik [/bib_ref] and requires high levels of metabolic and mechanical energy from the carrier, even for short-term walking [bib_ref] Effect of load and speed on the energetic cost of human walking, Bastien [/bib_ref] [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref] [bib_ref] Metabolic cost of generating muscular force in human walking: insights from load-carrying..., Griffin [/bib_ref] [bib_ref] Predicting energy expenditure with loads while standing or walking very slowly, Pandolf [/bib_ref]. In addition, these walking efforts are generally prolonged for hours or even days and can be associated with sleep rhythm disturbance or deprivation, caloric restriction, as well as environmental and psychological stresses, especially in the military context [bib_ref] Physiological and metabolic aspects of very prolonged exercise with particular reference to..., Ainslie [/bib_ref] [bib_ref] Combat rations and military performance -do soldiers on active service eat enough?, Booth [/bib_ref] [bib_ref] Physiological and psychological fatigue in extreme conditions: the military example, Weeks [/bib_ref]. Taken together or separately, these factors have delete-rious physiological consequences [bib_ref] Physiological decrements during sustained military operational stress, Henning [/bib_ref] [bib_ref] Physical performance responses during 72 h of military operational stress, Nindl [/bib_ref] and may induce neuromuscular (NM) fatigue.
NM fatigue is defined as an exercise-related decrease in the maximal strength or power of a muscle, whether or not the task can be sustained [bib_ref] Changes in muscle contractile properties and neural control during human muscular fatigue, Bigland-Ritchie [/bib_ref]. Fatigue potentially involves processes at all levels of the motor pathway from the motor cortex to skeletal muscle [bib_ref] Electrical stimulation for testing neuromuscular function: from sport to pathology, Millet [/bib_ref]. Classically, alterations of neuromuscular function due to fatigue are classified as central (neural) or peripheral (muscular) in origin. Central fatigue corresponds to a failure of the central nervous system to drive the motoneurons adequately and appears with exercise [bib_ref] Spinal and supraspinal factors in human muscle fatigue, Gandevia [/bib_ref] while peripheral fatigue may involve all processes located distal to the neuromuscular junction/sarcolemma, in particular the fatigue-induced alterations in excitationcontraction coupling [bib_ref] Muscle fatigue: from observations in humans to underlying mechanisms studied in intact..., Place [/bib_ref]. Central and peripheral origins are mutually dependent since recruitment of motoneurons depends on the descending drive from supraspinal sites and central drive is controlled through a combination of factors including excitatory and inhibitory reflex inputs from different peripheral afferents. Fatigue is known to depend on exercise duration and intensity, and type of muscle contraction [bib_ref] Exercise and fatigue, Ament [/bib_ref] [bib_ref] Neurobiology of muscle fatigue, Enoka [/bib_ref].
Previous studies have investigated the effects of exercises with heavy load carriage of short-to-medium duration on NM function. Clarke et al. [bib_ref] Strength decrements from carrying various army packs on military marches, Clarke [/bib_ref] have reported decreases in trunk, knee and ankle flexor/extensor maximal voluntary contraction (MVC) in soldiers after ,3-h walking (12.1 km at 4 km.h 21 ) while carrying loads up to 27 kg. More recently, Blacker et al. [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] showed that the carriage of a 25-kg pack during 2-h treadmill walking induced a 15% loss in knee extensor (KE) MVC and was associated with moderate central and peripheral fatigue. Furthermore, slight lowfrequency fatigue (LFF, which has been linked to excitationcontraction coupling failure and muscular damage [bib_ref] High-and low-frequency fatigue revisited, Jones [/bib_ref] was detected in this study for both level and downhill walking [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref].
Other studies have investigated the NM consequences of exercises without load carriage (or with light equipment) but of extreme duration. Very large KE and plantar flexor (PF) MVC declines were observed in ultra-marathon runners after a 24-h treadmill run (241% and 230%, respectively [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] and a ,40-h mountain ultra-marathon (235% and 239%, respectively [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref]. In both studies, MVC declines were associated with large central activation deficits (depending on the muscle group tested) and peripheral fatigue. LFF was also observed after the mountain ultra-marathon, likely due to long downhill sections (,9 500 m of total negative elevation) that involved intense eccentric muscular actions known to induce muscular damage [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref] [bib_ref] Molecular adaptations of neuromuscular disease-associated proteins in response to eccentric exercise in..., Feasson [/bib_ref].
To date, the question of the combined effects of heavy load carriage and exercise of extreme duration on NM fatigue remains however unresolved. Since military efforts naturally and severely combine these two features [bib_ref] Physiological and psychological fatigue in extreme conditions: the military example, Weeks [/bib_ref] , we thought it justified and relevant to use a military mission context to address this question. Indeed, beyond their scientific novelty, such data obtained in a real-world context could be applied to strategic planning and used in the military theater (e.g. management of forces during actual missions). Therefore, the main purpose of the present study was to investigate the effects of a prolonged mission with heavy military load carriage on the NM fatigue of two major muscle groups involved in human walking and load carriage [bib_ref] Backpack load affects lower limb muscle activity patterns of female hikers during..., Simpson [/bib_ref]. For this purpose, the KE and PF NM function of experienced infantrymen was assessed before and after a 21-h simulated military mission (SMM) specifically designed to represent the current operational reality.
In addition, in view of the existing links between NM fatigue, load carriage and the risks of slips, falls or injuries during locomotion [bib_ref] Load carriage using packs: a review of physiological, biomechanical and medical aspects, Knapik [/bib_ref] [bib_ref] Backpack load affects lower limb muscle activity patterns of female hikers during..., Simpson [/bib_ref] [bib_ref] Biomechanical analysis of fatigue-related foot injury mechanisms in athletes and recruits during..., Gefen [/bib_ref] [bib_ref] Effect of a high intensity quadriceps fatigue protocol on knee joint mechanics..., Murdock [/bib_ref] [bib_ref] Effects of quadriceps fatigue on the biomechanics of gait and slip propensity, Parijat [/bib_ref] [bib_ref] Plantar pressure changes after long-distance walking, Stolwijk [/bib_ref] , this study design allowed us to investigate the effects of extreme-duration heavy load carriage on walking mechanics and energetics. Indeed, although acute consequences of load carriage have been widely considered [bib_ref] Effect of load and speed on the energetic cost of human walking, Bastien [/bib_ref] [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref] [bib_ref] Metabolic cost of generating muscular force in human walking: insights from load-carrying..., Griffin [/bib_ref] [bib_ref] Predicting energy expenditure with loads while standing or walking very slowly, Pandolf [/bib_ref] , very little is known about the effects of fatigue and its potential interaction with load carriage on walking characteristics. Moreover, the few studies that have hitherto explored similar issues mostly used fatiguing laboratory protocols (e.g. treadmill run [bib_ref] Effects of quadriceps fatigue on the biomechanics of gait and slip propensity, Parijat [/bib_ref] [bib_ref] The effects of fatigue on plantar pressure distribution in walking, Bisiaux [/bib_ref] [bib_ref] Effects of load carriage and fatigue on gait characteristics, Qu [/bib_ref] [bib_ref] Influence of fatigue and load carriage on mechanical loading during walking, Wang [/bib_ref] , not real exercises in the field, and this may limit the representativity of their observations [bib_ref] Interactive processes link the multiple symptoms of fatigue in sport competition, Knicker [/bib_ref]. Therefore, the secondary aim of the present study was to investigate the consequences of this 21-h SMM on the energy cost, mechanical work and spatiotemporal parameters of walking.
Concerning the primary aim of the present study, we hypothesized that both central and peripheral NM function alterations would be larger after the SMM than after loadedwalking exercises of intermediate duration, i.e. from 2 to 3 h [bib_ref] Strength decrements from carrying various army packs on military marches, Clarke [/bib_ref] [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref]. This is because of (i) the much longer exercise duration, (ii) the heavier loads carried, and (iii) the additional stressors inherent to the mission-related effort (e.g. sleep rhythm and comfort disturbances, hilly terrain). However, NM function changes of lower magnitude than after ultra-marathon runs [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref] were expected (especially at the peripheral level) because, contrary to running, walking is not associated with repeated impacts or intense muscular eccentric actions [bib_ref] Why does knee extensor muscles torque decrease after eccentric-type exercise?, Martin [/bib_ref] [bib_ref] Changes in running mechanics and spring-mass behavior induced by a mountain ultra-marathon..., Morin [/bib_ref]. Regarding the secondary aim of this study, we hypothesized that walking characteristics would change after the 21-h SMM, especially the mechanical parameters [bib_ref] Effects of load carriage and fatigue on gait characteristics, Qu [/bib_ref] [bib_ref] Influence of fatigue and load carriage on mechanical loading during walking, Wang [/bib_ref] that have been shown to be more discriminating than energetical ones at moderate walking speed [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref].
# Methods
## Subjects and ethics statement
Ten males volunteered to participate in this study after being informed about the procedure and risks associated with the protocol. They were all involved in regular physical activities (5.462.7 hours per week) and were not presenting recent muscular, joint or bone disorders or receiving any medication that could interfere with their NM responses or walking pattern or influence their energetic metabolism. The subjects were recentlyretired infantrymen (seven from the French Foreign Legion) with a career of [bib_ref] Spinal and supraspinal factors in human muscle fatigue, Gandevia [/bib_ref] Written informed consent was obtained from the subjects, and the study was conducted according to the Declaration of Helsinki. The protocol was approved by the local ethics committee (Comité de Protection des Personnes, Sud-Est 1, France) and registered at http://clinicaltrial.gov (reference: NCT01127191).
## Protocol overview
The subjects were included in the study one week before the beginning of the specific research protocol. Inclusion sessions consisted of (i) a complete medical examination with anthropometric data collection, (ii) a standardized incremental maximal aerobic test, and (iii) a complete familiarization with the different devices/protocols used in the experimentation (see details in the section subject characterization and familiarization below).
The specific research protocol lasted 24 h and consisted of two (pre/post) 90-min laboratory measurement sessions separated by the 21-h SMM. Heart rate was the only parameter monitored continuously throughout the protocol using telemetric cardiofrequencemeters (Polar RS800CX, Polar Electro Oy, Kempele, Finland). Chronologically, the pre-SMM session (PRE) began with an evaluation of the subjects' instant rating of perceived fatigue (RPF), and a 6-min moderate cycling warm-up immediately followed by a standardized NM function evaluation (see details in the sections relating to fatigue assessment below). After 5 min of rest, expired gases were collected over 10 min of unloaded standing. Then the subjects performed three 3-min level-walking trials at 4 km.h 21 on an instrumented treadmill, during which walking energetics and mechanics were assessed in three loading conditions (i.e. one condition per trial, see details in the sections equipment/load characteristics and walking assessment below). Before starting the SMM at 12:00 am, the subjects checked their equipment and supplies. The 21-h SMM was performed in a low-to middle-mountain environment close to the laboratory (see details in the section SMM characteristics and figure 1 below). After the SMM, the subjects immediately performed the post-SMM session (POST). POST began with an evaluation of instant RPF and rating of perceived exertion regarding the entire SMM (RPE). Then, after taking off their equipment, the subjects performed the NM function evaluation within 963 min after the SMM. Finally, the course of POST (i.e. unloaded standing gas collection and walking assessments on the treadmill in three loading conditions) was identical to that of PRE.
For logistical reasons (duration of PRE and POST, follow-up and safety of the subjects during the SMM, etc.), the subjects were divided into two subgroups (n = 5) and performed the protocol two days apart. Similarly, in order to avoid a period of rest between the end of the SMM and the NM function evaluations at POST, the subjects in a same subgroup performed the protocol at 20-min intervals.
## Subject characterization and familiarization
During the inclusion sessions, the subjects were first examined by the same medical doctor that also performed anthropometric data collection. Stature was measured to the nearest 0.5 cm using a standardized wall-mounted height board. Leg length was measured from the great trochanter to the floor in a standing position with a tape measure. BM was measured to the nearest 0.1 kg with subjects standing in undergarments without shoes on a mechanical column weighing scale (Bascule type 286, Chollet, La Talaudière, France). BM Index was calculated as BM divided by height (in m) squared. Body fat percentage was estimated using skinfold thickness values and Durnin & Womersley standard equations [bib_ref] Body fat assessed from total body density and its estimation from skinfold..., Durnin [/bib_ref]. Skinfold thickness values were measured to the nearest millimeter in triplicate at the biceps, triceps, subscapular, and suprailiac points on the left and right side of the body using a Harpenden skinfold caliper (British Indicators, West Sussex, UK). At each of these four points, the mean value for the six skinfold thicknesses was calculated.
After this examination, the subjects performed an incremental running maximal test on a level treadmill. This test consisted of progressive 3-min stages separated by 1 min of rest. Running speed was 10 km.h 21 at the first stage and the increment of speed was 1.5 km.h 21 per stage until exhaustion. Heart rate (HR) was monitored continuously using a three-channel electrocardiogram (Cardiotest EK51, Hellige GMBH, Freiburg, Germany). Expired gases were collected during the last 30 s of each 3-min stage to determine oxygen consumption ( _ V VO 2 ). The subjects breathed through a two-way nonrebreathing valve (series 2700; Hans Rudolph, Kansas City, MO) connected to a three-way stopcock leading to a 100-L Douglas bag. The volume of the expired gas was measured by means of a Tissot spirometer (Gymrol, Roche-la-Molière, France), and fractions of expired gases were determined with a paramagnetic O 2 analyzer (cell 1155B; Servomex, Crowborough, England) and an infrared CO 2 analyzer (Normocap Datex, Helsinki, Finland). The analyzers were calibrated with mixed gases of known composition. HR max and _ V VO 2max corresponded to the highest values obtained at steady state during the last running stage.
Particular attention was paid to familiarize the subjects with the devices/protocols used during the specific experimentation, especially the MVC and electrical stimulation of the KE and PF muscles. The subjects repeated trials of the procedures until the results were reproducible.
## Equipment/load characteristics
For a detailed description and illustration of the equipment, the reader should refer to a recent article from our research group [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref]. Briefly, during the laboratory walking trials, three equipment conditions were tested, a Sportswear condition taken as reference (SP, mass #1 kg, corresponding to #1.4% of the subjects' BM) and two configurations of the new French infantry combat system During the SMM, both military BT and RM equipment were carried alternately (according to the military actions performed, e.g. RM while marching and BT while patrolling, see [fig_ref] Figure 1: Typical heart rate [/fig_ref] with an additional demilitarized rifle weighting 5.01 kg (FA-MAS FELIN, France) to create conditions comparable to those found in the military theater. The total loads carried during the SMM were therefore 27.461.1 kg in BT and 42.961.4 kg in RM, corresponding to 33.462.6% BM and 52.264.2% BM respectively. BT and RM configurations were designed to represent and meet the common necessities of a prolonged patrol and reconnaissance mission. Each participant was familiarized with all of the equipment tested prior to the experiment.
## Simulated military mission (smm) characteristics
An overview of the 21-h SMM design, intensity and integration in the general protocol is given in [fig_ref] Figure 1: Typical heart rate [/fig_ref]. The SMM was specifically designed by former officers to represent a typical prolonged patrol and reconnaissance mission. It began with a 4-h road march to simulate quietly approaching a hostile zone represented by a fragment of forest located in a middle-mountain environment 15 km away from the laboratory (uphill walking, 570 m of positive and 240 m of negative elevation change, mean walking speed of ,4 km.h 21 ). On arrival near the hostile zone, the subjects had a 1-h period to rest and ration. Each subject had a French war ration (''RCIR''; 3200 kcal, protein: 13%, lipid: 32%, carbohydrate: 55%) in his pack to provide sustenance over the entire SMM. After this reconditioning period, the subjects moved through the enemy forest (,30 min) in order to find and install (,30 min) an advanced camp. The objective of the following 2-h recon period was to progressively secure the zone that surrounded the camp while approaching a strategic enemy lane for observation. During the night hours, the subjects were equipped with night vision devices. Moreover, during the recon periods of the SMM, experimenters drove and moved through the forest to simulate enemy activity. The subjects were also told to try and intercept the experimenters as they tried to reach the advanced camp, just as enemies would do. After the aforementioned 8 h of various activities, the subjects were allowed two 3-h sleeping periods separated by another 2-h recon period. Sleeping periods were set and imposed according to the common turnover observed in missions and in previous reports [bib_ref] Physiological and psychological fatigue in extreme conditions: the military example, Weeks [/bib_ref]. Then one last 105-min recon was performed prior to a 15-min period during which the subjects gathered their equipment prior to walking back to the laboratory. The return road march was the exact reverse of the first one (downhill walking, 570 m of negative and 240 m of positive elevation change) except it was performed in 3 h because of the major negative slope (mean walking speed of ,5.5 km.h 21 ).
## Fatigue assessment: neuromuscular function
Experimental materials, preparation and setting. In the present study, we used the exact same (i) ergometers, positioning and instructions to subjects, (ii) electrical stimulation system and neural detection procedure, (iii) electrode models, electromyographic recording system and skin preparation procedure as in a recent study from our research group [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref].
Maximal voluntary contractions: the subjects were strongly encouraged during all the MVCs. For the KE testing, the subjects were seated in the frame of a Cybex II (Ronkonkoma, NY) and Velcro straps were strapped across the chest and hips to avoid lateral and frontal displacements. Subjects were also instructed to grip the seat during the MVC to further stabilize the pelvis. The KE muscular mechanical response was recorded with a strain gauge (SBB 200 Kg, Tempo Technologies, Taipei, Taiwan) located at the level of the external malleolus. All measurements were taken from the subject's right leg with the knee and hip flexed at 90 degrees from full extension. PF muscles were tested with an instrumented pedal (CS1060 300 Nm, FGP Sensors, Les Clayes Sous Bois, France). For the PF testing, the subjects were seated in the frame of a Cybex II similar to that used for KE. Velcro straps were also strapped across the chest and hips to avoid lateral and frontal displacements, and across the forefoot to limit heel lift during the MVC. The hip, knee and ankle angles were set at 90 degrees from full extension.
Electrical stimulation: after femoral (for KE) and posterior tibial nerve (for PF) detection with a ball probe cathode pressed into the femoral triangle and the popliteal fossa, respectively, electrical stimulation was applied percutaneously to the motor nerve via a self-adhesive electrode pressed manually (10-mm diameter, Ag-AgCl, Type 0601000402, Contrôle Graphique Medical, Brie-Comte-Robert, France). The anode, a 1065 cm self-adhesive stimulation electrode (Medicompex SA, Ecublens, Switzerland), was located either in the gluteal fold (for KE) or on the patella (for PF). A constant current stimulator (Digitimer DS7A, Hertfordshire, United Kingdom) was used to deliver a square-wave stimulus of 1000-ms duration with maximal voltage of 400 V. The stimulation intensity (70.6615.6 mA at PRE and 72.4616.9 mA at POST in KE, and 64.6617.2 mA at PRE and 65.3615.1 mA at POST in PF) was determined from maximal mechanical response to single twitch delivered to the relaxed muscle. This stimulation intensity was supramaximal and corresponded to 130% of the optimal intensity.
Electromyographic recordings: the EMG signals of the right vastus lateralis (VL) and soleus (SOL) were recorded using bipolar silver chloride surface electrodes of 10-mm diameter (Type 0601000402, Contrôle Graphique Medical, Brie-Comte-Robert, France) during the MVCs and electrical stimulation. The recording electrodes were taped lengthwise on the skin over the muscle belly, with an interelectrode distance of 25 mm. The position of the electrodes was marked directly on the skin with a permanent marker so that they could be placed in the exact same position before and after the SMM. The reference electrode was on the patella (for VL EMG) or malleolus (for SOL EMG). Low impedance (Z,5 kV) at the skin-electrode surface was obtained by abrading the skin with fine sand paper and cleaning with alcohol. EMG data were recorded with PowerLab system (16/30 -ML880/P, ADInstruments, Bella Vista, Australia) with a sampling frequency of 2000 Hz. The EMG signal was amplified with octal bio-amplifier (Octal Bioamp, ML138, ADInstruments), with a bandwidth frequency ranging from 5 to 500 Hz (input impedance = 200 MV, common mode rejection ratio = 85 dB, gain = 1000), transmitted to the computer and analyzed with LabChart 6 software (ADInstruments).
Experimental procedure. The present procedure was also identical to that used in our recent study [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref]. Briefly, the NM function evaluation consisted of determining the isometric KE and PF MVC. During MVC, when the torque had reached a plateau, a high-frequency (100 Hz) doublet was superimposed on the contracted muscle. Finally, ,2 s after the end of the MVC, evoked stimuli consisting of a high-frequency (100 Hz) doublet, a lowfrequency (10 Hz) doublet and a single twitch were delivered to the relaxed muscle in a potentiated state. This experimental set (MVC with superimposed doublet + evoked stimuli to the relaxed muscle) was repeated three times for both muscle groups with recovery of 1 min between repetitions.
Experimental variables and data analysis. M-wave: for both VL and SOL, M-wave peak-to-peak amplitude (in mV) and duration (in ms) were averaged from the EMG data from the three single potentiated twitches delivered to the relaxed muscle. Mwave characteristics provided information on action potential propagation.
Mechanical responses to nerve stimulation: for both KE and PF, the amplitude of the potentiated high-frequency doublet (PDb100), the ratio of paired-stimulus peak forces at 10 Hz to 100 Hz (Db10:100) and the amplitude of the potentiated peak twitch (Pt) were averaged from the values computed during the three experimental sets. Potentiated twitch contraction time (CT, in ms) and half-relaxation time (HRT, in ms) were also determined and were also calculated as the mean values of the three single twitches. Mechanical responses to nerve stimulation were used to determine the extent and origin of peripheral fatigue. In particular, Db10:100 was used to assess LFF [bib_ref] Comparison of electrical and magnetic stimulations to assess quadriceps muscle function, Verges [/bib_ref].
Maximal voluntary contraction and maximal voluntary activation level: the highest value of the three MVC was determined for both KE (in N) and PF (in Nm). MVC provided a global index of fatigue. Activation level (%VA) was calculated as follows:
[formula] %VA~(1{Superimposed Db100=Potentiated Db100)|100 [/formula]
%VA was used as an indicator of central fatigue. Finally, the root mean square (RMS) values of the VL and SOL EMG activity were calculated during the best MVC trial over a 0.5-s period after the MVC had reached a plateau and before the superimposed stimulus was delivered. This RMS value was then normalized to the maximal peak-to-peak amplitude of the M-wave to obtain RMS.M 21 . The latter variable provided complementary information about central fatigue.
Fatigue Assessment: Perceived Fatigue and Global Exertion RPE was measured at POST only using the 6 to 20 Borg scale [bib_ref] Psychophysical bases of perceived exertion, Borg [/bib_ref]. The subjects were asked to quantify the exertion characterizing the entire SMM. RPF was measured at PRE and POST, using the same 6 to 20 Borg scale applied to the sensation of fatigue. The subjects were asked to quantify their instant sensation of general fatigue. RPE and RPF were administered individually to subjects by the same experimenter for all subjects.
## Walking assessment
The three equipment conditions, SP, BT and RM, were performed in a randomized and counterbalanced order, and trials were separated by 5 to 10 min during which subjects rested and changed their equipment. The 4 km.h 21 walking speed was chosen for its economical feature in normal adult locomotion [bib_ref] Effect of load and speed on the energetic cost of human walking, Bastien [/bib_ref] [bib_ref] External work in walking, Cavagna [/bib_ref] and its consistency with the average walking speed currently used during military missions and experimentations [bib_ref] Optimum load for carriage by soldiers at two walking speeds on level..., Pal [/bib_ref]. Finally, rifle carriage was excluded to allow comparisons of our results with non-military studies since walking kinetics are altered by a limitation of the arm swing during rifle carriage [bib_ref] The influence of rifle carriage on the kinetics of human gait, Birrell [/bib_ref]. Complete details of the physiological and biomechanical methods employed have been recently presented [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref]. Thus, only the main principles and parameters are summarized below.
## Walking energetics
Energetic data were obtained from indirect calorimetry (Douglas bag method, vide supra). Expired gases were collected during the last 30 s of each 3-min walking trial. Unloaded standing metabolic rate and gross metabolic rate of walking (both in W) were determined from the steady-state _ V VO 2 and _ V VCO 2 using Brockway's standard equation [bib_ref] Derivation of formulae used to calculate energy expenditure in man, Brockway [/bib_ref]. The metabolic rate measured during unloaded standing was subtracted from all gross walking values to compute the net metabolic rate (in W) [bib_ref] Metabolic cost of generating muscular force in human walking: insights from load-carrying..., Griffin [/bib_ref]. Gross and net metabolic rates (in W) were divided by walking speed (in m.s 21 ) to obtain the gross and net energy costs of walking (C W , in J.m 21 ). Gross and net C W were also divided by the total mass in motion on the treadmill (TM, in kg), i.e. subject plus equipment, to obtain mass-relative gross and net C W (C W.TM , in J.kg 21 .m 21 ).
## Walking mechanics
Walking mechanics were analyzed using an instrumented 3-D force treadmill (ADAL, HEF Tecmachine, Andrézieux-Bouthéon, France) consisting of two left-right frames and belts allowing separate measurements of the left-and right-foot ground reaction forces (for complete description and validation, see ref. [bib_ref] A treadmill ergometer for three-dimensional ground reaction forces measurement during walking, Belli [/bib_ref]. Parameters were recorded over 20 s, 1.5 min after the beginning of each trial in order to ensure the stabilization of the gait pattern and avoid disturbances from the metabolic measurements. All data were sampled at 200 Hz and low-pass filtered at 30 Hz. Mechanical analyses were performed over five consecutive strides, one stride being defined as the period between two consecutive right heel strikes. Mechanical parameters were computed for each stride and then averaged to describe a typical mean stride.
Spatio-temporal parameters of walking were calculated from vertical ground reaction force signals. A duty factor (in %) was calculated as the ratio of stance duration to stride duration, and double support duration (in %) as the ratio of stance duration to stride duration. Finally, step frequency (in Hz) was computed as: (stride duration/2) 21 .
Kinetic parameters were computed from vertical, anteroposterior, and medio-lateral ground reaction forces. The external mechanical work (W ext , in J.m 21 ), i.e. the work done by the muscles to lift and accelerate the center of mass (COM), was calculated according to Cavagna's standard method [bib_ref] Force platforms as ergometers, Cavagna [/bib_ref]. W ext was also normalized by the total moving mass (TM) to obtain massrelative W ext (W ext.TM , in J.kg 21 .m 21 ). The inverted pendulum recovery of mechanical energy of the COM (in %) was calculated according to Schepens et al. [bib_ref] Mechanical work and muscular efficiency in walking children, Schepens [/bib_ref]. The internal work done during the double contact phase (W int,dc , in J.m 21 ), i.e. the work done by one leg against the other during the transfer from one foot to the other, was calculated from the forces exerted by each lower limb on the ground measured separately, as proposed by Bastien et al. [bib_ref] The double contact phase in walking children, Bastien [/bib_ref]. W int,dc was also divided by TM to obtain mass-relative W int,dc (W int,dc.TM , in J.kg 21 .m 21 ). Finally, locomotor efficiency (in %) was calculated as the ratio of W ext plus W int,dc to net C W (all in J.m 21 ).
## Statistical analyses
All descriptive data are presented as mean 6 SD. Normal distribution of the data was checked by the Shapiro-Wilk normality test and variance homogeneity between samples was tested by the F-Snedecor test. When conditions of t-test and analysis of variance (ANOVA) application were respectively met, each variable studied was compared (i) between the different times of measurements (i.e. PRE vs. POST) using paired t-tests for the NM data, or (ii) in the three different conditions of equipment across times of measurement (time 6 equipment) using two-factor within subjects ANOVAs for the data of locomotion. Newman-Keuls multiple comparison post-hoc tests were used to determine between-means differences if the ANOVA revealed a significant main effect. For the few NM variables that did not meet normality (i.e. KE MVC, SOL M-wave peak-to-peak duration, and PF %VA), Wilcoxon tests were used. Statistical significance was accepted at P,0.05. Effect size was calculated using Cohen's d by dividing the mean difference between PRE and POST (in absolute value) by the between-subject standard deviation at PRE. Effect size was therefore used and considered as a supplementary index of the importance of the effect for the variables showing significant differences or statistical trends between PRE and POST. A Cohen's d value of 0.2 was considered as a small effect, a value of 0.5 was considered as a moderate effect, and a value of 0.8 was considered as a large effect.
# Results
## Heart rate, global exertion and perceived fatigue
## Neuromuscular fatigue and its central and peripheral components
As shown in figures 2 and 3, MVC declined significantly by 210.263.6% for KE (P,0.01, d = 0.50) and a strong trend (210.7616.1%) was observed for PF (P = 0.06, d = 0.82) after the SMM. Concerning the peripheral aspect of fatigue induced by the SMM, PDb100 decreased significantly for KE (26.3968.01%, P,0.05, d = 0.43) and even more for PF (218.269.8%, P,0.001, d = 1.14). A trend towards Db10:100 decline (i.e. LFF index) was observed for KE (25.4669.35%, P = 0.08, d = 0.53) but not for PF (P = 0.21) after the SMM. There was no significant correlation between KE and PF changes in MVC or PDb100. shows the characteristics of the mechanical (KE and PF) and EMG (VL and SOL) responses to single electrical stimuli of the femoral and tibial motor nerves to the relaxed muscle. In particular, this table shows that the SMM induced considerable effects on the potentiated Pt and CT for both KE and PF. Concerning the central aspect of fatigue, the SMM did not induce change in %VA for PF (P = 0.73), but a decreasing trend was observed for KE 1862.96%, P = 0.05, d = 0.96). RMS.M 21 did not change significantly from PRE to POST for either KE (from 6.8962.81% to 6.1762.01%, P = 0.14) or PF (from 3.5860.63% to 3.5861.09%, P = 0.99).
## Walking energetics and mechanics
Metabolic and mechanical parameters of walking, taking into account the main effects of SMM-related fatigue, acute load carriage and their interaction are presented in tables 2 and 3, respectively. ANOVAs showed that acute military equipment (i.e. load) carriage induced significant increases in both gross and net C W (P,0.0001). A trend toward increased net C W.TM was also observed when load increased (P = 0.098) while gross C W.TM decreased significantly in this condition (P = 0.012). Concerning the mechanics of walking, all spatio-temporal parameters were altered by acute load carriage (all P,0.05) and both absolute and mass-relative W int,dc and W ext increased (all P,0.01). However, the inverted pendulum recovery of mechanical energy and the locomotor efficiency did not change while carrying loads. In contrast to these acute effects of load carriage, neither the time between PRE and POST (i.e. the fatigue induced by the SMM), nor the interaction between time and equipment (i.e. fatigue 6 load) had a significant effect on the parameters describing subjects' walking energetics and mechanics.
# Discussion
The Main Purpose of the Present Study was to Investigate the combined effects of heavy load carriage and exercise of extreme duration on NM fatigue. Specifically, we hypothesized that both central and peripheral NM function changes due to a 21-h military mission would be higher than after walking with load carriage for intermediate durations, i.e. up to ,3 h [bib_ref] Strength decrements from carrying various army packs on military marches, Clarke [/bib_ref] [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] , but lower (especially at the peripheral level) than after ultra-marathon runs lasting up to ,40 h and inducing extensive muscular damage [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref]. The main results of this study are: (i) MVC declined by ,10% after the SMM and the origin of fatigue was essentially
## Exertion and perceived fatigue due to the military mission
HR data monitored over the entire SMM showed that the most severe period of the mission was the first 4-h march (,7367% of HR max ), very likely due to the long and sometimes steep uphill sections (figure 1) with heavy load carriage (,43 kg). In contrast, the 14-h-battle phase was the least demanding period of the SMM (,4565% of HR max ), notably because of the two 3-h sleeping periods permitted during this phase. Nevertheless, HR peaked around 8366% of HR max during the battle phase, showing that the recons and patrols were quite intense. Finally, although the 3-h return march was performed at a high speed (,5.5 km.h 21 on . Energy cost of walking in Sport (SP), Battle (BT) and Road March (RM) conditions, before (PRE) and after (POST) the Simulated Military Mission. average) considering the load carried and the preceding efforts, the major presence of downhill sections resulted in a relatively moderate HR (,6066% of HR max ). The overall profile of the SMM (figure 1) was consistent with the observations reported in the military literature, namely the major presence of endurance activities combined with short periods of intense efforts [bib_ref] Physiological decrements during sustained military operational stress, Henning [/bib_ref]. After exercise, the subjects rated the entire SMM as ''very hard'' (16.762.4 on the 6-20 Borg's scale) despite their substantial experience with this type of effort. Moreover, their RPF increased significantly from PRE (8.362.2, ''very very low fatigue'') to POST (15.962.1, ''very high fatigue''; P,0.01, d = 3.47). Interestingly, the subjects' sensation of fatigue increased by almost 100% from PRE to POST, whereas their MVC (i.e. the global index of NM fatigue) declined only moderately (,10%, see below). However, it is likely that the peripheral NM fatigue observed in this study and the environmental stressors induced by the mission (e.g. sleep and food rhythms disturbances), which are known as drivers of subjective fatigue [bib_ref] Interactive processes link the multiple symptoms of fatigue in sport competition, Knicker [/bib_ref] [bib_ref] Can neuromuscular fatigue explain running strategies and performance in ultra-marathons?: the flush..., Millet [/bib_ref] [bib_ref] Sleep deprivation: effects on work capacity, self-paced walking, contractile properties and perceived..., Rodgers [/bib_ref] , have contributed to this result.
## Neuromuscular fatigue induced by the military mission
Comparison with ultra-endurance exercise without load carriage. As hypothesized, the SMM, which mainly consisted of walking for extreme duration with heavy loads, induced lower NM function alterations than ultra-marathons, which consist of running and high-speed walking for extreme duration with very light equipment. Indeed, the average extent of force loss from PRE to POST (,10% for both KE and PF, see figures 2 and 3) was 3 to 4 times lower in the present study than previously observed after 24-h [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] and mountain [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref] ultra-marathons for the same muscle groups. Moreover, central mechanisms, which are known as major components of NM fatigue after ultra-endurance runs [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] [bib_ref] Can neuromuscular fatigue explain running strategies and performance in ultra-marathons?: the flush..., Millet [/bib_ref] [bib_ref] Alterations of neuromuscular function after an ultramarathon, Millet [/bib_ref] , were only slightly affected in KE after the SMM, as shown by the 22.263.0% decline in %VA (P = 0.05, d = 0.96). A notable difference between ultra-marathons and the SMM is that in the former subjects perform ultra-endurance exercise as fast as they can, while the military subjects of the present study followed instructions and velocity plans. Therefore, results about MVCs and %VA after the SMM should be considered, among others, in light of this important difference with ultra-marathons. Nevertheless, the present results were in line with previous studies showing that KE are more prone to central fatigue than PF [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref].
The results of this study indicate that the impact of extremeduration heavy load carriage on peripheral fatigue should not be underestimated. Indeed, contrary to what was hypothesized, peripheral NM function alterations were large after the SMM, i.e. almost as large as in ultra-marathons, especially for PF muscles. For instance, PDb100 decreased by 26.468.0% for KE (P,0.05, d = 0.43) and 218.269.8% for PF (P,0.001, d = 1.14) after the SMM, while the mountain ultra-marathon study reported mean declines of about 212% for KE and 220% for PF [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref]. Furthermore, single stimuli to the relaxed muscles showed large decreases in potentiated Pt and CT after the SMM , corresponding to specific alterations comparable to those reported after ultra-endurance runs, especially for PF [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref] [bib_ref] Alterations of neuromuscular function after an ultramarathon, Millet [/bib_ref]. This general result regarding peripheral fatigue was not expected since, for conditions of duration and environment comparable to the SMM, extreme ultra-marathons are associated with much larger impacts underwent by the locomotor system at each step [bib_ref] Changes in running mechanics and spring-mass behavior induced by a mountain ultra-marathon..., Morin [/bib_ref]. However, the fact remains that load conditions are very different between these two types of exercise. The severe loads carried during the SMM may therefore have contributed to the substantial peripheral fatigue after the SMM. Finally, the presence of LFF in KE muscles after the SMM (Db10:100 declined by 25.569.3%, P = 0.08, d = 0.53) suggests specific failures in excitation-contraction coupling (i.e. reduction in Ca2+ release and/or decreased myofibrillar Ca2+ sensitivity [bib_ref] Electrical stimulation for testing neuromuscular function: from sport to pathology, Millet [/bib_ref] and muscular damage [bib_ref] High-and low-frequency fatigue revisited, Jones [/bib_ref]. Similarly, for PF muscles the decrease in paired stimuli peak force evoked at 10 Hz, yet in the same proportion as PDb100 since Db10:100 did not change, suggests alterations inside the muscle cell [bib_ref] Muscle fatigue: from observations in humans to underlying mechanisms studied in intact..., Place [/bib_ref]. This trend toward LFF after the SMM therefore indicates that the belief that low-speed locomotion is insufficient to promote the development of LFF [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] is not valid when subjects carry heavy loads, especially in KE muscles, which were more prone to LFF in the present study.
Unlike ultra-marathons after which lower amplitude and longer duration of M-wave were found [bib_ref] Neuromuscular consequences of an extreme mountain ultra-marathon, Millet [/bib_ref] , the SMM induced a 22.2621.2% increase in the SOL M-wave amplitude (P,0.01, d = 0.48; see . This result constitutes the most striking difference in peripheral NM function alterations between extremeduration heavy load carriage and extreme-duration runs. This also suggests that muscle excitability was preserved for both VL and SOL after the SMM. Therefore, processes located distal to the action potential propagation/transmission, involving Ca 2+ and/or cross-bridge kinetics [bib_ref] Muscle fatigue: from observations in humans to underlying mechanisms studied in intact..., Place [/bib_ref] , were implicated in the reduction of the evoked mechanical responses, in line with the trend toward LFF discussed previously.
Comparison with load-carrying exercise of moderate duration. We are not aware of any data regarding PF NM fatigue after load-carrying exercise. It is unfortunate since PF are more solicited than KE during walking, at least on flat terrain. The results of this study (i.e. PF MVC decline of ,11% associated with large peripheral fatigue after the 21-h SMM) therefore constitute to our knowledge the first data for future discussions about this muscle group. Concerning KE, it was surprising to see that the 21h SMM induced an overall fatigue (MVC decrease of ,10%, [fig_ref] Figure 2: Neuromuscular parameters measured in the knee extensors [/fig_ref] similar to that reported after much shorter load-carrying exercises. Indeed, after 12.1-km road marches performed at 4 km.h21 with loads up to 27 kg, Clarke et al. [bib_ref] Strength decrements from carrying various army packs on military marches, Clarke [/bib_ref] observed peak torque declines of about 8% in military subjects. Blacker et al. [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] even showed MVC decrease of ,15% (i.e. larger than in the present study) after bouts of 2-h treadmill walking performed at 6.5 km.h21 with a 25-kg backpack. In addition, beyond this observation about force loss, they also pointed out KE central activation deficit (4.2%) and LFF (4.5%) [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] , which represent central and peripheral NM function changes of KE similar to the present ones. This comparison therefore contradicts our initial hypothesis that the SMM would induce larger NM function alterations than load-carrying exercises of intermediate durations.
Although, at first glance, this finding could seem surprising because of the differences among studies in terms of exercise durations, environmental stressors and loads carried [bib_ref] Strength decrements from carrying various army packs on military marches, Clarke [/bib_ref] [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] , factors such as subjects' experience and mission profile should be considered. In fact, Blacker et al. [bib_ref] Neuromuscular function following prolonged load carriage on level and downhill gradients, Blacker [/bib_ref] studied recreational hikers while the present study investigated very experienced soldiers highly trained in load carriage exercise [bib_ref] A systematic review of the effects of physical training on load carriage..., Knapik [/bib_ref]. Furthermore, when looking at the SMM profile (figure 1), the distribution of the sleep, rest and rationing periods throughout the military mission likely allowed infantrymen to recover from or limit NM fatigue. Investigating the NM function of experienced soldiers after a similar mission without rest and sleep would therefore be interesting for future research. Nevertheless, the present results indicate that the typical management of soldiers' efforts during missions seems beneficial to preserve their NM function.
Effects of load and military mission-related fatigue on soldier locomotion. The results regarding the effects of acute military load carriage on locomotion (tables 2 and 3) were in line with our recent study [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref]. Briefly, the energy cost, mechanical work and spatio-temporal pattern of walking were significantly altered by load carriage [bib_ref] Effect of load and speed on the energetic cost of human walking, Bastien [/bib_ref] [bib_ref] Energy cost and mechanical work of walking during load carriage in soldiers, Grenier [/bib_ref] [bib_ref] Metabolic cost of generating muscular force in human walking: insights from load-carrying..., Griffin [/bib_ref] [bib_ref] Predicting energy expenditure with loads while standing or walking very slowly, Pandolf [/bib_ref]. Only the results relating to the effects of the SMM are discussed below.
Contrary to our initial hypothesis, neither the fatigue induced by the SMM, nor the interaction between SMM-related fatigue and acute load carriage induced significant changes in walking energetics or mechanics. Of these, the absence of fatigue effect (vs. interaction effect) is the most surprising result, especially from a mechanical standpoint [bib_ref] Effects of load carriage and fatigue on gait characteristics, Qu [/bib_ref] [bib_ref] Influence of fatigue and load carriage on mechanical loading during walking, Wang [/bib_ref]. Indeed, recent studies have shown that foot pressures, gait kinematics, peak ground reaction forces and loading rates are affected by fatiguing laboratory protocols and prolonged outdoor walking [bib_ref] Biomechanical analysis of fatigue-related foot injury mechanisms in athletes and recruits during..., Gefen [/bib_ref] [bib_ref] Effects of quadriceps fatigue on the biomechanics of gait and slip propensity, Parijat [/bib_ref] [bib_ref] Plantar pressure changes after long-distance walking, Stolwijk [/bib_ref] [bib_ref] The effects of fatigue on plantar pressure distribution in walking, Bisiaux [/bib_ref] [bib_ref] Effects of load carriage and fatigue on gait characteristics, Qu [/bib_ref] [bib_ref] Influence of fatigue and load carriage on mechanical loading during walking, Wang [/bib_ref]. Consequently, after such an extreme-duration exercise with severe loads, we expected changes in the walking gait. Nevertheless, the choice of mechanical parameters may partly explain the lack of alteration in the present study. Indeed, among the researchers that have investigated the effects of fatigue on the spatio-temporal pattern of walking, Qu & Yeo [bib_ref] Effects of load carriage and fatigue on gait characteristics, Qu [/bib_ref] , for instance, detected alterations but the significant differences they observed only concerned gait variability. Moreover, the other studies that have explored the effects of fatigue on walking mechanics mainly focused on plantar pressures [bib_ref] Biomechanical analysis of fatigue-related foot injury mechanisms in athletes and recruits during..., Gefen [/bib_ref] [bib_ref] Plantar pressure changes after long-distance walking, Stolwijk [/bib_ref] and peak forces during foot-ground contact [bib_ref] Influence of fatigue and load carriage on mechanical loading during walking, Wang [/bib_ref] , which represent different mechanical aspects of human walking than the absolute spatio-temporal parameters and mechanical work studied here (see e.g. ref. [bib_ref] The effects of fatigue on plantar pressure distribution in walking, Bisiaux [/bib_ref].
Nevertheless, in light of the relatively moderate NM function alterations observed after the SMM, it is also likely that our experienced soldiers were able to compensate for this KE and PF fatigue by biomechanical and NM adjustments [bib_ref] A systematic review of the effects of physical training on load carriage..., Knapik [/bib_ref]. Moreover, even if their sensation of general fatigue was high, the subjects did not report or mention lower-limb pain, which could have resulted in spatio-temporal or kinetic locomotor adaptations after the SMM, as shown after extreme runs [bib_ref] Changes in running mechanics and spring-mass behavior induced by a mountain ultra-marathon..., Morin [/bib_ref] [bib_ref] Changes in running kinematics, kinetics, and spring-mass behavior over a 24-h run, Morin [/bib_ref]. Such compensation for NM fatigue would however be unlikely in subjects performing an isokinetic fatiguing protocol inducing KE MVC decreases of 60% for instance [bib_ref] Effects of quadriceps fatigue on the biomechanics of gait and slip propensity, Parijat [/bib_ref]. Consequently, the present results raise the question of the scope of fatiguing laboratory protocols, which may induce artificial effects as compared with field protocols that reflect the reality of exercise [bib_ref] Interactive processes link the multiple symptoms of fatigue in sport competition, Knicker [/bib_ref]. Indeed, if we consider our subjects' locomotor responses after the SMM as representative of those of soldiers/walkers in real mission/trekking contexts, we can conclude that walking pattern and mechanical work do not appreciably change after overall heavy load carriage exercises of extreme-duration. This could also suggest that factors other than biomechanical and NM ones may better explain the accidentrelated musculoskeletal injuries reported after missions/trekking [bib_ref] Physiological and metabolic aspects of very prolonged exercise with particular reference to..., Ainslie [/bib_ref]. Decreased attention and cognitive fatigue are potential explanations but further study is necessary.
One possible limitation of the present study was the absence of a control group to rule out the potential effects of sleep rhythm disturbance on the NM function (i.e. only two 3-h sleeping periods separated by a 2-h period were allowed). However, in a previous study from our research group [bib_ref] Central and peripheral contributions to neuromuscular fatigue induced by a 24-h treadmill..., Martin [/bib_ref] , we showed that after 24 h of total sleep deprivation the subjects did not show any NM change. Thus, we assumed that this methodological choice did not alter the present data. Incidentally, the study of Rodgers et al. [bib_ref] Sleep deprivation: effects on work capacity, self-paced walking, contractile properties and perceived..., Rodgers [/bib_ref] showed no NM alteration in control groups subjected to 48 h without sleep, which reinforces our assumption on this point. Another potential limitation was the delay between the end of the SMM and the walking measurements, due to the NM function evaluation, which may have permitted some recovery. Nevertheless, when subjects present moderate NM function alterations such as those observed in the present study, metabolic factors are minimally involved in fatigue. Therefore, this delay in performing the measures does not seem critical (especially as the NM evaluation involved muscular contractions) and we can reasonably assume that this methodological choice did not change the results obtained in walking. Finally, although the SMM was designed to represent a military mission, it is accepted that such a simulation differs from the operational reality (i.e. absence of stress, fear, or operational goals). Thus, the mission environment represented here must be considered as a military-like physical context rather than an operational theater-like one. However, we are convinced that such a real-world approach remains more representative and beneficial than fatiguing laboratory protocols.
# Conclusion
The results of the present study showed that the central and peripheral NM function alterations were not larger after the 21-h military mission than after load-carrying exercises lasting from 2 to 3 h, contrary to our first hypothesis. Moreover, the NM function changes were lower after the 21-h military mission performed with heavy load carriage than after ultra-marathon runs, in accordance with our second hypothesis. Consequently, extreme-duration heavy load carriage induced overall moderate central and peripheral fatigue in experienced carriers for the two major muscle groups implicated in human walking (i.e. KE and PF). The lack of substantial central fatigue observed in the present study might be attributed, at least in part, to the beneficial resting and sleep times realistically distributed throughout exercise. Thus, NM fatigue was mainly attributable to peripheral alterations although it came with a significant perceived fatigue. It is therefore recommended to exploit each minute of resting/sleep periods to recover from and/or limit fatigue during such exercises, as observed in the very experienced soldiers of this study.
1 50 -Concerning the secondary aim of the present study, contrary to our third hypothesis, results indicated that the mechanical and metabolic parameters of walking do not appreciably change after an exercise lasting 21 h and involving severe load carriage. Therefore, assuming a link between perceived fatigue and attention/cognitive capabilities, it is expected that a large sensation of fatigue may decrease carriers' attention and, thus, partly explain the accident-related injuries arising at the end of prolonged load-carrying exercise (since the objective parameters reported here do not allow explanation of these kind of injuries).
[fig] Figure 1: Typical heart rate (HR) of a subject throughout the protocol.Figure includes the 21-h simulated military mission (SMM) and the pre-SMM (PRE) and post-SMM (POST) measurement sessions. Altitude, chronology and equipment conditions are inserted on and under the HR graph as indicative data. BT: battle equipment (27.461.1 kg corresponding to 33.462.6% of the subjects' BM), RM: road march equipment (42.961.4 kg, corresponding to 52.264.2% BM). doi:10.1371/journal.pone.0043586.g001 [/fig]
[fig] Figure 2: Neuromuscular parameters measured in the knee extensors (KE) before (PRE) and after (POST) the mission. a. Maximal voluntary contraction (MVC); b. Potentiated high-frequency doublet (PDb100); c. Ratio of paired stimulation peak forces at 10 Hz to 100 Hz (Db10:100); d. Voluntary activation level (%VA). *P,0.05, **P,0.01. doi:10.1371/journal.pone.0043586.g002 [/fig]
[fig] Figure 3: Neuromuscular parameters measured in the plantar flexors (PF) before (PRE) and after (POST) the mission. a. Maximal voluntary contraction (MVC); b. Potentiated high-frequency doublet (PDb100); c. Ratio of paired stimulation peak forces at 10 Hz to 100 Hz (Db10:100); d. Voluntary activation level (%VA). ***P,0.001. doi:10.1371/journal.pone.0043586.g003 Table 1. Potentiated peak twitch of knee extensors (KE) and plantar flexors (PF), and M-wave characteristics of vastus lateralis (VL) and soleus (SOL) muscles, before (PRE) and after (POST) the Simulated Military Mission. [/fig]
[table] Table 3: Spatio [/table]
|
Research on the Current Situation and Countermeasures of Inpatient Cost and Medical Insurance Payment Method for Rehabilitation Services in City S
Objective: This study aimed to introduce bed-day payment for rehabilitation services in City S, China, and analyze the cost of inpatient rehabilitation services. Key issues were defined and relevant countermeasures were discussed.Methods: The data about the rehabilitation cost of 3,828 inpatient patients from June 2018 to December 2019 was used. Descriptive statistics and the Kruskal-Wallis test were employed to describe sample characteristics and clarify the comparity of cost and length of stay (LOS) across different groups. After normalizing the distribution of cost and LOS by Box-Cox transformation, multiple linear regression was used to explore the factors influencing cost and LOS by calculating the variance inflation factor (VIF) to identify multicollinearity. Finally, 20 senior and middle management personnel of the hospitals were interviewed through a semi-structured interview method to further figure out the existing problems and countermeasures.Results: (1) During 2015-2019: both discharges and the cost of rehabilitation hospitalization in City S rose rapidly. (2) The highest number of discharges were for circulatory system diseases (57.65%). Endocrine, nutritional, and metabolic diseases were noted to have the longest average length of stay (ALOS) reaching 105.8 days. The shortest ALOS was found to be 24.2 days from the diseases of the musculoskeletal system and connective tissue. Neurological, circulatory, urological, psychiatric, infectious, and parasitic diseases were observed to be generally more costly. (3) The cost of rehabilitation was determined to mainly consist of the rehabilitation fee (23.63%), comprehensive medical service fee (22.61%), and treatment fee (19.03%). (4) Type of disease, age, nature of the hospital, and grade of the hospital have significant influences both on cost and LOS (P < 0.05). The most critical factor affecting the cost was found to be the length of stay (standardized coefficient = 0.777). (5) The key issues of City S's rehabilitative services system were identified to be the incomplete criteria, imperfections in the payment system, and the fragmentation of services.Conclusions: Bed-day payment is the main payment method for rehabilitation services, but there is a conflict between rapidly rising costs and increasing demand Tang et al.Research on Bed-Day Payment Method for rehabilitation. The main factors affecting the cost include the length of stay, type of disease, the grade of the hospital, etc. Lack of criteria, imperfections in the payment system, and the fragmentation of services limit sustainability. The core approach is to establish a three-tier rehabilitative network and innovate the current payment system.
# Introduction
Rehabilitation allows individuals with health problems to improve their functional status and reduce disability through interventions that interact with the environment (1). Thus, improving the accessibility and affordability of rehabilitation services is essential for maintaining the population's health (2). With the accelerated aging of people and changes in the disease spectrum, the demand for rehabilitation continues to increase. Alarcos Cieza et al. [bib_ref] Needs and unmet needs for rehabilitation services: a scoping review, Kamenov [/bib_ref] estimated that onethird of the world's population will need rehabilitation. The rapid growth in demand will inevitably create challenges for health care systems and health insurance, hence, the WHO in 2017 called to establish ten priority action areas, including the inclusion of rehabilitation in the universal health coverage, the establishment of a comprehensive rehabilitation services model, and the expansion of financing. Developed countries are at the forefront of the world's response to this issue [bib_ref] Rehabilitation for cancer survivors: how we can reduce the healthcare service inequality..., Anwar [/bib_ref] [bib_ref] Effectiveness of multidimensional cancer survivor rehabilitation and cost-effectiveness of cancer rehabilitation in..., Mewes [/bib_ref]. In 2002, the United States introduced a prospective payment system taking various factors, including patient diagnosis, functional status, age, and co-morbidities into consideration, to limit the cost increase and improve the quality of rehabilitation [bib_ref] Change in inpatient rehabilitation admissions for individuals with traumatic brain injury after..., Hoffman [/bib_ref]. Thus, a new payment system named FRGs based on the patient's functional status was developed. Another prospective payment system developed by GMS is Resource Utilization Groups (RUGs). Patients with similar resource utilization characteristics are divided into a group, and the case portfolio index or payment weight is calculated for each group. The facilities get payment based on their resource utilization. To constrain costs more effectively, Patient-Driven Payment Model (PDPM) was developed further. Patients are divided into groups according to clinical characteristics, such as main disease type and complications, and disability and dementia situations [bib_ref] Reforming medicare payments to skilled nursing facilities to cut incentives for unneeded..., Carter [/bib_ref]. Australia and the UK have also developed different case groups and strategies for rehabilitation to better promote rehabilitation development [bib_ref] Development of function-related groups version 20: a classification system for medical rehabilitation, Stineman [/bib_ref].
In the dual context of China's comprehensive promotion of the construction of a healthy China and the implementation of a strategy to actively cope with an aging population [bib_ref] Healthy China 2030: moving from blueprint to action with a new focus..., Chen [/bib_ref] , rehabilitation services are also receiving increased attention from the Chinese policymakers. The National Health Commission of the People's Republic of China pointed out the need to improve the rehabilitation services delivery system and enhance the capacity of rehabilitation services in its "Opinions on Accelerating the Development of Rehabilitation". The Chinese medical insurance is a typical public contract model, so the reform in payment methods from the administration will have great leverage [bib_ref] Brief introduction of medical insurance system in china, Chen [/bib_ref]. Governments from central to local levels are exploring reasonable payment systems for rehabilitation services to achieve the dual effect of improving quality and efficiency [bib_ref] Department of human resources and social security of Guangxi. Notice on the..., Zhao [/bib_ref]. Currently, pay for service (PFS) is the main payment option for rehabilitation services in most provinces in China, which easily results in waste of resources and accelerating increase in costs [bib_ref] Realignment of incentives for health-care providers in China, Yip [/bib_ref]. In 2017, the General Office of the central government suggested that bed-day payment could be applied to those rehabilitation services with a long length of stay and relatively stable cost. Zhejiang province began to set specific groups in the DRGs system for long-term rehabilitation hospitalization in 2019, taking average length of stay, average daily cost, and quality into account.
City S lies to the southeast coast of China, with a population of about 17.56 million in 2020. It is a relatively developed city, hospitals beds per 1,000 persons were 3.58 and physicians per 1,000 persons was 2.43. Now, it is in a dilemma where the medical resources especially quality resources are insufficient and the unmet need of residents for medical service is growing rapidly [bib_ref] Identifying spatial matching between the supply and demand of medical resource and..., Wu [/bib_ref]. The Health Security Bureau of City S Municipality (HSBS) realized the importance of rehabilitation earlier. In order to further reduce the economic burden of patients who need long-term rehabilitation, HSBS carried out a series of reforms, which could provide valuable lessons to other places. It started to sign rehabilitation service contracts with general or rehabilitative hospitals to better meet residents' rehabilitation needs in 2009. As of December 2020, HSBS has contracts with 13 hospitals: three of them are public, and ten are private, including three tertiary hospitals, three secondary hospitals, and seven primary hospitals. Patients with the following two conditions are eligible for the contract: vegetative people in a stable condition and needing long-term rehabilitative inpatient care; patients suffering from advanced tumors or cerebrovascular accident sequelae and trauma or needing hospice care [bib_ref] Study on the control and payment methods of long-term care coststaking long-term..., Chen [/bib_ref]. From 2015 to 2019, a total of 10,872 patients enjoyed this long-term rehabilitation care service.
The implementation of the program has primarily met the needs of residents for the long-term rehabilitation. Still, it is also facing the dilemma of much unmet demand from residents and excessive cost increases. Improving efficiency and quality has become a pressing problem for policymakers. This study aimed to analyze the cost of inpatient rehabilitation services in City S and discuss ways to improve the medical insurance policy and rehabilitative service delivery system.
# Materials and methods
# Quantitative research methods
# Materials
A retrospective study using data from HSBS was performed, selecting patients under the long-term rehabilitation service contracts. The information about cost, disease types (categorized by ICD 10), and length of stay (LOS) of all the discharges between 2015 and 2019 was provided by HSBS. With the help of HSBS, the data (N = 3,828) from June 2018 to December 2019 was extracted from the patients' electronic medical records of 13 hospitals under the contract after filtering out sensitive information (i.e., name, address). Information about sex, age, detailed cost categories, disease types (categorized by ICD 10), and other information in the medical records were included. Furthermore, the data were processed as follows: (1) exclude cases missing information about cost and length of stay; (2) exclude cases inconsistent with the actual situation, such as cases with length of stay greater than the duration from June 2018 to December 2019. Therefore, 59 cases were excluded and the final sample size was 3,769.
# Data analysis
The source data were entered using EXCEL 16.0 software to establish an independent database. Descriptive statistics were used to analyze the costs and sample characteristics. Due to the non-parametric distribution of data, the Kruskal-Wallis test was used to clarify the comparity of costs and length of stay in the different groups of patients; the multiple linear regression method was used to analyze the influencing factors of hospitalization cost and length of stay. Considering the hospitalization cost and length of stay were not normally distributed according to skewness and kurtosis, the Box-Cox transformation was performed. Variance inflation factor (VIF) was used to identify multicollinearity for the multivariate regression model. No VIF >10 was accepted [bib_ref] Cerebrospinal fluid shunt placement in the pediatric population: a model of hospitalization..., Lam [/bib_ref]. P < 0.05 (twotailed) was regarded as statistically significant. All the statistical analyses were implemented using the Stata 16.0.
# Qualitative research methods
To know more about current situation and explore countermeasures, semi-structured interviews were conducted with the middle and senior leaders of five hospitals that had contracts with HSBS. Two of the hospitals interviewed were tertiary hospitals, one was secondary, and two were primary. Besides, three of them were public and two were private. The interviewees included the president, vice president, director of the medical insurance department, and director of the rehabilitation department, for a total of 20 people. A separate interview with each person was conducted face to face and the total time was 720 min. The main contents were about the development status of the rehabilitation business of the hospitals, opinions on the current payment method, and the problems and countermeasures of the rehabilitation system in City S. The whole interviews were recorded, and after the data were compiled and analyzed, key issues were defined.
# Results
## Sample characteristics
Among all the 3,769 patients, 66.92% (N = 2,419) were male and 39.98% (N = 1,507) were 40-60 years old. The circulatory system diseases accounted for the largest proportion of 64.02% (N = 2,413). In total, 80.84% (N = 3,047) of patients were from private hospitals, 65.11% (N = 2,454) were from primary hospitals, followed by secondary hospitals with 25.76% (N = 971), and finally tertiary hospitals with 9.13% (N = 344). The cost and length of stay were found to be significantly different (P < 0.05) across groups of age, type of disease, nature of the hospital, and grade of the hospital (see [fig_ref] TABLE 1 |: Sample characteristics [/fig_ref].
## Cost
Cost From 2015 to 2019
The bed-day payment method was adopted under the rehabilitation service contract. In 2015, all facilities shared the same payment rate and were paid by 80$ per bed-day. In 2016, the payment rate began to take medical service price level into consideration. In City S, the price of inpatient medical services was divided into 3 levels: level 1, which was the standard price and applied to tertiary hospitals; level 2, 95% of level 1 and applied to secondary hospitals; level 3, 90% of level 1 and applied to primary hospitals. Therefore in 2016, the payment rate was divided into 3 classes: 105.6, 112, and 116.8 ($/bed-day). Tertiary hospitals enjoyed a higher rate than secondary or primary hospitals for the same service. The rate was adjusted dynamically and the formula was "payment rate = average inpatient cost of each bed-day in recent 3 years * (1 + basic growth rate) * (1 + inpatient price growth rate) * (1 + price level rate)". In 2019, the payment rates reached 130.4$ for tertiary hospitals, 123.8$ for secondary hospitals, and 117.3$ for primary hospitals.
During 2015-2019, the total discharges in City S for inpatient rehabilitation was 10,782 and tended to increase with an average growth rate of 166% each year. During this period, the cost of inpatient rehabilitation increased year by year with an average growth rate of 79.83%, and the growth rates in 2018 and 2019 were both over 100%. In 2015, the cost for inpatient rehabilitation accounted for and 4% of the total expenditure of the city's medical insurance fund, but it became 8.69% in 2019. The average cost per capita in 2015 was only 6,248.56$, while it rose to 7,930.864$ in 2019 (see [fig_ref] TABLE 2 |: Cost from 2015 to 2019 [/fig_ref].
From 2015 to 2019, 57.65% of the discharges were from circulatory system diseases (including cerebral infarction, hypertension, cerebral hemorrhage, and heart failure); neurological system diseases (namely, Alzheimer, enceph alomyelitis, hemiplegia, and epilepsy) accounted for 10.42%; musculoskeletal system and connective tissue diseases (namely, arthrosis, fracture, lumbar disc herniation, and cervical spondylosis), accounting for 6.18%. Overall, neurological, circulatory, urological, psychiatric, infectious, and parasitic diseases were observed to be generally more costly than bone and joint and sports-related diseases. In 2019, certain infectious and parasitic diseases were the costliest diseases and the average cost had reached 9,253.31$. Digestive system diseases were the least costly and the average cost was 2,491.65$ (see [fig_ref] TABLE 3 |: Average costs for different types of diseases [/fig_ref].
## Cost structure
The per capita rehabilitation cost for 3,769 patients was calculated to be 7,873.96$ and it mainly consisted of rehabilitation fee (23.66%), comprehensive medical service fee (22.61%), treatment fee (19.03%), Chinese medicine fee (12.54%), medicine fee (12.42%), and diagnosis fee (5.46%). The comprehensive medical service fee mainly contained a general treatment operation fee and a general medical service fee. Non-surgical treatment cost-dominated treatment fee. Overall, the cost for inpatient rehabilitation was mainly consisted of rehabilitation fee and service fee that reflects the labor value of medical and nursing staff. While the cost of drugs, consumables, and diagnostic fee were observed to account for a smaller proportion (see [fig_ref] TABLE 4 |: Composition of per capita rehabilitation medical costs [/fig_ref].
## Influencing factors of cost
The multiple linear regression analysis was performed using inpatient rehabilitation cost (transformed by the Box-Cox model) as the dependent variable. Independent variables were: age, length of stay (LOS), type of disease, nature of the hospital, grade of the hospital, and the out-of-pocket ratio of the cost. The results showed that F = 721.491, P < 0.001, and adjusted R 2 = 0.713, indicating that the regression equation held and fit well. The highest VIF was 5.710 and the model waived the risk of multicollinearity (VIF > 10). LOS, tertiary hospital (compared to primary hospital), circulatory system diseases (compared to tumors), and private hospital (compared to public hospital) were all found to be positively associated with inpatient rehabilitation cost. And the out-of-pocket ratio had a negative effect on inpatient rehabilitation cost. Furthermore, the standardized coefficient of LOS is 0.777, which was the biggest one, and the Pearson's correlation between LOS and inpatient rehabilitation cost was as high as 0.836, indicating LOS was the most influencing factor (see [fig_ref] TABLE 5 |: Multiple linear regression analysis of costs for inpatient rehabilitation [/fig_ref]. The length of stay varied in different types of diseases. Endocrine, nutritional, and metabolic diseases were found to have the longest average length of stay, reaching 105.8 days. Genitourinary diseases reached 72.8 days and the LOS of neurological disorders were 71.9 days, which were relatively longer than other types of diseases. The shortest average length of stay was found to be 24.2 days from the diseases of the musculoskeletal system and connective tissue (see [fig_ref] TABLE 6 |: Different diseases' length of stay [/fig_ref].
## Influencing factors of length of stay
Length of stay (transformed by the Cox-Box model) was set as the dependent variable in the multiple regression model. Independent variables were age, out-of-pocket ratio, type of disease, nature of the hospital, and grade of the hospital. The results showed higher age, smaller out-of-pocket ratio, public, and tertiary hospital resulted bigger length of stay. Different types of diseases were all found to be associated with LOS. But, what could not be ignored was that the adjusted R 2 was only 0.252, suggesting other critical factors omitted. The highest VIF was 5.595, and there was no risk of multicollinearity (see [fig_ref] TABLE 7 |: Multiple linear regression analysis of length of stay [/fig_ref].
## Key issues
Of all the interviewees, 75% were male, 35% were 40-45 years old, and 30% had 15-20 working years. There were separately 40% of total interviewees from primary and tertiary hospitals, and 60%
were from public hospitals (see [fig_ref] TABLE 8 |: Demographic information of interviewees [/fig_ref]. Key issues were defined as followed.
Rehabilitation-related criteria were identified to be incomplete, including patient admission and discharge criteria and grading and classification criteria, which was one of the reasons for the rapid increase in cost. Currently, the admission criteria applicable to the contract were vague, without clear regulations on the type and severity of the disease. It resulted in a mixed composition of patients. A significant number of longterm care and hospice patients, less costly but having a longer stay, were also included. But in the context of bed-day payment, the insurance needed to pay the same amount according to the length of stay, which was much more than the real cost. On the other hand, some patients had long been occupying beds due to the lack of clear discharge criteria, resulting in not only a tight bed capacity but high cost. Grading and classification criteria were needed to strengthen the management of patients and provide a reference for payment. But there was a lack of unified criteria, which brought a great barrier to the refinement of payment and quality improvement.
Single bed-day payment was insufficient to reflect the real resource usage and guarantee equity. Patients with different types of diseases or with the same disease but different degrees of severity were applicable to the same payment rate. For example, patients in critical rehabilitation usually required some surgical treatment and relied on medical instruments so that the cost of whom was much more than other kinds of patients. But the long-term care patients were much less costly, which was the main source of hospitals' profits. Therefore, hospitals might face pressure of cost control or room for profit. The equity of payment would be weakened or there might be a risk of patient selection or malpractice that reduced the quality of care. Besides, considering City S is building a separate long-term care insurance, the longterm care patients will be excluded from the rehabilitation service contract. Therefore, hospitals would face greater challenges as the average cost increase greatly. Rehabilitation service was fragmented and failed to meet the requirements of integrated care. Most rehabilitation facilities interviewed did not have referral agreements with other hospitals, which meant that patients could not get continued care. The linkage between acute care and rehabilitation was not smooth, resulting in some patients missing the best period of rehabilitation. Besides, some rehabilitation patients still had the need for acute care. But the payment of the medical insurance only covered rehabilitation services in single hospitalization, which resulted in the insufficient care for the sake of controlling cost.
# Discussion
The data from City S showed that there was a rapid growth in the number of discharges and expenditures from 2015 to 2019, which reflected the strong demand for rehabilitation and suggested that the reform successfully satisfied part of the unmet need. City S encouraged patients in need of rehabilitation services to go to secondary or primary rehabilitation facilities. Thus, there were only 18.51% of patients in tertiary hospitals and the average out-of-pocket ratio was only 8.48%. This provided a good example for other cities or countries to release tight capacity of tertiary hospitals and improve the availability of rehabilitation services. The bed-day payment was also an innovative method compared with paying for service, which was widely used in China [bib_ref] On perfecting the payment and cost control system of social health insurance..., Zhuo-Fan [/bib_ref]. It could reduce the waste of medical resources. However, the excessive increasing rate of expenditure, less specific payment system, and fragmented services delivery system cast a doubt on the sustainability. Therefore, there is still much room for improvement.
First, the length of stay should be reasonably controlled. The results showed that the length of stay was the most important factor influencing the growth of rehabilitation expenditure, which is consistent with other studies in China [bib_ref] Risk selection and cost shifting in a prospective physician payment system: evidence..., Kantarevic [/bib_ref] [bib_ref] How to use the international classification of functioning, disability and health as..., Stucki [/bib_ref] [bib_ref] Grey correlation analysis of average length of stay in general hospitals, Li [/bib_ref]. To contain the LOS and increase resources' mobilization, diagnosisrelated groups (DRGs)-based payments were wildly used in the US and other high-income countries. Hospitals were paid within the predefined scale according to classifications of DRG [bib_ref] Review of diagnosisrelated group-based financing of hospital care, Mihailovic [/bib_ref]. A new per diem inclusive payment system called the DPC/PDPS (diagnosis procedure combination/per diem payment system) were adopted in Japan. In total, three periods were specified for each disease along with standardized per diem payments for each period and the payments diminish with increasing LOS (34, [bib_ref] Comparison of the length of stay and medical expenditures among Japanese hospitals..., Nawata [/bib_ref]. Italy linked reimbursement to effective stay based on the time to reach peak improvement for different groups of conditions [bib_ref] Toward a new payment system for inpatient rehabilitation-Part II: reimbursing providers, Saitto [/bib_ref] [bib_ref] Building a people-centred integrated care model in urban China: a qualitative study..., Liang [/bib_ref]. A weighted blended payment model was designed for rehabilitation by Australia, which applies a mixture of the episode and per diem rates. The whole rehabilitation was divided into four episodes: short stay (1-3 days), low outliers, inlier range (ALOS +/-4 days), high outliers, and every episode was attached to a separate per diem rate according to the resources use [bib_ref] The development of version 2 of the AN-SNAP casemix classification system, Green [/bib_ref]. Thus, experience from the developed countries could be taken by City S that set a predefined payment rate based on diagnosis and interventions and makes the rate diminish over LOS.
The payment system needs to be innovated to drive up efficiency. Now, worldwide health systems are increasingly moving toward payment systems based on a fixed tariff structure for each episode of treatment and case-mix classification was adopted to drive up efficiency and to contain costs [bib_ref] International casemix and funding models: lessons for rehabilitation, Turner-Stokes [/bib_ref] [bib_ref] Casemix classification payment for sub-acute and non-acute inpatient care, Thailand, Khiaocharoen [/bib_ref] [bib_ref] Subacute casemix classification for stroke rehabilitation in an Australian setting, Kohler [/bib_ref] [bib_ref] Inpatient rehabilitation facilities under the prospective payment system: lessons learned, Zorowitz [/bib_ref]. For example, medicare beneficiaries are assigned to case-mix groups (CMGs) considering the diagnosis, age, level of motor, and cognitive function. Medicare would pay rehabilitation facilities predetermined per discharge rates based on the CMGs and market area wages. Therefore, bed-day payment rates could also be adjusted by diagnosis, grade of the hospital, function, and some other factors related to resources use. Hospitals have the right to negotiate with the medical insurance agency to set the final basic payment rate. Besides, more payment methods could be adopted for different kinds of patients [bib_ref] International casemix and funding models: lessons for rehabilitation, Turner-Stokes [/bib_ref]. For certain diseases with relatively stable resource usage level, a fixed tariff could be used. For simple services lasting a long time, capitation payment can be used. For some complex services, pay for service is more appropriate. Medicare also made outlier payments when a rehabilitation facility's estimated total costs for a case exceeded a cost threshold. The outlier payment for a case was equal to 80% of costs above this threshold. Therefore, for some rehabilitation patients with high-medical demand that are extraordinarily costly compared with the other cases, a outlier payment is also needed. The over costly medical service provided during the hospitalization could be assigned to DIP (diagnosis-intervention packet) payment, which is a novel casemix with a global budget payment system developed by China [bib_ref] The pilot of a new patient classification-based payment system in China: the..., Qian [/bib_ref]. Finally, a quality-based payment system for rehabilitation services should be further explored after various standards and assessment systems being improved. Various criteria need to be developed and insurance management should be strengthened. Clear admission and discharge criteria are necessary to increase efficiency and enhance effectiveness. For example, CMS specified 13 qualifying medical conditions for inpatient rehabilitation facilities to distinguish rehabilitation services from acute care, including stroke, spinal cord injury, amputation, etc.. A scientific function assessment system is also needed because it is the basis of clarifying patients' needs and providing services reasonably [bib_ref] Building a people-centred integrated care model in urban China: a qualitative study..., Liang [/bib_ref]. Establishing a third-party assessment committee to independently conduct a dynamic assessment, and the results could be used in the admission and discharge criteria or to evaluate hospitals' service quality. It also could be an important reference for payment. Finally, clinical paths could be introduced to efficiently manage hospitalization schedules [bib_ref] Evaluation of the inclusive payment system based on the diagnosis procedure combination..., Nawata [/bib_ref] , ensuring standardization of service.
To promote the integration of the rehabilitation service, a three-tier rehabilitative network should be built, including tertiary hospitals, secondary general hospitals or stand-alone rehabilitative centers, and community health facilities or primary hospitals. Although City S made some meaningful explorations about integrated care [bib_ref] Payment models in primary health care: a driver of the quantity and..., Chami [/bib_ref] , the reforms were centered in acute care with less attention paid to the rehabilitation service delivery system. The UK has developed a three-tiered model of local, district, and regional services. More specifically, Level 1 services are discrete tertiary specialized rehabilitation services. Level 2 services are discrete specialist rehabilitation services. Level 3 services are local non-specialist rehabilitation services [bib_ref] International casemix and funding models: lessons for rehabilitation, Turner-Stokes [/bib_ref]. In China, the tertiary hospitals could be the regional center delivering early rehabilitative services and responsible for the training of talents. Secondary hospitals or standalone rehabilitative centers provide post-acute rehabilitation. Primary hospitals or community health facilities are mainly for patients in stable conditions or needing long-term care and provide continuous services. Different levels of facilities should cooperate with each other and form a rehabilitative consortium.
These facilities as a whole can provide whole-life services for patients under the guidance of patient-centered philosophy on the basis of two-way referral programs. What is more, now in China there are too many patients staying in tertiary or secondary hospitals and reluctant to go to primary hospitals, exacerbating the strain on rehabilitative resources. The policy must guide the flow of patients to primary rehabilitation facilities. Some measures could be taken, such as increasing the reimbursement rate of primary rehabilitative facilities and so on.
# Conclusion
City S has set a good example for other low-and middleincome countries to satisfy the unmet need for rehabilitation services. It also provided evidence for how the bed-day payment method worked to drive up efficiency. But there is a conflict between rapidly rising costs and increasing demand for rehabilitation. After identifying length of stay as the most important factor affecting cost, key issues were defined about current payment method and rehabilitation service delivery system. To contain cost and drive-up efficiency, the core approaches are to establish a three-tier rehabilitative network and innovate current payment system through introducing a classification system based on diagnosis and interventions, making the payment rate diminish over LOS and biding a mixed payment system.
# Limitations and prospects
Only service providers were interviewed in our study, while other stakeholders' opinions especially the patients, did not receive enough attention. Due to the quality of data, the longitudinal changes of fees in different categories could not be analyzed and some key factors influencing LOS were omitted.
In the future, we hope to comprehensively analyze the problems of the current rehabilitation system from the perspectives of management, supply and demand sides. Besides, more factors influencing cost and LOS would be explored if high-quality data are available.
# Data availability statement
The datasets presented in this article are not readily available because of privacy and ethical restrictions.
Requests to access the datasets should be directed to [email protected].
# Author contributions
DT and DZ contributed to the conception of the study. DT, MH, JB, and DZ wrote the first draft of the manuscript and collected and analyzed the samples. DZ and NY consulted on data collection and analysis. DZ reviewed the manuscript and polished it. All authors contributed to the article and approved the submitted version.
[table] TABLE 1 |: Sample characteristics. [/table]
[table] TABLE 2 |: Cost from 2015 to 2019. [/table]
[table] TABLE 3 |: Average costs for different types of diseases ($). [/table]
[table] TABLE 4 |: Composition of per capita rehabilitation medical costs ($). [/table]
[table] TABLE 5 |: Multiple linear regression analysis of costs for inpatient rehabilitation. [/table]
[table] TABLE 6 |: Different diseases' length of stay. [/table]
[table] TABLE 7 |: Multiple linear regression analysis of length of stay. [/table]
[table] TABLE 8 |: Demographic information of interviewees. [/table]
|
Actomyosin Contractility in the Generation and Plasticity of Axons and Dendritic Spines
Actin and non-muscle myosins have long been known to play important roles in growth cone steering and neurite outgrowth. More recently, novel functions for non-muscle myosin have been described in axons and dendritic spines. Consequently, possible roles of actomyosin contraction in organizing and maintaining structural properties of dendritic spines, the size and location of axon initial segment and axonal diameter are emerging research topics. In this review, we aim to summarize recent findings involving myosin localization and function in these compartments and to discuss possible roles for actomyosin in their function and the signaling pathways that control them.
# Introduction
Neurons are highly specialized cells with an exceptional degree of spatial compartmentalization. Despite of a large morphological and functional diversity of cell types, most neurons possess long, thin processes known as axons and branched dendrites that can extend for distances several orders of magnitude higher than the size of the cell body they emanate from. At the same time, this extreme shape can persist for decades virtually unchanged. Clearly, axons, which are thousands of times longer than they are in diameter, experience great mechanical stress. They must be sufficiently stiff to resist mechanical tensions and not tear, but remain flexible enough to accommodate for structural plasticity that may be required for their functional adaptability. Such mechanical and structural properties are generated and maintained by the cytoskeleton in conjunction with force-generating molecular motors. An especially prominent role is played here by actomyosin, a network of interconnected actin filament bundles that are pulled together by myosin motors, especially non-muscle myosin 2 (NMII, Box 1). Actin and NMII are evolutionarily old molecules and both are ubiquitously expressed, including the vertebrate central nervous system. NMII has a well-described role in neurite elongation, axonal outgrowth and neuronal polarization, as is abundantly present in neuronal growth cones, where it controls microtubule bundling and regulates the actin-rich lamellipodium and filopodia. More recently, NMII has been localized to dendritic spines and the axonal initial segment (AIS). The topic of this review is actomyosin functions in these important neuronal subdomains which in contrast to the growth cone are much less well understood. Box 1. Actomyosin in the nervous system.
The term actomyosin generally describes contractile bundles assembled of actin filaments that are interconnected by bipolar bundles of myosin II. The motor domain of myosin II can execute a power stroke after hydrolysis of ATP and release of phosphate, moving the myosin molecule relative to the actin cable towards its barbed end in the process. This pulls the actin filaments on both ends of the myosin bundle closer together, leading to contraction of the actomyosin structure. Binding of a new ATP molecule leads to unbinding of the motor domain from actin and new binding further upstream the actin filament for progression of the movement. This mechanism of contraction is the basis for muscle function and many mechanical processes in cells. Generally, myosin II isoforms for the skeletal, cardiac and smooth muscle, which are specialized to function in elongated thick bundles called sarcomers, are distinguished from three non-muscle myosin II (NMII) isoforms that execute many different mechanical functions in cell biology. NMII exists as a hexamer that consists of two copies each of elongated heavy chains bearing motor domains, two regulatory light chains and two essential light chains that stabilize the heavy chain structure. These hexameric units with two motors on one end and an elongated coiled coil on the other end further bundle both in a parallel and antiparallel manner into bipolar structures that can pull actin filaments together. The essential light chain stabilizes the hexamer and phosphorylation of the regulatory light chain at serine 19 is required for NMII to be able to execute its power stroke. The three NMII isoforms NMIIA, -B and -C are encoded by the heavy chain genes MYH9, MYH10 and MYH14, respectively. All these isoforms are expressed in most non-muscle cells and NMII A and B are expressed highly in the nervous system. There is no detectable difference in actin binding or activation between non-muscle myosin II A, B and C, but the isoforms differ in subcellular localization and some biochemical properties such as ATPase activity and duty ratio. Most mammals have six actin genes, four of which are expressed mostly in muscle. The cytoplasmic β-actin (ACTB), a complex locus with 22 introns and 23 splice isoforms, and γ-actin (ACTG1), with 16 introns and 19 mRNAs that encode for 15 different isoforms in a highly tissue-specific manner, are both expressed in the nervous system. In this review, when we use the term actomyosin, we refer to bundles and networks of NMII and β-actin or γ-actin.
## Neuronal morphology and compartmentalization
Most neurons in the central nervous system undergo a defined developmental program that starts with the growth of several processes from the cell body or soma. The microtubule cytoskeleton plays an essential role in providing structural support for growing neurites, whereas a dynamic, branched actin cytoskeleton enriched at their tips in so-called growth cones is important for giving the directionality and further differentiation of the neuron. One of these processes poised to become the axon then undergoes a period of quick continuous growth that requires the generation of bundles of microtubules that are generated through de novo polymerization and microtubule transport, and the activity of cdc42. As a result, the neuron is polarized into somato-dendritic and axonal compartments. After a period of axonal outgrowth, the dendrites start to develop more and more complex branches and form hundreds of contact sites with axons from other cells. Stabilization of these connections between neurons and the recruitment of post-synaptic components, as well as pre-synaptic vesicles and secretion machinery, leads to synapse formation and specialization in the membrane composition of the pre-and post-synaptic sites. In mature neurons the majority of excitatory post-synapses are located to the flattened tips of bulbous protrusions called dendritic spines, where ion channels, receptors and adhesion molecules supported by scaffolding proteins are enriched in a membrane domain called the post-synaptic density (PSD, Box 2). This PSD is organized in nanodomains and tightly apposed across the synaptic cleft to a corresponding synaptic vesicle release machinery in the pre-synapse. Dendritic spines can vary in their morphology, size and molecular composition, and the formation, plasticity and stability of the synapse bearing spines is thought to be the physical correlate to learning and memory formation. The shape and geometry of neurons are thus closely connected to their function in the size range of neuronal compartments varies from elongated axons that can span many centimeters to deliver action potentials to the low micrometer scale of dendritic spines. Their highly branched and elongated processes reflect the network architecture of the nervous system and both the tubular narrow shape of the axon as well as the thin neck of the dendritic spine are thought to play important biophysical roles in regulating neuronal electrical signaling via the ion flow and in creating biochemically compartmentalized subcellular domains. The spatial and geometric organization of signaling molecules and structures is thus tightly regulated, and these functions are executed by the cytoskeleton.
## Box 2. dendritic spines.
Dendritic spines are small protrusions of the dendritic shaft that are the postsynaptic site of excitatory glutamatergic synapses, which comprise a pre-synaptic terminal or bouton separated by the synaptic cleft from a specialized membrane domain called the post-synaptic density (PSD). While the pre-synaptic site contains numerous neurotransmitter vesicles spatially arranged for swift membrane fusion by proteins forming a cytomatrix of the active zone and highly sensitive vesicle release machineries, the PSD accommodates different types of glutamate receptors and calcium channels anchored to scaffolding proteins. Calcium signaling plays an essential role in evoking vesicle release upon neuronal depolarization at the pre-synapse as well as in triggering calcium-dependent kinase and phosphatase pathways, such as CaMKII or calcineurin at the PSD. Calcium signaling via the calcium binding proteins calmodulin and caldendrin is transduced to various actin modifiers, thereby directly regulating the morphology of dendritic spines in response to stimuli. Pre-and post-synaptic sites are tightly connected via cell adhesion molecules forming hetero-and homophylic interactions between each other. The most common type of dendritic spines in the adult brain are mushroom-like spines. Their shape, with a bulbous head and thin neck, is important for the compartmentalization of synaptic signaling and provides input specificity to this very synapse. The specific shape of dendritic spines and their ability to undergo rapid changes or a long-term stabilization of their structure to a large degree depends on the actin cytoskeleton. Thousands of dendritic spines can be found on branched dendrites and excitatory synaptic plasticity is accompanied by changes in a number of AMPA receptors and in the head size of dendritic spines that are thought to be the physical correlate of learning and memory formation.
While the entire cytoskeleton has important roles in neurons, the role of neurofilaments and microtubules have been extensively reviewed in. Here we will focus on actomyosin cytoskeleton in narrow neuronal compartments, whereas its role in the growth cone has been discussed elsewhere. We will summarize recent discoveries on the structure and components of the neuronal actomyosin network and discuss mechanisms involved in regulation of contractility and its relationship to neuronal functions, such us plasticity of the axonal initial segment (AIS) or dendritic spines.
## The membrane-associated periodic skeleton (mps) in neurons
In one of the first breakthrough discoveries of single molecule super-resolution microscopy (SMLM), a periodic arrangement of~200 nm spaced actin rings has been found along the axon that is interconnected by bipolar spectrin tetramers. This finding came unexpected, as this striking structure had never been observed in decades of investigation of axons with electron microscopy, which is most likely due the very frail nature of actin filaments that easily get disrupted upon a treatment with detergents. Since then, the existence of a periodic MPS has been seen by a number of microscopy methods and in several laboratories. Importantly, the nanoscopic organization of the components of the MPS has recently been verified in platinum replica electron microscopy of unroofed cultured neurons. Strikingly, a new organization of actin filaments into~1 µm long braids of two actin filaments has been found here. The MPS structure seems to be conserved from worm to humans, and by now it has been found in a variety of excitatory and inhibitory neurons of the central nervous system and motor neurons of the peripheral nervous system. While the MPS was initially only observed in neuronal axons, it has since been observed in neuronal stem cells and shown to be present in astrocytes and oligodendrocytes as well. It is also found in dendrites, specifically at the neck of dendritic spines.
The MPS emerges early in neuronal development and can be detected from two days in vitro (DIV) onward progressing from the proximal axon, where it precedes all axon initial segment makers. Braided double actin filaments of the MPS align the cylindric axon along its entire length and colocalize with adducin, tropomyosin and the phosphorylated form of the myosin light chain. These ring structures are interconnected by spectrin tetramers consisting of two α and two β spectrins, with the N-termini of the β spectrins connected to actin and the C-termini located in the center between actin rings. Spectrins are required for MPS assemblyand the mechanical stability of axons. Indeed, in spectrin knockout animals, axons can tear in response to mechanical stress induced by movement of the animal, suggesting a structural role of the MPS in axonal stability. The β spectrin isoform in the MPS spectrin tetramer seems to vary with the specific location in the cell with spectrin βIV in the AIS, spectrin βII in the distal axon and dendrites and βIII largely in dendrites. Although the periodic pattern of βIII spectrin in spines has not been directly confirmed, only a fraction of dendritic spine necks contains βII spectrin, suggesting that other spectrin isoforms might be responsible for spacing of the actin rings at this site. Indeed, spectrin βIII has been found at the base of dendritic spines and shRNA knock down of this isoform results in collapse of dendritic spines into shaft synapses. Together, these findings suggest that the neuronal MPS may exhibit locally distinct compositions and β spectrin isoforms may fine-tune its structural properties. The organization of the neuronal cytoskeleton. A two-dimensional (2D) actin-spectrin meshwork, similar to those found in other cell types (e.g., erythrocytes), spans the soma of the cell. In contrast, a one-dimensional (1D) periodic membrane cytoskeleton (MPS) is found in axons, in a fraction of dendrites and at the neck of dendritic spines. Top: The MPS consists of actin rings at a periodicity of ~200 nm, interspersed with spectrin tetramers. Each actin ring is formed by two braided actin filaments. The actin rings are further stabilized and regulated by the capping protein adducin and by tropomyosin. The spectrin tetramers are comprised of two αII spectrins and two compartment specific isoforms of β-spectrin. The axon initial segment (AIS) is a stretch of 50-100 µm at the beginning of the axon. The major scaffold in the AIS is AnkyrinG (AnkG), which binds to spectrin βIV and recruits the adhesion molecule neurofascin and ion channels. Phosphorylated myosin light chain (pMLC) is localized to actin rings in the axon. Microtubule bundles are stabilized by plus end binding proteins (EB) along the axon in both the AIS and the distal axon. In the distal axon the MPS is organized by AnkB, which in turn binds to βII-spectrin. AnkB is also arranged periodically, though the pattern is less prominent. Bottom: Dendritic spines are important for compartmentalization of synaptic signaling conferred by glutamate receptors and calcium channels. While the head of the spine contains branched actin filaments, the MPS is prominent in the neck region. Likely consisting of acting rings interspersed by αII and βII-spectrin tetramers. The MPS has also been observed in a sub-fraction of mature dendrites. Here, spectrin tetramers contain the βIII-spectrin isoform.
## Regulation of the mps
The unique organization of the MPS is perfectly suited to provide structural durability and resistance to mechanical forces that axons or dendritic spines encounter during intracellular trafficking, remodeling of neuronal morphology or a tissue movement. Indeed, axons break more readily when spectrin is absent. Filamentous actin and spectrins are the most important structural elements of the MPS as pharmacological studies using actin-depolymerizing drugs such as cytochalasin D and latrunculin or down-regulation of spectrin's expression resulted in disintegration. The organization of the neuronal cytoskeleton. A two-dimensional (2D) actin-spectrin meshwork, similar to those found in other cell types (e.g., erythrocytes), spans the soma of the cell. In contrast, a one-dimensional (1D) periodic membrane cytoskeleton (MPS) is found in axons, in a fraction of dendrites and at the neck of dendritic spines. Top: The MPS consists of actin rings at a periodicity of~200 nm, interspersed with spectrin tetramers. Each actin ring is formed by two braided actin filaments. The actin rings are further stabilized and regulated by the capping protein adducin and by tropomyosin. The spectrin tetramers are comprised of two αII spectrins and two compartment specific isoforms of β-spectrin. The axon initial segment (AIS) is a stretch of 50-100 µm at the beginning of the axon. The major scaffold in the AIS is AnkyrinG (AnkG), which binds to spectrin βIV and recruits the adhesion molecule neurofascin and ion channels. Phosphorylated myosin light chain (pMLC) is localized to actin rings in the axon. Microtubule bundles are stabilized by plus end binding proteins (EB) along the axon in both the AIS and the distal axon. In the distal axon the MPS is organized by AnkB, which in turn binds to βII-spectrin. AnkB is also arranged periodically, though the pattern is less prominent. Bottom: Dendritic spines are important for compartmentalization of synaptic signaling conferred by glutamate receptors and calcium channels. While the head of the spine contains branched actin filaments, the MPS is prominent in the neck region. Likely consisting of acting rings interspersed by αII and βII-spectrin tetramers. The MPS has also been observed in a sub-fraction of mature dendrites. Here, spectrin tetramers contain the βIII-spectrin isoform.
A special type of the MPS is present in the AIS which is a stretch of 50-100 µm at the beginning of the axon that separates the axonal and somatodendritic domains of neuronal cells. The AIS contains a specific complement of cytoskeletal and adaptor proteins that cluster adhesion proteins and several types of ion-channels, which are responsible for the generation of action potentials. The most prominent AIS-specific molecules are spectrin βIV and a large AnkyrinG (AnkG) isoform, which directly interact, and the cell adhesion protein neurofascin. These molecules are organized in an MPS of higher complexity than it is found in the distal axon. In analogy to axonal spectrin βII, in the AIS, spectrin βIV interconnects braided actin rings in a~200 nm periodicity. AnkyrinG binds to spectrin βIV between the actin rings and this scaffold recruits neurofascin and ion channels in a manner that is evolutionarily conserved from vertebrates onward. Such high local density of ion channels allows for action potential initiation. In addition to this, AnkyrinG and the AIS are required for the maintenance of neuronal polarity. Specifically, if AnkG is downregulated by shRNA in cultured cells or in vivo, axons lose their molecular identity and start to acquire dendritic features including postsynaptic densities that form adjacent to nearby presynaptic terminals of other axons. In a situation of hypoxia, such as in ischemia, the AIS breaks down, and consequently, neuronal polarity is lost, resulting in severe pathological consequences. This process is mediated by the calcium-dependent cysteine protease calpain. Even though the AIS persists for the entire lifetime of a neuronal cell, it is remarkably plastic and dynamic. First, as described above, it assembles in a dynamic process during the first week of neuronal development. Secondly, in response to chronic depolarization, the entire AIS structure can change in length and shift in its location along the axon, resulting in modified action potential generation. Recently, a prominent role of non-muscle myosin II and regulating molecules in the plasticity of the AIS and the MPS in response to Ca2+ signaling as discussed below has emerged.
## Regulation of the mps
The unique organization of the MPS is perfectly suited to provide structural durability and resistance to mechanical forces that axons or dendritic spines encounter during intracellular trafficking, remodeling of neuronal morphology or a tissue movement. Indeed, axons break more readily when spectrin is absent. Filamentous actin and spectrins are the most important structural elements of the MPS as pharmacological studies using actin-depolymerizing drugs such as cytochalasin D and latrunculin or down-regulation of spectrin's expression resulted in disintegration of the MPS. Interestingly, the MPS is more resistant to actin depolymerizers than actin filaments at dendritic spines, filopodia or axonal actin cables. Such resistance is increasing with neuronal maturation and correlates with developmental expression profile of the actin-capping protein α-adducin. Moreover, there are location-specific differences in MPS stability with the AIS being particularly resistant to high doses of cytochalasin D and latrunculin, suggesting that there might be further differences indicating further diversity in the composition of the membrane cytoskeleton.
Earlier experimental and modeling data suggested that the flexibility of the MPS along the axon might be limited because the spectrin filaments are already held under entropic tension. Therefore, for example, injury-induced disruption of the cortical axonal cytoskeleton cannot be spontaneously restored. However, the axon is an elastic compartment that can undergo radial deformations for instance during trafficking of a large cargo or neuronal activity. The length and the localization of the AIS are changed during neuronal development and as response to neuronal activity which would require lateral assembly/disassembly-based movement of its specialized MPS. Indeed, recent discoveries indicate that the MPS can undergo reorganization upon very specific types of stimuli and the role of calcium signaling and the actomyosin network are currently the central focus.
## Regulation of the mps by calcium
The AIS disappears in brain regions affected by ischemic injury and oxygen-glucose depletion, which lead to degradation of βIV-spectrin and AnkG by the calcium-dependent cysteine protease calpain. MPS βII-spectrin can likewise be digested by calpain-2 upon activation of ERK signaling. Calpain-1 and -2 are major cellular proteases that are critical for proper neuronal branching and dendritic spine complexity. Increased calcium concentrations activate calpains and are known to lead to the cleavage of several major cytoskeletal proteins, such as MAP2, spectrin, vimentin, internexin and others. Therefore, it is not surprising that calpains play an important role in cytoskeletal reorganization in neurons as well. The axonal MPS serves as a signaling platform for receptor tyrosin kinases (RTK), adhesion molecules and G-protein coupled receptors, which are corralled by the periodic actin ringsand immobilized by them in response to extracellular stimuli. Functional cross-talk between clustered receptors triggers the activation of ERK kinase signaling, which in turn activates calpain-2, resulting in a rapid MPS degradation. Removal of the constraints, limiting the diffusion of membrane proteins, contributes to the termination of RTK trans-activation signaling. It is possible that similar mechanisms of MPS re-organization will be present not only in the axon but in dendrites and the neck of dendritic spines. It is well-known that synaptic firing results in a calcium influx which triggers activation of calpainsand the disruption of the dendritic MPS by neuronal activity was found to be dependent of NMDAR-mediated calcium influx. Calpain-2 is also known as m-calpain because of its millimolar calcium binding affinity in vitro. Since intracellular calcium concentration rarely exceeds 1 µM (resting state below 100 nM, activated state below 1 µM, maximal synaptic concentration 50 µMat physiological conditions, it is possible that in vivo calpain-2 is associated with other proteins, which make it more sensitive to calcium. Accordingly, calpain-2 is implicated in a number of pathological states, including neurodegenerative disorders like Alzheimer's disease. Therefore, future studies should address to what extend calpain-2 contributes to the MPS remodeling in different neuronal compartment, what kind of stimuli can trigger calpain-2-dependent degradation of MPS, how high the intracellular calcium concentration should be and how local this process is. It seems unlikely that the axonal RTK signaling will be triggered along extended membrane areas in vivo since even few hours of reduced axonal stability could have detrimental consequences for neuronal function and survival.
## Non-muscle myosins in regulation of the mps
Non-muscle myosin II motor proteins (represented by myosin IIA, myosin IIB, and myosin IIC) have recently attracted strong interest in respect of MPS regulation. Functional myosin II motor protein complex consists of two myosin heavy chains (MHCs), two essential light chains (ELCs) and two regulatory light chains (MLCs, see Box 1). In contrast to myosin V and VI, which are processive myosins involved in cargo trafficking along actin filaments, myosin II family members are contractile myosins assembled into antiparallel bundles that pull actin filaments together. Powered by ATP, they participate in a multitude of neuronal processes including reorganization of actin filaments in the axonal growth cone dynamics, structural plasticity of dendritic spines and the AIS, signaling to RhoGTPases and many others. Structurally, the MHC is subdivided into three distinct regions called motor, neck and tail domains. The globular motor domains transduce the energy released during hydrolysis of ATP into active mechanical force and processive movement along actin filament. The neck region connects the motor domain to the elongated coiled coil that forms the tail domain and has an important regulatory role. The two conserved IQ motifs in the neck domain control NMII switching between a folded (inactive) or open (active) conformation. These IQ motifs can bind calmodulin, but in the case of NMII they are usually occupied by ELC and MLCBinding of the ELC to the neck region provides stability to the neck domain of the MHC whereas MLC binding also allows regulation of motor activity through phosphorylation at two subsequent threonine and serine residues (T18/S19). The tail region coiled-coil assembles two NMII molecules into the constitutive homodimer. Activated myosin II dimers form an elongated shape and such hexameric complexes can bundle in a parallel and antiparallel way into bipolar filaments of approximately 300 nm in length. The myosin II motor domains are flanking both sides of the filament and their stepping motion in the barbed-end direction of oppositely oriented actin filaments results in actomyosin contraction. The structure and regulation of myosin II filaments in muscle and non-muscle cells have been recently reviewed by Dasbiswas et al.. Interestingly, in neurons actomyosin together with the MPS are implicated in regulation of the AIS and axonal diameter and action potential firing.
It has been known already for some time that repeated neuronal depolarization can lead to a reversible repositioning and length change of the AIS, which is crucial for homeostatic control and plasticity of neuronal excitability. This type of AIS structural plasticity occurs within a few hours after stimulation and involves AIS disassembly in a more proximal part resulting in the AIS shortening following more long-lasting extension towards the distal part of the axon. Myosin II was implicated both in rapid disassembly as well as relocation and extension, since the selective non-muscle myosin II inhibitor blebbistatin completely blocked any activity-dependent morphological alterations. In recent work, Berger et al. provided deeper insight into the mechanisms involved in AIS relocation and identified actomyosin as a key element. Myosin light chain (MLC) can activate myosin II when it is mono-or di-phosphorylated at S18 and Ser19 residues. The di-phosphorylated form of the MLC (ppMLC) is highly enriched in the AIS, where it associates with the actin rings of the MPS. MLC dually phosphorylated in this way stimulates myosin II contractility, which in turn is required for AIS assembly and the faithful distribution of AIS components. During neuronal depolarization, ppMLC is rapidly lost, which results in the destabilization of actin filaments, which in turn allows for a remodeling of the AIS. Interestingly, ppMLC was hardly detected in other axonal compartments and dendrites, suggesting that this mechanism of regulation is very specific for the AIS. Myosin light chain phosphorylation and AIS relocation both are calcium dependent, as calcium chelation and the inhibition of L-type calcium channels completely prevented loss of ppMLC and AIS disassembly. One calcium signaling pathway underlying this process seems to involve calpains, as the calpain-1 and -2 inhibitor MDL28170 partially rescued ppMLC levels and prevented AnkG loss. Furthermore, calcineurin is a calcium-and calmodulin-dependent serine/threonine protein phosphatase, which was implicated in the regulation of the AIS relocation upon neuronal depolarization. Indeed, the calcineurin inhibitor CsA reduced the levels of ppMLC and AnkG to varying degrees under resting conditions. This suggests that various calcium signaling cascades converge in the regulation of different steps of the AIS remodeling upon induction of plasticity.
The role of actomyosin in the neuronal cytoskeleton goes beyond regulation of the AIS plasticity. Two very recent studies reported that the axonal MPS is a contractive actomyosin network, which facilitates structural stability of the axon, regulates axon diameter and radial contraction as well as cargo trafficking and axonal propagation of electrical signals. Thin axons with an inner diameter of less than 1 µm are very abundant in the mammalian central nervous system. This poses obvious challenge for the transport of large cargoes, such as autophagosomes, mitochondria, endosomes or lysosomes, as their diameter can exceed the diameter of the axon. Using live super-resolution imaging in combination with pharmacological approaches and electron microscopy, it could be shown that the passage of large axonal cargoes causes a transient radial expansion of the axon followed by constriction which depended on myosin II activity. A short-term inhibition of myosin II activity did not change the periodicity of actin rings, but led to an increase in axon diameter instead, suggesting that myosin II activity keeps the MPS under constant tension to control ring size. SIM imaging revealed that the motor domain of myosin II but not the tail-domain involved in dimerization colocalized with the periodic, braided actin rings.
Work from the Sousa laboratory provided further molecular insights on the regulation of myosin II contractility along the axon. Similar to the AIS, they found that phosphorylation of the MLC plays a critical role in the contractility of the axonal actomyosin network. Of note, they used pharmacological inhibition of MLC phosphorylation and antibodies detecting di-phosphorylated S18/19 MLC, which cannot distinguish between either mono-or the di-phosphorylated forms of MLC. It remains therefore unclear, whether different states of phosphorylation lead to different outcomes. What is clear is that MLC phorphorylation by myosin light chain kinase triggered conformational changes and self-assembly of myosin II complexes in filaments, leading to a constriction of the axonal diameter. It was found that changes in axonal width did not affect the angle of actin rings in relation to the axonal axis. This suggests that that myosin II probably does not contract between adjacent rings.
Another example of narrowed compartments are dendritic spines where the thin spine neck also contains an MPS. NMII has been found in dendritic spines, and is found in about 90% of dendritic filopodia. It localizes mostly in the neck and head region. In the neck region, myosin II molecules were found to be organized in linear clusters, suggesting the assembly of myosin bundles, whereas in the head, single molecules were found in immunogold EM. NMII is required for shortening of the spine stalk in spine maturation and spine developmentin a process that involves RLC phosphorylation via Rho-kinase (ROCK). NMII activity is required both for NMDAR-dependent LTP and for the LTP-dependent spine actin polymerization required for structural plasticity of dendritic spinesas evident from knockout animals. These effects are dependent of MLC phosphorylation, which is controlled by NMDAR signaling. Indeed, myosin is required for memory formation and consolidation. For a detailed review of synaptic myosins, see. Unlike in the AIS, a role for NMII in the organization of the MPS at the neck of dendritic spines has not been yet investigated.
## Perspectives
Since the initial discovery of the periodic membrane cytoskeleton in axons, the molecular composition of the MPS, its distribution over neuronal compartments, its appearance during developmentas well as its presence across different cell types cells of the nervous systemand its evolutionary conservation from worm to mammalhas been described in quick succession. In a study using the nematode Caenorhabditis elegans as a model, it has been shown that the MPS plays an essential role during tissue movement: it provides mechanical support and elasticity to the axon. However, only very recently the regulation of MPS assembly and disassembly and its cellular functions have begun to emerge. In this respect the degradation of the MPS via calcium/calpain-2 is an efficient and elegant mechanism how the neuron could locally reorganize its membrane cytoskeleton and terminate RTK signaling in axons. Fundamental questions remain unanswered. It remains unclear whether this is an axon-specific pathway or whether the MPS in other neuronal compartments, such as the AIS, dendrites or the neck of dendritic spines, may undergo a similar type of regulation. A possible tuning of MPS sensitivity to degradation could stem from differences in susceptibility to calpain-cleavage of different β-spectrin isoforms. The availability of calpain-2 and the possible contribution of other calpain proteases could be additional factors in the local modification of MPS degradation. It is still an open question how the MPS can recover. It will be important to understand what proteins and signaling pathways can catalyze its assembly and what forces can bring spectrin tetramers into the energetically unfavorable extended arrangementin order to bridge neighboring actin rings.
A new exciting aspect of the biology of the MPS is its inclusion into a contractile actomyosin network. This recent finding offers an explanation as to how narrow cellular processes could dilate to allow the passage of large cargo or could scale their diameter in response to neuronal activity. The exact spatial relationship between the actin rings and the position of myosin II motor complexes remains unclear. Taking into account that the length of an active two-headed myosin motor complex is around 300 nm and that the distance between actin rings is 190 nm, actin rings may connect to myosin head groups in different ways. First, the myosin II bundle could crosslink two neighboring rings with an angle deviant from 90 - . This could be achieved if myosin motors connected to neighboring actin rings in a one-dimensional lattice (similar to spectrin) or across the axonal volume. In such a configuration, the force generated by myosin II would induce filament sliding within the braid that would results in constriction or expansion of the ring. However, as myosin II steps towards the barbed end, the orientation of filaments within one braid and filament polarity with respect to the neighboring rings will be of critical importance. Such a mechanism can work if (i) two filaments within the same braid have an opposite polarity, (ii) when filaments assembling the same braid are parallel but the adjacent braids have opposite polarity, or (iii) when the rings are composed of not one but several non-uniformly oriented braids and there is at least one filament present in the opposite orientation in two neighboring rings. In an alternative scenario, myosin bundles could cross inside a ring in a radial manner. For the myosin bundle to change ring diameter, actin filaments with opposite polarity would be required to allow for ring extension and contraction. A sarcomere-like organization of neuronal MPS is also unlikely because this would imply that multiple myosins are organized in a filament with motor domains pointing outside. The diameter of such filament is approximately 30 nm, which is much less than the spacing between actin rings. Of note, a replica EM also did not reveal such myosin structures in the axon. When considering the radial myosin organization model, it should be kept in mind that the cytosolic content of the axon, especially microtubules, intermediate filaments and the endoplasmic reticulum, can become a physical obstacle that restricts a number of myosin II complexes linked to the MPS. Cryo-electron microcopy could help in addressing some of the questions. For instance, what is the orientation of individual actin filaments in a braid; are all braids built uniformly and how are they made? Perhaps it would be possible to visualize the myosin II complex together with the MPS. Understanding the mechanisms of braid assembly will shed further light on their regulation and resistance to depolymerization. One can speculate that in the analogy with microtubules, actin braids can be made by the capability of myosin II to cross-link and slide the actin filaments. This could cause helical motion of overlapping filaments around each other and thus allow formation of actin braids. Arrangement of filaments into a helix could influence a pitch angle of the individual filaments and thus change their affinity for cofilin binding, thus making them more resistant to a severing. As incorporation of different actin isoforms and the post-translational modifications can also change the stability of actin filaments, it remains to be tested whether some of those are particularly enriched at the MPS. Furthermore, actin braids are a very unusual type of filament organization. It will be very interesting to reproduce and explore biophysical properties of such filaments in in vitro reconstitutes systems. leads to proximal shortening of the AIS and subsequent extension towards the distal axon. NMII activity is necessary for this process as blebbistatin completely blocks the activity. (a, bottom) NMII activity controls the axonal diameter. Inhibition of NMII by blebbistatin leads to an increase in axon diameter, indicating that NMII holds the membrane-associated cytoskeleton (MSK) under constant tension. In addition, NMII has been shown to be implicated in cargo trafficking along the AIS. The size of large cargo (e.g., autophagosomes, mitochondria, endosomes or lysosomes) can exceed the diameter of the axon. Passage of this large axonal cargo causes a transient radial expansion of the axon followed by constriction, which depend on myosin II activity. (b, top) Schematic organization of NMII. NMII exists as a hexamer that consists of two copies each of elongated heavy chains, two regulatory light chains (RLC) and two essential light chains (ELC) that stabilize the heavy chain structure. The heavy chain is composed of an N-terminal motor domain, a neck domain, which interacts with both light chains, an α-helical rod domain and a C-terminal tail. (b, middle) The hexameric units further bundle both in a parallel and antiparallel manner into bipolar structures that can pull actin filaments together. (b, bottom) Crystal structure of the motor and neck domains of NMII interacting with ELC and RLC. The motor domain contains the actin binding cleft where NMII interacts with actin. Shown in red is ADP bound at the nucleotide binding site. Cycling from ATP to ADP at the nucleotide binding site leads to conformational changes in the actin binding cleft, which modulate the interaction of NMII with actin. (c) Models of a possible spatial relationships between NMII and actin rings. The length of an active two-headed myosin motor complex is around 300 nm, while the distance between actin rings is only 190 nm. (c, left) NMII crosses the diameter of a single actin ring. Alternatively, NMII could cross link two neighboring rings with an angle deviant from 90 - , which can be achieved when myosin motors connect neighboring actin rings in a one-dimensional lattice (as spectrin) (c, middle) or across the axonal volume (c, right). (d) Polarity of actin filaments in ring-forming actin braids. Force generated by NMII induces filament sliding within the braid that results in constriction or expansion of the ring. However, as myosin II steps towards the barbed end of an actin filament, the orientation of filaments within a single braid and filament polarity with respect to the neighboring rings are important. The NMII mechanism can work when filaments within a single ring are parallel (d, left) but the adjacent braids have opposite polarity (d, middle) or when two filaments within the same braid have an opposite polarity (c, right).
Actomyosin and organelle trafficking in narrow compartments, such as axons of the neck of dendritic spines, is another very interesting area of research that has recently emerged. Future studies using improved live super-resolution microscopy techniques could highlight the spatio-temporal kinetics of actomyosin response. Research in this direction is warranted as it might be applied to improving transport properties in axons, which can be relevant for aggregation clearance, which in turn might be beneficial in a number of neurodegenerative disorders, such as Parkinson's disease or Alzheimer's disease.
|
Hyperviscosity syndrome in COVID-19 and related vaccines: exploring of uncertainties
Hyperviscosity syndrome (HVS) recently emerged as a complication of coronavirus disease 2019 and COVID-19 vaccines. Therefore, the objectives of this critical review are to establish the association between COVID-19 and COVID-19 vaccines with the development of HVS. HVS may develop in various viral infections due to impairment of humoral and cellular immunity with elevation of immunoglobulins. COVID-19 can increase blood viscosity (BV) through modulation of fibrinogen, albumin, lipoproteins, and red blood cell (RBC) indices. HVS can cause cardiovascular and neurological complications in COVID-19 like myocardial infarction (MI) and stroke. HVS with or without abnormal RBCs function in COVID-19 participates in the reduction of tissue oxygenation with the development of cardio-metabolic complications and long COVID-19. Besides, HVS may develop in vaccine recipients with previous COVID-19 due to higher underlying Ig concentrations and rarely without previous COVID-19. Similarly, patients with metabolic syndrome are at the highest risk for propagation of HVS after COVID-19 vaccination. In conclusion, COVID-19 and related vaccines are linked with the development of HVS, mainly in patients with previous COVID-19 and underlying metabolic derangements. The possible mechanism of HVS in COVID-19 and related vaccines is increasing levels of fibrinogen and immunoglobulins. However, dehydration, oxidative stress, and inflammatory reactions are regarded as additional contributing factors in the pathogenesis of HVS in COVID-19. However, this critical review cannot determine the final causal relationship between COVID-19 and related vaccines and the development of HVS. Prospective and retrospective studies are warranted in this field.
# Introduction
Hyperviscosity syndrome (HVS) is a group of symptoms induced by high blood viscosity (BV) including bleeding, headache, visual disturbances, seizure, vertigo, and coma. HVS is characterized by the triad of mucosal bleeding, visual changes, and neurological deficits. The main cause of HVS is Waldenstrom macroglobulinemia, which is an abnormal proliferation of plasma cells and lymphoplasmacytoid cells [bib_ref] Acalabrutinib monotherapy in patients with Waldenströmmacroglobulinemia: a single-arm, multicentre, phase 2 study, Owen [/bib_ref]. HVS is also caused by polycythemia, leukemia, multiple myeloma, sepsis, and sickle cell anemia [bib_ref] Evidence-based focused review of management of hyperviscosity syndrome, Stone [/bib_ref]. As a result, HVS is caused by an increase in the number of red blood cells (RBCs) or a deformity in RBC shape, as well as an increase in serum proteins [bib_ref] Evidence-based focused review of management of hyperviscosity syndrome, Stone [/bib_ref]. Normal BV is typically between 1.4 and 1.8 centipoise (cp), and symptoms of HVS develop when BV exceeds 4.0 cp.
HVS recently emerged as a complication of coronavirus disease 2019 (COVID- [bib_ref] A rare cause of cardiac ischemia: systemic lupus erythematosus presenting as the..., Corriganiii [/bib_ref] and COVID-19 vaccines [bib_ref] Expected viscosityafter COVID-19 vaccination, hyperviscosity and previous COVID-19, Joob [/bib_ref]. Therefore, the objectives of this critical review are to determine the association between COVID-19 and/or COVID-19 vaccines with the development of HVS.
## Hyperviscosity syndrome and viral infections
HVS can occur in a variety of viral infections, including human immunodeficiency virus type 1 (HIV-1) infections, as a result of impaired humoral and cellular immunity and an increase in immunoglobulin (IgG) [bib_ref] Hyperviscosity syndrome in an HIV-1-positive patient, Garderet [/bib_ref]. The underlying mechanisms of HVS in patients with HIV-1 are related to the direct activation of B cells by HIV-1, alteration of T cellsmediated B cell regulation, chronic exposure to the antigens of HIV-1 and high IL-6 [bib_ref] Hyperviscosity syndrome in an HIV-1-positive patient, Garderet [/bib_ref]. However, hyper-activation of B cells with high production of IgG could be the main mechanistic pathway of HVS in HIV-1 infection [bib_ref] Mechanisms of hypergammaglobulinemia and impaired antigen-specific humoral immunity in HIV-1 infection, De Milito [/bib_ref]. According to Jin et al., HVS was linked to the formation of myelomaassociated IgG1paraprotein against HIV-1 p24 antigen in HIV-1 patients [bib_ref] Hyperviscosity syndrome secondary to a myeloma-associated IgG1κparaprotein strongly reactive against the HIV-1..., Jin [/bib_ref].
As well, HVS has been demonstrated in patients with acute respiratory viral infections, including influenza complicated by pneumonia [bib_ref] Microcirculation and hemostasis in influenza and acute viral respiratory infections complicated with..., Bogomolov [/bib_ref]. A study involving 232 patients with influenza and acute respiratory viral infections showed significant alterations in the microcirculation, intravascular homeostasis, hypercoagulation, augmentation of fibrinolytic activity, and an increase in BV [bib_ref] Microcirculation and hemostasis in influenza and acute viral respiratory infections complicated with..., Bogomolov [/bib_ref]. Generally, Sloop and colleagues revealed that severe infections increase BV with the development of HVS due to inflammation-induced hypergammaglobulinemia and elevation of acute-phase reactants that increase BV [bib_ref] The role of blood viscosity in infectious diseases, Sloop [/bib_ref]. High BV or HVS fosters aggregation of RBCs with an increasing risk of thrombosis due to augmentation of vascular resistance, which impedes peripheral tissue perfusion [bib_ref] The role of blood viscosity in infectious diseases, Sloop [/bib_ref]. Of note, previous acute infection and chronic bronchitis within two months caused by influenza infection predispose to the risk of acute ischemic stroke, and influenza vaccine did not offered a protection against the development of acute ischemic stroke [bib_ref] Chronic bronchitis and acute infections as new risk factors for ischemic stroke..., Piñol-Ripoll [/bib_ref]. This observation suggests that HVS could be a possible risk factor for the development of acute ischemic stroke in patients with influenza infection.
Furthermore, indices of blood viscosity are increased in patients with hepatitis B virus (HBV) infection who are at risk for the development of HVS. A study of 55 patients with HBV infection illustrated that RBCs aggregation index, hematocrit, and whole BV were higher compared with control groups and unrelated to the state of oxidative stress and hemorheology indices. Of interest, HVS is implicated in the pathogenesis of septic shock in parallel with high fibrinogen levels. Van et al. reported that soluble fibrinogenlike protein 2 (sFGL2) is increased in patients with HBV infection [bib_ref] Soluble fibrinogen-like protein 2 levels in patients with hepatitis B virus-related liver..., Van Tong [/bib_ref]. Therefore, high sFGL2 plasma levels could be the potential cause of HVS in HBV infection. These findings indicated that HVS may be developed in various viral infections and contribute to the development of complications.
## Hyperviscosity syndrome and immunoinflammatory disorders
It has been reported that HVS is linked with acute inflammatory disorders since BV is sensitive to acute-phase reactants [bib_ref] Whole blood viscosity assessment issues IV: Prevalence in acute phase inflammation, Nwose [/bib_ref]. Therefore, HVS is high in subpopulations with high C-reactive protein and erythrocyte sedimentation rate (ESR) as compared with subpopulations with low CRP and ESR [bib_ref] Whole blood viscosity assessment issues IV: Prevalence in acute phase inflammation, Nwose [/bib_ref]. HVS has been shown to develop in patients with rheumatoid arthritis due to the formation of immunocomplexes which affect RBCs deformability and vascular resistance [bib_ref] Whole blood viscosity assessment issues IV: Prevalence in acute phase inflammation, Nwose [/bib_ref]. HVS in rheumatoid arthritis can be developed with a level of IgG less than in Waldenstrom macroglobulinemia [bib_ref] Hyperviscosity syndrome in rheumatoid arthritis, Miller [/bib_ref]. It developed in patients with rheumatoid arthritis due to the formation of an intermediate complex from the aggregation of Ig, RBCs aggregation, and high fibrinogen levels [bib_ref] Serum IgM level as predictor of symptomatic hyperviscosity in patients with Waldenströmmacroglobulinaemia, Gustine [/bib_ref]. However, HVS in rheumatoid arthritis is rare in treated patients, so treating with plasmapheresis and immunosuppressive agents can reduce the risk of development of HVS [bib_ref] Serum IgM level as predictor of symptomatic hyperviscosity in patients with Waldenströmmacroglobulinaemia, Gustine [/bib_ref]. Likewise, HVS in rheumatoid arthritis is significantly correlated with high activity of rheumatoid factor [bib_ref] Therapeutic plasma exchange for hyperviscosity syndrome secondary to high rheumatoid factor, Lokhandwala [/bib_ref].
Furthermore, HVS may be the presenting feature in patients with systemic lupus erythematosus (SLE) due to hyper-paraproteinemia and monoclonal gammopathy [bib_ref] A rare cause of cardiac ischemia: systemic lupus erythematosus presenting as the..., Corriganiii [/bib_ref]. Besides, HVS is also developed in IgG4-related disorders, which are systemic fibro-inflammatory disorders characterized by elevation of Ig, including IgG4 [bib_ref] Polyclonal hyperviscosity syndrome in IgG4-related disease and associated conditions, Chen [/bib_ref].
Of interest, CD169 macrophages contribute to the process of bone marrow erythropoiesis and maturation of RBCs. Over-activation of CD169 macrophages may be associated with the development of polycythemia [bib_ref] CD169+ macrophages provide a niche promoting erythropoiesis under homeostasis and stress, Chow [/bib_ref]. Thus, depletion of CD169 macrophages reduces bone marrow erythroblasts and prevents erythropoietic recovery from anemia [bib_ref] CD169+ macrophages provide a niche promoting erythropoiesis under homeostasis and stress, Chow [/bib_ref]. According to Asano et al., CD169 macrophages control and modulate immunological responses in the circulating fluid by recruiting monocytes and producing chemokines [bib_ref] CD169 macrophages regulate immune responses toward particulate materials in the circulating fluid, Asano [/bib_ref]. CD169 macrophages are activated during immunological disorders, tumor growth, and viral infections to produce immunological tolerance and antiviral effects [bib_ref] Functions of CD169 positive macrophages in human diseases (Review), Liu [/bib_ref]. As a result, in immunological diseases, activated CD169 macrophages may increase BV via boosting erythropoiesis.
Indeed, there is a close relationship between HVS and inflammation due to the increase in acute-phase reactant fibrinogen, whose level is correlated with increasing blood viscosity [bib_ref] Blood electrical impedance closely matches whole blood viscosity as parameter of hemorheology..., Pop [/bib_ref]. Gordy et al. revealed that fibrinogen-related proteins are increased during the immune response to various inflammatory stimuli [bib_ref] The role of fibrinogen-related proteins in the gastropod immune response, Gordy [/bib_ref]. Fibrinogen and fibrinogenrelated proteins play a critical role in neutralizing invading pathogens [bib_ref] Antimicrobial activity of fibrinogen and fibrinogenderived peptides-a novel link between coagulation and..., Påhlman [/bib_ref]. In turn, exaggerated immune responses and high levels of fibrinogen-related proteins are associated with the development of HVS.
These observations indicated that high BV or HVS is linked with underlying immunoinflammatory disorders.
## Hyperviscosity syndrome and covid-19
## Effects of covid-19 on blood viscosity
COVID-19 is a pandemic disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) leading to worldwide crisis with high morbidity [bib_ref] The potential role of neopterin in Covid-19: a new perspective, Al-Kuraishy [/bib_ref]. Till late January 2022, the total number of infected cases reached more than 370 million, with about 5 million confirmed deaths. In general, the clinical presentation of COVID-19 is mild in the majority of cases, though 15% of COVID-19 patients presented with pulmonary and extra-pulmonary manifestations including headache, fever, dry cough, sweating, and fatigue [bib_ref] Niclosamide for Covid-19: bridging the gap, Al-Kuraishy [/bib_ref] [bib_ref] Role of leukotriene pathway and montelukast in pulmonary and extrapulmonary manifestations of..., Al-Kuraishy [/bib_ref]. About 5% of COVID-19 patients may develop severe and critical outcomes due to the development and propagation of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) that require intensive care unit admission [bib_ref] Levamisoletherapy in COVID-19, Al-Kuraishy [/bib_ref].
SARS-CoV-2 exploits different receptor types to enter the affected cells. The angiotensin-converting enzyme 2 (ACE2) is a pioneer one related in the pathogenesis of SARS-CoV-2 infection [bib_ref] Pleiotropic effects of tetracyclines in the management of COVID-19: emergingperspectives, Al-Kuraishy [/bib_ref]. This interaction induces downregulation of ACE2, which is necessary for conversion of pro-inflammatory/vasoconstrictor angiotensin II (AngII) to vasodilator/anti-inflammatory Ang1-7 [bib_ref] Possible mechanistic insights into iron homeostasis role of the action of 4-aminoquinolines..., Batiha [/bib_ref]. In severe cases, SARS-CoV-2 infection may result in an exaggerated immune response, hyperinflammation, and hypercytokinemia, as well as a cytokine storm [bib_ref] The possible role of ursolic acid in Covid-19: a real game changer, Al-Kuraishy [/bib_ref] [bib_ref] The looming effects of estrogen in Covid-19: a rocky rollout, Al-Kuraishy [/bib_ref]. Therefore, SARS-CoV-2-induced upregulation of AngII may provoke the development of HVS in COVID-19 through the induction of inflammatory changes and vasoconstriction.
It has been shown that SARS-CoV-2 infection is associated with microcirculation failure in hospitalized patients with severe COVID-19 characterized by weak peripheral pulses, cold extremities, and metabolic acidosis [bib_ref] Impact of COVID-19 on red blood cell rheology, Renoux [/bib_ref]. Microcirculation dysfunction/failure has been linked to severe sepsis because of increased RBC aggregation, decreased RBC deformability, and alterations in RBC physiology/morphology [bib_ref] Microcirculatory dysfunction in sepsis: A pathogenetic basis for therapy?, Lehr [/bib_ref] [bib_ref] Microcirculatory dysfunction in sepsis, Lundy [/bib_ref]. Endothelial dysfunction, coagulation problems, and cytokine storm were observed to contribute to the development of microcirculation failure in septic COVID-19 patients by Colantuoni et al. [bib_ref] COVID-19 sepsis and microcirculation dysfunction, Colantuoni [/bib_ref]. A study that included 7 hospitalized COVID-19 patients compared to 7 non-COVID-19 septic patients and 7 healthy control illustrated that RBCs deformability was reduced in both COVID-19 patients and non-COVID-19 septic patients compared to the controls (P < 0.05) [bib_ref] Blood viscosity of COVID-19 patient: a preliminary report, Joob [/bib_ref]. Moreover, RBCs aggregation was higher in COVID-19 patients compared to 7 non-COVID-19 septic patients without significant changes in BV and fibrinogen levels [bib_ref] Impact of COVID-19 on red blood cell rheology, Renoux [/bib_ref]. This small sample study does not give any concrete clues about normal BV and fibrinogen levels in COVID-19. A retrospective study involving 41 COVID-19 patients revealed that estimated BV was higher in COVID-19 patients than in the control group [bib_ref] Blood viscosity of COVID-19 patient: a preliminary report, Joob [/bib_ref].
Enhanced RBCs aggregation with reduction of RBCs deformability in COVID-19 is increased in both stasis and low-shear flow [bib_ref] Impact of COVID-19 on red blood cell rheology, Renoux [/bib_ref] that together with increasing fibrinogen level may increase BV and development of HVS. Of note, acute viral infections are linked with development of HVS due to hypergammaglobulinemia and elevation of acute-phase reactants which might cause thromboembolic disorders and cardiovascular complications [bib_ref] Is chronic HIV infection associated with venous thrombotic disease? A systematic review, Klein [/bib_ref]. Increasing of BV and development of HVS in COVID-19 could be related to different mechanisms including exaggerated immune response, endothelial dysfunction, hypoxia, coagulation disorders [bib_ref] Blood viscosity of COVID-19 patient: a preliminary report, Joob [/bib_ref].
Similarly, changes in RBCs morphology/function, platelet hyper-reactivity, high ferritin, and P-selectin activity in COVID-19 could contribute in the development of HVS [bib_ref] Erythrocyte, platelet, serum ferritin, and p-selectin pathophysiology implicated in severe hypercoagulation and..., Venter [/bib_ref]. As well, psychological stress, fever, and dehydration may increase BV and compensatory increment in the release of arginine vasopressin in COVID-19 patients [bib_ref] Arginine vasopressin and pathophysiology of COVID-19: an innovative perspective, Al-Kuraishy [/bib_ref]. High arginine vasopressin triggers release of pro-inflammatory cytokines through activation of nuclear factor kappa B (NF-κB) and nod-like receptor pyrin 3 (NLRP3) inflammasome which participate in increasing of BV [bib_ref] Arginine vasopressin and pathophysiology of COVID-19: an innovative perspective, Al-Kuraishy [/bib_ref]. Both of NF-κB and NLRP3 inflammasome induce asymmetry of RBCs membrane with reduction of RBCs deformability in normal and sickle RBCs [bib_ref] Stimulation of calcium influx and CK1α by NF-κB antagonist [6]-gingerol reprograms red..., Alamri [/bib_ref] [bib_ref] The red blood cell-inflammation vicious circle in sickle cell disease, Nader [/bib_ref]. In addition, NF-κB and NLRP3 inflammasome are highly activated in COVID-19and could a potential causes for reduction of RBCs deformability in COVID-19.
Of note, CD169 macrophages are involved in the maturation of RBCs and development of polycythemia [bib_ref] CD169+ macrophages provide a niche promoting erythropoiesis under homeostasis and stress, Chow [/bib_ref]. CD169 monocytes are expressed in 93.7% of COVID-19 patients and could be of diagnostic benefits [bib_ref] Monocyte CD169 expression as a biomarker in the early diagnosis of coronavirus..., Bedin [/bib_ref]. Therefore, SARS-CoV-2-induced CD169 macrophages/monocytes may cause polycythemia and elevation of BV in COVID-19.
Exaggerated immune response and release of pro-inflammatory cytokines mainly IL-6 are linked with development of cytokine storm and multi-organ injury [bib_ref] Neutrophil extracellular traps (NETs) and Covid-19: a new frontiers for therapeutic modality, Al-Kuraishy [/bib_ref]. Panigada et al. observed that IL-6 is regarded as a powerful activator for synthesis of fibrinogen in COVID-19 [bib_ref] Hypercoagulability of COVID-19 patients in intensive care unit: a report of thromboelastography..., Panigada [/bib_ref]. Also, dysregulation of RAS with high AngII in COVID-19 may induce expression and synthesis of fibrinogen [bib_ref] The possible role of ursolic acid in Covid-19: a real game changer, Al-Kuraishy [/bib_ref] [bib_ref] Efficacy of serumangiotensin IIlevels in prognosis of patients with coronavirus Disease, Ozkan [/bib_ref]. Fibrinogen activates RBC membrane integrinαvβ3 receptors resulting in the activation of RBCs aggregation with subsequent development of HVS [bib_ref] Hypercoagulability of COVID-19 patients in intensive care unit: a report of thromboelastography..., Panigada [/bib_ref]. As well, hypoalbuminemia is linked with increasing of blood viscosity and development of HVS [bib_ref] Effects of whole blood viscosity and plasma NOx on cardiac function and..., Buyan [/bib_ref]. Serum albumin is inversely correlated with D-dimer and CRP, and hypoalbuminemia is associated with an increased risk of development of coagulopathy in COVID-19 patients by a reduction in the anticoagulant and antiplatelet effects of albumin [bib_ref] Hypoalbuminemia, coagulopathy, and vascular disease in COVID-19, Violi [/bib_ref]. A retrospective study involving 113 COVID-19 patients illustrated that a high fibrinogen/albumin ratio was associated with a high risk of thrombosis, disease severity, and poor clinical outcomes [bib_ref] Prediction of severe illness due to COVID-19 based on an analysis of..., Bi [/bib_ref]. As a result, the BV is augmented and reaches 4.2 cp. Therefore, hyperfibrinogenemia and hypoalbuminemia may increase BV and participate in the development of HVS and thrombotic events in COVID-19. Notably, severe COVID-19 is associated with the development of arterial and venous thromboembolic events due to direct SARS-CoV-2 cytopathic effects and associated endothelial dysfunction, platelet activation, coagulation activation, and inhibition of the fibrinolytic pathway [bib_ref] COVID-19 and thrombosis: from bench to bedside, Ali [/bib_ref]. Furthermore, downregulation of ACE2 with dysregulation of RAS together with exaggerated pro-inflammatory cytokines may initiate endothelial dysfunction by reduction of prostacyclin and nitrous oxide (NO) [bib_ref] Endothelial cell infection and endotheliitis in COVID-19, Varga [/bib_ref]. Felicetti et al. recently illustrated that thrombotic events may increase the risk of development of HVS [bib_ref] A molecular communications system for live detection of hyperviscosity syndrome, Felicetti [/bib_ref]. These observations suggest a mutual interaction between HVS and thrombotic events in COVID-19.
Moreover, SARS-CoV-2 may directly affect RBCs morphology through binding of membrane CD147 receptor and Band3 protein on the RBCs [bib_ref] CD147 as a target for COVID-19 treatment: suggested effects of azithromycin and..., Ulrich [/bib_ref] [bib_ref] RRM prediction of erythrocyte Band3 protein as alternative receptor for SARS-CoV-2 virus, Cosic [/bib_ref]. These changes hamper functional capacity for oxygen transport by RBCs leading to development of tissue hypoxia [bib_ref] RRM prediction of erythrocyte Band3 protein as alternative receptor for SARS-CoV-2 virus, Cosic [/bib_ref]. Besides, Foy and colleagues revealed that RBC distribution width and other indices were severely affected in SARS-CoV-2 infection and linked with COVID-19 severity and poor clinical outcomes [bib_ref] Association of red blood cell distribution width with mortality risk in hospitalized..., Foy [/bib_ref]. In addition, extreme hypoxia and acidosis induce alteration in RBCs morphology [bib_ref] The effect of flunarizine on erythrocyte suspension viscosity under conditions of extreme..., Kavanagh [/bib_ref]. These observations suggest that direct SARS-CoV-2-induced RBCs dysmorphology and associated metabolic acidosis and hypoxia may induce progression of HVS in COVID-19.
Furthermore, lipoproteins can affect blood viscosity since low-density lipoprotein (LDL) is positively correlated, while high-density lipoprotein (HDL) is negatively correlated with BV [bib_ref] The effects of low-density lipoprotein and high-density lipoprotein on blood viscosity correlate..., Sloop [/bib_ref]. HDL is necessary for RBCs morphology and deformability; thus, reduction of HDL may reduce RBCs life span by increasing osmotic fragility and reduction of RBCs deformability [bib_ref] HDL cholesterol levels are an important factor for determining the lifespan of..., Meurs [/bib_ref]. In COVID-19, there is a noteworthy alteration in lipoprotein serum levels, and low HDL levels are associated with COVID-19 severity [bib_ref] Lipoprotein concentrations over time in the intensive care unit COVID-19 patients: results..., Tanaka [/bib_ref] [bib_ref] Declined serum high density lipoprotein cholesterol is associated with the severity of..., Hu [/bib_ref]. Therefore, reduction of HDL in SARS-CoV-2 infection can increase BV with the development of HVS in COVID-19.
Moreover, SARS-CoV-2 infection-induced oxidative stress may trigger elevation of BV [bib_ref] SARS-CoV-2 mediated endothelial dysfunction: the potential role of chronic oxidative stress, Chang [/bib_ref]. It has been reported that high oxidative stress can induce abnormal hemorheological changes with a reduction of RBCs deformability and the induction of thrombotic changes [bib_ref] Erythrocyte oxidative stress is associated with cell deformability in patients with retinal..., Becatti [/bib_ref]. In COVID-19, severe oxidative stress triggers endothelial dysfunction and thromboembolic complications [bib_ref] SARS-CoV-2 mediated endothelial dysfunction: the potential role of chronic oxidative stress, Chang [/bib_ref]. Thus, alterations in RBC fragility and deformability together with endothelial dysfunction by SARS-CoV-2 infection-induced oxidative stress could cause HVS in COVID-19.
Interestingly, RBCs morphology and functions are affected in COVID-19 with the development of abnormal erythrocrine function [bib_ref] Silent hypoxia: higherNO in red blood cells of COVID-19 patients, Mortaz [/bib_ref]. Development of abnormal RBCs may contribute to the progression of endothelial dysfunction and vascular injury by increasing oxidative stress [bib_ref] Erythrocytes induce vascular dysfunction in COVID-19, Mahdi [/bib_ref]. RBCs from COVID-19 patients induce expression and upregulation of endothelial arginase with the production of reactive oxygen species (ROS), reduction of endothelial NO and development of endothelial dysfunction [bib_ref] Erythrocytes induce vascular dysfunction in COVID-19, Mahdi [/bib_ref]. Therefore, SARS-CoV-2 infection-induced oxidative stress could in part be mediated by the development of abnormal RBCs in COVID-19.
Moreover, COVID-19 is commonly associated with psychological stress and sympathetic outflow [bib_ref] Psychological health, sleep quality, and coping styles to stress facing the COVID-19..., Fu [/bib_ref]. Of interest, psychological stress increases circulating AngII as well, AngII promotes psychological stress through augmentation of sympathetic activation [bib_ref] Angiotensin II receptor blocker attenuates stress pressor response in young adult African..., Jeong [/bib_ref]. Likewise, AngII receptor blockers attenuate stress pressor in young adults [bib_ref] Angiotensin II receptor blocker attenuates stress pressor response in young adult African..., Jeong [/bib_ref]. Thus, COVID-19-induced psychological stress may augment the dysregulated RAS by increasing AngII with subsequent development of HVS.
Taken together, COVID-19 can increase BV through modulation of fibrinogen, albumin, lipoproteins, and RBC indices .
## Complications of hyperviscosity in covid-19
COVID-19 HVS has been linked with various cardiovascular and neurological complications such as myocardial infarction (MI) and stroke [bib_ref] Relationship between apical thrombus formation and blood viscosity in acute anterior myocardial..., Tak [/bib_ref] [bib_ref] Effect of warfarin versus aspirin on blood viscosity in cardioembolic stroke with..., Lee [/bib_ref]. The incidence of MI in COVID-19 has increased by up to 5% [bib_ref] Prevalence and impact of myocardial injury in patients hospitalized with COVID-19 infection, Lala [/bib_ref]. That could be due to the development of HVS. In addition, increasing of RBCs aggregation and SARS-CoV-2 infection-induced endothelial dysfunction and immunothrombosis may elevate BV in COVID-19 [bib_ref] Therapeutic plasma exchange for COVID-19-associated hyperviscosity, Truong [/bib_ref]. These changes increase the risk of the development of MI in surviving COVID-19 patients due to the development of coronary microangiopathy [bib_ref] COVID-19, microangiopathy, hemostatic activation, and complement, Song [/bib_ref].
HVS in COVID-19 causes poor tissue perfusion, peripheral vascular resistance, and thrombosis [bib_ref] The detrimental role of elevated blood viscosity in patients with COVID-19, Sloop [/bib_ref]. Lowshear areas are susceptible to thrombosis due to reduction in dispersion of clotting factors and attenuation of shearinduced release of antithrombotic molecules like NO and prostacyclin [bib_ref] The detrimental role of elevated blood viscosity in patients with COVID-19, Sloop [/bib_ref]. Remarkably, most of the COVID-19 patients with BV greater than 3.5cp had coagulation disorders [bib_ref] COVID-19-associated hyperviscosity: A link between inflammation and thrombophilia?, Maier [/bib_ref]. Herein, there is a close relationship between HVS and thrombotic events in COVID-19. Maier and coworkers reported 15 critical COVID-19 with thrombotic complications. All patients had a BV greater than 3.5cp (the normal range is 1.4-1.8 cp) as tested by a traditional capillary viscometer. The high BV was correlated with thrombotic events (r = 0.84, P < 0.01) [bib_ref] COVID-19-associated hyperviscosity: A link between inflammation and thrombophilia?, Maier [/bib_ref]. Further, a case series reported by Truong et al. showed symptoms of HVS were more evident in COVID-19 patients with BV greater than 4.2 cp [bib_ref] Therapeutic plasma exchange for COVID-19-associated hyperviscosity, Truong [/bib_ref]. These findings suggest that higher BV is linked with more severe HVS in COVID-19.
Furthermore, HVS may lead to complications like acute kidney injury, glucose intolerance, skeletal muscle ischemia, and myocardial necrosis. As well, HVS leads to pulmonary hypoperfusion and the development of ventilation-perfusion mismatch. These changes cause silent hypoxemia with the propagation of high pulmonary vascular resistance [bib_ref] Pneumolysis and "silent hypoxemia" in COVID-19, Zubieta-Calleja [/bib_ref].
Indeed, HVS is linked with the development of post-COVID-19 syndrome (long COVID-19), which is the persistence of symptoms like dyspnea, fatigue, cognitive dysfunction, and headache following recovery from COVID-19 [bib_ref] Post-COVID 19 Neurological Syndrome (PCNS); a novel syndrome with challenges for the..., Wijeratne [/bib_ref]. Long-term COVID-19 is associated with immunosuppression and cardio-pulmonary fibrosis due to upregulation of transforming growth factor beta (TGF-β) [bib_ref] A review of persistent post-COVID syndrome (PPCS), Oronsky [/bib_ref]. Prolonged inflammatory changes and high blood viscosity in patients with long COVID-19 can reduce tissue perfusion and cellular metabolism [bib_ref] Post COVID-19 syndrome associated with orthostatic cerebral hypoperfusion syndrome, small fiber neuropathy..., Novak [/bib_ref]. As mentioned above, prolonged abnormal RBCs function following COVID-19 may cause tissue hypoxia and subnormal cell metabolism with accentuation of long COVID-19 [bib_ref] Erythrocytes induce vascular dysfunction in COVID-19, Mahdi [/bib_ref].
Taken together, HVS with or without abnormal RBCs function in COVID-19 participates in reduction of tissue oxygenation with the development of cardio-metabolic complications and long COVID-19 .
## Hyperviscosity and covid-19 vaccination
COVID-19 vaccine was developed on the 8th of April 2020 to control the spread of SARS-CoV-2 infection and limit morbidity and mortality caused by COVID-19 [bib_ref] The COVID-19 vaccine development landscape, Thanh [/bib_ref]. Following COVID-19 vaccination, some reports showed that the BV was increased because of the induction of Ig. HVS may develop after COVID-19 vaccination, leading to immunopathological changes [bib_ref] High-dose intravenous immunoglobulin and serum viscosity: risk of precipitating thromboembolic events, Dalakas [/bib_ref]. HVS is correlated with the concentration of Ig, though the lowest normal Ig concentrations are 545 mg/dl, while the lowest BV is 1.5 cp [bib_ref] High-dose intravenous immunoglobulin and serum viscosity: risk of precipitating thromboembolic events, Dalakas [/bib_ref]. The BV will be 2.6 cp when Ig concentrations reach 6160 mg/dl. Of note, symptoms of HVS are developed when BV exceeds 4.0 cp.
Surprisingly, HVS can develop in vaccine recipients who have previously received COVID-19 due to higher underlying Ig concentrations, and only rarely in those who have never received COVID-19. Thus, screening for previous COVID-19 is essential before induction of COVID-19 vaccination to prevent the development of HVS and related hemorheological adverse effects. Alongside, use of contraceptives may augment the risk of development of HVS after COVID-19 vaccination [bib_ref] COVID-19 vaccine, contraceptive, viscosity and safety margin change, Yasri [/bib_ref]. Therefore, we suggest taking the risk into consideration for patients taking contraceptives at the time of COVID-19 vaccination.
Different studies revealed that metabolic alterations in patients with metabolic syndrome increase BV and the risk for development of HVS [bib_ref] Plasma and blood viscosity in metabolic syndrome, Irace [/bib_ref]. Metabolic syndrome Proposed mechanism of hyperviscosity syndrome in COVID-19: SARS-CoV-2 through induction of the downregulation of angiotensinconverting enzyme 2 (ACE2), psychological stress, hyperinflammation, oxidative stress (OS), abnormal morphology of red blood cells (RBCs), and reduction of high density lipoprotein (HDL). These changes increase fibrinogen, angiotensin II (AngII), and the induction of erythrocrine dysfunction with the subsequent development of hyperviscosity syndrome (HSV) is linked with systemic inflammation and oxidative stress which affect the microcirculation by increasing of BV due to reduction of RBCs deformability [bib_ref] Whole-blood viscosity and metabolic syndrome, Gyawali [/bib_ref]. Therefore, patients with metabolic syndrome are at the highest risk for propagation of HVS after COVID-19 vaccination. Joob and Wiwanitkit confirmed that COVID-19 vaccination increases the risk of development of HVS in patients with metabolic syndrome [bib_ref] Change of blood viscosity after COVID-19 vaccination: estimation for persons with underlying..., Joob [/bib_ref]. The BV is increased by 2.7 times in healthy subjects and by 2.99 in patients with metabolic syndrome following COVID-19 vaccination [bib_ref] Microviscometry reveals reduced blood viscosity and altered shear rate and shear stress..., Long [/bib_ref]. This increment in BV did not reach the state of HVS in both healthy subjects and patients with metabolic syndrome, which might be due to the validity of the method in the assessment of blood viscosity [bib_ref] Microviscometry reveals reduced blood viscosity and altered shear rate and shear stress..., Long [/bib_ref].
Generally, BV in healthy COVID-19 vaccine recipients is increased by 2.4 cp [bib_ref] A review of the progress and challenges of developing a vaccine for..., Sharma [/bib_ref]. However, COVID-19 vaccineinduced HVS is common in patients with metabolic syndrome due to high underlying metabolic disorders which increase BV [bib_ref] COVID-19 vaccine, immune thrombotic thrombocytopenia, jaundice, hyperviscosity: concern on cases with underlying..., Sookaromdee [/bib_ref]. proposed that underlying chronic liver diseases with high bilirubin levels may cause HVS after COVID-19 vaccination since hyperbilirubinemia is linked with the development of HVS [bib_ref] COVID-19 vaccine, immune thrombotic thrombocytopenia, jaundice, hyperviscosity: concern on cases with underlying..., Sookaromdee [/bib_ref]. Patients with underlying metabolic disorders have a higher chance of developing HVS following COVID-19 vaccination. Thus, close monitoring of blood viscosity in COVID-19 vaccine recipients is necessary to prevent post-vaccine complications [bib_ref] Short-term safety profile of Sars-Cov2 vaccination on glucose control: continuous glucose monitoring..., D'onofrio [/bib_ref].
Interestingly, oxidative stress can induce a reduction in RBCs deformability with a subsequent elevation of BV [bib_ref] Serum bilirubin and lipoprotein-a: how are these associated with whole blood viscosity?, Nwose [/bib_ref]. In obesity, high oxidative stress and fibrinogen together with prolonged low-grade inflammation are linked with the development of HVS [bib_ref] Increased oxidative stress in obesity and its impact on metabolic syndrome, Furukawa [/bib_ref]. Therefore, depending on these findings, obese patients are at high risk for the development of HVS after COVID-19 vaccination. Pivonello and colleagues suggested that the immune response in obese patients against the COVID-19 vaccine is low due to impaired reactivity of T and B cells. Therefore, a delay in immune response may reduce Ig concentrations following COVID-19 vaccination, and this may affect the development of HVS in obesity. Of note, the immune response in obese patients was low following the influenza vaccine [bib_ref] The weight of obesity on the human immune response to vaccination, Painter [/bib_ref]. These findings are premature to draw a final association between COVID-19 vaccination and the risk of HVS, and thus, prospective and retrospective studies are warranted in this regard.
The present review had many limitations, including the rarity of prospective studies that evaluate BV in COVID-19 at the time of admission and discharge. Also, most studies were speculative in their explanation of HVS in COVID-19 and COVID-19 vaccination. Despite these limitations, the present critical review revealed that HVS is an important mechanistic pathway in the development of complications in COVID-19 and related vaccines.
# Conclusions
COVID-19 and related vaccines are linked with the development of HVS mainly in patients with previous COVID-19 and underlying metabolic derangements. The possible mechanism of HVS in COVID-19 and related vaccines is increasing levels of fibrinogen and immunoglobulins. Dehydration, oxidative stress, and inflammatory reactions are regarded as additional contributing factors in the pathogenesis of HVS in COVID-19. However, this critical review cannot determine the final causal relationship between COVID-19 and related vaccines and the development of HVS. Prospective and retrospective studies are warranted in this field.
[fig] Figure 2: Complications of hyperviscosity in COVID-19: Hyperviscosity (HVS) in COVID-19 induces the development of endothelial dysfunction, microangiopathy, and hypoperfusion with the development of thrombosis and tissue hypoxia, which ultimately cause organ dysfunction [/fig]
|
Effect of neoadjuvant chemotherapy on overall survival of patients with T2-4aN0M0 bladder cancer: A systematic review and meta-analysis according to EAU COVID-19 recommendation
PurposeIn the context of the COVID-19 outbreak, the European Association of Urology (EAU) guidelines Rapid Reaction Group provided recommendations to manage muscle invasive bladder cancer (MIBC) based on priority levels: neoadjuvant chemotherapy (NAC) should be avoided for patients with T2-3N0M0 MIBC. This meta-analysis aims to evaluate the efficacy of NAC compared with radical cystectomy (RC) alone in improving the overall survival (OS) of patients with T2-4aN0M0 MIBC.Materials and methodsA systematic review was performed according to the PRISMA guidelines. The PubMed/ Medline, EMBASE, and Cochrane Library databases were searched. The primary outcome was OS of patients with T2-4aN0M0 MIBC, and the secondary outcome was OS of patients with only T2N0M0 MIBC.ResultsEight studies were included in this meta-analysis. Overall, the quality of all studies was relatively high, and little publication bias was demonstrated. The OS was significantly better in the NAC with RC group than in RC alone (HR, 0.79; 95% CI, 0.68-0.92; p = 0.002). A subgroup analysis was performed on only patients with T2N0M0 MIBC, and five studies were included. There was no difference in the OS between the NAC with RC and the RC alone groups (HR, 0.83; 95% CI, 0.69-1.01 p = 0.06). Citation: Kang DH, Cho KS, Moon YJ, Chung DY, Jung HD, Lee JY (2022) Effect of neoadjuvant chemotherapy on overall survival of patients with T2-4aN0M0 bladder cancer: A systematic review and meta-analysis according to EAU COVID-19 recommendation. PLoS ONE 17(4): e0267410.
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
With the spread of coronavirus disease 2019 , the world is experiencing an unprecedented time in history, and the medical field is undergoing major changes. Many medical and surgical procedures have been postponed or omitted because of the COVID-19 pandemic [bib_ref] Treatment delays in oncology patients during COVID-19 pandemic: A perspective, Kumar [/bib_ref] [bib_ref] Integrated Survival Estimates for Cancer Treatment Delay Among Adults With Cancer During..., Hartman [/bib_ref]. The basis for this approach is to reduce the risk of COVID-19 transmission, increase bed availability for patients with COVID-19 in wards and intensive care units, reduce the workload of healthcare providers except for COVID-19 cases, and limit aerosol generation procedures. For urological diseases, urologists should also manage patients with COVID-19 and urological diseases in a balanced manner. Considering the COVID-19 pandemic, the European Association of Urology (EAU) guidelines Rapid Reaction Group presents a response policy for diagnosis, treatment, and follow-up according to the priority of urologic diseases, including urological cancer [bib_ref] European Association of Urology Guidelines Office Rapid Reaction Group: An Organisation-wide Collaborative..., Ribal [/bib_ref]. The guidelines were also presented for bladder cancer (BCa), and they recommend that neoadjuvant chemotherapy (NAC) should be omitted in patients with T2-3 focal N0M0 muscle-invasive bladder cancer (MIBC) during the COVID-19 pandemic.
Of all urogenital cancers, it is difficult to make timely decisions about BCa. Patients with BCa are generally older, more susceptible to COVID-19, and often have several comorbidities. In particular, MIBC often requires both chemotherapy and radical cystectomy (RC), which can cause morbidity and mortality. Cisplatin-based NAC followed by RC is the standard of care and recommended treatment for clinical stage II and IIIA BCa [bib_ref] European Association of Urology Guidelines on Muscle-invasive and Metastatic Bladder Cancer: Summary..., Witjes [/bib_ref]. Chemotherapy can be performed before or after surgery [bib_ref] Why consider neoadjuvant chemotherapy for muscle-invasive transitional cell carcinoma of the bladder?, North [/bib_ref] , but there are several criteria for NAC. The first is to treat micrometastatic disease at the time of diagnosis when the burden of disease is the lowest [bib_ref] Perioperative chemotherapy in muscle-invasive bladder cancer to enhance survival and/ or as..., Sternberg [/bib_ref]. In addition, patients are more likely to tolerate preoperative chemotherapy than postoperative chemotherapy [bib_ref] Why consider neoadjuvant chemotherapy for muscle-invasive transitional cell carcinoma of the bladder?, North [/bib_ref].
Several meta-analyses on the efficacy of NAC for MIBC reported that NAC is effective [bib_ref] Meta-analysis of neoadjuvant chemotherapy compared to radical cystectomy alone in improving overall..., Hamid [/bib_ref]. However, one study reported that there was no difference in NAC compared with cystectomy and/or radiotherapy alone [bib_ref] Effect of cisplatin-based neoadjuvant chemotherapy on survival in patients with bladder cancer:..., Li [/bib_ref]. Disagreements in these studies require a more comprehensive analysis to encourage the use of NAC in MIBC treatment. We intend to perform an updated meta-analysis by extracting studies targeting patients with T4a or lower without lymph node metastasis among MIBC and adding recently published papers. Furthermore, through this meta-analysis, we would like to examine the evidence for the omission of NAC in patients with T2-3N0M0 MIBC as recommended by the EAU guidelines Rapid Reaction Group in more detail.
# Materials and methods
## Literature search and study selection
A comprehensive literature search was conducted until October 31, 2021, using PubMed Central, Cochrane Central Controlled Register of Trials (CENTRAL), and Embase. We used search terms such as "muscle-invasive bladder cancer," "neoadjuvant chemotherapy," "cystectomy," and "overall survival" and relevant variants. Titles and abstracts were screened for relevance, followed by full-text screening. All duplicate articles were excluded, and related articles were finally searched from the reference list of the retrieved articles. This meta-analysis was registered on the PROSPERO website (number: CRD42021299238).
## Inclusion and exclusion criteria
This systematic review and meta-analysis followed the participants, interventions, comparators, outcomes, and study design (PICOS) approach and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The PICOS model for this metaanalysis consisted of Participants (patients with T2-4aN0M0 MIBC), Intervention (T2-4aN0M0 MIBC patients who underwent neoadjuvant chemotherapy), Comparison (T2-4aN0M0 MIBC patients who underwent RC alone), Outcome (comparison of overall survival (OS)), and Study type (randomized clinical trials, prospective, and retrospective studies). The following inclusion criteria were adopted: patients with MIBC proven by histological examination; two-arm studies that compared NAC with RC and RC alone; no lymph node and distant metastases; and studies with OS outcomes. The exclusion criteria were: studies that compared different NAC regimens without RC, single-arm studies, case reports, reviews, commentaries, animal studies, and non-English written studies.
## Data extraction and quality assessment
All the identified articles were independently screened by two reviewers (DHK and HDJ), and two other reviewers (DYC and YJM) independently analyzed all the details of each article to confirm that they met the inclusion criteria. Any discrepancy between the two reviewers was resolved through discussion until agreed, or via third-party adjudication performed by another reviewer (JYL). Once the final group of articles was agreed upon, two researchers independently examined the quality. The quality of each study was estimated using the Cochrane riskof-bias tool for randomized control trials (RCTs) [bib_ref] The Cochrane Collaboration's tool for assessing risk of bias in randomised trials, Higgins [/bib_ref] and the Newcastle-Ottawa Scale (NOS) for non-randomized controlled trials (www.ohri.ca/programs/clinical_epidemiology/oxford. asp) [bib_ref] Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality..., Stang [/bib_ref].
## Data synthesis and analysis
The primary outcome measures were OS of patients with T2-4aN0M0 MIBC, and secondary outcome measures were OS of patients with only T2N0M0 MIBC. Meta-analysis was conducted using R (R version 4.1.3, R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org) and its meta and metasens packages. The effect measures of the outcomes of interest were the hazard ratios (HRs) with 95% confidence interval (CI), which was obtained by extracting the proportion of patients with the outcome events. Pooled incidences were calculated using a fixed-effects or random-effects model according to the heterogeneity of the included studies. The Cochrane Q and I 2 statistics were used to assess the statistical heterogeneity. If no significant statistical inconsistency was observed (I 2 < 25%), the summary estimate was calculated using the fixed-effects model. When heterogeneity was observed, the summary estimate was calculated using the random-effects model.
# Ethics statement
The data and results used in this paper are all from published studies, and there is no ethical issue, so the approval of the ethics committee is not required.
# Results
## Study characteristics
The database searches identified 355 articles that were included in this meta-analysis. According to the inclusion and exclusion criteria, 302 articles were excluded after a brief review of the titles and abstracts of the articles. This left 53 studies that evaluated the OS. After reviewing the full-text articles of these studies, 45 were excluded due to irrelevant results. [fig_ref] Fig 1: Search strategy for a systematic review and meta-analysis [/fig_ref] the results of the comprehensive search, selection process, and number of excluded studies along with the relevant evidence. The search finally identified eight studies [bib_ref] Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder..., Grossman [/bib_ref] [bib_ref] Randomised phase III study of neoadjuvant chemotherapy with methotrexate, doxorubicin, vinblastine and..., Kitamura [/bib_ref] [bib_ref] Neoadjuvant chemotherapy versus cystectomy in management of stages II, and III urinary..., Osman [/bib_ref] [bib_ref] Persistent muscle-invasive bladder cancer after neoadjuvant chemotherapy: an analysis of Surveillance, Epidemiology..., Lane [/bib_ref] [bib_ref] Superior efficacy of neoadjuvant chemotherapy and radical cystectomy in cT3-4aN0M0 compared to..., Hermans [/bib_ref] [bib_ref] Neoadjuvant chemotherapy for muscle invasive bladder cancer: a nationwide investigation on survival, Russell [/bib_ref] [bib_ref] Neoadjuvant chemotherapy plus radical cystectomy versus radical cystectomy alone in clinical T2..., Soria [/bib_ref]. Two studies used methotrexate/vinblastine/doxorubicin/cisplatin (MVAC), one study used gemcitabine/ cisplatin (GC), one study used cisplatin or carboplatin-based regimen, and one study used cisplatin-based NAC, remaining three studies did not mention the chemotherapy regimen. The characteristics of the included studies are presented in [fig_ref] Table 1: Characteristics of included studies [/fig_ref].
## Primary endpoint: outcomes of t2-4an0m0 mibc patients
Heterogeneity assessment. Eight trials accounting for a total of 12,672 assessable patients were included in the analysis. Forest plots of patients with T2-4aN0M0 MIBC are shown in [fig_ref] Fig 2: Forest plots for patients with T2-4aN0M0 MIBC [/fig_ref] There was high heterogeneity (I 2 = 48%, p = 0.05), hence, a random-effects model was used. After selection of the effect models, little heterogeneity was observed in the radial plots
## Plos one
Effect of neoadjuvant chemotherapy in MIBC [fig_ref] Fig 3: Radial plots of the overall survival of patients with T2-4aN0M0 MIBC [/fig_ref]. We conducted sensitivity analysis for outcome reporting bias (ORB) to examine the degree of heterogeneity [fig_ref] Fig 4: Sensitivity analysis for the outcome reporting bias [/fig_ref]. The sensitivity of this meta-analysis was considered robust, as the results on OS were not affected until up to two studies were excluded. There was high heterogeneity (I 2 = 59%, p = 0.04), thus a random-effects model was used. After selection of the effect models, little heterogeneity was observed in the radial plots [fig_ref] Fig 3: Radial plots of the overall survival of patients with T2-4aN0M0 MIBC [/fig_ref]. We conducted sensitivity analysis for ORB to examine the degree of heterogeneity [fig_ref] Fig 4: Sensitivity analysis for the outcome reporting bias [/fig_ref]. The sensitivity of this meta-analysis was considered robust, as excluding the study did not affect the results on OS.
Overall survival. There was no difference in OS between the NAC with RC group and the RC alone group (HR, 0.83; 95% CI, 0.69-1.01; p = 0.06).
## Quality assessment
The results of the quality assessment based on the Cochrane risk-of -bias tool of the three included RCTs are shown in . Cancer treatment consisting of NAC and surgery informed both clinicians and patients of the treatment and provided written consent to all patients. Therefore, the high risk of bias in allocation concealment and blind processes could not be avoided. The results of the quality assessment using the NOS for the included nonrandomized studies are shown in . All five studies received a score of 6 points (indicating high quality).
## Publication bias
Funnel plots from these meta-analyses are shown in [fig_ref] Fig 6: Funnel plots for patients with T2-4aN0M0 MIBC [/fig_ref] Little publication bias was observed in the funnel plots.
# Discussion
The results of this study confirmed that NAC with RC was more helpful in the survival of patients than RC alone in patients with T2-4aN0M0 MIBC, and that there was no difference in survival between the two groups of RC alone and NAC with RC in patients with T2N0M0 MIBC. The main purpose of this study was to corroborate the evidence presented by the EUA guidelines Rapid Reaction Group on omitting NAC in patients with T2-3N0M0 MIBC during the COVID-19 pandemic [bib_ref] European Association of Urology Guidelines Office Rapid Reaction Group: An Organisation-wide Collaborative..., Ribal [/bib_ref] through a met-analysis. The authors identified solid evidence in patients with T2N0M0 MIBC. Few studies have targeted patients with only T2-3N0M0, thus it was difficult to substantiate the guidelines for T3N0M0. However, in patients with T2-4aN0M0 MIBC, it was statistically confirmed that NAC with RC helped the patients' survival compared with RC alone, but compared with T2N0M0, HR did not differ much (HR 0.79 vs 0.83). Therefore, it can be inferred from this study that there is a sufficient possibility that there is no statistical difference if only T2-3N0M0 is targeted.
COVID-19 is an acute respiratory infectious disease caused by SARS-CoV-2, a novel strain of coronavirus that was first reported in November 2019 [bib_ref] Comprehensive review of coronavirus disease 2019 (COVID-19), Chauhan [/bib_ref]. The COVID-19 pandemic has
## Plos one
Effect of neoadjuvant chemotherapy in MIBC had major effects on individuals and healthcare systems. Limited healthcare resources, as well as those infected with SARS-CoV-2, have contributed to the spread of the virus. Thus resulting in reduced capacity and rapid depletion of health systems and hospitals [bib_ref] Health systems resilience in managing the COVID-19 pandemic: lessons from 28 countries, Haldane [/bib_ref]. To prevent the
## Table 2. results of quality assessment by the cochrane risk-of-bias tool (a) and nos (b).
A. Quality assessment of a randomized controlled trial spread of the virus, hospital visits should be limited by delaying surgery or procedures, or by omitting low-priority treatments [bib_ref] Delay or Avoidance of Medical Care Because of COVID-19-Related Concerns-United States, Czeisler [/bib_ref]. However, medical delays or omissions can increase the morbidity and risk of death from treatable and preventable diseases. Since there are few international medical emergency situations historically, it is difficult to judge the pros and cons of delaying or avoiding treatment for any disease. Therefore, preparing individual guidelines for each disease is very important in the current COVID-19 pandemic.
There have been many changes in the field of urology due to COVID-19. Many urologists are making great efforts not only for the management of COVID-19, but also for the management of urologic diseases in the current situation. In response to the COVID-19 pandemic, the EAU Guidelines Office has been working with the Executive Committee, the Section offices, and others to set up a "Rapid Reaction Group" [bib_ref] European Association of Urology Guidelines Office Rapid Reaction Group: An Organisation-wide Collaborative..., Ribal [/bib_ref]. The protocol was divided into four large categories to provide recommendations: diagnosis, surgical treatment and medical therapy, follow-up/telemedicine, and emergency. The panel provides a table with recommendations according to priority level. Four color-separated risk stratification tools were created to help apply recommendations and are as follows: low priority, clinical harm very unlikely if postponed for 6 months (green); intermediate priority: clinical harm possible, but unlikely, if postponed for 3-4 months (yellow); high priority: clinical harm very likely if postponed for >6 weeks (red); emergency: life-threatening situation-cannot be postponed for >24 h (black).
All urological cancers, including BCa, were evaluated by the EAU guidelines Rapid Reaction Group. Among urological cancers, the managements for BCa can be the most diverse treatments. Depending on the stage or grade, there are a wide variety of treatments such as active surveillance, transurethral resection of bladder tumor, intravesical bacillus Calmette-Guérin instillation, RC, chemotherapy, and radiotherapy [bib_ref] Bladder cancer, Kaufman [/bib_ref]. During the COVID-19 pandemic, the EAU Guidelines Office presented a diverse set of guidelines depending on the status of BCa, including omitting NAC for patients with T2-3 focal N0M0 MIBC (green color by the EAU guidelines Rapid Reaction Group). While these guidelines have been provided by highlyexperienced and respected board and panel members, scientific evidence may be lacking as the guidelines were developed in a short period of time. Thus, we concluded that a meta-analysis of NAC with RC vs. RC alone in patients with MIBC was necessary to support the rationale for the guideline. In patients with stage II and IIIA BCa, post-NAC, RC may be the standard treatment, or RC alone may also be an option (especially for those who are not eligible for cisplatin-based chemotherapy) [bib_ref] European Association of Urology Guidelines on Muscle-invasive and Metastatic Bladder Cancer: Summary..., Witjes [/bib_ref]. The EAU Guidelines Office recommends omitting NAC in patients with stage II and stage IIA BCa during the COVID-19 pandemic. We paid attention to this recommendation and concluded that it is very important to confirm the extent to which NAC actually helps survival in patients with T2-3N0M0 through meta-analysis.
Prior to this study, several meta-analyses have analyzed the effect of NAC with RC on survival in patients with MIBC compared to RC alone. A meta-analysis based on 11 trials, 3005 patients; comprising 98% of all patients from known eligible RCTs, reported that the cisplatinbased combination NAC resulted in an absolute OS benefit of 5% (HR 0.86 [0.77-0.95]) at 5 years. Recently, Hamid et al. [bib_ref] Meta-analysis of neoadjuvant chemotherapy compared to radical cystectomy alone in improving overall..., Hamid [/bib_ref] conducted a meta-analysis of 17 studies with 13,391 patients and reported that the results showed similar HR (HR 0.82 [0.71-0.95]) to the aforementioned study and that NAC improved OS. However, another meta-analysis reported that there was no difference in NAC with RC compared with RC and/or radiotherapy (HR 0.92 [0.84-1.00]) [bib_ref] Effect of cisplatin-based neoadjuvant chemotherapy on survival in patients with bladder cancer:..., Li [/bib_ref]. The results of the studies have some discrepancies due to minor differences in patient groups and treatments in each meta-analysis, including the present study. However, the inconsistencies of these study results indicate that a comprehensive analysis is necessary to determine the effect of NAC on the treatment of MIBC, and we suggested that the results of present study can help reach a consensus to some extent. In addition, our study has the advantage of providing more specific information because it limited patients up to T4a among patients without lymph node metastasis. In particular, in the current crisis of medical systems during COVID-19, it is important to efficiently use the limited medical resources through a balanced approach between the management of infectious diseases and the treatment of cancer. Our study can be helpful to establish the treatment guidelines for this global emergency.
In this study, NAC had an OS advantage in T2-4aN0M0 patients, but no OS benefit in T2N0M0 patients. The role of preoperative NAC in MIBC is well known. Patients can tolerate chemotherapy better, and it also makes surgery easier by reducing the tumor burden during surgery [bib_ref] Correlation of pathologic complete response with survival after neoadjuvant chemotherapy in bladder..., Petrelli [/bib_ref]. In addition, if there is micrometastasis, it can be effectively treated [bib_ref] Perioperative chemotherapy in muscle-invasive bladder cancer to enhance survival and/ or as..., Sternberg [/bib_ref]. These points act as an advantage; and in this study, it was considered that the OS in patients with T2-4aN0M0 was also beneficial. On the other hand, several reasons can be assumed as to why the OS did not show any benefit when only T2N0M0 patients were targeted. First, since T2 BCa can be considered as an organ-confined status, reducing the tumor burden may not have a significant effect [bib_ref] Survival of Patients with Muscle-Invasive Urothelial Cancer of the Bladder with Residual..., Pfail [/bib_ref]. Also, since the presence of micrometastases is relatively lower, NAC may not be effective. In addition, in patients in whom NAC is ineffective, delay in surgery due to NAC may actually cause disease progression [bib_ref] Delaying Radical Cystectomy After Neoadjuvant Chemotherapy for Muscle-invasive Bladder Cancer is Associated..., Boeri [/bib_ref]. Appropriate additional research is needed in this regard.
This study has some limitations. First, only T2-3N0M0 as suggested by the EAU guidelines Rapid Reaction Group was not separately classified. In most clinical studies, it was difficult to separate patients with T2-3N0M0 from patients with lymph nodes positive, T2-4, T2-4a, T3-4, or T2. However, T4a is a case of invasion of the prostate, seminal vesicles, uterus, or vagina, which can still be defined as a surgically resectable status. T4b is a case in which the pelvic wall or abdominal wall is invaded, and complete surgical removal is considered impossible. Therefore, T4a was included in this study and T4b was excluded, and the patient group was reduced as close to the guidelines presented by the EAU guidelines Rapid Reaction Group as much as possible. In addition, subgroup analysis was performed on only patients with T2N0M0, and the evidence of the guidelines was clearly corroborated for T2N0M0. Additionally, we did not investigate the side effects or quality of life associated with NAC. In a prospective study, NAC was found to be associated with a 30%-40% rate of grade 3-4 toxicity [bib_ref] Bladder Cancer, Version 5.2017, NCCN Clinical Practice Guidelines in Oncology, Spiess [/bib_ref]. In addition, most patients are elderly and often have comorbidities, so it is appropriate to consider the side effects and quality of life associated with NAC as well as other oncologic outcomes. Moreover, detailed analysis according to the regimen or cycle of NAC was not performed. Incomplete chemotherapy has been recently associated with pathological progression during NAC and an inferior pathological response after RC [bib_ref] Optimization of Patient Selection for Neoadjuvant Chemotherapy in Muscle-invasive Urothelial Carcinoma of..., Hensley [/bib_ref] [bib_ref] Gemcitabine and cisplatin neoadjuvant chemotherapy for muscle-invasive urothelial carcinoma: Predicting response and..., Gandhi [/bib_ref]. Therefore, when evaluating patient outcomes, it is necessary to carefully consider the regimen and cycle of NAC. Hence, additional studies are needed in the near future. Lastly, the studies included in this meta-analysis showed a high degree of heterogeneity. This was probably due to the large difference in weight between the studies, indicating high heterogeneity. To overcome this issue, we analyzed heterogeneity through radial plots after applying the random-effects model, and observed almost no heterogeneity. In addition, it was analyzed to show robust sensitivity through sensitivity analysis for ORB.
Nevertheless, this study is the first analysis to substantiate the evidence presented by the EAU guidelines Rapid Reaction Group to omit NAC from the T2-3 focal N0M0 MIBC during the COVID-19 pandemic, and its impact can be unique. Unfortunately, the T2-3N0M0 group was not analyzed due to the limited number of studies, but strong evidence was found in the T2N0M0 group. In addition, we tried to derive the results by narrowing T stages down to T4a or less, and we think that this study can positively support the COVID-19 EAU guideline when the results of this study are comprehensively considered. In the future, several studies on patients with T2-3N0M0 MIBC should be published, and studies examining side effects, quality of life, and oncological outcomes will be needed.
# Conclusion
As recommended by the EAU guidelines Rapid Reaction Group, patients with T2N0M0 MIBC should strongly consider omitting NAC until the end of the COVID-19 pandemic. Whether to omit NAC in patients with T3-4aN0M0 MIBC needs further discussion, and studies targeting only T2-3N0M0 MIBC are expected to proceed further.
## Supporting information
[fig] Fig 1: Search strategy for a systematic review and meta-analysis. https://doi.org/10.1371/journal.pone.0267410.g001 [/fig]
[fig] Fig 2: Forest plots for patients with T2-4aN0M0 MIBC. The overall survival was significantly better in the NAC with RC group than in the RC alone group (HR, 0.79; 95% CI, 0.68-0.92; p = 0.002).Overall survival. The OS was significantly higher in the NAC with RC group than in RC alone group (HR, 0.79; 95% CI, 0.68-0.92; p = 0.002)(Fig 2).Secondary endpoint: Outcomes of T2N0M0 MIBC patientsHeterogeneity assessment. Five trials accounting for a total of 11,208 assessable patients were included in the analysis. Forest plots of patients with T2N0M0 MIBC are shown inFig 5. [/fig]
[fig] Fig 3: Radial plots of the overall survival of patients with T2-4aN0M0 MIBC (A) and patients with T2N0M0 MIBC (B). After selection of the effect models, little heterogeneity was observed in the radial plots. https://doi.org/10.1371/journal.pone.0267410.g003 [/fig]
[fig] Fig 4: Sensitivity analysis for the outcome reporting bias (ORB) of patients with T2-4aN0M0 MIBC (A) and patients with T2N0M0 MIBC(B). The sensitivity was considered robust in the sensitivity analysis for ORB. [/fig]
[fig] Fig 5: Forest plots for patients with T2N0M0 MIBC. There was no difference in overall survival between the NAC with RC group and RC alone group (HR, 0.83; 95% CI, 0.69-1.01; p = 0.06).https://doi.org/10.1371/journal.pone.0267410.g005 [/fig]
[fig] Fig 6: Funnel plots for patients with T2-4aN0M0 MIBC (A) and patients with T2N0M0 MIBC (B). There was little publication bias in funnel plots. https://doi.org/10.1371/journal.pone.0267410.g006 [/fig]
[table] Table 1: Characteristics of included studies. OS: overall survival, NAC: neoadjuvant chemotherapy, RC: radical cystectomy, RCT: randomized controlled trial, MVAC: methotrexate/vinblastine/doxorubicin/ cisplatin, GC: gemcitabine/cisplatin, PW: propensity-weighting, IPTW: inverse probability of treatment-weighting https://doi.org/10.1371/journal.pone.0267410.t001 [/table]
|
Four Japanese Patients with Congenital Nephrogenic Diabetes Insipidus due to the AVPR2 Mutations
# Introduction
Nephrogenic diabetes insipidus (NDI) is a rare disease that is characterized by resistance of the distal renal tubule and collecting ducts to arginine vasopressin [bib_ref] Pathophysiology, diagnosis and management of nephrogenic diabetes insipidus, Bockenhauer [/bib_ref] [bib_ref] Genetic forms of nephrogenic diabetes insipidus (NDI): vasopressin receptor defect (X-linked) and..., Bichet [/bib_ref]. Vast majority of NDI is caused by mutations in the arginine vasopressin receptor 2 gene (AVPR2) on the X chromosome [bib_ref] Molecular cloning of the receptor for human antidiuretic hormone, Birnbaumer [/bib_ref]. At present, more than 250 mutations have been reported [bib_ref] Genetic forms of nephrogenic diabetes insipidus (NDI): vasopressin receptor defect (X-linked) and..., Bichet [/bib_ref]. Mutations in AVPR2 were classified into three types. Type-I mutants reach the cell surface but cannot bind its ligand, type-II mutant receptors have impaired intracellular transport and cannot reach the cell surface, and type-III mutants are inappropriately transcribed [bib_ref] Molecular biology of hereditary diabetes insipidus, Fujiwara [/bib_ref] [bib_ref] Characterization of vasopressin V2 receptor mutants in nephrogenic diabetes insipidus in a..., Robben [/bib_ref].
Common symptoms in male patients are polyuria, polydipsia, fever of unknown etiology, convulsions, and vomiting, which usually develop soon after birth [bib_ref] Clinical presentation and follow-up of 30 patients with congenital nephrogenic diabetes insipidus, Van Lieburg [/bib_ref]. On the other hand, female cases have only mild symptoms [bib_ref] Identification and characterization of a novel X-linked AVPR2 mutation causing partial nephrogenic..., Neocleous [/bib_ref]. Furthermore, some mutations in the AVPR2 are related to partial NDI [bib_ref] Vasopressin type 2 receptor V88M mutation: molecular basis of partial and complete..., Bockenhauer [/bib_ref] [bib_ref] V2 vasopressin receptor (V2R) mutations in partial nephrogenic diabetes insipidus highlight protean..., Takahashi [/bib_ref].
In this study, we assessed the clinical and biochemical parameters and AVPR2 status in four NDI cases of three unrelated Japanese families.
# Subjects and methods
2.1. Subjects. Clinical symptoms, age at diagnosis, biochemical data, and current treatment are summarized in [fig_ref] Table 1: Clinical characteristics of 4 patients with congenital nephrogenic diabetes insipidus and AVPR2... [/fig_ref]. All four patients had polyuria and polydipsia, and results of biochemical evaluations showed high plasma antidiuretic hormone (ADH) levels. Based on these findings, NDI was suspected. e Institutional Review Board Committee of Hokkaido University approved this study (approval number 13-061). e patients' parents provided written informed consent for their children's participation in this study.
## Case 1.
A 3-month-old Japanese boy was admitted because of poor body weight gain, vomiting, and fever that had persisted for one week. He was born as a full-term infant with no complications during pregnancy.
At the time of admission, he had polyuria with a urine volume of 700-800 mL/d. Results of laboratory examinations are shown in [fig_ref] Table 1: Clinical characteristics of 4 patients with congenital nephrogenic diabetes insipidus and AVPR2... [/fig_ref]. Findings of brain magnetic resonance imaging (MRI) were normal. Based on the polyuria and the high serum ADH level, the infant was diagnosed as having NDI, and hydrochlorothiazide was initiated. Spironolactone and potassium supplementation was added when he was 2 years old and 4 years old, respectively, and indomethacin and a protein-restricted diet were initiated when he was 6 years old. He is currently 13 years old. His height is 150 cm (−0.8 SD), and his weight is 37 kg (−0.6 SD). His urine volume is approximately 7 L/day. He has mild hydronephrosis in the right kidney. His mother is asymptomatic. e family tree of Case 1 is shown in .
## Case 2.
In Case 2, poor weight gain was pointed out at the age of 4 months in this male Japanese infant. Polydipsia and polyuria were noted when he was 17 months of age. At that time, his water intake volume was approximately 3 L/d. Previously, he had experienced recurrent mild to moderate fevers of unknown etiology. e laboratory examinations results are shown in Table 1. e water deprivation test showed elevated serum Na + , plasma osmolality, and urine osmolality [fig_ref] Table 2: Results of the water deprivation test in Case 2 [/fig_ref]. However, the subcutaneous injection of vasopressin did not greatly increase urine osmolality. Six and a half hours after the test started, his body weight was reduced by 4.1%. Finally, his plasma ADH elevated to 110.1 pmol/L. Brain MRI findings were normal. Based on these findings, a diagnosis of partial NDI was confirmed when he was 19 months of age. Trichlormethiazide was initiated in combination with spironolactone and sodium restriction. is treatment has successfully decreased the patient's urine volume and water intake, and his body weight has caught up to near normal for his age. Now, he is 3 years old, and his height is 90.8 cm (−0.6 SD) and weight is 12.9 kg (−0.4 SD). His mother had also complained of mild polydipsia (2,000 mL/day) and polyuria from childhood, and her plasma ADH level was mildly elevated (5.90 pmol/L), but further examination has not been done. e pedigree of this family is shown in .
## Cases 3 and 4.
Case 3 is now a 14-year-old Japanese boy. Polydipsia and polyuria were noticed at 4 years of age. He had enuresis every day from infancy. Since he had an elevated plasma ADH level (53.1 pmol/L), he was diagnosed as having NDI. e laboratory data at the time of diagnosis are shown in [fig_ref] Table 1: Clinical characteristics of 4 patients with congenital nephrogenic diabetes insipidus and AVPR2... [/fig_ref]. He is being treated with hydrochlorothiazide, potassium supplementation, and indomethacin. Currently, his water intake is approximately 3 L/d. Case 4 is the younger sister of Case 3 and she is now 12 years old. Polydipsia and polyuria were noted when she was 4 years old after the diagnosis of NDI in Case 3. Her plasma ADH level (6.2 pmol/L) was also elevated. She was diagnosed as having NDI, and treatment with hydrochlorothiazide, potassium supplementation, and indomethacin was initiated. eir mother also complained polydipsia and polyuria since her childhood. However, her latest daily urine volume Case Reports in Pediatrics (2,000 mL/day) did not meet the diagnostic criteria for NDI. e family tree is shown in .
## Sequence analysis of avpr2 and study of x chromosome
Inactivation. Genomic DNA was extracted from peripheral blood leukocytes of the cases and female carriers. e AVPR2 exon was amplified by polymerase chain reaction (PCR) using the primers as reported previously [bib_ref] Novel mutations associated with nephrogenic diabetes insipidus. A clinical-genetic study, Garcia Castano [/bib_ref] , and PCR products were purified and sequenced directly using an Applied Biosystems 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). e X-inactivation patterns of female carriers were investigated by studying the polymorphic trinucleotide (CAG) repeats in the first exon of the human androgen receptor gene as reported previously [bib_ref] Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in..., Allen [/bib_ref].
# Results
Hemizygous mutations of AVPR2 were identified in all three male patients [fig_ref] Figure 2: Results of sequencing of AVPR2 mutations [/fig_ref]. In Case 1, one base insertion caused a frame shift, generating a premature stop codon at codon 191 in exon 2 (c.299_300insA; p. K100KfsX91, designated as p.K100KfsX91). His mother was heterozygous for this mutation. Case 2 had a c.316C > T; p.R106C (designate as p.R106C) in exon 2, which was previously reported [bib_ref] Novel mutations associated with nephrogenic diabetes insipidus. A clinical-genetic study, Garcia Castano [/bib_ref] [bib_ref] Nature and recurrence of AVPR2 mutations in X-linked nephrogenic diabetes insipidus, Bichet [/bib_ref] [bib_ref] Nephrogenic diabetes insipidus: functional analysis of new AVPR2 mutations identified in Italian..., Albertazzi [/bib_ref] [bib_ref] Functional characterization of the molecular defects causing nephrogenic diabetes insipidus in eight..., Pasel [/bib_ref] [bib_ref] Identification of mutations in the arginine vasopressin receptor 2 gene causing nephrogenic..., Chen [/bib_ref] [bib_ref] Novel vasopressin type 2 (AVPR2) gene mutations in Brazilian nephrogenic diabetes insipidus..., Boson [/bib_ref] [bib_ref] Clinical overview of nephrogenic diabetes insipidus based on a nationwide survey in..., Fujimoto [/bib_ref]. His mother was heterozygous for this mutation. Cases 3 and 4 had a nucleotide change of G to A at position 296, resulting in the nonsense substitution (c.296G > A; p.W99X, designated as p.W99X). eir mother and Case 4 were heterozygous for the mutation. ese mutations were previously reported. e values of relative X-inactivation for the normal allele in Case 4 was 67.0%, and the values of mothers of all four patients were 65.0% (Case 1 mother), 62.0% (Case 2 mother), and 58.0% (Case 3 and 4 mother), respectively . Skewed X-inactivation is defined as inactivation of 75-80% of cells in the same allele [bib_ref] A skewed view of X chromosome inactivation, Minks [/bib_ref]. erefore, they had random X-inactivation. Case 2 * † * * * * : Family trees of 4 Japanese patients with congenital nephrogenic diabetes insipidus due to the AVPR2 mutations. Family trees of (a) Case 1, (b) Case 2, and (c) Cases 3 and 4. Index cases, also indicated by arrows, are represented by filled boxes and carriers as half-filled circles. † Xinactivation patterns were analyzed; * polydipsia and polyuria; * * mild polydipsia; ‡ maternal grandfather of Case 1 had tendency of polydipsia, but the detail is not clear; * * as maternal grandfather of Case 2 is dependent on alcohol, and polydipsia and polyuria are not clear; § maternal grandfather of Cases 3 and 4 had a tendency of polydipsia, but the detail examination is not performed; # maternal uncle of Cases 3 and 4, who had been undergone artificial dialysis for renal failure, died at the age of forty-seven. e detail for renal failure was not clear; NA, not accessed.
# Discussion
Currently, over 250 mutations in the AVPR2 have been described as the cause of NDI [bib_ref] Genetic forms of nephrogenic diabetes insipidus (NDI): vasopressin receptor defect (X-linked) and..., Bichet [/bib_ref]. In our study, three previously reported mutations (p.K100KfsX91, p.W99X, and p.R106C) were identified [bib_ref] Nature and recurrence of AVPR2 mutations in X-linked nephrogenic diabetes insipidus, Bichet [/bib_ref] [bib_ref] Clinical overview of nephrogenic diabetes insipidus based on a nationwide survey in..., Fujimoto [/bib_ref]. e mutations p.K100KfsX91 and p.W99X produced a premature stop codon, resulting in a truncated protein. Regarding p.R106C, that mutation was identified in 7.7% (5/65) of Japanese NDI patients [bib_ref] Clinical overview of nephrogenic diabetes insipidus based on a nationwide survey in..., Fujimoto [/bib_ref] and also was identified in other Asian and ethnic populations [bib_ref] Genetic forms of nephrogenic diabetes insipidus (NDI): vasopressin receptor defect (X-linked) and..., Bichet [/bib_ref] [bib_ref] Nature and recurrence of AVPR2 mutations in X-linked nephrogenic diabetes insipidus, Bichet [/bib_ref] [bib_ref] Identification of mutations in the arginine vasopressin receptor 2 gene causing nephrogenic..., Chen [/bib_ref] [bib_ref] Novel vasopressin type 2 (AVPR2) gene mutations in Brazilian nephrogenic diabetes insipidus..., Boson [/bib_ref]. e p.R106C mutation occurs at CpG nucleotides, which are mutation hot spots for genetic diseases.
Although most cases of NDI are diagnosed within the first year of life, some are diagnosed later because of milder symptoms. Especially, several missense mutations have been related to mild phenotypes of NDI [bib_ref] Functional characterization of the molecular defects causing nephrogenic diabetes insipidus in eight..., Pasel [/bib_ref] [bib_ref] Clinical overview of nephrogenic diabetes insipidus based on a nationwide survey in..., Fujimoto [/bib_ref] [bib_ref] AVPR2 variants and mutations in nephrogenic diabetes insipidus: review and missense mutation..., Spanakis [/bib_ref]. Among our cases, Case 2 with p.R106C had poor weight gain from 4 months of age. Although he did not develop severe dehydration, he had frequent episodes of mild to moderate fever until the diagnosis was made. Pediatricians should keep in mind the possibility of mild NDI during the differential diagnosis of fever with unknown etiology.
As a previous in vitro study showed that p.R106C retained a slight capacity for production of cAMP in response to AVP, p.R106C is thought to cause less severe NDI : Analysis of X-chromosome inactivation. Arrows indicate the peak fluorescence intensity of the androgen CAG repeat on each X chromosome. Samples were treated with or without HpaII, and fluorescence intensities between treated and untreated samples were compared. e calculated X-inactivation percentage is shown at the bottom of each column. [bib_ref] Identification of mutations in the arginine vasopressin receptor 2 gene causing nephrogenic..., Chen [/bib_ref]. Two patients with p.R106C reported by Pasel et al. [bib_ref] Functional characterization of the molecular defects causing nephrogenic diabetes insipidus in eight..., Pasel [/bib_ref] had high basal urine osmolality and partial response to AVP administration. Chen et al. [bib_ref] Identification of mutations in the arginine vasopressin receptor 2 gene causing nephrogenic..., Chen [/bib_ref] also reported that an NDI patient with p.R106C had normal urine and plasma osmolality and plasma electrolytes, similar to our patient (Case 2).
It is thought that the development of NDI in female carriers is due to skewed X-chromosome inactivation of the normal allele. In one Japanese study, 25% of female carriers developed NDI [bib_ref] Hereditary nephrogenic diabetes insipidus in Japanese patients: analysis of 78 families and..., Sasaki [/bib_ref]. A recent Spanish study showed a frequency of NDI in female carriers of 50% [bib_ref] Novel mutations associated with nephrogenic diabetes insipidus. A clinical-genetic study, Garcia Castano [/bib_ref]. In our studies, skewed X-inactivation was not found in Case 4. is can be explained by different degrees of X-inactivation ratios in each organ as suggested previously [bib_ref] Age-and tissuespecific variation of X chromosome inactivation ratios in normal women, Sharp [/bib_ref] [bib_ref] Correlation between clinical phenotypes and X-inactivation patterns in six female carriers with..., Satoh [/bib_ref].
In conclusion, we report Japanese NDI patients with AVPR2 mutations (p.K100KfsX91, p.W99X, and p.R106C).
ere are broad phenotypic differences among the patients with the same type of mutations. In female patients, skewed X-inactivation is not always detected in peripheral blood lymphocytes. As some NDI cases do not show severe dehydration, they should be checked not only electrolyte but also their urine and serum osmolality once they suspected NDI. Low-grade fever and poor body weight gain in infant can be the clue of diagnosing mild NDI.
## Conflicts of interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.
# Authors' contributions
Noriko Namatame-Ohta designed the study and wrote the initial draft of the manuscript. Shuntaro Morikawa and Toshihiro Tajima contributed to analysis and interpretation of data and assisted in the preparation of the manuscript. All other authors have contributed to data collection and interpretation and critically reviewed the manuscript. e final version of the manuscript was approved by all authors.
[fig] Figure 2: Results of sequencing of AVPR2 mutations. Sequence chromatograms of AVPR2 in patients and their mothers. Arrows indicate mutation sites. [/fig]
[table] Table 1: Clinical characteristics of 4 patients with congenital nephrogenic diabetes insipidus and AVPR2 mutations. [/table]
[table] Table 2: Results of the water deprivation test in Case 2. [/table]
|
Effects of Glucocorticoids on Postoperative Neurocognitive Disorders in Adult Patients: A Systematic Review and Meta-Analysis
Background: Postoperative neurocognitive disorders (PNDs) is common among surgical patients, however, the effect of glucocorticoids for preventing PNDs is not clear. This review aims to evaluate the effect of glucocorticoids on the incidence of PNDs in adult patients undergoing surgery.Methods: The databases of PubMed/Medline, Embase, the Cochrane Library, and Web of science were searched for all available randomized controlled trials (RCTs) from inception to April 30, 2022. RCTs comparing the effect of glucocorticoids with placebo on the incidence of PNDs in adult surgical patients (≥18 years old) were eligible. Subgroup analyses and meta-regressions were performed to evaluate sources of clinical heterogeneity. The level of certainty for main outcomes were assessed by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology.Results: Eleven trials with a total of 10,703 patients were identified. Compared with the control group, glucocorticoids did not reduce the incidence of PNDs (RR: 0.84, 95% CI: 0.67 to 1.06, P = 0.13, GRADE = moderate). Secondary analyses for primary outcome did not change the result. In addition, the length of ICU stay was decreased in glucocorticoids group (RR: −13.58, 95% CI: −26.37 to −0.80, P = 0.04, GRADE = low). However, there were no significant differences between groups with regards to the incidence of postoperative infection (RR: 0.94, 95% CI: 0.84 to 1.06, P = 0.30, GRADE = moderate), blood glucose level (RR: 1.05, 95% CI: −0.09 to 2.19, P = 0.07, GRADE = low), duration of mechanical ventilation (RR: −2.44, 95% CI: −5.47 to 0.59, P = 0.14, GRADE = low), length of hospital stay (RR: −0.09, 95% CI: −0.27 to 0.09, P = 0.33, GRADE = moderate) and 30-day mortality (RR: 0.86, 95% CI: 0.70 to 1.06, P = 0.16, GRADE = moderate). Xie et al.Glucocorticoids for Postoperative Neurocognitive DisordersConclusions: This meta-analysis suggests that perioperative administration of glucocorticoids may not reduce the incidence of PNDs after surgery. The effect of glucocorticoids on decreased length of ICU stay needs further researches. Future high-quality trials using acknowledged criteria and validated diagnostic tools are needed to determine the influence of glucocorticoids on long-term PNDs.Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_ record.php?ID=CRD42022302262, identifier: CRD42022302262.
# Introduction
Postoperative neurocognitive disorders (PNDs) is an overarching term that includes postoperative delirium and postoperative cognitive dysfunction (POCD) [bib_ref] Cognitive decline associated with anesthesia and surgery in older patients, Vacas [/bib_ref]. According to the Perioperative Cognition Nomenclature Working Group in 2018 [bib_ref] Recommendations for the nomenclature of cognitive change associated with anaesthesia and surgery-2018, Evered [/bib_ref] , postoperative delirium is an acute state of cognitive impairment occurring within days after surgery and up to 1 week or until discharge, while POCD is a prolonged cognitive decline usually detected between 30 days and 12 months postoperatively. It has been reported that postoperative delirium occurred in 10-60% of elderly surgical patients, varying by surgical procedures (American Geriatrics Society Expert Panel on Postoperative Delirium in Older , and the incidence of POCD is approximately 25-40% [bib_ref] The potential role of the NLRP3 inflammasome activation as a link between..., Wei [/bib_ref]. Old age, low educational levels, poor preoperative cognitive function, perioperative pain and complicated surgery process are thought to be risk factors of PNDs [bib_ref] New biomarkers of postoperative neurocognitive disorders, Xie [/bib_ref] [bib_ref] Preoperative frailty predicts postoperative neurocognitive disorders after total hip joint replacement surgery, Evered [/bib_ref] [bib_ref] Sleep, pain, and cognition: modifiable targets for optimal perioperative brain health, O'gara [/bib_ref]. PNDs are the very common and severe postoperative neurological complications with poor outcomes, including increasing the length of hospital stay, mortality, and the risk of long-term cognitive impairment. These would cause significant clinical, social, and financial burdens on the patients and their communities [bib_ref] Predictors of cognitive dysfunction after major noncardiac surgery, Monk [/bib_ref] [bib_ref] Delirium in elderly people, Inouye [/bib_ref] [bib_ref] Economic burden of postoperative neurocognitive disorders among US medicare patients, Boone [/bib_ref].
Improving cognitive outcome after surgery, therefore, is an important objective for anesthesiologists and surgeons. To date, there have been no compelling pharmacologic interventions to limit the incidence or severity of PNDs [bib_ref] State of the clinical science of perioperative brain health: report from the..., Mahanna-Gabrielli [/bib_ref] [bib_ref] Effect of early cognitive interventions on delirium in critically ill patients: a..., Deemer [/bib_ref]. Dexmedetomidine, an anesthetic agent with neural anti-inflammatory effects, has been found to show promise for PNDs prevention [bib_ref] The effect of the timing and dose of dexmedetomidine on postoperative delirium..., Lee [/bib_ref] [bib_ref] Perioperative dexmedetomidine supplement decreases delirium incidence after adult cardiac surgery: a randomized,..., Likhvantsev [/bib_ref]. However, it has common side effects such as bradycardia and hypotension [bib_ref] Perioperative dexmedetomidine reduces delirium after cardiac surgery: a meta-analysis of randomized controlled..., Wu [/bib_ref] [bib_ref] Dexmedetomidine improves early postoperative neurocognitive disorder in elderly male patients undergoing thoracoscopic..., Shi [/bib_ref] [bib_ref] The effect and optimal dosage of dexmedetomidine plus sufentanil for postoperative analgesia..., Zhao [/bib_ref] , and the evidence to support this effect is limited [bib_ref] Meta-analysis of randomised controlled trials of perioperative dexmedetomidine to reduce delirium and..., Sanders [/bib_ref]. For non-pharmacologic approaches, cognitive prehabilitation, physical activity, and management of hypertension and diabetes seem to be effective to improve cognitive function [bib_ref] Effect of cognitive prehabilitation on the incidence of postoperative delirium among older..., Humeidan [/bib_ref] , but there is still a gap in their integration into pathways of care for patients [bib_ref] Home-based cognitive prehabilitation in older surgical patients: a feasibility study, Vlisides [/bib_ref] [bib_ref] Adherence to recommended practices for perioperative anesthesia care for older adults among..., Deiner [/bib_ref].
Proposed potential mechanisms for PNDs, including mitochondrial dysfunction, oxidative stress [bib_ref] Oxidative stress and mitochondrial dysfunction contributes to postoperative cognitive dysfunction in elderly..., Netto [/bib_ref] , synaptic damage [bib_ref] PGE2-EP3 signaling exacerbates hippocampusdependent cognitive impairment after laparotomy by reducing expression levels..., Xiao [/bib_ref] , and neurotrophic support impairment [bib_ref] Enriched environment attenuates surgery-induced impairment of learning, memory, and neurogenesis possibly by..., Fan [/bib_ref] are speculative, among which neuroinflammation is the most significantly concerned [bib_ref] Postoperative cognitive dysfunction in the aged: the collision of neuroinflammaging with perioperative..., Luo [/bib_ref]. It has been reported that surgery and anesthesia could lead the peripheral immune system to produce proinflammatory signals [bib_ref] Identification of a novel mechanism of blood-brain communication during peripheral inflammation via..., Balusu [/bib_ref] [bib_ref] Self-extracellular RNA acts in synergy with exogenous danger signals to promote inflammation, Noll [/bib_ref]. These inflammatory mediators could transfer into the brain through paraventricular areas of the blood-brain barrier (BBB) and stimulate microglia to produce proinflammatory factors, destroying synapses and neurons, thus causing neurotoxic symptoms and cognitive disorders [bib_ref] Peripheral inflammation and cognitive aging, Lim [/bib_ref] [bib_ref] Emerging roles of immune cells in postoperative cognitive dysfunction, Liu [/bib_ref].
Glucocorticoids are commonly used in the perioperative period to attenuate the inflammatory response [bib_ref] Perioperative single-dose glucocorticoid administration: pathophysiologic effects and clinical implications, Holte [/bib_ref] [bib_ref] Perioperative glucocorticoids in hip and knee surgery -benefit vs. harm? A review..., Lunn [/bib_ref]. And they can alleviate the inflammation by inhibiting prostaglandin production [bib_ref] Antiinflammatory action of glucocorticoidsnew mechanisms for old drugs, Rhen [/bib_ref] , activating endothelial nitric oxide synthetase [bib_ref] Acute cardiovascular protective effects of corticosteroids are mediated by non-transcriptional activation of..., Hafezi-Moghadam [/bib_ref] , and decreasing the stability of mRNA for genes for inflammatory proteins [bib_ref] Decreased mRNA stability as a mechanism of glucocorticoid-mediated inhibition of vascular endothelial..., Gille [/bib_ref] [bib_ref] Dexamethasone causes sustained expression of mitogen-activated protein kinase (MAPK) phosphatase 1 and..., Lasa [/bib_ref] [bib_ref] Control of the expression of inflammatory response genes, Saklatvala [/bib_ref]. Evaluating whether the perioperative administration of glucocorticoids is helpful in preventing cognitive decline could promote targeted preventive and therapeutic interventions. Therefore, in recent years several studies have investigated the efficacy of glucocorticoids on cognitive disorders after anesthesia and surgery. However, their conclusions have been inconsistent. [bib_ref] Postoperative cognitive dysfunction after inhalational anesthesia in elderly patients undergoing major surgery:..., Qiao [/bib_ref] and [bib_ref] Effects of single low dose of dexamethasone before noncardiac and nonneurologic surgery..., Valentin [/bib_ref] investigated the effect of glucocorticoids on PNDs in elderly patients undergoing non-cardiac surgery. They found that the preventive administration of glucocorticoids could effectively reduce POCD. In contrast, [bib_ref] Intraoperative dexamethasone and delirium after cardiac surgery: a randomized clinical trial, Sauër [/bib_ref] demonstrated the opposite result, showing that intraoperative administration of glucocorticoids did not reduce the incidence of delirium after cardiac surgery. Besides, [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] studied the effect of glucocorticoids in patients suffering from facial spasms requiring microvascular decompression. They found that administering a higher dose of glucocorticoids increased the incidence of POCD in the early postoperative period. Therefore, we applied a systematic review and meta-analysis to explore the effect of perioperative glucocorticoids administration on the incidence of PNDs.
# Methods
This meta-analysis was conducted following the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [bib_ref] PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic..., Page [/bib_ref]. This study protocol was registered in PROSPERO database (CRD42022302262).
## Search strategy
The databases of Pubmed/Medline, Embase, the Cochrane Library/Central, and Web of science were systematically searched for all relevant studies from inception to April 30, 2022. The references of included researches were also examined. According to the search strategy, both MeSH terms and free terms were used. The following keyword search terms were used: glucocorticoids, cognitive disorders,
## Study selection criteria
Studies restricted to randomized controlled trials (RCTs) in adult surgical patients (≥18 years old). All published full-article RCTs compared the effect of glucocorticoids with placebo or equal volume of normal saline (NS) on the incidence of PNDs were eligible for inclusion. Language restriction was not applied. Pediatric surgery, non-intravenous administration of glucocorticoids, no available assessment tools, and animal experiments were excluded from this meta-analysis.
## Data extraction
Data extraction and quality assessment were completed by two authors (XX and RG) independently. One author (XX) entered the information into the table and checked for consistency and completeness. Disagreements on data extraction and quality assessment were handled by discussion or reviewed by the third author (CC). The extracted data and information were as follows: first author, year of publication, surgery type, patient age, glucocorticoids type, timing and dose, control group, PNDs type, and cognitive assessment timing and methods. In addition, the following adverse events were extracted as well, including PNDs, infection, blood glucose level, duration of mechanical ventilation, length of ICU and hospital stay, and 30-day mortality.
## Quality assessment
Quality assessment of included RCTs was performed according to the second version of the Cochrane risk-of-bias tool for RCTs (RoB 2.0) [bib_ref] RoB 2: a revised tool for assessing risk of bias in randomised..., Sterne [/bib_ref]. There are seven sections of this assessment, random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases. Each section was classified into the low, high, or unclear risk of bias.
## Endpoints
The primary endpoint of this meta-analysis was the incidence of PNDs. The secondary outcomes were the incidence of postoperative infection, blood glucose level, duration of mechanical ventilation, the length of ICU and hospital stay, and postoperative 30-day mortality.
# Statistical analysis
For dichotomous data (incidence of PNDs, postoperative infection, and 30-day mortality), the Mantel-Haenszel method was used to combine outcomes and risk ratio (RR) with 95% confidence intervals (CI) were calculated. Concerning continuous variables (blood glucose level, duration of mechanical ventilation, and the length of ICU and hospital stay), the Inverse-Variance method was used, and mean difference (MD) or standardized mean difference (SMD) with 95% CI were calculated. The I 2 statistics used to evaluate heterogeneity were divided into the following three levels [bib_ref] The effects of clinical and statistical heterogeneity on the predictive values of..., Melsen [/bib_ref] : low (I 2 < 50%), moderate (I 2 = 50-75%) and high (I 2 > 75%). When the heterogeneity was low, we used fixed effects model to pooled the data; otherwise, we chose random effects model.
To find sources of heterogeneity, subgroup analyses were conducted according to the type of PNDs (postoperative delirium and POCD), the type of glucocorticoids, surgery, dose, and age. In subgroup analysis for the dose of glucocorticoids, the trials were stratified into three broad dose groups: low dose group if the total dose of glucocorticoid used was ≤30 mg prednisolone or equivalent, medium dose group if the total dose used was between 30 and 100 mg prednisolone or equivalent, and high dose group if the total dose used was >100 mg prednisolone or equivalent. These cut points were chosen according to clinical practice [bib_ref] Pharmacokinetics and pharmacodynamics of systemically administered glucocorticoids, Czock [/bib_ref]. For studies that used dexamethasone, methylprednisolone, or hydrocortisone, the total dose of glucocorticoid used was converted to an equivalent dose of prednisolone with similar glucocorticoid effect. The dose conversion factors for dexamethasone, methylprednisolone, and hydrocortisone to prednisolone were 6, 1.25, and 0.25, Frontiers in Aging Neuroscience | www.frontiersin.org respectively (https://clincalc.com/corticosteroids/). In subgroup analysis for patient age, we calculated the mean age of the study population as the basis of classification. Besides, Hartung-Knapp adjusted meta-regressions were performed to find interactions between variables if subgroup analyses could not explain sources of heterogeneity.
Publication bias was assessed through visual inspection of funnel plots and Egger's test to evaluate the small-study effects. The influence of a potential publication bias on findings was explored by using the Duval and Tweedie trim-and-fill procedure. Sensitivity analyses were performed by excluding high-risk studies evaluated by RoB 2.0 or omitting one study each time to detect robustness of the pooled results. Finally, the level of certainty for main outcomes were assessed by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology [bib_ref] Assessing and presenting summaries of evidence in Cochrane reviews, Langendam [/bib_ref]. P < 0.05 was considered statistically significant for all tests. All data analyses were performed by Revman 5.3. and Stata 16.
# Results
## Study characteristics
According to the search strategy, a total of 466 trials were identified. Among them, 128 studies were removed due to duplication, and the other 327 studies were excluded based on inclusion and exclusion criteria. Ultimately, 11 RCTs, including 10,703 patients, were included in this meta-analysis. The selection process flow chart was shown in [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref] , and methodological quality assessment was conducted according to the RoB 2.0, and the result was summarized in [fig_ref] FIGURE 2 |: Risk of bias assessment was undertaken for each included trial according to... [/fig_ref]. The major characters of these eligible studies were extracted and presented in [fig_ref] TABLE 1 |: Characteristics of the 11 included trails [/fig_ref]. Seven trials were cardiac surgery [bib_ref] Glucocorticoid-endocannabinoid interaction in cardiac surgical patients: relationship to early cognitive dysfunction and..., Hauer [/bib_ref] [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Effects of dexamethasone on cognitive decline after cardiac surgery: a randomized clinical..., Ottens [/bib_ref] [bib_ref] Intraoperative dexamethasone and delirium after cardiac surgery: a randomized clinical trial, Sauër [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] [bib_ref] Impact of methylprednisolone on postoperative quality of recovery and delirium in the..., Royse [/bib_ref] , one trial was neurologic surgery [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] , and three trials were non-cardiac, non-neurologic surgery, including one laparoscopic gastrointestinal surgery [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] and two hip fracture surgery [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref]. Furthermore, six studies used dexamethasone as intervention [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] [bib_ref] Effects of dexamethasone on cognitive decline after cardiac surgery: a randomized clinical..., Ottens [/bib_ref] [bib_ref] Intraoperative dexamethasone and delirium after cardiac surgery: a randomized clinical trial, Sauër [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref] , four studies used methylprednisolone [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Impact of methylprednisolone on postoperative quality of recovery and delirium in the..., Royse [/bib_ref] [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] , and one study used hydrocortisone [bib_ref] Glucocorticoid-endocannabinoid interaction in cardiac surgical patients: relationship to early cognitive dysfunction and..., Hauer [/bib_ref]. All of the included studies used normal saline as a placebo. Besides, eight studies investigated postoperative delirium [bib_ref] Glucocorticoid-endocannabinoid interaction in cardiac surgical patients: relationship to early cognitive dysfunction and..., Hauer [/bib_ref] [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Intraoperative dexamethasone and delirium after cardiac surgery: a randomized clinical trial, Sauër [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Impact of methylprednisolone on postoperative quality of recovery and delirium in the..., Royse [/bib_ref] [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref] [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] while the other three trials investigated POCD [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] [bib_ref] Effects of dexamethasone on cognitive decline after cardiac surgery: a randomized clinical..., Ottens [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref]. The timing and dose of glucocorticoids administration and cognitive assessment methods were varied between included studies.
## Primary outcome
The overall pooled result showed that glucocorticoids did not decrease the incidence of PNDs compared to the controls (RR: 0.84, 95% CI: 0.67 to 1.06, P = 0.13, I 2 = 57%) [fig_ref] FIGURE 3 |: Forest plot for the incidence of PNDs [/fig_ref]. Sensitivity analyses were performed by excluding the high-risk study [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] or omitting one study each time from included studies, and the pooled result was still robust [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref]. Meanwhile, no significant publication bias was evidenced by visual inspection of funnel plot [fig_ref] FIGURE 4 |: Funnel plot of the primary outcome [/fig_ref] and Egger's test [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref] for the effect of glucocorticoid administration on PNDs.
Similarly, the finding was consistent in subgroup analyses between postoperative delirium (RR: 0.78, 95% CI: 0.61 to 1.01, P = 0.05, I 2 = 44%) and POCD (RR: 1.00, 95% CI: 0.51 to 1.96, P = 1.00, I 2 = 77%) [fig_ref] FIGURE 5 |: Subgroup analysis of the incidence of PNDs for the type of PNDs [/fig_ref] , between dexamethasone (RR: 0.87, 95% CI: 0.59 to 1.27, P = 0.46, I 2 = 65%), methylprednisolone (RR: 0.72, 95% CI: 0.47 to 1.09, P = 0.12, I 2 = 65%) and hydrocortisone [fig_ref] FIGURE 6 |: Subgroup analysis of the incidence of PNDs for the type of glucocorticoids [/fig_ref]. However, subgroup analyses for glucocorticoids dose, surgery type and patient age showed the inconsistent results. There were significant differences in medium dose group (RR: 0.49, 95% CI: 0.33 to 0.73, P = 0.0005, I 2 = 0%) [fig_ref] FIGURE 2 |: Risk of bias assessment was undertaken for each included trial according to... [/fig_ref] , noncardiac, non-neurologic surgery group (RR: 0.52, 95% CI: 0.33 to 0.80, P = 0.003, I 2 = 0%) [fig_ref] FIGURE 3 |: Forest plot for the incidence of PNDs [/fig_ref] and mean age ≥70 years group (RR: 0.60, 95% CI: 0.43 to 0.85, P = 0.004, I 2 = 0%) [fig_ref] FIGURE 4 |: Funnel plot of the primary outcome [/fig_ref]. However, further meta-regressions showed that when glucocorticoids type, surgery type, patient age, and their interactions were entered as covariates in models, there were no significant differences between glucocorticoid group and placebo group on the incidence of PNDs (Supplementary [fig_ref] TABLE 2 |: GRADE evidence for main outcomes [/fig_ref].
## Secondary outcomes
Of the 11 studies included in this meta-analysis, four studies [bib_ref] Glucocorticoid-endocannabinoid interaction in cardiac surgical patients: relationship to early cognitive dysfunction and..., Hauer [/bib_ref] [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] compared the length of ICU stay between groups and the glucocorticoid group significantly reduced the length stay in ICU (RR: −13.58, 95% CI: −26.37 to −0.80, P = 0.04, I 2 = 86%) [fig_ref] FIGURE 7 |: Forest plot of length of ICU stay [/fig_ref]. However, there were no significant differences in postoperative infection [five trials [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref] [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] ; RR: 0.94, 95% CI: 0.84 to 1.06, P = 0.30, I 2 = 26%] [fig_ref] FIGURE 8 |: Forest plot of postoperative injection [/fig_ref] , blood glucose level [two trials [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] ; RR: 1.05, 95% CI: −0.09 to 2.19, P = 0.07, I 2 = 61%] [fig_ref] FIGURE 9 |: Forest plot of postoperative blood glucose level [/fig_ref] , duration of mechanical ventilation [two trials [bib_ref] Glucocorticoid-endocannabinoid interaction in cardiac surgical patients: relationship to early cognitive dysfunction and..., Hauer [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] ; RR: −2.44, 95% CI: −5.47 to 0.59, P = 0.14, I 2 = 0%] [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref] , length of hospital stay [six trials [bib_ref] Prophylaxis of dexamethasone protects patients from further post-operative delirium after cardiac surgery:..., Mardani [/bib_ref] [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref] [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] ; RR: −0.09, 95% CI: −0.27 to 0.09, P = 0.33, I 2 = 11%] [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref] and 30-day mortality [four trials [bib_ref] Methylprednisolone in patients undergoing cardiopulmonary bypass (SIRS): a randomised, double-blind, placebocontrolled trial, Whitlock [/bib_ref] [bib_ref] Effect of a single pre-operative 125 mg dose of methylprednisolone on postoperative..., Clemmesen [/bib_ref] [bib_ref] Steroids to reduce the impact on delirium (STRIDE): a double-blind, randomised, placebo-controlled..., Kluger [/bib_ref] [bib_ref] The effect of preoperative methylprednisolone on postoperative delirium in older patients undergoing..., Xiang [/bib_ref] ; RR: 0.86, 95% CI: 0.70 to 1.06, P = 0.16, I 2 = 0%] [fig_ref] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials [/fig_ref].
## Level of certainty for outcomes (grade)
Basing on GRADE framework, we evaluated the level of certainty for our main outcomes. The quality of these outcomes varied from low to moderate and the detailed information were shown in [fig_ref] TABLE 2 |: GRADE evidence for main outcomes [/fig_ref].
# Discussion
This meta-analysis suggests that perioperative glucocorticoids administration does not reduce the incidence of PNDs. Subgroup analyses and meta-regressions considering potential variables such as PNDs type, glucocorticoids type, dose, surgery type and patient age remained no difference in the outcome. Besides, glucocorticoids infusion was associated with a shorter length of ICU stay, while the incidence of postoperative infection, blood glucose level, duration of mechanical ventilation, length of hospital stay, and 30-day mortality did not differ significantly between groups.
Neuroinflammation has become a key hallmark of neurological complications including PNDs [bib_ref] Neuroinflammation and perioperative neurocognitive disorders, Subramaniyan [/bib_ref]. Perioperative glucocorticoid administration has been used in different surgical settings to counter the detrimental effect of inflammation induced by surgery and anesthesia [bib_ref] High-dose preoperative glucocorticoid for prevention of emergence and postoperative delirium in liver..., Awada [/bib_ref]. Several studies have shown that the preoperative administration of glucocorticoids reduces peripheral inflammatory markers in hepatic surgery [bib_ref] Systematic review and meta-analysis of the effect of perioperative steroids on ischaemiareperfusion..., Orci [/bib_ref] [bib_ref] Use of pre-operative steroids in liver resection: a systematic review and meta-analysis, Richardson [/bib_ref]. However, the effect of glucocorticoids on PNDs was not observed despite pooling data for more than 10,000 randomized participants, with overall low risk of bias across studies in our review. Several reasons may account for this result. First, although excessive neuroinflammation leads to injury and death of neural elements, a large body of literatures have demonstrated that proper neuroinflammatory response can benefit outcomes of central nerve system [bib_ref] The benefits of neuroinflammation for the repair of the injured central nervous..., Yong [/bib_ref]. For example, neuroinflammation can promote neurogenesis [bib_ref] Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities..., Ziv [/bib_ref] , facilitate axonal regeneration [bib_ref] Macrophages can modify the nonpermissive nature of the adult mammalian central nervous..., David [/bib_ref] , and is critical for remyelination [bib_ref] A silver lining of neuroinflammation: Beneficial effects on myelination, Goldstein [/bib_ref]. It's crucial to suppress the excessive inflammation that mediates damage without inhibiting the repairment effect from preventing PNDs. Second, the genesis of PNDs is multifactorial that other factors except for neuroinflammation may play great roles in the development of neurocognitive deficit [bib_ref] Interventions for preventing delirium in hospitalised non-ICU patients, Siddiqi [/bib_ref]. Third, prolonged exposure to high concentrations of glucocorticoids can be toxic to neural structures, especially the glucocorticoid receptor-rich hippocampus [bib_ref] Glucocorticoids and hippocampal atrophy in neuropsychiatric disorders, Sapolsky [/bib_ref].
To explore sources of heterogeneity, we performed subgroup analyses based on PNDs type and glucocorticoids type, and the outcome remained no difference. Although subgroup analyses about glucocorticoids dose, surgery type and patient age seemed significantly different, meta-regressions were further conducted to understand the interactions between these variables on our outcomes and the results changed to no difference. Here what should be noted was that the subgroup of age was classified according to the mean age values in studies, which might induce some misclassification of accurate age. In general, these secondary analyses suggest that the genesis of PNDs is multifactorial that only administration of glucocorticoids may not significantly reduce the incidence of neurocognitive disorders after surgery. Besides, the impact of patient, surgery, or other variables, both measured and unmeasured, on the PNDs development likely far outweighs the impact of glucocorticoids.
PNDs is a summarized term encompassing postoperative delirium, a most pronounced and acute postoperative form, and POCD which is described as a long-term neurocognitive impairment [bib_ref] Recommendations for the nomenclature of cognitive change associated with anaesthesia and surgery-2018, Evered [/bib_ref]. Delirium and POCD previously were considered distinct entities, but recent data has suggested an underlying relationship between them [bib_ref] Postoperative neurocognitive disorders, Olotu [/bib_ref]. Several risk factors are common to both postoperative delirium and POCD, [bib_ref] Intraoperative high-dose dexamethasone for cardiac surgery: a randomized controlled trial, Dieleman [/bib_ref] were different and not included here.
and animal studies raise the possibility that neuroinflammation may play a role in both of these states [bib_ref] Postoperative delirium and postoperative cognitive dysfunction: two sides of the same coin?, Devinney [/bib_ref] [bib_ref] Postoperative delirium and postoperative cognitive dysfunction: overlap and divergence, Daiello [/bib_ref]. Therefore, in this metaanalysis we choose PNDs as the endpoint which in other words, combining postoperative delirium and POCD in the composite outcome, is acceptable from perspective of pathogenesis. The diagnosis of PNDs, especially POCD, is complex requiring neuropsychological tests which are varied in studies. A recent systematic review noted that in 274 existing studies of POCD, diagnosis was based on 259 different cognitive assessment tools [bib_ref] Methodology of measuring postoperative cognitive dysfunction: a systematic review, Borchers [/bib_ref]. Moreover, neuropsychological tests have been undertaken at variable time intervals after anesthesia and surgery. In 2018, [bib_ref] Recommendations for the nomenclature of cognitive change associated with anaesthesia and surgery-2018, Evered [/bib_ref] clarified that postoperative delirium was defined as occurring in hospital and up to 1 week post-procedure or until discharge, while POCD persisted for more than 30 days but <12 months following anesthesia and surgery. However, in trials included in this meta-analysis, [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] and [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] examined POCD in postoperative day 5 and 6, respectively, in which delirium is usually to be assessed; while only [bib_ref] Effects of dexamethasone on cognitive decline after cardiac surgery: a randomized clinical..., Ottens [/bib_ref] examined POCD at 1 month after surgery. In summary, heterogeneities of assessment tools, diagnostic criteria, and follow-up time limit the interpretation of existing data surrounding PNDs. Besides, there were only three studies [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] [bib_ref] Effects of dexamethasone on cognitive decline after cardiac surgery: a randomized clinical..., Ottens [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] examining POCD in our meta-analysis, and two of them [bib_ref] Higher dose dexamethasone increases early postoperative cognitive dysfunction, Fang [/bib_ref] [bib_ref] Effects of dexamethasone on early cognitive decline after cardiac surgery: a randomised..., Glumac [/bib_ref] assessing this entity in the early postoperative day. Thus, the effect of glucocorticoids on long-term PNDs is still unclear. PNDs is associated with long-term sequelae including ongoing impaired cognition, increased risk of dementia, increased mortality, and premature retirement from work [bib_ref] Long-term consequences of postoperative cognitive dysfunction, Steinmetz [/bib_ref]. Interventions to mitigate these sequelae may therefore provide clinical and economic benefit in the long run. Further trials are needed to assess POCD using uniform criteria and validated diagnostic tools and then evaluate the effects of glucocorticoids on the incidence of long-term PNDs.
In our meta-analysis, glucocorticoids significantly reduced the length of ICU stay. However, there were only four studies examining this outcome with high heterogeneity. Besides, it is still obscure whether administration of glucocorticoids could impact the risk of postoperative infection, blood glucose level, length of hospital stay, duration of mechanical ventilation, and 30-day mortality. In the Dexamethasone for Cardiac Surgery (DECS) trial, [bib_ref] Intraoperative high-dose dexamethasone for cardiac surgery: a randomized controlled trial, Dieleman [/bib_ref] reported that intraoperative administration of dexamethasone for cardiac surgery was associated with higher postoperative glucose level, lower infection rate, decreased duration of mechanical ventilation, and reduced length of hospital stay. Because of high heterogeneity and limited studies in this meta-analysis, future high-quality researches are still needed to confirm these outcomes.
To our knowledge, this is the first systematic review and metaanalysis to comprehensively evaluate effects of glucocorticoids on PNDs. A meta-analysis by [bib_ref] Effects of dexamethasone on post-operative cognitive dysfunction and delirium in adults following..., Li [/bib_ref] studied effects of dexamethasone on postoperative cognitive dysfunction and delirium in adults following general anesthesia, which did not take other types of glucocorticoids into consideration. Another systematic review and meta-analysis performed by [bib_ref] Effects of glucocorticoids on postoperative delirium in adult patients undergoing cardiac surgery:..., Liu [/bib_ref] was about effects of glucocorticoids on postoperative delirium in adult patients undergoing cardiac surgery. Similarly, POCD, another form of PNDs mentioned above, and non-cardiac surgery patients were not included in this study. In our review, we thought through PNDs type, glucocorticoids type, and surgery type to make conclusions as rigorous as possible.
There are still several potential limitations in this metaanalysis. First, we studied three types of glucocorticoids and two types of PNDs in cardiac, neurologic, and non-cardiac nonneurologic surgery, so there was potential heterogeneity such as methods of diagnosis and dosages of glucocorticoids, which may affect the precision and reliability of the results. Second, most of these included studies excluded patients with preexisting cognitive impairment, and children were also excluded from this meta-analysis. Therefore, the extrapolation of this meta-analysis was limited. Third, some studies that contained our interested second outcomes, but not PNDs data, were excluded, thus influencing this meta-analysis's completeness of secondary outcomes. Further, more structured and standardized perioperative glucocorticoids protocols and uniform definition and assessment tools of PNDs may be necessary to accurately evaluate the effect of glucocorticoids on PNDs.
# Conclusion
In summary, our findings suggest that perioperative administration of glucocorticoids does not reduce the incidence of PNDs, regardless of PNDs type, glucocorticoids type, dose, surgery type and patient age. The effect of glucocorticoids on decreased length of ICU stay needs further researches. Future high-quality trials using acknowledged criteria and validated diagnostic tools are needed to determine the influence of glucocorticoids on long-term PNDs.
# Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.
# Author contributions
XX and RG wrote the manuscript and collected the data. XX, RG, TZ, and CC chose the topic. HC, XZ, and XC contributed to the conception. CZ and CL searched the literature. XX, RG, HC, XZ, and XC analyzed the data. TZ and CC made final decisions. All authors have read and approved the final manuscript.
# Acknowledgments
We thank the authors of the primary studies for providing their data and other critical information. Additionally, we thank all participants for their valuable contributions to this article.
# Supplementary material
[fig] FIGURE 1 |: Flow chart of search strategy to identify the eligible randomized controlled trials. [/fig]
[fig] FIGURE 2 |: Risk of bias assessment was undertaken for each included trial according to the Cochrane Risk of Bias Methods. Frontiers in Aging Neuroscience | www.frontiersin.org [/fig]
[fig] FIGURE 3 |: Forest plot for the incidence of PNDs. [/fig]
[fig] FIGURE 4 |: Funnel plot of the primary outcome (the incidence of PNDs). [/fig]
[fig] FIGURE 5 |: Subgroup analysis of the incidence of PNDs for the type of PNDs. [/fig]
[fig] FIGURE 6 |: Subgroup analysis of the incidence of PNDs for the type of glucocorticoids. [/fig]
[fig] FIGURE 7 |: Forest plot of length of ICU stay. [/fig]
[fig] FIGURE 8 |: Forest plot of postoperative injection. [/fig]
[fig] FIGURE 9 |: Forest plot of postoperative blood glucose level. [/fig]
[fig] FIGURE 10 |: Forest plot of duration of mechanical ventilation. [/fig]
[fig] FIGURE 11 |: Forest plot of length of hospital stay. [/fig]
[fig] FIGURE 12 |: Forest plot of postoperative 30-day mortality. [/fig]
[fig] FUNDING: This work was supported by the National Natural Science Foundation of China (Nos. 82171185 and 81870858 to CC); The National Key R&D Program of China (No. 2018YFC2001800 to TZ) and the National Natural Science Foundation of China (No. 81671062 to TZ); China Postdoctoral Science Foundation (Grant No. 2020M673234 to RG), Post-doctoral Research Project, West China Hospital, Sichuan University (Grant No. 2020HXBH022 to RG). [/fig]
[fig] The: Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnagi. 2022.939848/full#supplementary-material Supplementary Material | Search strategy. Supplementary Figure 1 | Sensitivity analysis of the incidence of PNDs by excluding each study individually. Supplementary Figure 2 | Subgroup analysis of the incidence of PNDs for the dose of glucocorticoids. Supplementary Figure 3 | Subgroup analysis of the incidence of PNDs for the type of surgery. Supplementary Figure 4 | Subgroup analysis of the incidence of PNDs based on age classification.Supplementary [/fig]
[table] TABLE 1 |: Characteristics of the 11 included trails. [/table]
[table] TABLE 2 |: GRADE evidence for main outcomes. [/table]
|
Mental Health, Sexual Orientation, and Perceived Social Network Support in Relation to Hazardous Alcohol Consumption Among Active Duty Military Men
Knowledge surrounding perceived network support and alcohol consumption among active duty U.S. military personnel is limited, particularly among sexual minorities.We sought to determine the correlates of hazardous alcohol consumption and whether perceived network support moderated the relationship between sexual orientation and Alcohol Use Identification Test (AUDIT-C) score.The sample comprised cisgender men currently serving in the U.S. military (N = 292). Participants were recruited through respondent-driven sampling and completed an online survey. Logistic regression analysis evaluated associations between positive AUDIT-C with sociodemographic characteristics (including sexual orientation), military service, mental health, and perceived social network support. Interaction analysis assessed the moderating effect of perceived network support on sexual orientation and AUDIT-C.Among study participants, 52.7% (154/292) had positive AUDIT-C, while 65.4% (191/292) self-identified as heterosexual/straight and 34.6% (101/292) identified as gay or bisexual. In adjusted analysis, positive AUDIT-C was associated with increased post-traumatic stress disorder symptomatology (adjusted odds ratio [adjOR] 1.03; 95% CI [1.00, 1.06]; p = .019) and high perceived network support (adjOR 1.85; 95% CI [1.04, 3.29]; p = .036), while mental health service utilization had reduced odds of positive AUDIT-C (adjOR 0.40; 95% CI [0.20, 0.78]; p = .007). In interaction analysis, high perceived network support was associated with increased odds of positive AUDIT-C among sexual minority men (adjOR 3.09; 95% CI [1.21, 7.93]; p = .019) but not heterosexual men (adjOR 1.38; 95% CI [0.68, 2.81]; p = .37).Hazardous alcohol use was prevalent among all men in our sample. Perceived social network support may influence hazardous alcohol consumption, particularly among sexual minority servicemen. These findings suggest the potential role of tailored social network-based interventions to decrease hazardous alcohol use among military personnel.
# Introduction
Excessive alcohol use is a significant public health threat and has been associated with negative health-related outcomes as well as legal and workplace ramifications. Alcohol consumption is common among military servicemembers and largely integrated into military culture, for social purposes and to reward hard work. Combat exposure and deployments have been associated with increased alcohol consumption and the development of alcohol use disorders among military personnel. Alcohol may be used as a form of self-medication for underlying mental illness, such as depression or posttraumatic stress disorder (PTSD), or to cope with traumatic or stressful events.
Stressful military workplace dynamics may also contribute to excessive alcohol use. A subpopulation particularly vulnerable to such stressors are servicemembers identifying as sexual minorities (SM). Prior military policies, such as Don't Ask, Don't Tell, Don't Pursue (DADT), limited the ability for SM to openly serve, as sexual identity disclosure would result in a dishonorable discharge . This resulted in an environment where SM personnel would serve in secrecy, oftentimes concealing their sexual orientation in both the workplace and social settings. While DADT was repealed in 2011, many SM servicemembers continue to experience discrimination, bias, and accompanying mental health symptomologycausing this subpopulation to be vulnerable to alcohol use disorders and poorer mental health outcomes. Evidence demonstrates that psychological distress due to sexual orientation-based discrimination among civilian SM is associated with an increased risk of alcohol use. While data surrounding alcohol use among SM in the military is limited, a recent study has demonstrated that psychological distress related to sexual orientationbased discrimination mediates alcohol use among SM servicemembers. Among veteran populations, anxiety surrounding concealment of sexual orientation during military service has been associated with elevated rates of PTSD, depression, and alcohol use among SM veterans compared with their heterosexual counterparts.
Social networks represent an important form of social support and can be extremely influential in coping with stressful events, particularly among SM who often rely on social networks comprised of SM to cope with minority stress. Social networks can also reinforce negative behaviors. Studies have identified that group norms and attitudes can significantly influence behavior, particularly regarding alcohol use. Research within civilian social networks in the United States demonstrated that both the proportion of individuals who drink and the number of drinks consumed by network members strongly correlate with one's own alcohol use. In addition to actual consumption, perceived network drinking behaviors and attitudes are important determinants of alcohol use. This observation may be particularly relevant for military personnel. Overestimation of perceived drinking behaviors and norms has been associated with increased alcohol consumption among veteransand active duty military servicemembers.
Excessive alcohol consumption within the military is of significant public health importance as it results in negative outcomes that may reduce the overall fitness of the armed forces. A recent survey of U.S. military personnel demonstrated that 8.2% of servicemembers reported at least one serious consequence (e.g., arrest for driving under the influence, causing an automobile accident, having a physical altercation, intimate partner violence) and 6.1% reported work-related productivity loss due to alcohol use. SM may be more vulnerable to such negative consequences given higher levels of alcohol use reported in this population, with 37.6% SM and transgender active duty service members reporting binge drinking in the past month compared with 29.3% of their non-SM and nontransgender counterparts. While civilian social networks have an important impact on alcohol consumption, similar data involving the military is limited. Further, research involving SM military personnel has been hindered due to previous policies (e.g., DADT) that resulted in significant knowledge gaps surrounding health outcomes and experiences of this population. As a better understanding of factors associated with hazardous alcohol consumption is critical to the development of efficacious public health programs for harm reduction, this study seeks to evaluate the association of perceived network support, mental health, and sexual orientation with hazardous alcohol use among a cohort of active duty servicemen.
# Materials and methods
## Participants and procedures
Data for the present study come from the Military Acceptance Project, a Department of Defense-funded project designed to understand the acceptance and integration of active duty lesbian, gay, bisexual, and transgender (LGBT) military personnel. Respondent-driven sampling (RDS) was utilized to recruit LGBT and non-LGBT cohorts of active duty servicemembers. Initial seed participants were recruited through referrals from an expert advisory panel and military network contacts from primary investigators and study staff. When seed recruitment slowed, it was augmented by promotion through military-related social media and events, college campus organizations using flyers, and purchased advertising in diverse media outlets, including Facebook and the Military Times. Seeds were provided unique referral codes to share with peers, allow for tracking of recruitment chains, and ensure that no single group or platform yielded more than 20 eligible seeds at a time. All survey respondents were provided up to six unique referral codes to recruit additional participants. Recruitment lasted from August 2017 to March 2018.
Participants were eligible if they were at least 18 years old and active duty servicemembers of the U.S. Army, Navy, Marine Corps, or Air Force. Once screened eligible, participants were directed to a secure online survey using the Qualtrics platform (Qualtrics, Provo, UT, USA). A total of 991 individuals accessed the survey, of whom 709 were eligible and provided consent to participate. Of these, 165 individuals did not pass fraud detection measures or did not complete the survey, 220 individuals were excluded due to identifying as a woman or transgender, and three individuals were excluded due to missing RDS cluster data, resulting in a sample of 321 cisgender men. As the analysis was limited to cisgender men who responded to all variables of interest, 29 individuals were excluded due to missing data, resulting in an analytic sample of 292.
Women from the sample were not included in this analysis given gender differences in alcohol use and consequences. Studies suggest that the relative contribution of risk factors for alcohol use among men and women in the military may differ, such as gender differences observed with perceptions surrounding trauma, the prevalence of posttraumatic stress symptoms, the prevalence of depression, response to trauma, as well as the impact of combat exposures on PTSD and depressive symptoms. Due to concern that these gender-based differences in mental health and risk factors for alcohol use may confound the findings in this analysis, women were excluded from the present study.
Participants were provided a $25 e-gift card for completion of the survey (if off duty) and $10 e-gift card incentives for each referral who completed the survey. The study protocol was reviewed and approved by the Office of the Human Research Protection Program (OHRPP) at the University of California, Los Angeles (#18-000984) and the Human Subjects Protection Program (HSPP) at the University of Southern California (#UP-16-00070).
## Measures
Sociodemographic Characteristics. Sex assigned at birth was reported using one item: "What sex were you assigned at birth, (i.e., what sex is on your birth certificate)?" (response options: male or female). Gender identity was assessed using a single question: "What is your gender identity?" (response options: male, female, transgender male/trans man, transgender female/trans woman, genderqueer/gender nonconforming, and gender identity not listed-please specify). As the analysis was limited to cisgender men, individuals who responded "male" to the sex assigned at birth and gender identity questions were included. Sexual orientation was measured using a single item: "What is your sexual identity?" (response options: heterosexual or straight, gay or lesbian, bisexual, and sexual orientation not listed hereplease specify). Sexual orientation was recoded as a binary variable to assess those reporting a sexual minority identity (i.e., gay or bisexual) (reference group: heterosexual or straight).
Age was reported as a continuous variable. Racial and ethnic identity was assessed using a single question, with response options including Black or African American, Latino or Hispanic, White or Caucasian, Native American or Alaskan Native, Asian or Pacific Islander, multiracial, and other. Due to the small number of individuals reporting Native American/Alaskan Native, multiracial, or other, these responses were condensed into a single other category (reference group: White or Caucasian). Education was assessed using one item: "What is your highest level of education completed?" (response options: some high school, General Educational Development , high school diploma, some college, associate degree, bachelor's degree, master's degree, and doctorate . Due to the small number of individuals reporting doctoral degrees, education was recoded to GED/high school diploma (i.e., GED or high school diploma), some college/associate degree (i.e., some college or associate degree), bachelor's degree, and graduate school (master's degree or doctorate) (reference group: high school/GED).
## Characteristics of military service.
Respondents reported their military service branch (U.S. Air Force, U.S. Army, U.S. Marine Corps, or U.S. Navy) (reference group: U.S. Air Force). Current pay grade was assessed (responses ranged from E-1 to O-6) and listed by ascending order of rank based on prior work evaluating substance use among active duty personnel. Rank/paygrade for service members was grouped based on ascending order of rank and officer status: E1-E3 (junior enlisted), E4-E6 (middle enlisted), E7-E9 (senior enlisted), O1-O3 (junior commissioned officers), and O4-O6 (senior commissioned officers).
## Mental health: service utilization and ptsd symptomatology.
To assess receipt of mental health services, respondents were asked whether they had received counseling or mental health services by a military behavioral health provider in the past year (reference group: no). PTSD was measured using the PTSD Checklist for DSM-5 (PCL-5). The PCL-5 consists of a 20 item self-report scale assessing PTSD symptomatology, with a PCL-5 score of 33 or higher suggesting that the patient may benefit from PTSD treatmentNational Center for PTSD, n.d.). PCL-5 scores are reported as a continuous variable with higher PCL-5 scores consistent with higher levels of PTSD symptomatology.
Social Network Characteristics. A social network inventory was obtained, where participants provided information on up to five network members. Participants were asked to preferentially list social network members with military experience, followed by civilians if they did not have five network members who served. Average age of the respondent's social network members was calculated from the continuously reported age of each member. Network density was calculated from the proportion of actual connections divided by the total possible number of connections between network members. Frequency of interaction with network members was assessed with a single item: "Who do you talk to or see at least once per week?". As the number of nominated social network members varied by respondent, the frequency of interaction with network members was calculated as a proportion. To facilitate interpretation, responses were recoded as a binary variable based on whether the respondent interacted with the majority (i.e., greater than 50%) of network members at least once per week (reference category: less than 50%).
Perceived social network support was assessed with three questions: "Who do you talk to for mentorship or advice, such as about military or career issues?"; "Who can you count on to listen to you when you need to talk?"; and "Who do you go to when you need help or advice?". Responses to these questions were condensed into one variable by calculating the proportion of network members identified as potential sources of support (proportions were calculated as the number of nominated social network members varied by respondent). To facilitate interpretation of the calculated proportions of network support, proportions were dichotomized into a binary variable based on whether the respondent reported the majority (i.e., greater than 50%) of their network members as a potential source of support (reference group: less than 50%).
Alcohol Use. Alcohol use was evaluated using the Alcohol Use Identification Test (AUDIT-C), a validated screening tool to identify hazardous drinking. The AUDIT-C consists of three questions where responses are measured on a 12-point scale. Scores of 4 or more in men are considered positive and are suggestive of hazardous drinking behaviors. As the AUDIT-C is used as a clinical screening tool designed to screen for hazardous drinking based on a positive or negative screening result, responses were dichotomized based on positive or negative AUDIT-C.
# Data analysis
We described the association of sexual orientation, military service characteristics, mental health service utilization, and perceived social network support with our outcome of interest-hazardous alcohol use, defined as positive AUDIT-C. Descriptive statistics (frequency, median, interquartile range) were used to describe the sample. Distributional differences between predictor variables with the outcome of interest were evaluated using χ 2 analysis and Fisher's exact tests (where appropriate) for categorical predictors and Kruskal-Wallis tests for nonparametric, continuous variables. Bivariate analyses estimated the association of participant and network characteristics with positive AUDIT-C. Demographic characteristics and predictor variables with an alpha <0.05, or that were important conceptually based on the literature, such as military branch, rank/paygrade, receipt of mental health services, PTSD symptomatology, network density, age of network members, social network size, talking to or seeing network members at least once weekly, were included in multivariable regression models. Logistic regression analysis was used to calculate unadjusted odds ratios (unadj. OR) of selected variables with positive AUDIT-C. Mixed-effects multivariable logistic regression with 10 integration points was used to calculate adjusted ORs, controlling for RDS cluster membership. To evaluate the association of network support with hazardous drinking among SM men, an additional mixed-effects model was created that included an interaction term of network support with sexual orientation. All multivariable regression models adjusted for age, race/ ethnicity, education, sexual orientation, military branch, rank/paygrade, receipt of mental health services, PTSD, social network size, average age of network members, network density, interaction with network members at least weekly, and perceived network support. All analyses were conducted using Stata 15.1 (StataCorp, College Town, TX, USA).
# Results
## Sample characteristics
Of the 292 men included in the analytic sample, 52.7% (n = 154) had positive AUDIT-C, and 47.3% (n = 138) had negative AUDIT-C scores. Most men self-identified as heterosexual/straight (65.4%; n = 191) with 34.6% (n = 101) identifying as a SM (gay or bisexual). Median age was 26 years (range 19-52). Most men identified their race/ethnicity as White/Caucasian (59.3%; n = 173/292), with 17.1% (n = 50) identifying as African American/ Black, 13.0% (n = 38) Hispanic/Latino, 6.2% (n = 18) Asian or Pacific Islander, and 4.5% (n = 13) as multiracial/other. All four military branches were represented, with 35.6% (n = 104) serving in the U.S. Air Force, 36.3% (n = 106) in the Army, 12.7% (n = 37) in the Marine Corps, and 15.4% (n = 45) in the Navy. Most respondents were junior enlisted servicemen (28.4%; n = 83), middle enlisted servicemen (27.7%; n = 81), or junior commissioned officers (31.2%; n = 91). Descriptive statistics of the sample stratified by AUDIT-C score are listed in.
## Bivariate analysis of respondent and network characteristics
Individuals who identified their race/ethnicity as Asian or Pacific Islander had lower odds of positive AUDIT-C than White/Caucasian servicemen (unadj. OR 0.15; p = .004). Men serving in the U.S. Marine Corps had higher odds of hazardous drinking compared with those in the Air Force (unadj. OR 2.92; p = .011). Receipt of mental health services in the last year reduced the odds of hazardous alcohol use by almost half (unadj. OR 0.53; p = .029) compared with individuals who did not receive mental health care. Additionally, men who reported interacting with most network members at least weekly had higher odds of hazardous drinking (unadj. OR 1.91; p = .008) compared with those who did not. Results from the bivariate analysis are depicted in.
## Military experiences and mental health
Respondents who received mental health services in the last year had lower odds of hazardous alcohol use than those who did not (adj. OR 0.40; p = .007). Men with worsening PTSD symptomatology had 3% higher odds of hazardous drinking for each unit increase in PCL-5 score in both our main effects (adj. OR 1.03; p = .019) and interaction (adj. OR 1.03; p = .022) models. While respondents who served in the Marine Corps had higher odds of positive hazardous drinking than those in the Air Force in our unadjusted analysis, military branch was not associated with hazardous drinking after adjusting for covariates in our model. Rank/paygrade was not associated with hazardous drinking in any of our models. Age, education, network size, network density, and average age of network members were not associated with hazardous alcohol use in our main effects or interaction models. Results from the main effects and interaction multivariable regression models are in.
## Social network support
While significant in bivariate analysis, interacting with a majority of network members at least weekly was not associated with hazardous alcohol use in either multivariable model (adj. OR 1.37, p = .27; adj. OR 1.37, p = .27 for main effects and interaction, respectively) when compared with those with less frequent network interaction. In our main effects model, men who identified most network members as potential sources of support had almost twice higher odds of hazardous drinking compared with those with less supportive networks (adj. OR 1.85; p = .036). When evaluating the interaction of sexual orientation with social network support on hazardous alcohol use, SM men who perceived the majority of their network as supportive had over three times higher odds of hazardous drinking (adj. OR 3.09; p = .019) compared with heterosexual men who identified less than half of their network as supportive. Perceived support from most network members was not associated with hazardous drinking among men who identified as heterosexual/straight (adj. OR 1.38; p = .37) when compared with heterosexual men with less supportive networks.
# Discussion
We sought to determine if perceived social network support, sexual orientation, and mental health experiences were associated with hazardous alcohol use in this analysis of U.S. active duty servicemen. Our findings demonstrate a positive relationship between hazardous alcohol use and perceived social network support among all men in our sample. When evaluating for interaction between sexual orientation and network support, our study demonstrates that higher perceived support was associated with hazardous alcohol use among SM men, and yet, no association was observed among heterosexual men with our outcome of interest. .37
Greater than 50% × gay/bisexual 3.09 (1.21, 7.93) .019
Note: bold indicates p-value <.05. PTSD = posttraumatic stress disorder; GED = General Educational Development; OR = odds ratio.
## Influence of military culture and social networks
High rates of hazardous alcohol use were reported in this sample with over half (52.7%) having a positive AUDIT-C. It is well documented that military personnel experience higher rates of alcohol consumption compared with their civilian counterparts, with an estimated 33%-35% of active duty servicemembers having a positive AUDIT score compared with 6% of civilian adults estimated to have an alcohol use disorderSubstance Abuse and Mental Health Administration (SAMHSA), 2017). High rates of alcohol consumption within the military are partially attributed to a culture permissive toward alcohol as well as norms ingrained in military traditions. Alcohol may be used as a method to break down barriers, facilitate bonding, and improve unit cohesion, particularly after a stressful training exercise or deployment. Alcohol consumption is additionally used to cope with personal or occupational-related stressors, given accommodating attitudes toward drinking as a form of stress reduction within the armed forces. These cultural norms and attitudes may also influence perceptions regarding alcohol consumption within military social networks. Our findings demonstrate a positive relationship between perceived network support and hazardous alcohol use. This observation may initially seem counterintuitive as increased levels of social support have been associated with reduced alcohol use, particularly among individuals with PTSD and depressive symptoms. As respondents preferentially listed social network members with military experience, these findings may reflect the diffusion of military cultural practices into network behaviors. This finding is consistent with research demonstrating that maintaining largely military social networks is associated with increased alcohol consumption among veterans. Within these networks, alcohol may be consumed as a social activity: a form of bonding between peers. Individuals may have key network members from whom they receive support while consuming alcohol.
## Perceived network support among sexual minority men
SM men with high perceived network support had increased odds of hazardous alcohol use, and yet, no association was observed among heterosexual servicemen in the interaction analysis. This may result from existing military drinking attitudes that are reinforced by network dynamics unique to sexual minorities. Social networks and peer norms have an important influence on the behaviors of SM men, particularly as alcohol consumption in this population frequently occurs in social settings and within groups. Studies have demonstrated that SM men who have network members and/or close peers who drink are more likely to consume alcohol themselves. These dynamics influence binge drinking as well, with one study demonstrating that a strong predictor of binge drinking among civilian SM men was having a social network member who engaged in binge drinking. In addition to internal dynamics within SM networks shaping individual behavior, external influences on how these networks are formed and maintained likely have an impact, as well.
Historical norms and attitudes toward SM in the military have led to significant marginalization of this group. Exclusionary policies, such as DADT, may have led SM men to have smaller, tighterknit social networks compared with their heterosexual counterparts. While DADT was repealed, sexual minorities still face many barriers toward full equity and inclusion within the military and continue to be victims of discrimination and bias. Stress from these experiences may result in increased reliance on these tighter-knit networks as a form of social support. If permissive attitudes toward drinking are prevalent within these networks, increased alcohol consumption may result, particularly as the density and intensity of alcohol use among network members has a strong influence on individual drinking behaviors and norms surrounding alcohol consumption.
## Ptsd and receipt of mental health services
In adjusted analysis, having PTSD was associated with hazardous drinking which is expected given the high comorbidity of PTSD and alcohol use disorders, especially among military members with a history of deployments and combatrelated traumas. Among these servicemembers, alcohol consumption is used as a way to regulate emotions, particularly among those who rely upon avoidance coping strategies. Consuming alcohol to moderate negative emotions likely explains our finding that receipt of mental health care was negatively associated with hazardous drinking. Additionally, receiving care from a mental health provider may lead to the identification of alcohol use disorders, resulting in linkage to substance use counseling or treatment. Conversely, individuals with high levels of psychotherapy stigma may not seek mental health services resulting in increased drinking behaviors to self-regulate negative emotions .
# Limitations
As with any study, ours comes with limitations. Data were collected online and via self-report, and therefore, responses may be affected by recall and social desirability bias. However, we utilized attention control measures and assured participants that all responses were confidential throughout the survey to reduce social desirability bias. As complete case analysis was used for our analysis, participants with missing data were excluded, which may have introduced bias. However, we believe that potential bias introduced by using complete case analysis is small (n = 29 participants) and unlikely to significantly influence our results. As respondents were asked to preferentially list network members with military experience, bias could have been introduced as influential members may not have been captured. For example, participants may have had key network members outside of the military who were stronger forms of support than those listed. However, this also represents a potential strength as we evaluated predominantly military social networks of active duty servicemembers, given the paucity of data on this topic. While RDS is a useful sampling tool for difficult to reach populations, our sample may not be generalizable to the entire population of active duty military personnel, particularly as participants were recruited from within respondents' social networks potentially resulting in selection bias. Finally, given the crosssectional nature of our data, we are unable to make causal inferences with our analysis.
# Conclusions
As military personnel are disproportionately affected by alcohol use disorders, research into factors associated with hazardous alcohol consumption is critical to improve the health and well-being of this population. Our study demonstrates that hazardous drinking is associated with increased perceived network support, particularly among SM men, which have important implications for future program development. Specifically, our findings highlight the potential role of network-based interventions directed at changing attitudes and norms surrounding alcohol use by promoting moderation of drinking behaviors within networksand facilitating larger cultural changes regarding alcohol consumption within the military. Further, network dynamics unique to sexual minorities appear to influence alcohol use within this group. Minority stressors resulting from pursuing a career in a traditionally heteronormative workplace may contribute to these dynamics, suggesting the need for continued changes toward diversity and inclusion within the military. Finally, our study demonstrates that PTSD symptomatology is positively associated with hazardous alcohol use, while receipt of mental health services is protective. This highlights the need for improved screening, identification, and linkage to care of servicemembers with PTSD symptomatology and/or signs concerning alcohol use disorders.
## Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
# Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Department of Defense under Grant [W81XWH-151-0700] to CAC, JTG, and IWH. CSB was supported by the National Institute of Mental Health under Grant [T32MH080634].
## Orcid id
Cheríe S. Blair https://orcid.org/0000-0001-6535-0018
|
Medication Adherence in Type 2 Diabetes: The ENTRED Study 2007, a French Population-Based Study
Background: Adherence to prescribed medications is a key dimension of healthcare quality. The aim of this large population-based study was to evaluate self-reported medication adherence and to identify factors linked with poor adherence in patients with type 2 diabetes in France.Methodology: The ENTRED study 2007, a French national survey of people treated for diabetes, was based on a representative sample of patients who claimed reimbursement for oral hypoglycaemic agents and/or insulin at least three times between August 2006 and July 2007, and who were randomly selected from the database of the two main National Health Insurance Systems. Medication adherence was determined using a six-item self-administered questionnaire. A multinomial polychotomous logistic regression model was used to identify factors associated with medication adherence in the 3,637 persons with type 2 diabetes.Principal Findings: Thirty nine percent of patients reported good medication adherence, 49% medium adherence and 12% poor adherence. The factors significantly associated with poor adherence in multivariate analysis were socio-demographic factors: age ,45 years, non-European geographical origin, financial difficulties and being professionally active; disease and therapy-related factors: HbA 1c .8% and existing diabetes complications; and health care-related factors: difficulties for taking medication alone, decision making by the patient only, poor acceptability of medical recommendations, lack of family or social support, need for information on treatment, reporting no confidence in the future, need for medical support and follow-up by a specialist physician.Conclusions: In a country with a high level of access to healthcare, our study demonstrated a substantial low level of medication adherence in type 2 diabetic patients. Better identification of those with poor adherence and individualised suitable recommendations remain essential for better healthcare management.
# Introduction
Diabetes mellitus, a complex chronic disease, is a growing worldwide epidemic with the number of people with diabetes estimated to reach 330 million by 2030 [bib_ref] Global prevalence of diabetes: estimates for the year 2000 and projections for..., Wild [/bib_ref]. Given the high morbidity and mortality associated with the disease, primarily due to macrovascular complications, type 2 diabetes is a major publichealth concern. Most European countries have formulated evidence-based guidelines with clear targets, but actual care often falls far short of these targets [bib_ref] Improvements in quality of clinical care in English general practice 1998-2003: longitudinal..., Campbell [/bib_ref] [bib_ref] Trends in the quality of care for elderly people with type 2..., Pornet [/bib_ref].
A key dimension of healthcare quality is adherence to prescribed medications. According to the World Health Organization (WHO), adherence is the extent to which a person's behaviour -taking medication, following a diet, and/or executing lifestyle changes -corresponds with agreed recommendations from the health care provider. However, medication nonadherence is particularly common among patients with type 2 diabetes [bib_ref] A systematic review of adherence with medications for diabetes, Cramer [/bib_ref] and inadequate adherence compromises safety and treatment effectiveness, leading to increased mortality and morbidity with considerable direct and indirect costs to the healthcare system [bib_ref] Impact of medication adherence on hospitalization risk and healthcare cost, Sokol [/bib_ref] [bib_ref] Prevalence and economic consequences of medication adherence in diabetes: a systematic literature..., Lee [/bib_ref]. A recent WHO report states that, because the magnitude of non-adherence and the scope of its sequelae are so alarming, more health benefits worldwide would result from improving adherence to existing treatments than by developing new medical treatments.
Previously, numerous studies have explored potential risk factors of adherence to medicines across a variety of conditions. However, the majority of studies have explored largely unmodifi-able variables due to the retrospective databases that are often used to measure adherence. Frequently cited risk factors include age, sex, ethnicity, income, education, and comorbidity though their relationship to adherence has been inconsistent due to variations in study designs and sample populations [bib_ref] A systematic review of adherence with medications for diabetes, Cramer [/bib_ref] [bib_ref] Variations in patients' adherence to medical recommendations: a quantitative review of 50..., Dimatteo [/bib_ref] [bib_ref] Adherence to medication, Osterberg [/bib_ref] [bib_ref] Adherence to preventive medications: predictors and outcomes in the Diabetes Prevention Program, Walker [/bib_ref]. Therefore there is a continuing need to better identify factors related to medication adherence. Moreover, previous studies on medication adherence often included a limited number of diabetic patients and with selected patient population [bib_ref] Impact of dosage frequency on patient compliance, Paes [/bib_ref] [bib_ref] Neuropsychological correlates of suboptimal adherence to metformin, Rosen [/bib_ref] [bib_ref] Depression and diabetes: impact of depressive symptoms on adherence, function, and costs, Ciechanowski [/bib_ref] [bib_ref] Adherence to prescribed oral hypoglycaemic medication in a population of patients with..., Donnan [/bib_ref] [bib_ref] Adherence to oral antidiabetic therapy in a managed care organization: a comparison..., Melikian [/bib_ref] [bib_ref] Compliance with sulfonylureas in a health maintenance organization: a pharmacy recordbased study, Venturini [/bib_ref] [bib_ref] Effects of once-daily and twice-daily dosing on adherence with prescribed glipizide oral..., Dezii [/bib_ref] [bib_ref] Polypharmacy and medication adherence in patients with type 2 diabetes, Grant [/bib_ref] [bib_ref] Improving adherence and reducing medication discrepancies in patients with diabetes, Grant [/bib_ref] [bib_ref] Ten-year follow-up of antidiabetic drug use, nonadherence, and mortality in a defined..., Brown [/bib_ref] [bib_ref] Medication adherence and racial differences in A1C control, Adams [/bib_ref] [bib_ref] Racial differences in long-term adherence to oral antidiabetic drug therapy: a longitudinal..., Trinacty [/bib_ref] [bib_ref] Refill adherence and polypharmacy among patients with type 2 diabetes in general..., Van Bruggen [/bib_ref] [bib_ref] Factors that Affect Medication Adherence in Elderly Patients with Diabetes Mellitus, Park [/bib_ref] [bib_ref] Patient-reported tolerability issues with oral antidiabetic agents: Associations with adherence; treatment satisfaction..., Pollack [/bib_ref] , which limited the generalizability of the results. This large population based study, the largest European study to our knowledge, was conducted to evaluate medication adherence in people with type 2 diabetes, and to identify the risk factors for poor adherence and especially, modifiable factors.
# Methods
# Ethics statement
The French National Ethics Committee and the French Data Protection Authority Committee gave its approval to the ENTRED study. According to the French regulations, written consent for filling questionnaire was not required as no intervention was performed on the participants and no blood or human tissue was considered. Filling a self-reported questionnaire and mailing it back was thus considered as consent for all participants. For medical data filled by medical practitioners, participants who provided addresses of their care practitioners filled a form which was sent to the care practitioners to authorize them to provide further medical information.
## Study population and design
The ENTRED study was a French national public survey. A complex random sample was selected from all patients aged over 18 years who claimed reimbursement for oral hypoglycaemic agents (OHA) and/or insulin at least three times between August 2006 and July 2007, from the two main National Health Insurance Systems (NHIS), which cover all active and retired employees and their relatives-about 80% of the French population. To classify the different types of diabetes, we used an epidemiological algorithm: people diagnosed before the age of 45 years and treated with insulin within two years from diagnosis were classified as having type 1 diabetes and have not been considered for this study.
A detailed questionnaire with a total of 110 questions was sent to all patients (48% response rate; n = 3,973 with 3,637 type 2 diabetic patients) and a medical questionnaire was also sent to the medical-care providers of those among the responders who gave their medical provider's addresses (63% response rate; n = 2,485). Care providers reported the most recent clinical measurements.
Medication adherence, the dependant variable, was analysed in the subgroup of people with type 2 diabetes. The independent variables to explain the medication adherence included sociodemographic characteristics (age, gender, education level, geographical origin, marital status, residence, professional activity, financial level, complementary health insurance…), characteristics associated with disease and therapy (time since diabetes diagnosis, type of treatment, body mass index, hypertension, dyslipidaemia, smoking state, glycaemic control, microvascular or macrovascular complications…), and associated with medical care (decision making, follow-up by a specialist, acceptability of medical recommendations, ability for taking medicine alone, need for medical support or information on treatment…).
## Measure of medication adherence
In this study, medication adherence (referring to any medicine, not just for diabetes) was determined using a six item selfadministered questionnaire, drawing upon the works by Girerd et al. [bib_ref] Evaluation de l'observance du traitement anti-hypertenseur par un questionnaire: mise au point..., Girerd [/bib_ref]. Patients responded yes or no to each of the following questions: (1) do you sometimes forget to take your medicine,have you ever run out of your medicine, (3) do you sometimes take your medicine late, (4) do you sometimes decide not to take your medicine because someday you feel that your treatment do more harm than good, (5) do you think that you have too many pills to take, (6) when you feel better, do you sometimes stop taking your medicine. It has been shown that such a questionnaire has sufficient validity and reliability [bib_ref] Evaluation de l'observance du traitement anti-hypertenseur par un questionnaire: mise au point..., Girerd [/bib_ref]. Compared to a clinical evaluation of medication adherence, the values of kappa indices were 0.65 in ''good adherence'' when ''No'' was answered to the 6 items, 0.5 in ''medium adherence'' when 1 or 2 ''Yes'' were given and 0.56 in ''poor adherence'' when 3 or more ''Yes'' were given [bib_ref] Evaluation de l'observance du traitement anti-hypertenseur par un questionnaire: mise au point..., Girerd [/bib_ref].
## Statistical analyses
Descriptive analyses were first performed. In data reported by patients, missing data for medication adherence were excluded from analyses. To minimise potential non-response biases, all analyses were weighted to take into account the participation rate based on socio-demographic data and the type of antidiabetic treatment. Quantitative values are expressed as means 6 standard deviation, and were compared by Student's t test, analysis of variance or nonparametric test when appropriate, while qualitative values were compared by the x 2 test or Fisher's exact test. The outcome of interest, medication adherence, was classified into three categories; 'good', 'medium' and 'poor' and was treated as a nominal variable since the proportional odds assumption was rejected. A multinomial polychotomous logistic regression model was used to estimate the effect of each covariate on the odds of poor adherence and on the odds of medium adherence versus good adherence, while simultaneously adjusting for all other variables in the model. All statistical analyses took into account the sample survey design and were carried out using SAS version 9.1.3 (SAS Institute Inc, Cary, NC). The characteristics of those who responded to the detailed questionnaire were compared with those who did not, using the 2007 administrative data available for all people.
# Results
## Participants' characteristics
This study included a total of 3,637 type 2 diabetic patients (2,138 men and 1,499 women) with a mean age of 65 years (18 to 102 years). The main socio-demographic and clinical baseline characteristics of the responders are summarized in [fig_ref] Table 1: Main socio-demographic and clinical baseline characteristics of the 3,637 responders, Entred study [/fig_ref]. Eighty one percent were treated with OHA without insulin, mean HbA 1c was 7.0% and 41% had microvascular or macrovascular complications. Respondents to the survey were slightly younger than non-respondents (64 years versus 66 years on average; p,0.0001), most frequently male (59% versus 52%; p,0.0001), most frequently born in France (79% versus 70%; p,0.0001), less likely to be treated with OHA without insulin (74% versus 78%; p,0.0001) and had better medical follow-up (43% versus 35% had three HbA 1c tests per year; p,0.0001).
## Medication adherence
Thirty nine percent of patients had good adherence, 49% medium adherence and 12% poor adherence: 18% of patients reported sometimes forgetting to take their medicine, 9% running out of their medicine, 38% sometimes taking their medicine late, 4% sometimes deciding not to take their medicine because someday they felt that their treatment do more harm than good, 34% having too many pills to take and 5% sometimes stopping to take their medicine when they felt better.
In univariate analysis, many factors were associated with medication adherence (not detailed). On the contrary gender and duration since diagnosis did not affect medication adherence (p = 0.93 and 0.90, respectively). Patients with professional activity (who currently work) forgot more often to take their medicine (30% versus 15%, p,0.0001) and took more often their medicine late (51% versus 35%, p,0.0001).
In polychotomous logistic regression [fig_ref] Table 2: Multivariate analysis of medication adherence [/fig_ref] , socio-demographic factors significantly associated with poor (versus good) adherence were: age ,45 years (Odds Ratio (OR) = 5.2), non-European geographical origin (OR = 2.6), financial difficulties (OR = 1.7) and being professionally active (OR = 1.5). Disease and therapy-related factors significantly associated with poor adherence were: HbA 1c .8% (OR = 2.0) and existing diabetes complications (OR = 1.7). A trend was observed with self-reported hypertension (OR = 1.4, p = 0.08) and dyslipidemia (OR = 1.4, p = 0.08) while being treated with insulin and diabetes duration did not influence medication adherence. Lastly, health carerelated factors significantly associated with poor adherence were: difficulties for taking medication alone (OR = 3.8), decision making by the patient only (OR = 3.3), poor acceptability of medical recommendations (OR = 2.7), lack of family or social support (OR = 2.5), need for information on treatment (OR = 2.0), reporting no confidence in the future (OR = 1.6), need for medical support (OR = 1.6) and follow-up by a specialist physician (OR = 1.4).
# Discussion
Medication adherence is a key component of self-management for patients with diabetes. Our population-based study found a low rate of good adherence (39%), which means that for many patients, medication adherence could be improved. One of the most common challenges physicians face with a patient with poorly controlled diabetes is to try to figure out if the patient's hyperglycemia is due to poor adherence or is occurring despite proper medication use (i.e., therapy needs to be intensified). Since patients may be more willing to report suboptimal adherence (self-reports typically provide overestimates of adherence for several reasons: first, they may rely on patients' own interpretation or memory of what advice was given and, if accepted, how closely it has been followed; second, patients may tend to report higher levels of adherence in order to please health care providers or avoid embarrassment), probing the handful of strongly predictive factors we have identified is useful for two reasons. First, it can help identify those likely to be poor adherers. Second, it can direct the physician on aspects of diabetes and its management on which they should focus their patient education efforts [bib_ref] Predictors of adherence to diabetes medications: the role of disease and medication..., Mann [/bib_ref].
Our results are consistent with previous studies and particularly, with the DARTS study which found that adequate adherence (adherence index $90%) was found in only approximately one in three of those with type 2 diabetes receiving OHA [bib_ref] Adherence to prescribed oral hypoglycaemic medication in a population of patients with..., Donnan [/bib_ref]. Nonetheless, a recent meta-analysis found that medication adherence ranged from 36% to 93% depending on the definition applied [bib_ref] A systematic review of adherence with medications for diabetes, Cramer [/bib_ref]. However, the lack of standard measurements prevents comparison being made between studies and across populations. We also found that poor glycaemia control and presence of microvascular or macrovascular complications were more common among patients with poor adherence to medications.
In previous studies, many factors were inconsistently found as risk factors of poor adherence to drug therapy in type 2 diabetes. The statistical power of this study based on a large sample size was sufficient to detect small differences. Age [bib_ref] Adherence to prescribed oral hypoglycaemic medication in a population of patients with..., Donnan [/bib_ref] [bib_ref] Younger patients with type 2 diabetes need better glycaemic control: results of..., Rothenbacher [/bib_ref] , financial difficulties, ethnicity [bib_ref] Medication adherence and racial differences in A1C control, Adams [/bib_ref] [bib_ref] Racial differences in long-term adherence to oral antidiabetic drug therapy: a longitudinal..., Trinacty [/bib_ref] , psychological factors [bib_ref] Meta-analysis of correlates of diabetes patients' compliance with prescribed medications, Nagasawa [/bib_ref] , social support, quality of the relationship between patient and physician [bib_ref] The patient-provider relationship: attachment theory and adherence to treatment in diabetes, Ciechanowski [/bib_ref] were confirmed as risk factors of medication adherence. In this study, we particularly focused on modifiable factors associated with medical care (decision making, acceptability of medical recommendations, medical support …). Contrary to Lawton et al. who found that non-adherence was related more to patient forgetfulness than to specific concerns about medications or interaction with physicians [bib_ref] Patients' perceptions and experiences of taking oral glucose-lowering agents: a longitudinal qualitative..., Lawton [/bib_ref] , our results show the importance of shared decision-making in which the beliefs and preferences of the patient are taken into consideration. This patient centred approach should enhance adherence and improve outcomes [bib_ref] Participatory decision making, patient activation, medication adherence, and intermediate clinical outcomes in..., Parchman [/bib_ref] [bib_ref] Patient versus general practitioner perception of problems with treatment adherence in type..., Moreau [/bib_ref]. Other interesting risk factors of poor adherence found in this study were the presence of difficulties for taking medicine alone, need for information and poor acceptability of medical recommendations. These two last findings confirm the need to better inform patients about their disease and treatment and to individually adapt recommendations. This could underline the insufficiency of patient education especially since it is now well established that education enables the patient to acquire knowledge and understanding of diabetes, selfmanagement skills and psychosocial competencies [bib_ref] The effects of patient communication skills training on compliance, Cegala [/bib_ref] [bib_ref] The effect of an educational intervention on patients' knowledge about hypertension, beliefs..., Magadza [/bib_ref]. Social or family support is also crucial. Family members are frequently involved and recognized as supportive: they act as counsellors encouraging diet and exercise behaviours, facilitating adherence with medication, and altogether helping patients to ''live with the disease'' [bib_ref] How do patients with type 2 diabetes perceive their disease? Insights from..., Mosnier-Pudar [/bib_ref]. Lastly, the fact that professional activity was associated with poor adherence could be surprising considering a potential healthy worker effect. However, we clearly report that active patients more often forget to take their medicine and/or take their medicine late, and the relationship remains significant after adjustments. Ultimately, two categories of risk factors for poor adherence could be distinguished: unmodifiable factors (such as age, ethnicity…) or factors that are hardly modifiable in the context of the medical relationship (financial difficulties, presence of professional activity…) which may help physicians to better identify patients at high risk for poor adherence and to adapt their medical care ; and some modifiable factors (such as social support, quality of the relationship between patient and physician, need for information, poor acceptability of medical recommenda-tions…) on which physicians could focus their efforts to improve medication adherence and, as a consequence, to improve glycaemia control. In our study, no significant difference in adherence was found between males and females, which has already been shown [bib_ref] Adherence to prescribed oral hypoglycaemic medication in a population of patients with..., Donnan [/bib_ref]. However, it should be noted that the DIABASIS study evidenced clear gender differences in the perception and self-management of disease. Women took the disease more seriously, reported a higher impact on daily life and were more involved in self-management, while men relied more on family support. The authors suggested that physicians should take these differences in attitudes into account when counseling, educating and treating patients [bib_ref] How do patients with type 2 diabetes perceive their disease? Insights from..., Mosnier-Pudar [/bib_ref].
After identifying patients at high risk for poor adherence, the physician should try more than usual to apply multiple interventions in order to improve adherence: educational, behavioral, and affective interventions [bib_ref] Improving adherence to diabetes selfmanagement recommendations, Schechter [/bib_ref]. Educational interventions seek to improve adherence by providing information and/or skills. Education may take the form of individual instruction or group classes. In any event, a key element of successful educational strategies is providing simple, clear messages, hopefully tailored to the needs of the individual, and verifying that the messages have been understood. Behavioral approaches have their roots in cognitive-behavioral psychology and use techniques such as reminders, memory aids, synchronizing therapeutic activities with routine life events (e.g., taking pills before you shower), goalsetting, self-monitoring, contracting, skill-building, and rewards. For example, reminders may be mailed, e-mailed, or telephoned.
What is important is that the behavior in question has been negotiated with and accepted by individual patients so that adoption of the behavior has a chance of succeeding in the long term. Affective interventions seek to enhance adherence by providing emotional support and encouragement. Finally, it should be remembered that application of multiple interventions of different types is more effective than any single intervention [bib_ref] Improving adherence to diabetes selfmanagement recommendations, Schechter [/bib_ref].
Our results should be viewed with consideration of several limitations. One limitation was the use of self-report data on medication adherence, because of a resulting tendency to overestimate adherence due to recall biases and social desirability. However, self-reported questionnaires have frequently been used because they are low in both cost and time expenditure and appropriate for large population-based samples. Subsequent research suggests that the self-report methods provide a reasonably accurate estimate of adherence [bib_ref] The concordance of self-report with other measures of medication adherence: a summary..., Garber [/bib_ref]. Besides, our results based on self-report questionnaires were consistent with the literature, poor glycaemic control being more common among patients with low adherence to medications [bib_ref] The association between diabetes metabolic control and drug adherence in an indigent..., Schectman [/bib_ref] [bib_ref] Influence of oral antidiabetic drugs compliance on metabolic control in type 2..., Guillausseau [/bib_ref] and many well-known factors associated with poor adherence were also identified by our study. The total number of medications prescribed to the patient has not been assessed in our study. However, this factor has been recognized as a contributor of patient adherence for a long time [bib_ref] Drug compliance in type 2 diabetes: role of drug treatment regimens and..., Penfornis [/bib_ref] and does not need any more to be established. Biases linked to participation are a common limitation, although our response rate was in keeping with population-based surveys. It is possible that respondents were more concerned about their diabetes than the others and, as a result, this may have led to an overestimation of medication adherence. To account for potential non response biases, we weighted our results according to the participation rate, based on socio-demographic data and type of antidiabetic treatment, as these auxiliary variables were correlated with both the non-response process and the survey estimate [bib_ref] On weighting the rates in non-response weights, Little [/bib_ref] [bib_ref] National Survey of Family Growth, Cycle 6: sample design, weighting, imputation, and..., Lepkowski [/bib_ref]. Lastly, our study was cross sectional, where causal relationship between the independent and dependent variables cannot be fully established.
Despite these limitations, this study provides valuable information in support of the literature and has several major strengths. The number of people with type 2 diabetes was large, and to our knowledge larger than any European previously published study of adherence. The studied sample constituted a large nationally representative cohort of diabetic patients. Therefore, the generalizability of the results to countries with similar health care system is high. This study, combining multiple data sources (self-reported, medical-care providers, data from the two main National Health Insurance Systems), provided a large number of diabetes-related variables.
In summary, medication adherence is vital for effective diabetes management. Our findings point towards the interest of finetuning the primary care provider's approach to the individual patient by taking into account medication adherence. More evidence to support specific interventions that will be effective in overcoming adherence challenges for diabetes patients is needed. The patients should have a pivotal role in their diabetes management. Therefore, they need to acquire knowledge and skills, but also the ability for behavioural change, which often requires intensive patient-centred health education. In a country with a high level of access to healthcare, our study demonstrated a low level of medication adherence. Better identification of patients with poor adherence, who require a more specific and rigorous patient physician relationship and individualised suitable recommendations, remains essential to obtain better outcomes in type 2 diabetic patients.
[table] Table 1: Main socio-demographic and clinical baseline characteristics of the 3,637 responders, Entred study. [/table]
[table] Table 2: Multivariate analysis of medication adherence. Entred study, N = 3637. [/table]
|
Regulatory Eosinophils in Inflammation and Metabolic Disorders
# Introduction
Eosinophils are potent effector cells implicated in allergic responses and helminth infections and have cytotoxic granules containing major basic protein (MBP), eosinophil cationic protein (ECP), eosinophil peroxidase (EPO), and eosinophil-derived neurotoxin (EDN) [bib_ref] The Eosinophil, Rothenberg [/bib_ref]. Responding to stimuli such as an allergen, helminth infection, or tissue injury, they release their granule-derived cytotoxic proteins and are involved in inflammatory processes. However, the intestine (constantly in contact with the gut microbiota) is a major site where eosinophils are abundantly present under homeostatic conditions. Intestinal eosinophils rarely undergo degranulation and have a long lifespan as compared with eosinophils in the blood and lungs [bib_ref] Common g-chain-dependent signal confer selevtive survival of eosinophils in the murine small..., Carlens [/bib_ref]. A possible reason is that inhibitory receptor signal regulatory protein a (SIRPa)/CD172a, highly expressed on the intestinal eosinophils, inhibits degranulation of eosinophils and promotes their survival [bib_ref] SIRPa/CD172a regulates eosinophil homeostasis, Garcia [/bib_ref]. In support of this notion, the ligand for SIRPa/CD172a, CD47, is expressed in the intestine [bib_ref] SIRPa/CD172a regulates eosinophil homeostasis, Garcia [/bib_ref]. In addition, intestinal eosinophils highly express common g-chain, which is an integral part of the cytokine receptors for IL-2, IL-4, IL-7, IL-9, IL-15, Eosinophils are potent effector cells implicated in allergic responses and helminth infections. Responding to stimuli, they release their granule-derived cytotoxic proteins and are involved in inflammatory processes. However, under homeostatic conditions, eosinophils are abundantly present in the intestine and are constantly in contact with the gut microbiota and maintain the balance of immune responses without inflammation. This situation indicates that intestinal eosinophils have an anti-inflammatory function unlike allergic eosinophils. In support of this notion, some papers have shown that eosinophils have different phenotypes depending on the site of residence and are a heterogeneous cell population. Recently, it was reported that eosinophils in the small intestine and adipose tissue, respectively, contribute to homeostasis of intestinal immune responses and metabolism. Accordingly, in this review, we summarize new functions of eosinophils demonstrated in recent studies and discuss their homeostatic functions.
[Immune Network 2017;17(1):41-47] Keywords: Eosinophils, Anti-inflammation, IL-1R antagonist, Th17 cells, IL-4, IgA and IL-21; intestinal eosinophils are severely reduced in number in the common g-chain-deficient mice unlike eosinophils in blood and lungs; therefore, common g-chain-dependent signals seem to play an important role in selective survival of intestinal eosinophils [bib_ref] Common g-chain-dependent signal confer selevtive survival of eosinophils in the murine small..., Carlens [/bib_ref]. Intestinal eosinophils are produced from hematopoietic stem cells in bone marrow and migrate into the intestine under the influence of chemokine eotaxin-1 [bib_ref] Fundamental signals that regulate eosinophil homing to the gastrointestinal tract, Mishra [/bib_ref]. During the development and survival of eosinophils, signaling of the common b-receptor chain shared by IL-3, IL-5, and GM-CSF is required (4), and among these cytokines, IL-5 is particularly important; overproduction of IL-5 in transgenic mice results in eosinophilia (5) whereas deletion of IL-5 abrogates eosinophilia induced by an aeroallergen [bib_ref] Interleukin 5 deficiency abolishes eosinophilia, airways hyperreactivity, lung damage in a mouse..., Foster [/bib_ref]. According to a recent report, IL-5 is mainly produced by type 2 innate lymphoid (ILC2) cells, and its production in ILC2 cells is enhanced by vasoactive intestinal peptide [bib_ref] Type 2 innate lymphoid cells control eosinophil homeostasis, Nussbaum [/bib_ref]. Additionally, transcription factor GATA-1 is crucial for the development of eosinophils; deletion of the palindromic double GATA high-affinity binding sites in the GATA-1 promoter results in eosinophil-deficient mice (called Ddbl GATA-1 mice) [bib_ref] Targeted deletion of a highaffinity GATA-binding site in the GATA-1 promoter leads..., Yu [/bib_ref]. PHIL mice, which express diphtheria toxin A chain under the control of the eosinophil-specific EPO promoter, are also eosinophil-deficient mice [bib_ref] Defining a link with asthma in mice congenitally deficient in eosinophils, Lee [/bib_ref]. In recent studies on these eosinophil-deficient mice, it has been revealed why eosinophils are abundant in the intestine without inflammation and what their physiological roles are under homeostatic conditions. Some papers have shown that intestinal eosinophils promote formation and maintenance of IgA-expressing plasma cells (10) and abundantly produce IL-1 receptor antagonist (IL-1Ra), contributing to intestinal immune homeostasis. Several lines of evidence have shown that eosinophils also exist in adipose tissue and are involved in metabolic homeostasis [bib_ref] Eosinophils sustain adipose alternatively activated macrophages associated with glucose homeostasis, Wu [/bib_ref] [bib_ref] Eosinophils and type 2 cyotkine signaling in macrophages orchestrate development of functional..., Qiu [/bib_ref]. Therefore, in this review, we summarize recent findings regarding eosinophils and discuss the homeostatic functions of these cells.
## The anti-inflammatory function
Recently, it has been reported that small-intestinal eosinophils produce large amounts of IL-1Ra. The latter competes with IL-1b for receptor binding and inhibits inflammatory responses [bib_ref] The IL-1 family: regulators of immunity, Slims [/bib_ref]. Meanwhile, IL-1b is primarily produced by activated macrophages and functions as a key mediator in various inflammatory diseases including inflammatory bowel disease and rheumatoid arthritis [bib_ref] The IL-1 family: regulators of immunity, Slims [/bib_ref]. In this context, a decrease in the IL-1Ra to IL-1 ratio is linked to inflammatory bowel disease in humans [bib_ref] Mucosal imbalance of IL-1 and IL-1 receptor antagonist in inflammatory bowel disease...., Casini-Raggi [/bib_ref]. IL-1Ra-deficient mice spontaneously develop arthritis with a marked increase in the number of Th17 cells [bib_ref] IL-17 production from activated T cells is required for the spontaneous development..., Nakae [/bib_ref] [bib_ref] Interleukin-1 drives pathogenic Th17 cells during spontaneous arthritis in interleukin-1 receptor antagonist-deficient..., Koenders [/bib_ref]. In line with this finding, it has been reported that microbiota-induced IL-1b, but not IL-6, plays an important role in the development of Th17 cells in the intestine [bib_ref] Microbiota-induced IL-1b, but not IL-6, is critical for the development of steady-state..., Shaw [/bib_ref]. Therefore, small-intestinal eosinophils suppress the differentiation and maintenance of Th17 cells by constantly secreting large amounts of IL-1Ra, thus contributing to intestinal immune homeostasis (11) [fig_ref] Figure 1: Anti-inflammatory functions of eosinophils [/fig_ref]. Intestinal eosinophils also inhibit differentiation and/or proliferation of Th1 cells in the lamina propria (LP) though less strongly in comparison with the effects on Th17 cells. However, because IL-1Ra-deficient mice do not show a significant increase in the number of Th1 cells, intestinal eosinophils may suppress differentiation and/or proliferation of Th1 cells by secreting other unidentified cytokines. Unlike Th1 and Th17 cells, small-intestinal eosinophils do not affect frequencies of Foxp3 + regulatory T (Treg) cells and Th2 cells in the LP. Particularly in food allergy, intestinal eosinophils activate dendritic cells (DCs) through a release of EPO and promote their migration from the intestine to draining lymph nodes, thus initiating a primary Th2 immune response [bib_ref] Indigenous enteric eosinophils contraol DCs to initiate a primary Th2 immune response..., Chu [/bib_ref]. In contradiction to these findings, one group reported that the numbers of CD103 + DCs and CD103 + CD4 + T cells including Foxp3 + Treg cells are reduced in the LP of eosinophil-deficient mice, although they did not explain the mechanism in the paper [bib_ref] Eosinophils promote generation and maintenance of immunoglobulin-A-expressing plasma cells and contribute to..., Chu [/bib_ref] , and this phenomenon should be confirmed later.
Because small-intestinal eosinophils, but not eosinophils in blood and bone marrow, produce a large amount of IL-1Ra under homeostatic conditions, we can hypothesize that gut microbiota and its metabolites may be involved in IL-1Ra production. On the other hand, small-intestinal eosinophils in both SPF and GF mice produce similar levels of IL-1Ra, and metabolites such as retinoic acid and short-chain fatty acids do not induce IL-1Ra production in bone marrow eosinophils. Instead, GM-CSF, which is known to be involved in the development and survival of eosinophils, performs an important function in IL-1Ra production. In this regard, the high basal concentration of GM-CSF in the small intestine and high expression of GM-CSF receptor in small-intestinal eosinophils seem to explain the unique ability of smallintestinal eosinophils to continuously secrete a large amount of IL-1Ra under steady-state conditions. In addition, because Th17 cells can produce GM-CSF in an IL-1-and IL-23-dependent manner [bib_ref] The encephalitogenicity of T H 17 cells is dependent on IL-1-and IL-23-induced..., El-Behi [/bib_ref] , it is plausible that GM-CSF produced by Th17 cells endows eosinophils with the superior capacity for IL-1Ra production and regulates Th17 cell homeostasis via a negative feedback loop. In contrast, ironically, large-intestinal eosinophils, when activated by GM-CSF produced by Th17 cells, secret EPO as well as proinflammatory cytokines TNF and IL-13, and aggravate colitis [bib_ref] Granulocyte macrophage colony-stimulating factor-activated eosinophils promote interleukin-23 drive chronic colitis, Griseri [/bib_ref]. These findings imply that small-intestinal eosinophils are different from large-intestinal ones. In support of this notion, largeintestinal eosinophils produce much smaller amounts of IL-1Ra and show a distinctive expression pattern of surface proteins including CD11c, ST2, and Ly6C as compared with small-intestinal ones [bib_ref] Indigenous enteric eosinophils contraol DCs to initiate a primary Th2 immune response..., Chu [/bib_ref]. In addition, although eosinophils are mainly present throughout the intestine, large-intestinal eosinophils are scarce in a steady state [bib_ref] Indigenous enteric eosinophils contraol DCs to initiate a primary Th2 immune response..., Chu [/bib_ref] , and only under intestinal inflammatory conditions, does their frequency dramatically increase [bib_ref] Granulocyte macrophage colony-stimulating factor-activated eosinophils promote interleukin-23 drive chronic colitis, Griseri [/bib_ref]. Accordingly, large-intestinal eosinophils seem to be proinflammatory unlike their small-intestinal counterparts.
## Antibacterial defense
Most of the intestinal plasma cells are located in the LP and mostly produce IgA [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. IgA is secreted into the intestinal lumen and functions as the first line of defense through neutralization and clearance of enteric pathogens. Nevertheless, IgA-producing B cells are generally formed in Peyer's patches (PPs) and migrate into the LP [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. In PPs, antigen-experienced CD4 + T cells migrate into B-cell follicles via CXCR5 and activate B cells [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. These specialized CD4 + T cells are called follicular B helper T (T FH ) cells and are different from other CD4 + T cell lineages. They highly express costimulatory molecules for their development and function, e.g., CD40L, OX-40, and programmed cell death 1 (PD-1) rather than cytokines and transcription factors characteristic of Th1, Th2, and Th17 cells [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. Transcription factor B cell lymphoma 6 (Bcl-6) plays an important role in T FH differentiation, whereas transcription factor B lymphocyte maturation protein 1 (Blimp-1) inhibits it [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. The activated B cells induce activation-induced cytidine deaminase (AID) and provoke class switch recombination [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. Because TGF-b1, a key cytokine for IgA switching, is expressed by many cells in PPs, activated B cells preferentially switch their isotype from IgM to IgA [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref] [bib_ref] TGF-b receptor controls B cell responsiveness and induction of IgA in vivo, Cazac [/bib_ref]. Moreover, IL-21 secreted by T FH cells enhances the proliferation and differentiation of IgA plasma cell precursors in synergy with TGF-b1 [bib_ref] A T celldependent mechanism for the induction of human mucosal homing immunoglobulin..., Dullaers [/bib_ref]. Meanwhile, IgM + B cells in the LP are activated by polyclonal stimuli or by antigens presented by DCs in the LP (LP-DCs) and differentiate into IgA-producing plasma cells without the help of T FH cells [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. In this T cell-independent formation of IgA-producing B cells, important functions are performed by TGF-b1, B-cell activating factor of the tumor-necrosis factor family (BAFF), and a proliferation-inducing ligand (APRIL) produced by DCs and stromal cells in the LP [bib_ref] Adaptive immune regulation in the gut: T cell-dependent and T cell-independent IgA..., Fagarasan [/bib_ref]. In vitro, cytokines such as IL-6, TNF, IL-10, IL-4, and IL-5, also prolong the survival of plasma cells [bib_ref] Plasma cell survival is mediated by synergistic effects of cytokines and adhesion-dependent..., Cassese [/bib_ref]. In this context, bone marrow eosinophils have been reported to support the survival of plasma cells by secreting APRIL and IL-6 and contribute to the long-term maintenance of plasma cells in bone marrow [bib_ref] Eosinophils are required for the maintenance of plasma cells in the bone..., Chu [/bib_ref]. Although intestinal eosinophils are mainly in the LP, not in PPs, they are involved in the formation and maintenance of IgA-producing B cells in PPs as well as in the LP [bib_ref] Eosinophils promote generation and maintenance of immunoglobulin-A-expressing plasma cells and contribute to..., Chu [/bib_ref]. Intestinal eosinophils express BAFF, TGF-b1, and matrix metalloproteases (that are involved in the release of TGF-b1 from the large latent complex) as well as APRIL and IL-6 at the mRNA level [bib_ref] Eosinophils promote generation and maintenance of immunoglobulin-A-expressing plasma cells and contribute to..., Chu [/bib_ref]. Therefore, intestinal eosinophils appear to influence T-cell-independent formation and maintenance of IgA-producing B cells via production of these cytokines and enzymes. Accordingly, the amount of IgA is dramatically reduced in the intestine of eosinophildeficient mice; this change increases bacterial load in the gut and affects composition of the microbiota there [bib_ref] Eosinophils promote generation and maintenance of immunoglobulin-A-expressing plasma cells and contribute to..., Chu [/bib_ref]. Therefore, it seems likely that eosinophils indirectly exert their effects on T-cell-dependent formation of IgAproducing B cells in PPs. Recently, one paper revealed that eosinophils can be involved in IgA production through IL-1b [bib_ref] IL-1b in eosinophilmediated small intestinal homeostasis and IgA production, Jung [/bib_ref]. Moreover, eosinophils can release mitochondrial DNA in a catapult-like manner and kill bacteria [bib_ref] Catapult-like release of mitochondrial DNA by eosinophils contributes to antibacterial defense, Yousefi [/bib_ref]. Nevertheless, the release of mitochondrial DNA does not cause cell death of eosinophils. For the release of mitochondrial DNA, eosinophils should be activated by lipopolysaccharide under conditions where they are primed with IL-5 or IFN-g [bib_ref] Catapult-like release of mitochondrial DNA by eosinophils contributes to antibacterial defense, Yousefi [/bib_ref]. In the intestinal environment, gut bacteria can exert their effects on immune cells through bacteria-derived molecules, and immune-cell populations producing IFN-g (Th1) and IL-5 (ILC2) exist; therefore, it seems plausible that intestinal eosinophils reduce the bacterial load in the gut via the release of mitochondrial DNA under homeostatic conditions. Accordingly, intestinal eosinophils prevent the host from bacterial invasion and contribute to homeostasis of intestinal immunity.
## Metabolic homeostasis
Obesity induces chronic low-grade inflammation, consequently causing insulin resistance and type 2 diabetes [bib_ref] Type 2 diabetes as an inflammatory disease, Donath [/bib_ref] [bib_ref] Regulation of metabolism by the innate immune system, Lackey [/bib_ref]. In the obese state, a large number of monocytes migrate into adipose tissue and differentiate into classically activated (M1) macrophages expressing TNF-a and inducible nitric oxide synthase, which triggers inflammation and results in metabolic diseases [bib_ref] Type 2 diabetes as an inflammatory disease, Donath [/bib_ref] [bib_ref] Regulation of metabolism by the innate immune system, Lackey [/bib_ref]. By contrast, resident adipose-tissue macrophages in the lean state express IL-10 and arginase 1, and show an anti-inflammatory phenotype of alternatively activated (M2) macrophages [bib_ref] Type 2 diabetes as an inflammatory disease, Donath [/bib_ref] [bib_ref] Regulation of metabolism by the innate immune system, Lackey [/bib_ref]. Activation of the antiinflammatory M2 macrophages requires Th2 cytokines such as IL-4 and IL-13 [fig_ref] Figure 1: Anti-inflammatory functions of eosinophils [/fig_ref] , whereas activation of M1 macrophages requires Th1 cytokine IFN-g [bib_ref] Type 2 diabetes as an inflammatory disease, Donath [/bib_ref] [bib_ref] Regulation of metabolism by the innate immune system, Lackey [/bib_ref]. Eosinophils are a major IL-4-expressing cell population in adipose tissue and help to sustain M2 macrophages (12) [fig_ref] Figure 1: Anti-inflammatory functions of eosinophils [/fig_ref]. High-fat diet-induced obesity significantly reduces the number of eosinophils in adipose tissue [bib_ref] Eosinophils sustain adipose alternatively activated macrophages associated with glucose homeostasis, Wu [/bib_ref]. Therefore, eosinophil-deficient mice show aggravation of obesity-induced metabolic diseases via a dramatic reduction in the number of M2 macrophages in adipose tissue, indicating that eosinophils are intimately involved in metabolic homeostasis [bib_ref] Eosinophils sustain adipose alternatively activated macrophages associated with glucose homeostasis, Wu [/bib_ref]. Besides, ILC2 cells promote accumulation of eosinophils and activate M2 macrophages in adipose tissue through production of IL-5 and IL-13 and are implicated in metabolic homeostasis [bib_ref] Innate lymphoid type 2 cells sustain visceral adipose tissue eosinophils and alternatively..., Molofsky [/bib_ref] [bib_ref] Group 2 innate lymphoid cells promote beiging of white adipose tissue and..., Brestoff [/bib_ref].
Eosinophils are also involved in the biogenesis of beige fat [bib_ref] Eosinophils and type 2 cyotkine signaling in macrophages orchestrate development of functional..., Qiu [/bib_ref] [bib_ref] Eosinophils in fat: pink is the new brown, Lee [/bib_ref]. The latter represents clusters of uncoupling protein 1 (UCP-1)-expressing adipocytes in white adipose tissue and functions in heat production under conditions of increased energy expenditure, such as cooling and exercise [bib_ref] Brown and beige fat: development, function and therapeutic potential, Harms [/bib_ref] [bib_ref] Adipose tissue browning and metabolic health, Bartelt [/bib_ref]. Thus, beige adipocytes contribute to suppression of obesity and alleviation of metabolic diseases. By contrast, white adipocytes do not express UCP-1 and are not involved in thermogenesis; instead, they participate in fat storage, contributing to obesity-induced metabolic diseases [bib_ref] Adipose tissue browning and metabolic health, Bartelt [/bib_ref]. IL-4 released by eosinophils induces the tyrosine hydroxylase expression of M2 macrophages under cold conditions and drives these macrophages to produce catecholamines such as norepinephrine [bib_ref] Eosinophils and type 2 cyotkine signaling in macrophages orchestrate development of functional..., Qiu [/bib_ref]. The macrophage-derived catecholamines trigger expression of the UCP-1 gene in subcutaneous white adipose tissue and are involved in the development of beige fat [bib_ref] Eosinophils and type 2 cyotkine signaling in macrophages orchestrate development of functional..., Qiu [/bib_ref]. In addition, meteorinlike (Metrnl), which is a hormone induced in muscle after exercise and in adipose tissue upon cold exposure, exerts its effect on eosinophils and increases their expression levels of Th2 cytokines such as IL-4 and IL-13, consequently promoting the development of (and thermogenesis in) beige fat. In particular, ILC2 cells contribute to the biogenesis of beige fat through production of methionine-enkephalin peptide as well as via activation by IL-33 receptor signaling [bib_ref] Group 2 innate lymphoid cells promote beiging of white adipose tissue and..., Brestoff [/bib_ref].
Increasing evidence suggests that obesity and other metabolic syndrome are associated with altered composition of the gut microbiota [bib_ref] The intestinal immune system in obesity and insulin resistance, Winer [/bib_ref]. In agreement with this notion, the transfer of gut microbiota from obese to germfree mice increases obesity more than gut microbiota from lean mice does [bib_ref] An obesity-associated gut microbiome with increased capacity for energy harvest, Turnbaugh [/bib_ref]. Microbiota depletion either by antibiotic treatment or in germ-free mice promotes beige fat biogenesis and alleviates obesity and other metabolic syndrome; these effects are mediated by eosinophil accumulation, enhanced Th2 cytokine signaling, and M2 macrophage polarization in white adipose tissue of microbiota-depleted mice [bib_ref] Microbiota depletion promotes browning of white adipose tissue and reduces obesity, Suárez-Zamorano [/bib_ref]. In this regard, intestinal eosinophils may inhibit obesity and other metabolic syndrome through a reduction of bacterial burden via IgA production [bib_ref] Eosinophils promote generation and maintenance of immunoglobulin-A-expressing plasma cells and contribute to..., Chu [/bib_ref] [bib_ref] IL-1b in eosinophilmediated small intestinal homeostasis and IgA production, Jung [/bib_ref]. In addition, because obesityinduced metabolic diseases show elevated levels of IL-1b [bib_ref] Type 2 diabetes as an inflammatory disease, Donath [/bib_ref] , intestinal eosinophils may alleviate inflammation by producing large amounts of IL-1Ra and thus alleviate metabolic diseases.
## Prospects
According to recent reports, eosinophils are a heterogeneous cell population and have different characteristics depending on the site of residence [bib_ref] Indigenous enteric eosinophils contraol DCs to initiate a primary Th2 immune response..., Chu [/bib_ref] [bib_ref] Granulocyte macrophage colony-stimulating factor-activated eosinophils promote interleukin-23 drive chronic colitis, Griseri [/bib_ref]. Although eosinophils are mainly distributed in the intestine, the phenotype of large-intestinal eosinophils is different from that of small-intestinal ones, and the number of largeintestinal eosinophils in a steady state is much smaller than that of their small-intestinal counterparts [bib_ref] Indigenous enteric eosinophils contraol DCs to initiate a primary Th2 immune response..., Chu [/bib_ref] [bib_ref] Granulocyte macrophage colony-stimulating factor-activated eosinophils promote interleukin-23 drive chronic colitis, Griseri [/bib_ref]. Moreover, small-intestinal eosinophils in response to GM-CSF produce large amounts of IL-1Ra and exert an anti-inflammatory function (11), whereas large-intestinal eosinophils increase production of proinflammatory cytokines TNF-a and IL-13 and promote colitis [bib_ref] Granulocyte macrophage colony-stimulating factor-activated eosinophils promote interleukin-23 drive chronic colitis, Griseri [/bib_ref]. These observations mean that eosinophils do not all have an identical function, but rather their function is likely controlled by local milieu. Thus, if anti-inflammatory eosinophils can be upregulated or inflammatory eosinophils can be converted into anti-inflammatory ones, then these modalities should help to combat inflammatory diseases such as allergies, inflammatory bowel disease, and obesity-related metabolic diseases. To this end, it would be worthwhile to identify the surface markers that can distinguish homeostatic and inflammatory eosinophils. The extract of Ulmus davidiana var. japonica Nakai increases the number of small-intestinal eosinophils and suppresses differentiation and/or proliferation of Th1 and Th17 cells [bib_ref] Ulmus davidiana var. japonica Nakai upregulates eosinophils and suppresses Th1 and Th17..., Lee [/bib_ref]. Regulatory eosinophil-recruiting prebiotics or probiotics may be useful for prevention of inflammatory and metabolic diseases. More studies should be conducted in this field for successful development of eosinophil-targeting treatments.
[fig] Figure 1: Anti-inflammatory functions of eosinophils. Small-intestinal eosinophils suppress differentiation and/or proliferation of Th17 cells via production of a large amount of IL-1Ra and are involved in homeostasis of intestinal immunity. Meanwhile, eosinophils in adipose tissue activate M2 macrophages through IL-4 expression and inhibit inflammation, thereby contributing to metabolic homeostasis. [/fig]
|
RNA-Seq with a novel glabrous-ZM24fl reveals some key lncRNAs and the associated targets in fiber initiation of cotton
Background: Cotton fiber is an important natural resource for textile industry and an excellent model for cell biology study. Application of glabrous mutant cotton and high-throughput sequencing facilitates the identification of key genes and pathways for fiber development and cell differentiation and elongation. LncRNA is a type of ncRNA with more than 200 nt in length and functions in the ways of chromatin modification, transcriptional and post-transcriptional modification, and so on. However, the detailed lncRNA and associated mechanisms for fiber initiation are still unclear in cotton.Results:In this study, we used a novel glabrous mutant ZM24fl, which is endowed with higher somatic embryogenesis, and functions as an ideal receptor for cotton genetic transformation. Combined with the high-throughput sequencing, fatty acid pathway and some transcription factors such as MYB, ERF and bHLH families were identified the important roles in fiber initiation; furthermore, 3,288 lncRNAs were identified, and some differentially expressed lncRNAs were also analyzed. From the comparisons of ZM24_0 DPA vs ZM24_-2 DPA and fl_0 DPA vs ZM24_0 DPA, one common lncRNA MSTRG 2723.1 was found that function upstream of fatty acid metabolism, MBY25-mediating pathway, and pectin metabolism to regulate fiber initiation. In addition, other lncRNAs MSTRG 3390.1, MSTRG 48719.1, and MSTRG 31176.1 were also showed potential important roles in fiber development; and the co-expression analysis between lncRNAs and targets showed the distinct models of different lncRNAs and complicated interaction between lncRNAs in fiber development of cotton.Conclusions: From the above results, a key lncRNA MSTRG 2723.1 was identified that might mediate some key genes transcription of fatty acid metabolism, MYB25-mediating pathway, and pectin metabolism to regulate fiber initiation of ZM24 cultivar. Co-expression analysis implied that some other important lncRNAs (e.g., MSTRG 3390.1, MSTRG 48719.1, and MSTRG 31176.1) were also showed the different regulatory model and interaction between them, which proposes some valuable clues for the lncRNAs associated mechanisms in fiber development.
# Introduction
Cotton (Gossypium spp.) is one of the most important cash crops in the world because its main product fiber is the important natural source for the textile industry. In the four cultivars of Gossypium genus (G.hirsutum, G. barbadense, G. arboreum, and G.raimondii), G hirsutum (upland cotton) is the most widely planted due to its high yields and adaptability [bib_ref] Identifying QTL for fiber quality traits with three upland cotton (Gossypium hirsutum..., Shao [/bib_ref]. The period of cotton Open Access *Correspondence: [email protected]; [email protected] [bib_ref] A New Fuzzless Seed Locus in an Upland Cotton (Gossypium hirsutum L.)..., Bechere [/bib_ref] Zhengzhou Research Base, State Key Laboratory of Cotton Biology, School of Agricultural Sciences, Zhengzhou University, Zhengzhou 450001, China Full list of author information is available at the end of the article fiber development has been classified into four stages: initiation, elongation, secondary cell wall deposition, and maturity of fiber [bib_ref] Cotton fiber: a powerful singlecell model for cell wall and cellulose research, Haigler [/bib_ref]. The first two stages could determine the number and length of fibers, further affecting fiber yields. Consequently, many studies have been documented to explore the underlying genetic mechanisms related to fiber initiation and elongation, contributing to cotton production improvement [bib_ref] Functional analysis of the seed coat-specific gene GbMYB2 from cotton, Huang [/bib_ref] [bib_ref] Recent Advances and Future Perspectives in Cotton Research, Huang [/bib_ref] [bib_ref] The MYB transcription factor GhMYB25 regulates early fibre and trichome development, Machado [/bib_ref] [bib_ref] PAG1, a cotton brassinosteroid catabolism gene, modulates fiber elongation, Yang [/bib_ref].
Cotton mutants with fibreless, fuzzless, and lintless are good materials for studying the mechanism of fiber initiation development. With the auxin and gibberellin (GA) application in two fibreless mutants of Asian cotton in vitro culture, it showed that fiber cells differentiated from ovule epidermis at a temperature lower than 30 degrees, but not above 32 degrees, which indicated the important roles of auxin and GA in fiber development promotion at some specific conditions [bib_ref] Temperature-dependent Response to Indoleacetic Acid Is Altered by NH (4) in Cultured..., Beasley [/bib_ref]. SNPs comparison obtained by RNA-Seqs showed that glabrous mutant Xu142fl may be the progeny of G. barbadense. Based on the F 2 and BC 1 population between TM-1 and Xu142fl, the Li 3 gene encoding an MYB-MIXTA-like transcription factor was mapped and adjacent to MYB25-like in the D12 chromosome [bib_ref] Genetics and evolution of MIXTA genes regulating cotton lint fiber development, Wu [/bib_ref]. The inheritance evaluation of fuzzless seed in segregation population suggested that the interaction of three loci (N 1 , n 2 and n 3 ) contributed to fuzzless seed [bib_ref] Identification of a third fuzzless seed locus in upland cotton, Turley [/bib_ref] , among which two loci, N 1 and n 2 , located on a pair of homologous chromosomes A12/D12 [bib_ref] Identification of ten chromosome deficiencies of cotton, Endrizzi [/bib_ref]. The plants of N 1 N 1 homozygous and N 1 n 1 heterozygous produced fuzzless seeds [bib_ref] Identification of a third fuzzless seed locus in upland cotton, Turley [/bib_ref]. The n 3 locus that could produce the fibreless seed was identified by genetic analysis of cross progeny between N 1 N 1 and n 2 n 2 [bib_ref] Identification of a third fuzzless seed locus in upland cotton, Turley [/bib_ref]. The fourth locus, named n t 4 n 4 t , was identified from ethyl methanesulfonate (EMS) induced mutation analysis, whose homozygous seed exhibited a partially naked phenotype [bib_ref] A New Fuzzless Seed Locus in an Upland Cotton (Gossypium hirsutum L.)..., Bechere [/bib_ref]. All these fiber development defect mutants provide suitable materials for fiber development study.
With the advantage of Next Generation Sequencing (NGS), RNA-Seq as one of the NGS has been widely used to reveal expressions of genes and transcripts, among which some transcripts have been identified as non-coding RNA (ncRNA) because of their limitation of coding proteins. NcRNA includes microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and so on, which have emerged as key regulators of gene expression through their direct and indirect actions on chromatin [bib_ref] Gene regulation by the act of long non-coding RNA transcription, Kornienko [/bib_ref] [bib_ref] RNA discrimination, Kowalczyk [/bib_ref] [bib_ref] Long noncoding RNAs: past, present, and future, Kung [/bib_ref]. In Oryza sativa, 1,254 differentially expressed lncRNAs (DELs) were identified from BIL progenies. Another RNA-Seq showed that 328 of 444 DELs were associated with meiosis and the low fertility in autotetraploid rice [bib_ref] Global identification and analysis revealed differentially expressed lncRNAs associated with meiosis and..., Li [/bib_ref]. The lncRNAs were also involved in abiotic stress such as drought and rewatering in Brassica napus [bib_ref] Genome-wide analysis of long non-coding RNAs (lncRNAs) in two contrasting rapeseed (Brassica..., Tan [/bib_ref] , and osmotic and salt stress in Medicago truncatula [bib_ref] Identification and characterization of long non-coding RNAs involved in osmotic and salt..., Wang [/bib_ref]. The differences in genes expressions and regulations between fibreless mutants and wild-type have been investigated using omics methods [bib_ref] RNA-Seq-Mediated Transcriptome Analysis of a Fiberless Mutant Cotton and Its Possible Origin..., Ma [/bib_ref] [bib_ref] Differential expression of micro-RNAs during fiber development between fuzzless-lintless mutant and its..., Sun [/bib_ref] [bib_ref] A comparative miRNAome analysis reveals seven fiber initiation-related and 36 novel miRNAs..., Wang [/bib_ref]. With fiberless mutant Xu142fl and its counterpart Xu142, a previous comparative small RNAome analysis uncovered a possible network of fiber initiation-related miRNAs in cotton ovules, which comprises seven miRNAs expressed in cotton ovules, and each of them bears functional specific targets [bib_ref] A comparative miRNAome analysis reveals seven fiber initiation-related and 36 novel miRNAs..., Wang [/bib_ref]. Another work showed that 54 miR-NAs are differentially expressed in fiber initiation between Xu142fl and its wild-type, which are potentially targeted to TFs such as MYB, auxin response factor, and Leucine repeat receptor [bib_ref] Differential expression of micro-RNAs during fiber development between fuzzless-lintless mutant and its..., Sun [/bib_ref]. Using multiomics, the differentially expressed genes (1,953), proteins (187), and phosphoproteins (131) were identified by the comparison of Xu142 and Xu142fl [bib_ref] RNA-Seq-Mediated Transcriptome Analysis of a Fiberless Mutant Cotton and Its Possible Origin..., Ma [/bib_ref]. Genetic markers including 302 SNPs for fiber development were also developed and validated based on a deep sequencing between Xu142 and Xu142fl [bib_ref] RNA-Seq-Mediated Transcriptome Analysis of a Fiberless Mutant Cotton and Its Possible Origin..., Ma [/bib_ref]. In particular, a transcriptomic repertoire revealed that 645 and 651 lncRNAs were preferentially expressed in Xu142fl and Xu142, respectively. Further study showed that down-regulating two lncRNAs XLOC_545639 and XLOC_039050 in Xu142 fl increased the fiber initials on the ovules, while silencing XLOC_079089 in Xu142 shortened the fiber length, indicating the important and diverse roles of lncRNAs in fiber development.
LncRNA is a type of ncRNA with more than 200 nt in length and without protein-coding abilities [bib_ref] Integrative annotation of human large intergenic noncoding RNAs reveals global properties and..., Cabili [/bib_ref] [bib_ref] Genomewide identification of long intergenic noncoding RNA genes and their potential association..., Zhou [/bib_ref]. Lots of evidence have shown that lncRNAs could regulate genes functions in the ways of chromatin modification, transcriptional and post-transcriptional modification, etc. [bib_ref] Long non-coding RNAs: insights into functions, Mercer [/bib_ref] , through which the lncRNAs play vital roles in plant growth, development [bib_ref] Long noncoding RNAs: new regulators in plant development, Zhang [/bib_ref] , and response to biotic [bib_ref] Long noncoding RNAs involve in resistance to Verticillium dahliae, a fungal disease..., Zhang [/bib_ref] and abiotic stresses [bib_ref] RNA discrimination, Kowalczyk [/bib_ref] [bib_ref] Genome-wide identification and functional prediction of novel and drought-responsive lincRNAs in Populus..., Shuai [/bib_ref] [bib_ref] Identification and characterization of wheat long non-protein coding RNAs responsive to powdery..., Xin [/bib_ref]. LncRNAs could be classified into long intergenic non-coding RNAs (lincRNAs), natural antisense non-coding RNAs (lnc-NAT), sense non-coding lncRNAs, and intronic lncR-NAs according to their location in the genome [bib_ref] Long non-coding RNAs and their functions in plants, Chekanova [/bib_ref] [bib_ref] Discovery and annotation of long noncoding RNAs, Mattick [/bib_ref] [bib_ref] Genome-wide profiling of long noncoding RNAs from tomato and a comparison with..., Wang [/bib_ref]. In cotton fiber development, the detailed regulation mechanism of lncRNA is still ambiguous. In this research, we introduced a natural lintless-fuzzless (ZM24fl) mutant from zhongmiansuo24 (ZM24) cultivar, which is easy for transformation [bib_ref] iTRAQ protein profile differential analysis between somatic globular and cotyledonary embryos reveals..., Ge [/bib_ref] [bib_ref] Differential gene expression of cotton cultivar CCRI24 during somatic embryogenesis, Wu [/bib_ref]. Further, strand-specific transcriptome sequencing was conducted to reveal the differential expression profile of genes in the fiber initiation periods including lint and fuzz initiation stages. The lncRNAs/PC-genes pairs are identified and analyzed to explore the potential key lncRNAs and the corresponding targets contributing to the fiber development differences between the ZM24fl and ZM24.
# Results
ZM24 (G. hirsutum) is a high-quality commercial cotton cultivar bred by the Chinese Cotton Research Institute; this cultivar exhibits excellent somatic embryogenesis potential, which makes it an ideal receptor for cotton genetic transformation [bib_ref] Differential gene expression of cotton cultivar CCRI24 during somatic embryogenesis, Wu [/bib_ref]. We identified a spontaneous mutation in ZM24 that resulted in the production of fuzzless and lintless seed and designated it as ZM24 fuzzless-lintless (ZM24fl), which would be indicated as fl in the following study.
## The difference of the fiber cell development between fl mutant and zm24
Firstly, we observed the vegetative development of fl and ZM24; the results showed that the development of the organs and tissues such as branches, leaves, bolls, and epidermal hair of leaves, and stems is normal in fl, and similar with ZM24, except that the boll size is smaller in fl due to fiber development defects [fig_ref] Figure 1: The microscope of ovules during fiber initiation stages in ZM24 and the... [/fig_ref].
To reveal the difference of fiber development between fl and ZM24 during fiber initiation stages, ovules at -2 day-post-anthesis (DPA), 0 DPA, 1 DPA and 2 DPA from two lines were stripped for observation using scanning electron microscopy. The scanning results showed that there were no significant differences on the epidermis of -2 DPA ovules in both lines [fig_ref] Figure 1: The microscope of ovules during fiber initiation stages in ZM24 and the... [/fig_ref]. The epidermis of ovules from 0 to 2 DPA, which indicated the fiber initiation stage, presents the clear differences of fibrous protuberance between ZM24 and fl. The fibrous protuberance appeared on the epidermis of 0 DPA ovules in ZM24, and they slowly elongated over time. However, no fiber was observed on the epidermis of ovules of 0 DPA, 1 DPA, and 2 DPA from fl [fig_ref] Figure 1: The microscope of ovules during fiber initiation stages in ZM24 and the... [/fig_ref] , which indicated that fl is an excellent line for fiber cell initiation study.
## Identification and characterization of lncrnas in cotton fiber initiation
LncRNAs have been shown the important roles in fiber development. To investigate the potent lncRNAs responsible for the fiber initiation, we performed strandspecific RNA sequencing on 6 ovule samples (-2 DPA, 0 DPA, and 5 DPA from ZM24 and fl) with twice replicates. A total of 362.86 Gb clean reads were obtained by removing low-quality reads, and all Q30 of RNA-Seq results are greater than 85%, indicating that clean reads were qualified for downstream analysis (Additional file 1). To assess the correlations between biological replicates, the Pearson correlation was calculated using FPKM assays and the r 2 between two replicates of each material was above 0.88, indicating the high correlation and reliable data . An integrated pipeline was used to identify lncRNA in these tissues (see details in materials and methods). Finally, 3,288 lncRNA transcripts were obtained, among which the lincRNAs (2,618), lncNATs (559), sense lncRNAs (78) and intronic lncRNAs [bib_ref] CUT1, an Arabidopsis gene required for cuticular wax biosynthesis and pollen fertility,..., Millar [/bib_ref] accounted for 79.6 %, 17 %, 2.4 %, and 1 %, respectively . The pattern of exons in lncRNA transcripts showed that single-exon lncRNAs represented the largest proportion, accounting for 43.8% . The physical positions and the class codes of lncRNAs were detailed in Additional file 2.
## Differentially expressed lncrnas and protein-encoding genes between fl and zm24
Furthermore, the clean reads were mapped to ZM24 genomes, and FPKM and counts of transcripts and genes were identified for further analysis with a threshold of |log 2 (FC)| > 1 and p-value < 0.05. In summary, a total of 371 differentially expressed lncRNAs (DELs) and 12,971 DEGs were identified in vertical (fl vs ZM24 at the same developmental stages) and horizontal (different development stages in the same line) comparisons, respectively. The FPKM values of the DELs and DEGs were shown in Additional files 3 and 4.
In detail, 87 (32 up-regulated, 55 down-regulated), 27 (5 up-regulated, 22 down-regulated), and 201 (133 up-regulated, 68 down-regulated) lncRNAs were differentially expressed in the -2, 0, and 5 DPA ovules of fl compared with that in ZM24, respectively , which indicated that many down-regulated lncRNAs in the -2 and 0 DPA ovules of fl genes play potential roles in fiber initiation. Correspondingly, there were 452 (347 up-and 105 down-regulated), 389 (25 up-and 364 downregulated), and 3,986 DEGs (1,378 up-and 2,608 downregulated) in fl compared to ZM24 at the three time points of fiber initiation stage, respectively , these results indicated that the lncRNAs might regulate target genes encoding proteins either negatively or positively.
To further investigate the DEGs between fl and ZM24 at different development stages, the DELs and DEGs of "0 DPA vs -2 DPA" and "5 DPA vs 0 DPA" were identified, and the Venn diagram established showed that 21 (10 upand 11 down-regulated) and 60 (30 up-and 30 down-regulated) DELs were identified in fl and ZM24, respectively, in the comparison of 0_DPA vs -2 DPA, of which only 3 common DELs (1 up-regulated, 2 down-regulated)) were identified in both comparisons. Similarly, 218 DELs were identified in comparison of 5 DPA vs 0 DPA, in which, 34 and 163 DELs were unique to fl and ZM24, respectively. Interestingly, in the other 21 common DELs, 17 and 3 DELs were up-and down-regulated at 5 DPA compared to 0 DPA; only one DELs, MSTRG.3394.1, was down regulated in fl_5DPA compared with fl_0DPA, but upregulated in that of ZM24. These results indicated that more lncRNA were involved in the fiber development than that in the ovule development. Consistently, a lot of DEGs were identified from the comparisons , suggesting that the dramatic transcription regulation differences between two lines might contribute to the development of the fiber in ZM24. These results proposed that the expression of genes including protein-encoding genes and lncRNA-encoding genes changed dramatically in the two lines, which may contribute to the formation of the glabrous seed of fl, however, the underlying mechanism is unclear.
## Go enrichment of degs
To investigate the functions of DEGs between fl and ZM24 in fiber initiation and elongation, the GO enrichment analysis of DEGs was performed to explore the causal pathways or genes responsible for the phenotype differences in two lines. According to the results of GO enrichment, the most significant (corrected p-value < 0.05) classes were identified. On -2 DPA, the biological process of regulation activity and exogenous and endogenous signals such as cold (GO:0009409), fungus (GO:0050832), jasmonic acid (GO:0009753), and chitin (GO:0010200) were mainly enriched in up-regulated genes, whereas the down-regulated DEGs were enriched in the cellular process such as "cytosol" (GO:0005829), "plastid" (GO:0009536), "postembryonic development" (GO:0009791) in fl compared with ZM24 [fig_ref] Figure 3: Gene ontology classifications of DEGs between fl vs ZM24 at -2 DPA... [/fig_ref]. On the day of flowering, the downregulated genes were enriched in terms of "wax biosynthetic process" (GO:0010025), "cuticle development" (GO:0042335), "3-oxo-lignoceronyl-CoA synthase activity" (GO:0102338), "3-oxo-cerotoyl-CoA synthase activity" (GO:0102337), and "fatty acid biosynthetic process" (GO:0006633), which were related to lipid metabolism [fig_ref] Figure 3: Gene ontology classifications of DEGs between fl vs ZM24 at -2 DPA... [/fig_ref] , indicating that cell wall plasticity regulated by lipid metabolism might play important roles in fiber cell initiation. On 5 DPA, the earlier fiber elongation stage, the DEGs of two lines exhibited more significant functions in cellular components [fig_ref] Figure 3: Gene ontology classifications of DEGs between fl vs ZM24 at -2 DPA... [/fig_ref]. For example, the down-regulated genes in "fl_5 vs ZM24_5" comparison was mostly enriched in terms of "plasma membrane" (GO:0005886), "cytoplasm" (GO:0005737), "integral component of membrane" (GO:0016021), etc., indicating that fiber development requires the expression and regulation of a large number of genes to bring cellular component effect.
## Identification of potential genes for fiber initiation
Generally, the phenotype is determined by genotype. Therefore, the reason why the fl exhibits a fibreless phenotype is mainly due to the differences in transcriptional regulation that happened during a few days before the flowering. To analyze and obtain genes sets related to fiber initiation, venn diagrams were conducted to show the common target between "ZM24_0 vs fl_0" and "ZM24_0 vs ZM24_-2". Subsequently, one common DEL, MSTRG2723.1, was identified between "ZM24_0 vs fl_0" and "ZM24_0 vs ZM24_-2" [fig_ref] Figure 4: The potential lncRNAs and genes for fiber initiation [/fig_ref] , which showed significantly higher expression in ovules of ZM24_0 DPA compared with fl_0 DPA and ZM24_-2 DPA, and targeted to 36 DEGs (Additional files 3 and 5). Moreover, 273 common DEGs between the up-regulated genes in ZM24_0 vs fl_0 and ZM24_0 vs ZM24_-2 comparisons were identified [fig_ref] Figure 4: The potential lncRNAs and genes for fiber initiation [/fig_ref]. The potential lncRNAs and genes for fiber initiation. a and b Venn diagrams showed the common lncRNAs and genes in DELs and DEGs from comparisons of "ZM24_0 vs fl_0" and "ZM24_0 vs ZM24_-2", respectively. gU represented the up-regulated genes. c The KEGG pathways of the commonly targeted genes of lncRNA and the common DEGs. d qPCR experiments confirmed the expression profiles of the common lncRNA (MSTRG.2723.1), its target genes, and some DEGs that enriched in fatty acid elongation and phenylpropanoid biosynthesis. The GhHistone3 (AF024716) gene was used as a reference gene, and the data were shown as mean ± SD. The student's t-test was used for the significance statistic
Hence, KEGG enrichment showed that these targeted genes and DEGs were significantly enriched in "fatty acid elongation" (12 genes), "cutin, suberine, and wax biosynthesis" (9 genes), "phenylpropanoid biosynthesis" (5 genes), "starch and sucrose metabolism" (4 genes), "ubiquinone and another terpenoid-quinone biosynthesis" (3 genes) and "fatty acid metabolism" (3 genes) pathways that possibly be related to fiber initiation [fig_ref] Figure 4: The potential lncRNAs and genes for fiber initiation [/fig_ref]. To validate the RNA-Seq results, the specific primers of the lncRNA MSTRG2723.1, and the randomly selected 11 genes from the targeted genes and DEGs were designed and used for Q-PCR. Relative expression levels of these genes in -2, 0, 2, and 5 DPA ovules were identified and consistent with the RNA-Seq. Of these, the transcription factor coding genes GhMYB25-like and GhMYB25 that were positive to fiber initiation were down-regulated in fl. In addition, the genes that were annotated as pectin lyase-like, pectinesterase, and a leucine-rich repeat (LRR) were down-regulated in fl compared to ZM24. Furthermore, the encoding 3-ketoacyl-CoA synthase genes Ghicr24_A01G004900 (KCS9) and Ghicr24_ D10G234500 (KCS19), targeted by MSTRG.2723.1, were significantly down-regulated in fl during the initiation stage, and enriched in the fatty acid elongation pathway [fig_ref] Figure 4: The potential lncRNAs and genes for fiber initiation [/fig_ref]. The down-regulated genes such as CUT1 and LONG CHAIN ACYL-COA SYNTHASE 1 (LACS1) are required for elongation of C24 very-longchain fatty acid (VLCFA) with a function in wax production [bib_ref] CUT1, an Arabidopsis gene required for cuticular wax biosynthesis and pollen fertility,..., Millar [/bib_ref]. The gene GhPAS2, enriched in the fatty acid metabolism pathway, was also characterized as the enzyme that catalyzes VLCFA production [bib_ref] Quantitative proteomics and transcriptomics reveal key metabolic processes associated with cotton fiber..., Wang [/bib_ref]. The genes 4-COUMARATE-COA LIGASE 1 (4CL1) and O-HYDROXYCINNAMOYLTRANSFERASE (HST) that were down-regulated in fl and enriched in phenylpropanoid biosynthesis pathway, affected the lignin composition [bib_ref] Silencing of hydroxycinnamoyl-coenzyme A shikimate/quinate hydroxycinnamoyltransferase affects phenylpropanoid biosynthesis, Hoffmann [/bib_ref]. These results indicated that the potential targeted genes involved in VLCFA by lncRNAs may play critical roles in fiber initiation.
## Identification of genes encoding transcription factors from the degs
Transcription factors (TFs) play vital roles in gene expression regulation and plant development. A considerable number of researches have shown that some TFs play key roles in fiber initiation and elongation [bib_ref] Functional analysis of the seed coat-specific gene GbMYB2 from cotton, Huang [/bib_ref] [bib_ref] The MYB transcription factor GhMYB25 regulates early fibre and trichome development, Machado [/bib_ref] [bib_ref] The R2R3 MYB transcription factor GhMYB109 is required for cotton fiber development, Pu [/bib_ref] [bib_ref] GhMYB25-like: a key factor in early cotton fibre development, Walford [/bib_ref] ]. An analysis of TFs from the DEGs set showed that 1,258 genes encode TFs, accounting for 9.7 % (1,258/12,971) DEGs, which were classified into 48 families. The largest group of TFs was the MYB (135) family, followed by the bHLH (120), ERF (110), C2H2 (78), and HD-ZIP (66) families, etc. [fig_ref] Figure 5: Transcription factors [/fig_ref]. Of these differentially expressed Given that MYB TFs play key roles in fiber initiation, hierarchical clustering analysis was performed to show the expression profiles of the 135 MYB TFs encoding genes in two lines [fig_ref] Figure 5: Transcription factors [/fig_ref]. These MYBs were grouped into 5 main clusters (A to E). The clusters C and A harbored 14 TFs each and genes in them are specifically expressed in -2 and 0 DPA, respectively. Furthermore, 6 genes in A2 were showed higher expression in ZM24 but low expression in fl on 0 DPA. Twenty-seven genes clustered in B1 were highly expressed in -2 and 0 DPA, whereas the other 8 MYB TFs in B2 showed high expression in all samples except at 5 DPA of ZM24. In clusters D and E, 11 genes in D1 and 22 genes in E2 were specifically expressed in 5 DPA. In addition, 18 MYBs in group D2 showed highest expression at 5 DPA of ZM24, with the remaining 21 MYBs clustered in E1 showed the highest expression at 5 DPA of fl. Several known positive MYB genes that have been reported to regulate fiber initiation [bib_ref] Functional analysis of the seed coat-specific gene GbMYB2 from cotton, Huang [/bib_ref] [bib_ref] The MYB transcription factor GhMYB25 regulates early fibre and trichome development, Machado [/bib_ref] [bib_ref] GhMYB25-like: a key factor in early cotton fibre development, Walford [/bib_ref] were selected for the further qPCR experiment, and most of which showed a down-regulated profile in fl [fig_ref] Figure 5: Transcription factors [/fig_ref]. In addition, two HD-ZIP TFs, GhHD1, and GhHOX3 that promoted fiber initiation and elongation by mediating accumulation of ethylene and reactive oxygen together [bib_ref] Control of cotton fibre elongation by a homeodomain transcription factor GhHOX3, Shan [/bib_ref] were down-regulated in fl at -2 DPA, 2 DPA, and 5 DPA. Furthermore, two TRANSPARENT TESTA GLABRA loci (TTG1 and TTG4) that were positive to fiber initiation also showed lower expression in fl than that of ZM24 during fiber initiation [bib_ref] Updates on molecular mechanisms in the development of branched trichome in Arabidopsis..., Wang [/bib_ref]. These results indicated that different MYBs and other TFs might exert specific roles during the different fiber developmental stages, which help to the identification of distinct TFs and the detailed gene transcription regulation mechanisms underlying the fiber initiation and primary development.
## Functions analysis of dels based on lncrna-mrna co-expression network
To further explore the roles of DELs, cis-targeted genes and trans-targeted genes of DELs were identified by calculating PCC values and distances between lncRNAs and mRNAs. Finally, 249 DELs were co-expressed with 5,501 DE_mRNAs. Thereby 70,987 pairs of lncRNA-mRNAs were identified in all the comparisons (Additional file 7). According to the expression profile, the targeted DEGs were summarized as shown in [fig_ref] Figure 6: Number of up-and down-regulated targeted genes of DELs in horizontal and vertical... [/fig_ref]. In ovules of -2, 0, and 5 DPA, 49, 170, and 2,102 targeted mRNAs were down-regulated in fl compared with ZM24. To comprehensively visualize the relationship between lncRNAs and the targeted mRNAs, the co-expression network among DELs, DE-mRNAs, and TFs at 0 DPA were shown using Cytoscape software . The center lncRNAs MSTRG.2723.1 and MSTRG.3390.1 are more important and independent lncRNAs that regulate many gene transcriptions including some known key TFs coding genes (e.g., MYB25, MYB25-Like, bZIP) in fiber initiation. Other types of lncRNAs such as MSTRG.31176.1 and MSTRG.48719.1 regulate genes transcription reciprocally, in which other different lncRNAs are also involved. Above these imply the diversity and complexity of molecular mechanisms by lncRNAs and provide a possible lncRNA-TFs regulation model in the fiber cell initiation.
# Discussion
With the glabrous mutant of cotton and high-throughput technology, generous coding-and non-coding RNAs and associated mechanisms have been identified in the fiber development including TFs, miRNAs and lncRNAs and so on [bib_ref] RNA-Seq-Mediated Transcriptome Analysis of a Fiberless Mutant Cotton and Its Possible Origin..., Ma [/bib_ref] [bib_ref] A comparative miRNAome analysis reveals seven fiber initiation-related and 36 novel miRNAs..., Wang [/bib_ref] [bib_ref] Identification and profiling of upland cotton microRNAs at fiber initiation stage under..., Zhao [/bib_ref]. Many studies on non-coding RNAs in cotton have been limited to small RNAs until now. For instance, a lot of miRNAs specifically expressed during anther development or callus were identified in male sterile cotton as well as cotton somatic embryogenesis [bib_ref] Comparative expression profiling of miRNA during anther development in genetic male sterile..., Wei [/bib_ref] [bib_ref] Identification of miRNAs and their targets using high-throughput sequencing and degradome analysis..., Yang [/bib_ref]. Gong et al. revealed the 33 conserved miR-NAs families between the A and D genomes [bib_ref] Composition and expression of conserved microRNA genes in diploid cotton (Gossypium) species, Gong [/bib_ref]. On the genomic level, the expression of 79 miRNAs families was studied and 257 novel miRNAs were identified related to cotton fiber elongation [bib_ref] Genome-wide analysis of small RNAs reveals eight fiber elongation-related and 257 novel..., Xue [/bib_ref]. In addition, two key miR828 and miR858 were proved the roles in the regulation of homoeologous MYB2 (GhMYB2A and GhMYB2D) in G. hirsutum fiber development [bib_ref] miR828 and miR858 regulate homoeologous MYB2 gene functions in Arabidopsis trichome and..., Guan [/bib_ref].
As a kind of long non-coding RNA, lncRNA provides more regulatory mechanisms for gene expression, protein synthesis, chromatin remodeling etc., while it is not clear about the detailed lncRNAs and the underlying mechanism in fiber development. A previous study identified 30,550 lincRNAs loci and 4,718 lncNATs loci, which are rich in repetitive sequences and preferentially expressed in a tissue-specific manner with weak evolutionary conservation. Further, lncRNAs showed overall higher methylation levels, and their expression was less affected by gene body methylation [bib_ref] Long noncoding RNAs and their proposed functions in fibre development of cotton..., Wang [/bib_ref]. Using the epidermal cells from the ovules at 0 and 5 DPA from Xu142 and Xu142fl, 35,802 lncRNAs and 2,262 circular RNAs (circRNAs) were identified, of which 645 lncRNAs were preferentially expressed in the fibreless mutant Xu142fl and 651 lncRNAs were preferentially expressed in the fiber-attached lines; three lncRNAs XLOC_545639, XLOC_039050, and XLOC_079089 all showed the solid function in fiber development by VIGS assay. Here, a novel glabrous mutant-ZM24fl, which showed excellent somatic embryogenesis induction was used to identify the key lncRNA involved in fiber initiation development [bib_ref] Differential gene expression of cotton cultivar CCRI24 during somatic embryogenesis, Wu [/bib_ref].
Totally, 3,288 lncRNA transcripts were identified from the -2 DPA, 0 DPA and 5 DPA ovules of ZM24 and fl, which is significantly different from the number of identified lncRNA in Xu142fland G. barbadense L. cv 3-79 [bib_ref] Long noncoding RNAs and their proposed functions in fibre development of cotton..., Wang [/bib_ref]. To identify the causal lncRNAs for fiber initiation, some comparisons were built to analyze the differentially expressed genes including lncRNAs and mRNAs The co-expression network between DELs, the targeted DE-mRNA and TFs. Four lncRNAs showing significant expression difference between ZM24 and fl were identified and presented here. The blue rectangle and rhombus indicate the TFs and potential associated lncRNA. Other targeted mRNA and lncRNA were shown with green and yellow circles. The network was visualized by Cytoscape 3.6.1 during fiber initiation and earlier elongation. The identified DELs and DEGs in comparisons of 0 DPA vs -2 DPA and 5 DPA vs 0 DPA of ZM24 and fl indicated that many lncRNAs and coding genes are involved in the fiber initiation and primary development, while few lncRNAs and coding genes may involve the ovule development. The analysis of the DEGs further showed that fatty acid metabolism, very long strain fatty acid synthesis and sugar metabolism play important roles in the fiber initiation of ZM24, supporting the previous results [bib_ref] AKR2A participates in the regulation of cotton fiber development by modulating biosynthesis..., Hu [/bib_ref] [bib_ref] The delayed initiation and slow elongation of fuzz-like short fibre cells in..., Ruan [/bib_ref]. Moreover, some MYB family, bHLH type TFs encoding genes were also identified the important roles in fiber initiation, which is in agreement with the function of these TFs in previous research [bib_ref] The HD-Zip IV gene GaHOX1 from cotton is a functional homologue of..., Guan [/bib_ref] [bib_ref] Functional analysis of the seed coat-specific gene GbMYB2 from cotton, Huang [/bib_ref] [bib_ref] The MYB transcription factor GhMYB25 regulates early fibre and trichome development, Machado [/bib_ref] [bib_ref] The R2R3 MYB transcription factor GhMYB109 is required for cotton fiber development, Pu [/bib_ref] [bib_ref] Control of cotton fibre elongation by a homeodomain transcription factor GhHOX3, Shan [/bib_ref] [bib_ref] GhMYB25-like: a key factor in early cotton fibre development, Walford [/bib_ref]. To uncover the upstream factors such as lncRNAs, we focused on the comparisons of ZM24_0 DPA vs fl_0 DPA and ZM24_0 DPA vs ZM24_-2 DPA to find the common lncRNAs which should be a key regulator for fiber initiation. Consequently, one lncRNA MSTRG 2723.1 was obtained, which locates on the A02G (84218766-84219942) encoding a lncNAT and covering the most coding region and partial 3'-terminal untranslated region of Ghicr24_A02G147600 [fig_ref] Figure 4: The potential lncRNAs and genes for fiber initiation [/fig_ref]. The co-expression analysis further identified its potent targets including 3-ketoacyl-CoA synthase, MYB family proteins, phosphatase 2C family proteins, pectin lysase, and some uncharacterized proteins, which may are involved in fiber initiation through fatty acid pathway, cell wall plasticity, MYB-mediated signaling etc. These results provide important clues for the upstream regulatory lncRNAs in fiber initiation and novel information associated with the fiber development regulation network. In addition, MSTRG 3390.1, MSTRG 48719.1, and MSTRG 31176.1 were also identified some positive correlation between fiber development and ovule development. The sequence analysis indicated that these lncRNAs are different from the previous lncRNAs XLOC_545639, XLOC_039050, and XLOC_079089. The target analysis also implied the possible interaction between different lncRNAs through mediating the common targets, which provide novel clues to explore the regulatory lncRNAs and underlying mechanisms in fiber development. Even with some achievement of lncRNAs, the understanding of the underlying mechanism of lncRNAs regulating targets or chromosome remodeling still needs more work to disclose.
# Conclusion
Here, a novel glabrous cotton mutant ZM24fl was identified and applied to study the potential lncRNAs for fiber development with high-throughput sequencing. ZM24fl is derived from an elite cultivar of ZM24, which posses high callus induction and somatic embryogenesis ability, and is endowed with the valuable receptor for cotton genetic transformation [bib_ref] Differential gene expression of cotton cultivar CCRI24 during somatic embryogenesis, Wu [/bib_ref]. Through the RNA-Seq and analysis in different ovules of ZM24 and fl, 3,288 lncR-NAs were identified and some differentially expressed lncRNAs responsible for fiber (lint and fuzz) initiation and fiber earlier elongation were showed. Collectively, four lncRNAs MSTRG.2723.1, MSTRG.3390.1, MSTRG.48719.1 and MSTRG.31176.1 were showed potential important roles in fiber development, and the analysis of the target implied that MSTRG 2723.1 may function upstream of fatty acid metabolism, MBY25mediating pathway, and pectin metabolism to regulate fiber initiation; the co-expression analysis between lncR-NAs and targets further indicated the distinct models of different lncRNAs and interaction between lncRNAs, which provide precious information for illumination of the molecular mechanism of lncRNAs in fiber development of cotton.
# Materials and methods
# Plant materials
Gossypium hirsutum L. acc. Zhongmiansuo24 (ZM24) and a natural fuzzless-lintless (fl) mutant from ZM24 were used and grown under standard field conditions in the Institute of Cotton Research of the Chinese Academy of Agricultural Sciences (Zhengzhou research base, Henan). The ovule tissues were collected from cotton bolls on -2, 0, and 5 DPA using a sterile knife. All materials were frozen in liquid nitrogen immediately and stored at -80 °C for the following experiments.
## Microscopic observation of fiber initiation on ovules epidermis
To study the fiber initiation phenotypes of ZM24 and fl, the cotton bolls of two lines on -2, 0, 1 and 2 DPA were collected. Then, the ovules were stripped from the bolls in the middle region. Immediate Scanning electron microscopy (Hitachi) was performed to observe the ovule epidermis as described previously.
## Strand specific libraries construction and sequencing
Total RNAs of each ovule sample was extracted using the RNAprep Pure Plant Kit (Tiangen, Beijing, China) following the manufacturer's instruction. Total RNAs of each sample was quantified and qualified by Agilent 2100 Bio-analyzer (Agilent Technologies, Palo Alto, CA, USA), Nanodrop 2000 (Thermo Fisher Scientific Inc.), and 1% agarose gel. RNA with RIN value above 7 was used for following library construction. The rRNA was removed using the Ribo-Zero ™ rRNA removal Kit. The ribosomal depleted RNA was then used for sequencing library preparation according to the manufacturer's protocol (NEBNext ® Ultra ™ Directional RNA Library Prep Kit for Illumina ® ). The cDNA libraries with different indices were multiplexed and loaded on an Illumina Hiseq2500 with 150 base pair (bp) paired-end (PE150) raw reads according to the manufacturer's instruction (Illumina, San Diego, CA, USA). RNA-Seq raw data with accession number SRP285346 was uploaded in the NCBI sequence read archive (http://www.ncbi. nlm.nih.gov/sra/) and the accession numbers of the twenty-four runs are SRR12710181-SRR12710192, and SRR12718970-SRR12718981.
## Mapping to the reference genome and lncrnas identification
The raw data in fastq format were filtered with cutadapter (v1.9.1) software [bib_ref] Cutadapt removes adapter sequences from high-throughput sequencing reads, Martin [/bib_ref]. Clean data were obtained by removing reads that contained adapter, poly-N and base with Phred quality < 20 in 3' or 5' end, and the reads of length < 75 bp were removed after filtering. Finally, the GC percentage and Q30 of each sample were calculated using FastQC software (https://www. babraham.ac.uk/) and shown in . Clean data were mapped to the ZM24 genome (https://github.com/ gitmalm/Genome-data-of-Gossypium-hirsutum/)
[66] using HISAT(v2.1.0) [bib_ref] HISAT: a fast spliced aligner with low memory requirements, Kim [/bib_ref] [bib_ref] Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype, Kim [/bib_ref] software with the parameter "--rna-strandness RF". Transcriptomes of each sample were assembled based on mapped reads and were merged by StringTie software (v2.0) [bib_ref] StringTie enables improved reconstruction of a transcriptome from RNAseq reads, Pertea [/bib_ref]. Transcripts annotation was performed using Cuffcompare [bib_ref] Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching..., Trapnell [/bib_ref]. Long non-coding RNA was identified as following steps: 1) transcripts with class codes of "i", "u", "x", "j" representing the intronic transcripts, long intergenic noncoding RNAs (lincRNAs), long noncoding natural antisense transcripts (lncNAT), and the sense transcripts, respectively, were selected. 2) Transcripts with length > 200 bp, coverage > 1, FPKM > 0.5; 3) The CNCI [bib_ref] Utilizing sequence intrinsic composition to classify protein-coding and long non-coding transcripts, Sun [/bib_ref] , CPC [bib_ref] CPC: assess the protein-coding potential of transcripts using sequence features and support..., Kong [/bib_ref] and Pfam-Scan software were used to assessed protein-coding ability [bib_ref] The Pfam protein families database: towards a more sustainable future, Finn [/bib_ref] , with the parameter of (CPC score < 0, CNCI score < 0).
# Differential expression analysis
The FPKM values and counts of genes and lncRNAs in each sample were calculated using StringTie and Ballgon. Differential expression analyses were conducted by edgeR in R package [bib_ref] Moderated statistical tests for assessing differences in tag abundance, Robinson [/bib_ref]. The DEGs and DELs were identified with an expression FPKM > 1.0, FDR (false discovery rate < 0.001), and |log2( fold change value)| ≥1 between each pairwise comparison.
## Co-expression analysis between lncrna and mrna
To unveil the potential functions of DELs between the two genotypes, two interaction models of lncRNAs and protein-coding genes (lncRNAs/PC-genes) including cis-and trans-target were analyzed: 1) the Pearson correlation coefficient (PCC) between differentially expressed lncRNAs and mRNAs were calculated using the OmicShare tools (https://www.omicshare.com/) with the expression profiles (FPKM). The lncRNA-mRNA pairs with |PCC| > 0.95 and p-value < 0.01 were regarded as trans interaction between lncRNAs and mRNAs. 2) Protein-coding genes with a distance less than 20 kb from the upstream or downstream of lncRNAs were putative cis interaction. The co-expression networks were visualized by Cytoscape 3.6.1 [bib_ref] Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks, Shannon [/bib_ref].
## Go and kegg
To explore the functions of DEGs and lncRNAs between ZM24 and fl, the gene ontology (GO) enrichment was performed using the BLASTP program [bib_ref] Basic local alignment search tool, Altschul [/bib_ref] and GO databases (http:// archi ve. geneo ntolo gy. org/ latest-lite/) and (http:// ftp. ncbi. nlm. nih. gov/ gene/ DATA/). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed at KOBAS 3.0 website [bib_ref] KOBAS server: a web-based platform for automated annotation and pathway identification, Wu [/bib_ref] [bib_ref] KOBAS 2.0: a web server for annotation and identification of enriched pathways..., Xie [/bib_ref] (http:// kobas. cbi. pku. edu. cn/ kobas3).
# Q -pcr analysis
Ovules from bolls at -2, 0, 2, and 5 DPA were collected, and then total RNAs were extracted using the RNAprep Pure Plant Kit (Polysaccharides & Polyphenolics-rich, Tiangen, Beijing, China) following the manufacturer's instruction. Each reverse-transcribed reaction was performed with 1 μg RNA using a transScript ® First-Strand cDNA Synthesis SuperMix (AT301-02, TransGen). The real-time PCR was performed on Roche 480 PCR system with a SYBR-Green Real-time PCR SuperMix (AQ101-01, TransGen). The 20 uL reaction volumes in each well contain 1 μL cDNA, 8.2 μL sterile water, 10 μL Mix, and 0.4 μL each of the forward and reverse primers. The Q-PCR procedures were as: pre-incubation of 30 s at 95 °C; followed by denaturation at 95 °C for 10 s, primer annealing at 55 °C for 10 s, and then extension at 72 °C for 30 s; finally, a melting curve at 95 °C for 30 s to check the primer specificity. The GhHistone3 (AF024716) gene was used as a reference gene. The 2 -∆Ct method was used to calculate the relative expression of each gene, with three technical repetitions and three biological repetitions. Data were shown as mean ± SD. The student's t-test was used for the significance statistic. The primer sequences used in the presented study are listed in Additional file 9.
[fig] Figure 1: The microscope of ovules during fiber initiation stages in ZM24 and the fl mutant. The SEM photographs of ovules in ZM24 (left) and ZM24 fl (right) on -2, 0, 1 and 2 DPA. All ovules were taken from the same position of the bolls in a similar position on each plant. bars, 40 μm (magnified) and 150 μm [/fig]
[fig] Figure 2 61, Figure 2: An integrative computational pipeline for the bioinformatics analysis and summary of DELs and DEGs in the RNA-Seq. a Bioinformatics methods for the identification of lncRNA. b The number of lincRNA, lncNAT, sense lncRNA, and ntronic lncRNA. c Exon number pattern of four kinds of lncRNA. d and e Number of differentially expressed lncRNAs and genes (up-and down-regulated) between fl and ZM24, respectively, in three time points (-2, 0, and 5 DPA) during fiber initiation. f and g Venn diagrams showing the common DELs and DEGs between comparisons of '0 DPA vs -2 DPA' and '5 DPA vs 0 DPA' of fl and ZM24, respectively. DELs and DEGs were identified with log 2 | (fold change)| > 1, FDR < 0.001 (See figure on next page.) Zou et al. BMC Plant Biology (2022) 22:See legend on previous page.) [/fig]
[fig] Figure 3: Gene ontology classifications of DEGs between fl vs ZM24 at -2 DPA and 0 PDA. The most highly enriched GO terms showed the 105 down-and 347 up-regulated genes in ovules of -2 DPA (a), and the 364 down-and 25 up-regulated genes in ovules of 0 DPA of ZM24 comparing with fl (b), respectively [/fig]
[fig] Figure 4: The potential lncRNAs and genes for fiber initiation. a and b Venn diagrams showed the common lncRNAs and genes in DELs and DEGs from comparisons of "ZM24_0 vs fl_0" and "ZM24_0 vs ZM24_-2", respectively. gU represented the up-regulated genes. c The KEGG pathways of the commonly targeted genes of lncRNA and the common DEGs. d qPCR experiments confirmed the expression profiles of the common lncRNA (MSTRG.2723.1), its target genes, and some DEGs that enriched in fatty acid elongation and phenylpropanoid biosynthesis. The GhHistone3 (AF024716) gene was used as a reference gene, and the data were shown as mean ± SD. The student's t-test was used for the significance statistic [/fig]
[fig] Figure 5: Transcription factors (TFs) identification and analysis in DEGs. a The classifications of the 1,258 TFs. b Hierarchical clustering heatmap of the 135 MYB encoding genes, which were grouped into five clusters according to their expression profile. c qRT-PCR of the eight genes encoding different TFs that have been documented to affect fiber initiation and elongation. The GhHistone3 (AF024716) gene was used as reference gene, and data were shown as mean ± SD. The student's t-test was used for the significance statisticIn the comparison of "in "fl_0 vs fl_5" and "ZM24_0 vs ZM24_5", respectively. Thus, the KEGG enrichment of targeted DE-mRNAs was performed to investigate the functions of DELs. As a result, of the down-regulated targeted mRNAs in fl related to ZM24, genes were significantly enriched in fatty acid elongation and lipid metabolism at 0 and 5 DPA. Of the up-regulated targeted mRNAs, DE-mRNAs at 5 DPA related to 0DPA in fl were mainly enriched in amino acid metabolism, phenylpropanoid biosynthesis, fatty acid degradation and flavonoid biosynthesis pathways; however, the pathways of fatty acid elongation, lipid metabolism, carbohydrate metabolism, energy metabolism were enriched by DE-mRNAs of that in ZM24 (Additional file 8). [/fig]
[fig] Figure 6: Number of up-and down-regulated targeted genes of DELs in horizontal and vertical comparisons between fl and ZM24. Seven comparisons were established and the differentially expressed lncRNAs were identified and summarized [/fig]
|
Support for Local Tobacco Policy in a Preemptive State
Policy at the local level is a critical component of comprehensive tobacco control programs. This study examined the relationships of individual and social factors with support for tobacco-related public policy using cross-sectional data (n = 4461) from adults participating in a statewide survey. Weighted multivariate, multinomial logistic regression examined associations between individual and social factors and support for tobacco-free city properties and support for limiting the number of stores that sell tobacco near schools. Oklahomans were more likely to favor policies that create tobacco-free city properties than policies that limit the number of stores that sell tobacco near schools. While non-smokers were most likely to favor both policies, support for both policies was greater than 50% among current smokers. Knowledge of secondhand smoke (SHS) exposure harm and female gender were predictors of support for both policies and among current, former, and never smokers. Rural-urban status was a predictor of support among former smokers and never smokers. Tobacco use among friends and family was only a predictor among never smokers' support for limiting the sale of tobacco near schools. This study demonstrates that level of support differs by policy type, individual smoking status, as well as among subpopulations, and identifies critical elements in the theory of change for tobacco control programs.
# Introduction
Oklahoma has one of the highest tobacco prevalence rates in the nation. According to 2017 BRFSS data, Oklahoma's smoking prevalence was 20.1% compared to the national average of 14.0% . Best practices, informed by years of research, have established the critical need for comprehensive state tobacco control programs that include local policy measures proven to reduce tobacco use and its associated morbidity and mortality.
"Oklahoma has been slow to experience statewide tobacco control and prevention policy wins that have been associated with reduced prevalence in other states. This is due in part to the powerful presence and influence the tobacco industry lobby has had in state government". Oklahoma laws governing clean indoor air and youth access include tobacco industry supported preemptive language. The state law related to clean indoor air includes exemptions for bars, restaurants, outdoor seating areas of restaurants, hotels, in-home child care centers when children are not present, workplaces with incidental public access, and allows for all workplaces to provide smoking rooms in which no work is performed. The law preempts municipalities from enacting laws more stringent than state law but does include clarifying language that states local municipalities have the discretion to prohibit smoking in or on property owned or operated by their own governing body. State law related to preventing youth access to tobacco prohibits municipalities from enacting any law related to the sale, purchase, distribution, advertising, sampling, promotion, display, possession, licensing, influence related to tobacco; (4) knowledge of the harm related to secondhand smoke (SHS) exposure; and (5) other demographic factors are associated with policy support.
This study showed that the majority of Oklahomans favor policies that establish city property as tobacco-free and that limit the sale of tobacco near schools. While non-smokers were most likely to favor each of the policies, support was also high among former smokers, and support for both policies was greater than 50% among current smokers. Female gender and knowledge of the harm associated with SHS exposure were consistent, strong predictors of support for both policies and among current, former, and never smokers, indicating a continued need for education about the adverse effects of exposure to SHS.
# Materials and methods
The study design is cross-sectional with two waves of data collection through a telephone survey of land line and cell phones. This paper includes baseline data from the first wave of data collection in 2016. The target population for the survey consisted of non-institutionalized adults residing in one of the 63 counties served by the TSET HLP. Random samples were used to complete surveys via landlines (n = 1097) and cell (n = 3364) telephone numbers, for a total sample size of 4461. A stratified sampling design was used to ensure that each of six pre-defined regions in Oklahoma was equally represented. The response rate was 10% (AAPOR RR1). Weighted estimates were calculated and adjusted for non-coverage and non-response, creating estimates more representative of the Oklahoma population residing in the HLP counties. All subjects gave their oral verbal informed consent for inclusion before they participated in the study. The study was approved by the Institutional Review Board of the University of Oklahoma Health Sciences Center (#6351) on 01/22/2016.
## Outcome measures
Participants were asked, "Do you favor, oppose, or neither favor nor oppose the following government policies?" a Policies that prohibit tobacco use on city-owned properties. b
Policies that limit the number of stores that sell tobacco near schools.
A 5-point Likert measurement scale was used. Respondents who answered "strongly favor" or "somewhat favor" were categorized as favoring the policy. Those that answered "strongly oppose" or "somewhat oppose" were categorized as opposing the policy. A third category of those answering "neither favor nor oppose" was also used.
## Smoking status
Current smoking status was determined by the answers to two questions. "Have you smoked at least 100 cigarettes in your life? And, "Do you now smoke cigarettes every day, some days, or not at all?" Current smokers were defined as those that had smoked at least 100 cigarettes in their life and now smoke every day or some days. Former smokers were defined as those that had smoked at least 100 cigarettes in their life and now do not smoke at all. Never smokers were defined as those that had not smoked at least 100 cigarettes in their life.
## Sociodemographic factors
Six geographical regions of the state were included: Northwest, Northeast, Tulsa, Central, Southwest, and Southeast. These regions reflect those used in statewide surveillance efforts such as implementation of the Behavioral Risk Factor Surveillance System. Rural-urban status was assigned at the county level. The urban category includes the three most populous counties in the state and accounts for about 43% of the state's population. The remaining TSET HLP counties were categorized as rural. Several sociodemographic measures were included, such as gender (female or male), age , race/ethnicity (White, Black or African American, American Indian or Alaska Native, Hispanic, Multiracial), education (high school or less, some college or more), income (≤$30,000, $30,000-<$45,000, $45,000-<$60,000, $60,000-<$80,000, $80,000-<$100,000, $100,000-<$150,000, ≥$150,000) and marital status (married, widowed, divorced or separated, never married, other).
## Social influence
Respondents were asked the question "How many of the people that are important to you smoke or use other tobacco products?" Answer options included: none of them, less than half of them, about half of them, more than half of them, and all of them. Based on the distribution of responses and methods used in previously published research examining the influence of family and friends smoking, a dichotomous version of this variable was created-none and any. Respondents that answered "none" were categorized as having none of the people important to them using tobacco. All other respondents were placed in the "any" category.
## Knowledge, perceptions, and awareness
For this analysis, a knowledge summary score was created based on responses to four questions. Respondents' scores ranged from 0 to 4 based on the number of correct responses ("Yes") to the knowledge-based questions. "Don't know" responses were combined with the "No" responses. The four knowledge questions reflect questions used in prior surveys and priorities within the state's health communication campaign, Tobacco Stops with Me. Does breathing smoke from other people's cigarettes cause asthma?
The summary score was used in the multivariate, multinomial analysis, and the mean score was reported among those who favor, oppose, and neither favor nor oppose each policy.
Respondents were asked the question "How serious of a problem is smoking and tobacco use in your community?" A 4-point Likert scale was used. Answer options included: not a serious problem, only a little serious, fairly serious, and very serious. A dichotomous version of this variable was created based on the distribution of responses and methods used in previous research assessing perceptions of tobacco use. Due to small sample sizes for "not a serious problem" and "only a little serious" as well as no significant differences between these response options and the "fairly serious" option, we grouped these as one category. Those answering "very serious" were placed in one category and those answering "not a serious problem, only a little serious, and fairly serious" were placed in the other category.
Awareness of local, community efforts addressing the issue of tobacco use was also included as a measure. Awareness was measured by the item, "Are you aware of any programs, activities, services, or policies to decrease tobacco use or exposure to secondhand smoke in your community?" Answer options were "yes" and "no". Some participants volunteered the answer "don't know". "Don't know" responses were combined with "no" responses.
# Statistical methods
Data were analyzed using SAS, version 9.4 (SAS Institute Inc., Cary, NC, USA). Weighted estimates with 95% confidence intervals for support of the two policies were calculated and stratified by covariates. Weighted multivariate, multinomial logistic regression models were constructed with three levels of responses to the policy questions: favor, neither favor nor oppose, and oppose. Oppose served as the reference category. Interaction was present by smoking status; thus, all models were stratified by three levels of smoking status: current smokers, former smokers and never smokers. All covariates were included as potential confounders in the models, using purposeful selection and based on previous research. Adjusted odds ratios (OR) and 95% confidence intervals (CI) were reported. Assessment of multicollinearity included comparison of standard errors (univariate vs. multivariate models) obtained from the SAS SurveyLogistic procedure, evaluation of variance inflation factor (VIF), and Tolerance obtained from linear regression model. All showed no sign of multicollinearity effect.
# Results
## Prevalence of support
## Support for tobacco-free city property policies
Three-fourths (75.3%) of respondents favored policies that prohibit tobacco use on city property. As shown in, the prevalence of favoring policies that make city property tobacco-free was higher among females (81.5%) than males (68.8%); among residents residing in the Central region of the state (79.5%) than those residing in the Northwest region (69.6%); among urban residents (78.8%) than rural residents (72.2%); and among those with some college or more (78.0%) than those with a high school or less education (71.7%). Perceptions, social influence, and knowledge were also associated with favorable attitudes toward policies that prohibit tobacco use on city property. Prevalence was higher among respondents that perceived smoking and tobacco use as a "very serious" problem for people in their community (78.9%) than those who reported smoking and tobacco use as a "fairly serious/only a little serious/not a serious problem" within their community (71.6%); among respondents that reported that "none" of the people that were important to them use tobacco (82.3%) than those that reported that "any" of the important people in their lives use tobacco (73.0%); among those that correctly answered all four knowledge questions (84.4%) than those that answered two questions correctly (75.3%), those that answered one question correctly (67.8%) and those that answered none of the knowledge questions correctly (51.8%). Additionally, the mean knowledge score among those that favored the policy was significantly higher than those that "opposed" or "neither favored nor opposed the policy" (2.5 vs. 2.0, 1.9).
By smoking status, prevalence of favorable attitudes toward policies that make city property tobacco-free was higher among never smokers (82%) than former smokers (72.7%) or current smokers (56.4%). Prevalence of reporting "neither favor nor oppose" the policy was higher among current smokers (17.2%) than former smokers (9.6%) and never smokers (7.7%).
No meaningful differences were observed in the prevalence of favoring polices that make city property tobacco-free among race and ethnicity groups, age, income, marital status groups, and awareness of efforts in the community to decrease tobacco use and prevent secondhand smoke exposure.
## Support for policies that limit the number of stores that sell tobacco near schools
Two-thirds (66.7%) of respondents favored policies that limit the number of stores that sell tobacco near schools. As shown in, favoring the policy was higher among females (72.7%) than males (60.6%); among those residing in the Central (most populous) region of the state (72.8%) than those residing in the Northwest region (56.9%), and those residing in the Southwest region (62.5%); and among urban residents (71.3%) compared to rural residents (62.9%). Favor for policies that limit the number of stores that sell tobacco near schools was lowest among those in the highest income group, more than $150,000 (60.5%).
Social influence and knowledge were also associated with favorable attitudes toward policies that limit the number of stores that sell tobacco near schools. Support was higher among respondents that reported that "none" of the people that were important to them used tobacco (72.1%) than those that reported that "any" of the important people in their lives use tobacco (65.0%); among those that correctly answered all four knowledge questions (79.5%) than those that answered two questions correctly (68.4%), those that answered one question correctly (56.9%) and those that answered none of the knowledge questions correctly (41.5%). Additionally, the mean knowledge score among those that favored the policy was higher than those that "opposed" or "neither favored nor opposed" the policy (2.54 vs. 2.0, 2.01).
By smoking status, prevalence of favorable attitudes toward policies that limit the number of stores that sell tobacco near schools was higher among never smokers (72.7%) than former smokers (61%) or current smokers (54.4%). Prevalence of reporting "neither favor nor oppose" the policy was higher among current smokers (17.9%) than never smokers (11.5%).
No meaningful differences were observed in the prevalence of favoring polices that limit the number of stores that sell tobacco near schools among race and ethnicity groups, age, marital status groups, perceptions of the seriousness of smoking and tobacco use in their community, and awareness of efforts in the community to decrease tobacco use and prevent secondhand smoke exposure.show the results from the multivariate, multinomial logistic regression. Oppose is the reference category; 2 Awareness of any local programs, activities, services, or policies to decrease tobacco use or exposure to SHS; 3 How serious of a problem is smoking and tobacco use for people in your community.
## Multivariate, multinomial logistic regression
## Support for tobacco-free city property policies
## Current smokers
Among current smokers, females had higher odds of favoring tobacco-free city property policies compared to males (OR 1.58, CI 1.00-2.49), although the association was borderline significant. For each one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.34 (CI 1.10-1.63). None of the covariates were statistically significantly associated with "neither favoring nor opposing the policy".
## Former smokers
Among former smokers, females had higher odds of favoring tobacco-free city property policies compared to males (OR 2.34, CI 1.36-4.01); and for each one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.51 (CI 1.24-1.85). Rural residents had lower odds of favoring the policy than urban residents (OR 0.51, CI 0.29-0.90). None of the covariates were significantly associated with "neither favoring nor opposing" the policy.
## Never smokers
Among never smokers, odds of favoring tobacco-free city property policies were higher among females than males (OR 1.86, CI 1.28-2.71); and those with some college education versus those with high school education or less (OR 1.55, CI 1.05-2.27). For every one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.44 (CI 1.23-1.68). Never smokers that reported that "any" of the people important to them used tobacco had lower odds of favoring the policy than those that reported "none" of the people important to them used tobacco (OR 0.65, CI 0.44-0.96). None of the covariates were significantly associated with "neither favoring nor opposing" the policy. Oppose is the reference category; 2 Awareness of any local programs, activities, services, or policies to decrease tobacco use or exposure to SHS; 3 How serious of a problem is smoking and tobacco use for people in your community.
## Support for policies that limit the number of stores that sell tobacco near schools
## Current smokers
Among current smokers, the only statistically significant association with favoring policies that limit the number of stores that sell tobacco near schools was for SHS knowledge. For every one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.39 . None of the covariates were associated with "neither favoring nor opposing" the policy.
## Former smokers
Among former smokers, the only statistically significant association with favoring policies that limit the number of stores that sell tobacco near schools was for SHS knowledge. For every one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.54 (CI 1.29-1.85).
Odds of "neither favoring nor opposing" policies that limit the number of stores that sell tobacco near schools were higher among females than males (OR 1.93, CI 1.05-3.56); among those that reported that "any" of the people important to them used tobacco versus former smokers that reported "none" of the people important to them used tobacco (OR 2.11, CI 1.05-4.25); and among those with some college than those with a high school education or less (OR 1.95, CI 1.05-3.64). Former smokers that perceive smoking and tobacco use as a "serious problem" in their community had lower odds of reporting that they "neither favored nor opposed" the policy than former smokers who perceive smoking and tobacco use as "not a serious, only a little serious, or a fairly serious problem in their community" (OR 0.49, CI 0.27, 0.91).
## Never smokers
Among never smokers, for every one-point increase in SHS knowledge score, the odds of favoring the policy increased by a factor of 1.38 . The odds of favoring policies that limit the number of stores that sell tobacco near schools were higher among females than males (OR 1.74, CI 1.28, 2.36); and among those that perceived smoking and tobacco use to be a "very serious problem" in their community than those who perceive smoking and tobacco use as "not a serious, only a little serious, or a fairly serious problem" in their community . Odds of favoring policies that limit the number of stores that sell tobacco near schools were lower among rural residents than urban residents (OR 0.69, CI 0.51, 0.94); and the association was borderline significant among those aged 55+ compared to those aged 18 to 34 years (OR 0.68, CI 0.46-1.00). Those aged 55+ also had lower odds of reporting that they "neither favored nor opposed" the policy than those aged 18 to 34 years (OR 0.46, CI 0.26-0.82).
Seefor descriptive statistics of all study variables.
# Discussion
Data from the TSET HLP Community Survey show that three out of four (75%) Oklahomans favor policies that establish city property as tobacco-free and two out of three (66%) favor policies that would limit the sale of tobacco near schools. Greater levels of support for tobacco-free city properties, as compared to limiting the sale of tobacco near schools, may be influenced by Oklahoma's conservative leaning ideology and values. Conservative values that favor an individualistic ethos over social or governmental responsibility and free enterprise in this Right to Work state are deeply rooted. According to the Pew Research Center 38% of Oklahomans identify as conservative as compared to 19% identifying as liberal and 37% as moderate. Furthermore, 59% of Oklahomans would rather have smaller government as compared to a bigger government. While there is a growing norm that smoking in areas open to the public is no longer socially acceptable; government policy that may be perceived as impinging on business owners' rights may be viewed as less desirable. The political ideology of free enterprise often gives rise to the anti-tobacco regulation argument of "government should 'butt out' and let the market give consumers what they want" is a sentiment that resonates with many. Indeed, the Tobacco Industry has a long-standing history in Oklahoma of exploiting this value in appealing to and funding business groups such as the Chamber of Commerce, the Oklahoma Restaurant Association, the Oklahoma Grocers Association, and ubiquitous convenience store QuikTrip to fight against attempts to enact tobacco-related legislation in Oklahoma. This history and Oklahomans' conservative roots may result in less favorable views of government policy that could be framed as interference in a business owner's right to profit by selling a legal product. It is also possible that favoring or opposing tobacco-related policy could be influenced more by beliefs about government intervention than by the merits of the policy itself.
Overall, these data are consistent with the Truth Initiative ® finding that residents within Tobacco Nation support tobacco-related policy at similar rates to the rest of the Nation. However, level of support differs by policy type, individual smoking status, as well as among subpopulations.
While never smokers were most likely to favor each of the policies, support was also high among former smokers, and support for both policies was greater than 50% among current smokers. This is consistent with other research that shows that non-smokers are more likely to favor tobacco-related policy than current smokers. Current smokers were also significantly more likely to select the neutral response of "neither favoring nor opposing" for both policies. Further research is warranted to explore whether the selection of this response option by smokers is due to social desirability effect or if this group of people represents an important opportunity for education and advocacy in an effort increase policy support.
Knowledge of the harm associated with SHS exposure was a consistent and strong predictor of support for both policies and among current, former, and never smokers, indicating a continued need for education about the adverse effects of exposure to SHS. In addition to our finding that SHS knowledge is a predictor for policy support, the Centers for Disease Control and Prevention identified increasing knowledge of harm related to SHS as a key outcome indicator for tobacco control programs because of its association with taking action to reduce SHS exposure and increased intentions to quit and quitting tobacco. Best Practices for Comprehensive Tobacco Control emphasize the need to increase knowledge among groups of people experiencing greater exposure to SHS and disparities in chronic diseases related to tobacco as well as by using mass-reach health communication campaigns.
With the exception of current smokers' support for government policies that limit the sale of tobacco near schools, female gender was also a consistent predictor of support for both policy types and across the smoking status continuum. This finding has been reported in other research looking at tobacco-related policy support. Additionally, the Pew Research Center found that females are much more likely to support bigger government as compared to men, 58% and 37%, respectively. A recent report requested by the World Health Organization (WHO) Framework Convention on Tobacco Control defines gender as a "social construct referring to the roles, behaviors, activities, attributes and opportunities" that a population attributes to people of various gender identities. It is distinct from biological sex. The WHO posits that "gender-blind" approaches miss key opportunities to define risk as well as develop appropriate interventions. Therefore, future efforts to create and develop policies that seek to reduce the harm of tobacco use should integrate gender responsive actions. Future research is needed to better understand how gender interacts with support for tobacco-related policy.
When considering rural-urban status, research shows that many rural populations experience higher rates of tobacco use and harm associated with tobacco use, and that states with large rural populations are less likely to have strong, protective tobacco-related laws and policies in place. As interest grows in the exploration of rural-urban differences in tobacco use within Tobacco Nation, this analysis contributes insight into policy support. In our analysis, prevalence of support for both policies was lower among rural residents than urban residents. Adjusted odds ratios produced through multivariate, multinomial analysis indicate that rural-urban status remained a predictor for former smokers' support for tobacco-free city properties and for both former and never smokers' support for limiting the sale of tobacco near schools. Rural-urban status was not a predictor of support among current smokers.
Tobacco control programs often count on the support of non-smokers when implementing policy initiatives. However, this study found that never smokers who report that important people in their lives use tobacco are less likely to favor the policy of limiting the sale of tobacco near schools than their peers who report that none of the people important to them use tobacco. Further qualitative research is needed to explore the influence of social influence on policy support by smoking status.
Strengths of this study include a large, representative sample of Oklahomans, inclusion of both landline and cell phones, and the ability to look at a variety of covariates. In addition to traditional sociodemographic characteristics and community readiness, this study also considered questions related to local tobacco control efforts and the influence of people the respondents perceived as important to their lives. The study also occurred within a unique tobacco-related state context which includes high smoking prevalence, near complete preemption, and a large rural population. Limitations of the study include the cross-sectional, self-report nature of the survey. It is also possible that participants' responses were influenced by social acceptability bias. Additionally, the study sample was limited to non-institutionalized adults residing in in one of the 63 funded counties. Important to future tobacco control policy planning, it cannot be assumed from these data that participants' support for policies that establish tobacco-free city properties equates support for clean indoor and outdoor air for all properties. However, other data related to Oklahomans' attitudes toward policy showed that 77% of Oklahomans favor a law making all public places smokefree.
# Conclusions
As efforts continue to create comprehensive clean indoor air laws and protect constituents from the harm related to tobacco use, policy makers should consider the clear support for tobacco-related policy, even among current smokers. Other studies have shown that policy support continues to increase after the implementation of protective, tobacco-related policy due to social norm change. Efforts to educate the public about the harm of tobacco use and SHS exposure should also continue as this study reinforces its important role along with recent research that shows harmful perceptions of tobacco use is decreasing. This study identifies information that may be useful for education related to tobacco harm and advocacy planning, specifically audience segmentation for key messaging to promote the program-defined outcomes of 100% tobacco-free city-owned property policies and policies that limit the sale of tobacco near schools. Mass-reach health communication interventions, a component of comprehensive tobacco control programs, are one way to educate the population. This study underscores the need for mass-reach campaigns to increase the salience of tobacco use as a serious community problem as well as a means to educate the public about its harm. Oklahoma's Tobacco Stops With Me campaign has been an integral part of comprehensive tobacco control in Oklahoma. The campaign should continue its strategy of ensuring media placement in rural locales as well as developing key messages formulated for males, females, and those that identify as part of other gender groups. Media and advocacy education should also craft messages that support the commonly held, and often exploited by the Tobacco Industry, political ideologies of freedom, fairness, and free enterprise. Educational and advocacy interventions should also be tailored to communities' level of readiness to address tobacco control and prevention . The findings of this study also contribute to knowledge about factors that influence support for policy among residents of a Tobacco Nation state.
Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/16/18/3378/s1,: Descriptive statistics of study variables.
|
Service‐user experiences of an integrated psychological intervention for depression or anxiety and tobacco smoking in improving access to psychological therapies services: A qualitative investigation into mechanisms of change in quitting smoking
Introduction: High smoking prevalence leads to increased morbidity and mortality in individuals with depression/anxiety. Integrated interventions targeting both smoking and mood have been found to be more effective than those targeting smoking alone, but the mechanisms of change of these interventions have not been investigated.This qualitative study aimed to understand participants' experiences of the mechanisms underlying change in smoking behaviour following an integrated cognitive behavioural technique-based intervention for smoking cessation and depression/anxiety.Methods: This study was embedded within an ongoing randomized-controlled acceptability and feasibility trial (http://www.isrctn.com/ISRCTN99531779). Semistructured interviews were conducted with 15 IAPT service users. Data were analysed using thematic analysis. During the interviews, participants were asked open-ended questions about their quitting experience and perception of how the intervention aided their behaviour change.Results: Five themes were identified. Acquiring an increased awareness of smoking patterns: participants described an increased understanding of how smoking was contributing towards their mental health difficulty. Developing individualized strategies: participants described acquiring 'a couple of tricks up your sleeve' that were helpful in making smoking cessation feel more 'manageable'. Practitioner style as 'supportive but not lecture-y': participants expressed how important the Health Expectations. 2023;26:498-509. 498 | wileyonlinelibrary.com/journal/hexThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
# | introduction
Individuals with depression/anxiety are twice as likely to smoke than those without depression/anxiety 1 ; this disparity increases mortality in people with depression/anxiety compared to the general population (mortality rate ratio, 1.92 [95% confidence interval (CI):
1.91-1.94]). [bib_ref] A comprehensive analysis of mortality-related health metrics associated with mental disorders: a..., Plana-Ripoll [/bib_ref] Despite being equally as motivated to quit, this population smokes more heavily, are more addicted and are less likely to successfully quit than the general population. [bib_ref] Assessing motivation to quit smoking in people with mental illness: a review, Siru [/bib_ref] [bib_ref] Past major depression and smoking cessation outcome: a systematic review and meta-analysis..., Hitsman [/bib_ref] There are many reasons for the high smoking rates in this population; for example, they are less likely to be prescribed smoking medicines, [bib_ref] Prescribing prevalence, effectiveness, and mental health safety of smoking cessation medicines in..., Taylor [/bib_ref] and it is widely believed that smoking cessation can worsen mental health. [bib_ref] In the shadow of a new smoke free policy: a discourse analysis..., Johnson [/bib_ref] [bib_ref] A survey of staff attitudes to smoking-related policy and intervention in psychiatric..., Mcnally [/bib_ref] However, a recent Cochrane review found evidence that smoking cessation can improve anxiety/depression compared to continuing smoking (standardized mean difference, −0.31 [95% CI: −0.40 to −0.22]). [bib_ref] Smoking cessation interventions for smokers with current or past depression, Taylor [/bib_ref] Another Cochrane review found that smoking cessation interventions offered alongside mood management support led to higher cessation rates compared to smoking cessation interventions alone for people with depression (risk ratio 1.47, 95% CI: 1.13-1.92), 10 indicating the importance of integrated interventions to improve smoking and mood outcomes. However, this review did not shed light on the mechanisms that led to behavioural change.
Integrated interventions may be more effective for quitting smoking for various reasons. For example, cognitive behavioural techniques (CBT) could alter unhelpful beliefs about the relationship between smoking and depression/anxiety (e.g., 'smoking helps my mood'), [bib_ref] Addressing concerns about smoking cessation and mental health: theoretical review and practical..., Taylor [/bib_ref] which could promote cessation and prevent relapse. Such an intervention could also promote alternative strategies for managing depression/anxiety to smoking. [bib_ref] Views about integrating smoking cessation treatment within psychological services for patients with..., Taylor [/bib_ref] For example, behavioural activation aims to increase pleasurable activities.It is also possible that the therapeutic alliance in psychological interventions could help facilitate behaviour change. [bib_ref] Early therapeutic alliance as a predictor of treatment outcome for adolescent cannabis..., Diamond [/bib_ref] [bib_ref] The clients' perspective on change during treatment for an alcohol problem: qualitative..., Orford [/bib_ref] There are some evidence-based models that we can use to investigate mechanisms of behavioural change. The Capability, Opportunity, Motivation-Behaviour model (COM-B model)suggests that a person's capability (i.e., a person's physical and psychological capacity), opportunity (i.e., external factors that facilitate behaviour) and motivation are involved in behavioural change. The Smoking, Not smoking, Attempting to quit, Planning to quit model (SNAP) [bib_ref] The multiple facets of cigarette addiction and what they mean for encouraging..., West [/bib_ref] suggests that smoking cessation involves moving through the four stages of (1) smoking, [bib_ref] A comprehensive analysis of mortality-related health metrics associated with mental disorders: a..., Plana-Ripoll [/bib_ref] attempting to quit, (3) planning to quit and (4) not smoking. The misattribution hypothesis suggests that smokers misattribute nicotine withdrawal symptoms of stress or anxiety/depression and believe that smoking alleviates symptoms of mental health difficulties. [bib_ref] Smoking cessation leads to reduced stress, but why?, Parrott [/bib_ref] [bib_ref] An application of the stress-diathesis model: a narrative review about the association..., Taylor [/bib_ref] Understanding mechanisms of behavioural change and how they fit into evidence-based frameworks could improve our understanding of the active intervention components and help identify therapist characteristics that optimize therapeutic benefits, potentially leading to more effective and streamlined interventions.
Addiction research has been criticized for excluding the patient's view and focussing on intervention techniques rather than intervention mechanisms. [bib_ref] The clients' perspective on change during treatment for an alcohol problem: qualitative..., Orford [/bib_ref] Therefore, as part of a wider trial, [bib_ref] IntEgrating Smoking Cessation treatment As part of usual Psychological care for dEpression..., Taylor [/bib_ref] In this qualitative investigation, we aimed to use evidence-based models of behaviour change and behavioural intervention development [bib_ref] The multiple facets of cigarette addiction and what they mean for encouraging..., West [/bib_ref] [bib_ref] Smoking cessation leads to reduced stress, but why?, Parrott [/bib_ref] [bib_ref] An application of the stress-diathesis model: a narrative review about the association..., Taylor [/bib_ref] to explore participants' subjective experience of mechanisms of change in smoking cessation.
# | methods
## | design
Our study was preregistered (https://osf.io/nfgu4/) and was part of an ongoing RCT (ESCAPE, http://www.isrctn.com/ISRCTN995 31779). The preprint is available via medRxiv (https://doi.org/10. 1101/2022.03.23.22272703). We followed COREQ reporting guidelines in writing this manuscript. [bib_ref] Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews..., Tong [/bib_ref] We used qualitative in-depth interviews to explore participants' perceptions of change. [bib_ref] Mediators and mechanisms of change in psychotherapy research, Kazdin [/bib_ref]
## | participants
We approached 19 patients and conducted interviews with 15 who took part in the ESCAPE trial intervention arm and had attended three or more intervention sessions. Reasons for nonparticipation were not recorded. Participants in the intervention arm received a CBT-based smoking cessation intervention that was integrated into routine IAPT care for depression/ anxiety 9,20 (Supporting Information: A). Improving Access to Psychological Therapies (IAPT) is a primary care service in the UK National Health Service providing evidence-based psychological therapies for depression/anxiety. Trial inclusion criteria were self-reported daily smokers of at least 1 year, aged ≥18 years, met thresholds for depression and/or anxiety (clinician-administered PHQ-9 score ≥ 10 and/or GAD-7 ≥ 8 scores) and were about to start IAPT treatment. Individuals were excluded if they did not have the capacity to give informed consent, or if they were pregnant or breastfeeding at trial entry.
## | procedure and recruitment
Recruitment procedures for the ESCAPE trial can be found in the trial protocol (https://osf.io/nfgu4/). Purposive sampling was used to recruit participants for follow-up interviews about the intervention.
During 3-and 6-month telephone ESCAPE trial follow-ups, participants were asked if they would like to be interviewed about their experiences in the study and attempting to quit. We recruited participants until information power was reached. [bib_ref] Sample size in qualitative interview studies:guided by information power, Malterud [/bib_ref] Information power is more suitable for pragmatic applied health research than data saturation. 'Data saturation' was originally developed for grounded theory analysis. [bib_ref] Sample size in qualitative interview studies:guided by information power, Malterud [/bib_ref] Participants gave oral consent before the interview, and again as an audio-recorded consent statement.
## | interviews
The interview schedule (Supporting Information: B) aimed to explore participants' experiences of mechanisms of change. Interviews lasted 30-45 min and were embedded within a longer interview schedule, lasting no more than 60 min, which also investigated the acceptability of the intervention and trial procedures (the additional findings will be presented elsewhere). Interviews were conducted by K. F. S. and K. S.
# | analysis
Data were transcribed using a University-approved service. Fifty percent of the audio data were checked against the transcripts to ensure fidelity.
The data were analysed using a reflexive thematic analysis following the steps outlined by Braun and Clarke. [bib_ref] Using thematic analysis in psychology, Braun [/bib_ref] [bib_ref] Reflecting on reflexive thematic analysis, Braun [/bib_ref] Reflexive thematic analysis was used as it is not tied to theoretical or epistemological approaches and can be used both inductively and deductively. A critical realist approach was adopted; meaning was viewed as both socially constructed and relating to individuals' experiential reality. [bib_ref] Critical realism in discourse analysis: a presentation of a systematic method of..., Sims-Schouten [/bib_ref] Braun and Clarke's 25 guidance for the six phases of thematic analysis was followed: (1) familiarization with the data, , which has shown good reliability for categorizing components of behaviour change interventions. [bib_ref] The behaviour change wheel: a new method for characterising and designing behaviour..., Michie [/bib_ref] Coding was conducted manually. Participant IDs were used throughout, and any potentially identifying information was removed.
Once the data were coded, all relevant coded data extracts were collated and organized into potential themes and subthemes. A series of tables were developed to explore possible relationships between codes, themes and subthemes. Potential themes and subthemes were then reviewed, refined and assessed according to the criteria of internal homogeneity and external heterogeneity.Analysis was viewed as an iterative process; the researcher at times returned to previous stages rather than following a rigidly linear process. A self-reflexive stance was used throughout data collection and analysis to increase awareness of and limit the impact of the researcher's potential biases and assumptions.
## | patient and public involvement
# | results
Fifteen participants were recruited .
We identified 5 themes and 12 subthemes presented them with illustrative quotes [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref].
## | increased awareness of smoking patterns
Participants described an increased awareness of smoking patterns as a key step in facilitating quitting smoking. They described a shift in their awareness, from smoking being automatic, where they would 'smoke and not necessarily know I was smoking' (Participant 2), to being more 'analytical' about their smoking patterns (Participant 4).
This 'understanding why you smoke' (Participant 3) seemed to facilitate participants' sense of psychological capability to not smoke [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref] , Quote 1).
One aspect of increased awareness was an improved awareness of smoking triggers [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref] , Quote 2). One of the most common triggers for smoking was strong negative emotions.
Participants expressed that through the intervention, they gained increased awareness of the link between strong negative emotions and smoking: 'when I talk about it, I do smoke because I'm unhappy and because it's a distraction' (Participant 11). Some expressed the realization that smoking was often an attempt to cope with negative emotions: 'before, I felt that if I was stressed, I could step away and have a fag and that would help' (Participant 5). This response reflects the misattribution hypothesis. [bib_ref] Smoking cessation leads to reduced stress, but why?, Parrott [/bib_ref] The increased awareness of this misattribution seemed to result in a change in thinking about the relationship between smoking and stress, as well as providing alternative options for managing stress: 'I recognised that and that's more just stepping out of the situation and just chilling for a bit' (Participant 5). Therefore, increased awareness of the need for alternatives to smoking for managing emotions appears to be a mechanism of change in reducing smoking. Increased awareness of the need for alternatives Increased awareness of smoking as a maintaining factor in vicious cycles Quote 3: 'It was like a coping mechanism that, in the short term, seemed to help, but in the long-term then I'd worry about it (…), it was making me feel worse' (Participant 3).
Quote 4: 'You have the cigarette, you feel a bit better but when the nicotine is less effective (…) you feel worse again and you have the cravings and you feel more stressed and more anxious again after it's worn off, which I never really thought about before' (Participant 5).
Quote 5: 'I was conscious of my weight and thought, "If I smoke, I'm going to have less of an appetite" (…) I'm a healthy weight and the way I was thinking was, and acting when I was younger was making me underweight and unhealthy, and I think it was just like having a realisation of that' (Participant 13).
Theme 2: Developing individualized strategies: 'What's in your toolkit?'
Finding alternative ways to fulfil the function of smoking Quote 6: 'I'd be anxious when I came home from school and I still had a lot of work to do, and I'd have a wine and sit in the courtyard and smoke because that helped, but in the short term. So, we looked at what could you do differently, so could you go for a run, could you go for a swim, could you play the piano, and things like that. So, then it gave me a bit more, changed the behaviour, rather than just stopped the smoking' (Participant 3).
Strategies enhance sense of capability Quote 7: 'The strategies are really helpful-rather than just going, "OK, I just will stop this habit I've had since I've been, 21 years," that's a bit impossible, well, I found it a bit impossible before' (Participant 3).
Quote 8: 'You could see that you were doing better, when they did the carbon monoxide test' (Participant 1).
Quote 9: 'It was like, "wow, look how much money I've saved but also how much time I've saved"' (Participant 3).
Quote 10: 'When I had the Champix, because it made the actual smoking make me feel sick, you couldn't smoke, it completely put you off it' (Participant 2).
Quote 11: 'Because I wasn't really craving it because I sort of had the nicotine, but yeah, it was a lot easier and easier to manage at work. I was calmer, not trying to give up and work and get stressed, so that was really helpful' (Participant 1).
Quote 12: 'I would not smoke for a couple of days but then I'd just find myself getting drawn back into doing it, through having colleagues that smoked or friends that smoked when we were out' (Participant 13).
## T a b l e 3 (continued)
Theme Subtheme Quote(s) Quote 13: 'Then they're going to be aware of me doing it, and not be likely to offer me a cigarette (…) it did work and I think it did help having that' (Participant 13). Quote 17: 'As soon as I turned around and said, "Yeah, I haven't had one" she said, "Oh that's fantastic!" Even just that little bit of encouragement was really good' (Participant 5).
Confronting avoidance Quote 18: 'She wasn't avoiding talking to me about, which was brilliant, because, "Let's talk about the smoking," so I knew I had to confront it, it wasn't, "You said you'd stop smoking last week and you didn't so we're just going to ignore it," it was, "What happened this time that made you feel…?," so it was examining it, which is great' (Participant 6).
Guided discovery rather than directive Quote 19: 'I feel it was subtly getting me to make the decisions and getting me to make the choices (…) it was very much guiding rather than leading' (Participant 4).
Quote 20: 'I guess the only thing I would suggest is if they had a bit more time for calls so they didn't feel … because you know that they're getting stressed too. The person who's supposed to be counselling you is getting stressed because they've got to get onto the next thing, and you can sense that they're trying to interpret what you're saying so they can get onto the next because they have six more questions and there's only four minutes left' (Participant 10). Although smoking cessation medication was not viewed as helpful by everyone, some participants found it to be an important tool [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref] , Quote 10). It seemed particularly helpful for managing the physiological nicotine cravings, which enhanced some participants' sense of their physical capability to quit [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref] , Quote 11).
The strategy of communicating with others about quitting was viewed as important by several participants. They described being with others who were smoking as a trigger [fig_ref] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental... [/fig_ref] , Quote 12).
However, telling others about the quit attempt also helped (
## | 'having someone that's checking on you': regularity of sessions
Most participants expressed the importance of the regularity of sessions, which supported the development and maintenance of the therapeutic alliance. One helpful aspect of regular sessions was having the space to solve problems as they arose (
# | discussion
## | summary of findings
In this qualitative study, we investigated participants' subjective experiences of mechanisms underlying change in smoking behaviour.
Participants reported that the integrated smoking cessation and mood intervention helped facilitate reductions in smoking through increased self-awareness of smoking patterns, and supported them in developing individualized behavioural and cognitive strategies to aid cessation. Participants stated that the regularity of support, the supportive, 'non-lecture-y' therapeutic style of IAPT practitioners and being offered the smoking intervention at the 'right time' (i.e., integrated with mental health support) all contributed to participants' sense of being able to make changes to the smoking. These findings further our understanding of the active ingredients and processes for behaviour change in this integrated intervention.
## | study strengths and limitations
Participants were all white British; this limits the study's transferabil- participants. In addition, as participants were also part of a larger study that was not designed to investigate mechanisms of smoking cessation, it is unknown if these findings are transferable to the general IAPT population. However, studies of participants who were not selected to take part in an RCT are predictive of some of the findings in this study. For example, in another study of IAPT users, participants prospectively predicted that knowing about the mental health benefits of smoking cessation could help them to quit smoking. [bib_ref] Views about integrating smoking cessation treatment within psychological services for patients with..., Taylor [/bib_ref] When interpreting qualitative results, it is important to acknowledge the researcher's background [bib_ref] Qualitative quality: eight "Big-Tent" criteria for excellent qualitative research, Tracy [/bib_ref] Participant ability to quit smoking coinciding with the opportunity to access support for smoking cessation whilst receiving mental health therapy is consistent with the COM-B model.The presence of mental health problems did not prevent participants from feeling ready to make changes to their smoking. This is important because individuals with mental health problems are not given the same opportunities for smoking cessation support as the general population, [bib_ref] Confronting a neglected epidemic: tobacco cessation for persons with mental illnesses and..., Schroeder [/bib_ref] and there is a widely held belief among healthcare professionals that smoking cessation should only be attempted after mental health has improved. 7,8 These findings reinforce previous research suggesting that individuals with depression or anxiety can be motivated and supported to successfully reduce or quit smoking when given the opportunity. 10,41 However, this did not apply to all participants, with some expressing that life felt too difficult to quit.
Nevertheless, the intervention appeared to help some participants move closer to 'the right time', into the 'preparing to quit' stage of smoking cessation, [bib_ref] The multiple facets of cigarette addiction and what they mean for encouraging..., West [/bib_ref] suggesting that the intervention may assist individuals to progress in their change process.
## | research and practice implications
These findings further our understanding of the active ingredients and processes for behaviour change in integrated smoking and mental health interventions. In terms of clinical practice, we have outlined the strategies and the therapeutic stance that could be embedded and emphasized in manualized interventions to optimize therapeutic benefits. Findings indicate that practitioners should not assume that having anxiety or depression means that individuals are not ready or motivated to quit smoking. These processes are currently being implemented in the ESCAPE trial; however, a fullscale trial is required to investigate the effectiveness of these components. Future research in this area should strive to recruit people from diverse cultural backgrounds and include a longer-term follow-up.
# | conclusions
Several key factors were identified by participants to be important in
[fig] 2: coding, (3) generating themes, (4) reviewing themes, (5) defining and naming themes and (6) writing up. A combined inductive and deductive approach to coding was adopted, whereby codes and themes were developed both from the existing theory and the data. The framework of deductive codes was constructed based on the COM-B model (Supporting Information: C and [/fig]
[fig] FREDMAN: ity to people of different backgrounds. Furthermore, the lack of a longer-term follow-up means that we could not show whether intervention mechanisms were maintained long-term. The sampling method may have contributed towards bias. Participants consisted of those from the ESCAPE trial who had attended three or more sessions, completed at least one follow-up and had self-selected to attend interviews. This population may have had more positive experiences of the intervention compared to those not sampled, which may not represent the experiences of all intervention arm [/fig]
[table] Theme 3: Practitioner style as 'supportive but not like lecture-y' Empathy and a nonjudgemental stance enabled the disclosure of setbacks Quote 14: 'If you've got somebody that you're not connecting with (…) And that doesn't seem to be understanding of you, you're not going to take their advice because it's going to be: you don't understand where I'm coming from, how can you expect me to do something when you don't understand? Showing no empathy for what's going on. It makes you put a wall up, I suppose, and you get quite defensive. Whereas if you've got somebody that you get on with, you can open up to, that seems to understand what you're going through, the wall comes down so you're more likely to take on-board what's said' (Participant 2). Quote 15: 'She was very understanding of what I was going through as a whole anyway but it just meant that when it came to the smoking bit, it helped make that a more comfortable environment to discuss it' (Participant 13). Quote 16: 'The temptation is to be very self-critical when something you try doesn't work, and actually having somebody say, "It doesn't matter, these things happen, it's absolutely fine for it not to have worked this time, let's just have a reboot," something about what it was that caused the hiccup last time, "See if we can try and avoid that next time" (…) There was no judgement or condemnation' (Participant 4). [/table]
[table] Table 3 ,: Quote 13). This suggests that changing the social environment can Several participants found it helpful that the IAPT practitioner helped them to confront their avoidance of talking about smoking or attempting to quit(Table 3, Quote 18). [/table]
|
Living Matter Observations with a Novel Hyperspectral Supercontinuum Confocal Microscope for VIS to Near-IR Reflectance Spectroscopy
A broad range hyper-spectroscopic microscope fed by a supercontinuum laser source and equipped with an almost achromatic optical layout is illustrated with detailed explanations of the design, implementation and data. The real novelty of this instrument, a confocal spectroscopic microscope capable of recording high resolution reflectance data in the VIS-IR spectral range from about 500 nm to 2.5 μm wavelengths, is the possibility of acquiring spectral data at every physical point as defined by lateral coordinates, X and Y, as well as at a depth coordinate, Z, as obtained by the confocal optical sectioning advantage. With this apparatus we collect each single scanning point as a whole spectrum by combining two linear spectral detector arrays, one CCD for the visible range, and one InGaAs infrared array, simultaneously available at the sensor output channel of the home made instrument. This microscope has been developed for biomedical analysis of human skin and other similar applications. Results are shown illustrating the technical performances of the instrument and the capability in extracting information about the composition and the structure of different parts or compartments in biological samples as well as in solid statematter. A complete spectroscopic fingerprinting of samples at OPEN ACCESS microscopic level is shown possible by using statistical analysis on raw data or analytical reflectance models based on Abelé s matrix transfer methods.
have limited spatial resolution and limited time resolution, due to the slow acquisition time, or require the use of synchrotron radiation as a bright source [bib_ref] IR spectroscopic characteristics of cell cycle and cell death probed by synchrotron..., Holman [/bib_ref].
In this context, Hyperspectral Imaging (HSI), also known as Chemical or Spectroscopic Imaging, has emerged as a powerful technique that integrates conventional imaging and spectroscopy [bib_ref] Hyperspectral imaging: A novel approach for microscopic analysis, Schultz [/bib_ref]. Its introduction as a specific feature of LSCM, following the development of white laser sources, is quite recent and not common, and allows one to greatly enhance the performance and flexibility of the microscope [bib_ref] Supercontinuum ultra wide range confocal microscope for reflectance spectroscopy of living matter..., Selci [/bib_ref]. Unlike a typical reflectance confocal image, a hyperspectral image provides a complete spectrum of the sample at every pixel location, so that, in a confocal configuration, one can obtain 3D morphological features together with local spectral information.
Here we describe a novel confocal microscope aimed at integrating structural and morphological information with detailed spatially resolved spectroscopic properties, powered by a supercontinuum laser source in the visible and near infrared spectral ranges, and able to collect a complete spectroscopic image by the acquisition of a one-shot wide range reflectance spectrum for every image pixel in the three dimensional sets of data typical of a confocal microscope. The microscope has an achromatic light path which guarantees that each wavelength contribution is effectively related to the same point of the sample.
The microscope performance is shown here in detail including axial resolution, achromaticity, lateral resolution, as well as the potential applications in different research areas as a powerful technique for in situ real-time imaging and spectroscopic investigation. We prove how the hyperspectral confocal microscope provides structural and chemical analysis of surfaces and interfaces in multilayer materials. Furthermore, we report our results of hyperspectral reflectance analysis of visible and NIR radiation from human primary melanoma cells in culture. We will show that differences between spectra from cytoplasmic areas and nuclear areas allow distinguishing cellular compartments without any exogenous dye and using a multivariate analysis approach. The application of our microscope to the analysis of scattering and heterogeneous tissue samples is under investigation. Preliminary results show that it is possible to obtain images very similar to those acquired with commercial reflectance confocal microscopes used for dermatologic applications working with selectable laser lines. The spectroscopic analysis of tissues results is under development and will be the argument of our subsequent experimental activity.
## Concept and technical description
The confocal microscope presented here has been designed and realized by our group with a strong focus on the spectroscopy side. The instrument is completely conceived and realized from scratch to provide a versatile laboratory microscope able to explore different possibilities about the accessible spectral ranges, optical insertions, beam manipulation and sample access. The underlying idea is to realize simplified versions of the microscope to eventually address specific classes of samples for clinical or industrial applications.
A preliminary set-up and results have been previously reported [bib_ref] Supercontinuum ultra wide range confocal microscope for reflectance spectroscopy of living matter..., Selci [/bib_ref]. Since then, the electronic layout has been completely changed, and the full IR exploitation has been completed with the insertion of a suitable array detector. Therefore, we are now able to completely report about the axial and lateral performances in terms of resolution and achromaticity.
The following section describes the current set-up. Then, a discussion is made about the optical properties of the most relevant parts. Following the optical path described in [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref] , from right to left, we see different optical blocks. The source block S includes: the supercontinuum laser (SLAS; NKT SuperK Power+, NKT Photonics, Birkerød, Denmark), a unit able to deliver on average up to 1 mW/nm over the 0.5 μm-2.4 μm optical range, a beam shutter (SH; SR474 by Stanford Research Systems, Sunnyvale, CA, USA), a chopper (CH; Thorlabs 2000 crystal-stabilized unit, Newton, NJ, USA), and two coupled all-reflective beam-expanders (BEX; units BE02R/BE04R by Thorlabs), capable of overall achromatic reflectance >90% between the 0.5 μm and 20 μm wavelengths. The beam expander is needed to fill completely the objective entrance: the fiber output of the supercontinuum laser has, in fact, a fitted collimator with an output beam diameter of about 1 mm at 800 nm, actually depending linearly on the wavelength, and a divergence <0.5 mrad, unable to fill the objective input, as we will see later in the following section. The relay optics (RL1) section, which is only assembled with plane mirrors, is placed just after the source block. This section is used for beam alignment and to diaphragm the beam by additional iris; moreover, a moveable mirror can insert into the optical path another more conventional laser, to carry out, for instance, fluorescence measurements. The condenser/objective block (COB) is formed by only two components carefully aligned on the beam path: the beam splitter (BS1) and the reflective objective (ROB). We have available many different laminas to be used as beam splitters. Different materials and/or thicknesses have been tested to minimize spurious effects and modulate the power fed onto the sample. For the current application, we have determined that a Borosilicate Crown glass (BK7) lamina of 160 μm thickness is the best suited. The reflective objective is a reverse Cassegrain following the Schwarzschild design (50102-02 model, Newport, Irvine, CA, USA), with 36× magnification and a back focal length at infinity. The reflective aluminum internal coating is over coated with magnesium fluoride (MgF 2 ) allowing a 200 nm-20 μm useful optical range. The working distance is 10.4 mm with a numerical aperture (NA) of 0.52, an optical clearance of 5.6 mm, and an optical obscuration of 17%.
The beam is focused on the sample fixed to the piezoelectric-motor based stage XYZS. We are currently dithering the sample for greater ease of allocation of many different fixtures. The XY stage is a double M-664 stage (Physik Instrumente, Karlsruhe, Germany) with a 25 mm available range, 100 nm resolution, 200 nm unidirectional repeatability and a maximum speed of 400 mm/s. Under the XY stage, the Z movement is supplied by a Newport GTS30V stage that offers a 30 mm excursion with ±100 nm bidirectional repeatability and 100 mm/s maximum speed.
The specular reflection is then converted again to a parallel beam (of course, only if the sample is exactly at the focus point) and conveyed by a second relay group of plane mirrors (RL2) to the dichroic beam splitting and focusing group (B&F).
The second beam splitter (BS2) is a 2‖ diameter 300 μm thick Silicon wafer, with both faces polished. This component allows a good reflection of the visible (VIS) part of the spectrum above the silicon gap. Instead, all the infrared (IR) part of the spectrum is transmitted through the silicon lamina for energies below the gap, starting around the Nd:YAG 1.064 μm seed line of the supercontinuum laser. On both sides of the silicon beam splitter, two off-axis parabolic mirrors (Newport 50338AL) focus the parallel beam onto two pinholes (PH1 and PH2) at the entrance points of two different spectrometers, making simultaneously available the VIS and the IR spectral regions. The parabolic mirrors have a focal distance of 10 cm, and with their magnesium-fluoride-protected aluminum coating can be used without chromatic aberrations between 200 nm and 10 μm. While in the past we have used variable sized pinholes, the complexity of the alignment with the spectrometers optics made us opt for a fixed size pinhole of 25 μm diameter. In the future, more flexible solutions could be available.
The spectrometers SPCVIS and SPCIR are the final points of the two split optical paths. The VIS spectrometer consists in a PIXIS CCD detector (Princeton Instruments, Trenton, NJ, USA) mounted on a SP2150 f/4 spectrograph (Acton, Princeton Instruments, Trenton, NJ, USA) equipped with two gratings of 1,200 L/mm and 150 L/mm, both with 500 nm blaze; the high resolution grating covers about 100 nm dispersed on the CCD, with a maximum resolution of 0.074 nm, while, with the other, we can use the full VIS range, dispersing about 1,000 nm on the detector, with 0.8 nm of spectral resolution. The detector used is a thermoelectric Peltier cooled (−80 °C) back-illuminated CCD, with 1,340 × 100 pixels, each 20 μm × 20 μm wide. The peak quantum efficiency is more than 95% between 500 nm and 650 nm, and the readout noise is of the order of 10 e-rms, making the detector very sensitive also for more demanding tasks than reflectance.
The IR spectrometer is realized by coupling a Newport MS127i imaging spectrograph with one of the InGaAs array detectors by Hamamatsu (Hamamatsu City, Japan), that is G9211-256S and G9208-256W 256 pixels each, with spectral ranges sensitivities between 0.9 μm to 1.67 μm and between 0.9 μm to 2.5 μm, respectively. Both sensors have a peak sensitivity of more than 1 A/W and an associate dark current of the order of 2 pA and 500 pA, respectively, making possible direct measurement of photon intensity in the nW range. The measurement that we report here has been obtained with the first of the two detectors. We are using a grating (75 L/mm, 1.6 μm blaze) to cover all the available range at once. The resolution is, in this case, 4.85 nm/pixel. The controlling software, not shown here, has been written using custom C++ programming routines using the software drivers and libraries of all available instruments, realizing a virtual panel. Data are packed as 16 bit binary data, incorporated in a four-dimensional matrix that can be read cutting the hypercube along any possible interesting cutting plane.
Electronic signals are used to enable hardware synchronization of all different subsections, for instance to store, exactly at the same time, data from different detectors for the VIS and IR parts. The acquisition speed is determined by the limits imposed by both the array detectors, i.e., about 1,000 spectra/sec, with VIS and IR parts acquired simultaneously in the own buffer, and takes into account the high precision of 16 bit/pixel for both detectors. Reducing the precision, and using other types of detectors, the speed could be substantially increased. Data are downloaded from both detectors after every line, being the scanning between the single pixels completely governed by hardware triggering among the instruments.
Each spectral point is loaded into detectors by means of a master electronic trigger generated by an optical chopper for every image pixel. As usual for spectroscopy, the chopper is also used to make the correct sequence of light and dark on detectors, cleaning charges between subsequent acquisitions. More recently, a virtual chopper has been used, introducing several ghost acquisitions collected at the beginning of each row and then discarded, in order to clean the detector's channels as well. Thus, in our current configuration the master electronic trigger is generated by the PI C867 260 XY motors driver producing transistor-transistor logic (TTL) triggers at specific calibrated points during the fast axis scan.
## Optical considerations
The distinguishing feature of this apparatus is the ability to collect reflectivity with the assurance that, for all the accessible wavelengths, the single spectrum associated with a physical pixel can be related with a single point with space coordinates X, Y and Z. Moreover, the optical performances are to be not too far from the typical characteristics of a standard reflectance confocal microscope. Therefore, some considerations can be usefully drawn at this point. In any case, the only very critical point is to avoid the chromatic aberration of the focus on the sample, after which any possible residual chromatism on the return path, being part of the transfer function of the microscope, can be incorporated in data normalization.
## The beam splitter
The beam splitter (BS1 in [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref] is the only part of the instrument not realized as a pure reflective component, so extreme care has been taken to minimize any possible chromatic aberrations. By design, the beam reaches the sample after reflection on the beam splitter, with the result that also the overall intensity flux on the sample surface has been considered for the best choice of this component. In fact, while the overall intensity on the detector side is the product of the beam splitter reflection and transmission, therefore irrespective of the order, the dose on the sample is not. In below, a comparison between different solutions is given. A BK7 lamina has been adopted due to the fact that we often work with live cells, thus minimizing the irradiated dose on the sample. Other possibilities exist that can, instead, maximize the total through output.
Various adverse effects on the beam propagation are coming from multiple reflections as a function of lamina thickness and composition like: (i) a shift of the focus, due to a ghost filling of the objective input; (ii) a further shift that, after transmission, can cause an axial error on the return optical axis; (iii) a chromatic behavior that accompanies the two previous effects with an explicit dependence on the wavelength due to the material dispersion; (iv) interference effects within the slab that make the reflectivity measurement ill characterized. All those effects gain some importance because of the huge spectral range involved here, while they are almost negligible when dealing with a fixed frequency laser confocal microscope.
Upon hitting the beam splitter, and following the notation of Born and Wolf, the beam makes one reflection [bib_ref] Infrared spectroscopy and microscopy in cancer research and diagnosis, Bellisola [/bib_ref] r on the first interface air-material and, then, a second reflection 21 r between the material and the air. This second reflection is then refracted back, and comes out parallel to the first one, but shifted. The shift happens to be dependent on the slab thickness and on its index of refraction. In case of dispersion, such a shift displays a color composition on the beam side, instead of an otherwise white beam. Following the construction of [fig_ref] Figure 2: Optical paths of the reflected and the refracted rays on the first... [/fig_ref] , that obviously represents only the main rays that we are considering, the amount of the shift AB results to be:
[formula] 2 tan( ) cos( ) ti AB d (1) [/formula]
where the refraction angle θ t is related to the index of refraction of the material by the Snell's law. Some possibilities that we have explored are shown in . Two materials, both very well known for their wide IR transparency (down to 10 μm), i.e., calcium fluoride and a β form of zinc sulphide (CLEARTRAN™ by CVD, Dow Chemical Company, Midland, MI, USA), cause huge shifts with non-negligible chromatic aberration. To move away the second reflection, a thicker CaF 2 could be used, e.g., of about 10 mm, in order to extend the usability of the microscope farther in the IR, while for CLEARTRAN™ the required thickness is considerably more and not realistic. As it is clear, the thin BK7 lamina introduces a rigid shift AB of the order of 100 μm with a chromatic component of the order of 1 μm, that is absolutely negligible in relation to the 5.5 mm beam diameter, and has a transparency perfectly compatible with the current spectroscopic range of the microscope. In a similar way, it is possible to compute the shift CD of the output beam, considering only the reflection, and then the refraction of the main beam of amplitude 12 r :
[formula] sin( ) cos( ) it t CD d .(2) [/formula]
Again, the BK7 solution has a very good performance, with a very small chromatic aberration, while, generally, the magnitude effect is more dependent on the absolute value of the index of refraction.
The exact solution of the beam splitter reflectivity and transmittance takes into account multiple reflections, yet it can be expressedin terms of [bib_ref] Infrared spectroscopy and microscopy in cancer research and diagnosis, Bellisola [/bib_ref] r and 21 r :
[formula] 2 .(3) [/formula]
Using Equation (3) it is possible to evaluate the remaining quantities in , in particular the exact computation for BK7 sample irradiation, the overall beam splitter efficiency, and the residual interference pattern integrated over the spectral resolution: this pattern can be arbitrarily decreased by binning the detector channels, reducing spectral resolution. For other materials, for which the thickness rules out interference effects, the calculations merely report the corresponding Fresnel equations considering only the first reflective surface, because a realistic computation should evaluate the exact part of the shifted beam that eventually enters into the objective. Here, only the average between transverse electric (TE) and transverse magnetic (TM) radiations is considered. Polarization effects are indeed quite strong: for BK7 the ratio of TE polarization with respect to TM is about 10:1. . First beam splitter performances for various materials as a function of the wavelength λ. All distances are given in mm, while the irradiation and efficiency are given as percentage of the incoming beam, as well as the pk-pk amplitude of the interference effect. Here the values are averaged taking into account the spectral resolution as given in Section 2.1. The use of reflective objective in microscopy is well consolidated whenever infrared [bib_ref] Resolution limits for infrared microspectroscopy explored with synchrotron radiation, Carr [/bib_ref] or special solutions [bib_ref] Single-component reflecting objective for low-temperature spectroscopy in the entire visible region, Fujiyoshi [/bib_ref] are required. The main characteristics we are interested in, with the adoption of a reflective objective, are the absence of any chromatic aberration, the perfect spherical aberration cancellation, because of infinite conjugate points [bib_ref] Resolution limits for infrared microspectroscopy explored with synchrotron radiation, Carr [/bib_ref] , and a useful large working distance. Many FTIR commercial microscopes include this type of objectives, deserving detailed studies on their performances [bib_ref] Multi-beam synchrotron FTIR chemical imaging: Impacts of Schwarzschild objective and spatial oversampling..., Mattson [/bib_ref]. Usually, the lateral resolution measurement has to be compared with the theoretical point spread function (PSF). This is indeed not a trivial task, unless one adopts crude approximations or makes more detailed considerations [bib_ref] Restoration and spectral recovery of mid-infrared chemical images, Mattson [/bib_ref]. Here, we only want to compare the experimental resolution with the expected one, even if with some approximation.
## Beam splitter
The lateral resolution can be estimated by directly computing the corresponding annular aperture diffraction, coinciding with PSF if no aberration is included. We therefore write the wave amplitude at focus as: [bib_ref] Infrared spectroscopic imaging for histopathologic recognition, Fernandez [/bib_ref] (2 )
where the arguments of the first order Bessel functions are: the Numerical Aperture NA, a radial coordinate normalized with the wavelength, r, and the relative radius reduction of the useful area given as the square root of the obscuration area (OBS). The expression is somewhat similar to Equation (4) of reference [bib_ref] Restoration and spectral recovery of mid-infrared chemical images, Mattson [/bib_ref] that, however, seems wrong because of a parameter is placed outside the function argument, instead of inside. The focus intensity is shown for this objective with NA = 0.52 and obscuration area of 17% in, compared with a standard objective.presents the intensity after passing the collector, i.e., the same objective. The curves are very similar to well-known results [bib_ref] Image formation in confocal scanning microscopes, Sheppard [/bib_ref] , for which an annular objective produces an increase of the side lobes, also if the central one narrowing promises a better resolution than standard objectives. At the end, the lateral resolution computed for our NA is apparently slightly better, passing from the Airy radius 0.61 1.22
## Airy na
to about 0.8λ. However, the contrast of the image can be lower, because of the side lobes, and can be better evaluated by the modulation transfer function (MTF), notoriously worse for this type of objectives.
## The pinhole
The ability to discriminate the out-of-focus signal, typical of the confocal layout, is realized by focusing the returning beam on a pinhole. To retain some flexibility in the mechanical design, we realize the focus by using parabolic mirrors that rotate the beam by 90° without chromatic or spherical aberrations. Due to the focal length and the beam diameter, the parabolic mirrors are equivalent to an optical lens with NA = 0.025. Below, in [fig_ref] Figure 4: PSF focus radial intensity profile of parabolic mirrors expected from construction parameters [/fig_ref] , the produced focus is represented by an Airy intensity plot with an adimensional diameter of 49 r , measured between the two first minima. With a pinhole diameter of 25 μm, the focus completely enters the pinhole only at the wavelengths around 0.5 μm, being cropped at longer wavelengths. This type of chromatic aberration is only relevant for the signal intensity as measured by the detectors, but has no consequences on the achromaticity, as it is possible to deduce by the axial measurements made to evaluate the out of focus discrimination (see [fig_ref] Figure 5: Reflectance intensity profiles at different wavelengths as a function of focus axial... [/fig_ref].
# Results and discussion
In the following paragraphs, we present experimental results useful to define both the axial resolution and the lateral one. As explained in Section 2, we have realized a system that manages a rather large spectral range so to obtain performances related only to physical quantities, such as the specific wavelength, and not to accidental factors, like for instance, the difficulty to maintain exactly focused two different spectrometers that work with two different detectors in completely different spectral ranges.
It is easy to show the usefulness of this hyperspectral confocal microscope for solid state and other technological application. Here, we present the simple case of the silicon/silicon oxide calibration sample, where the explicit dependence of the local reflectivity on chemically different areas allows precise measurements of the local thickness. This example would be used as a template for a vast range of problems, mainly of a technological kind, for which local composition and local properties are of some value for a deep analysis of the sample properties.
The original purpose of the realized microscope is its application to the study of biological matter, like living cells or tissues. In this case, the large accessible spectral range can increase the inventory of label marking, without adding exogenous compounds, via only the spectral features of the specific biological state of the particular specimen. The final goal of a label-free spectroscopic fingerprinting has to be still fully demonstrated mainly because of lacking of clear biochemical patterns in the near IR spectral region. However, we show here that such a purpose can be fruitfully addressed by using statistical methods for data classification, as usual for hyper-spectroscopic results.
## Axial resolution
The simplest and common way to evaluate the axial resolution of a confocal microscope is to measure the output intensity reflected by a mirror surface as a function of an axial movement of the sample or the objective along the Z axis. This quantity is directly related to the out-of-focus discriminating ability, although less clearly defined when the beam focus is immersed in a multilayered or highly scattering medium. A not exhaustive literature exists for reflective objectives, the obtained results are nevertheless not very far from a classical formulation.
For this purpose we have used a really flat silicon sample covered by a 1 μm thick silicon nitride layer, coated with 100 nm of chromium and 100 nm of gold, with an overall roughness of few nanometers, with engraved a micrometric square. This last structure is used below for lateral resolution measurements, while the mirror surface is used for axial resolution. [fig_ref] Figure 5: Reflectance intensity profiles at different wavelengths as a function of focus axial... [/fig_ref] shows some selected wavelengths within more than one thousands collected wavelengths. The FWHMs of the shown curves have a value of 3.93 μm at a wavelength of 0.55 μm, 6.51 μm at 1.064 μm, reaching 11.16 μm as measured at a wavelength of 1.5 μm. We can compare the experimental results with the expression elaborated for wide field microscopyand suitable for standard objectives:
[formula] min 2 2 z NA .(5) [/formula]
The expression, valid in air, represents the distance between minima in a reasonable description of the axial focus. Taking into account that 0.52 NA , the measured values obey Equation (5) within ±5% on average. Therefore, despite the crude approximations, the axial resolution is aligned to the predictable performances of our microscope based on the optical properties of the used objective.
## Lateral resolution
We have evaluated the lateral resolution by imaging the stepped edges of a square hole realized on the sample referred in the previous section. The method has the advantage of simplicity: acquiring the image, and selecting one or more sectioning profiles, we simultaneously obtain a profile for all accessible wavelengths. In [fig_ref] Figure 6: Image at 550 nm wavelength of square hole in Au coated silicon... [/fig_ref] we show the entire structure within a 400 μm wide image obtained with our microscope, selecting a wavelength of 550 nm. We only measure the optical intensity, but we know that the surface roughness is very small, of the order of few nanometer, and that the Electron Beam Lithography (EBL) used to realize this structure is able to produce a sharpness of a few tens of nanometers, also including the chemical processes that follow the EBL part of work. Therefore, the measured step width is broadened essentially by the optical resolution, with the addition of the stage motion inaccuracies. The measured resolution is thus an upper limit that includes mechanical errors that, on the already given specifications of our stage, are likely to be not so important.
A normalized step profile is visible in [fig_ref] Figure 7: Reflected intensity measured on the edge of the square hole [/fig_ref]. Embedded in the figure is reported the 60 μm wide image, again selected at 550 nm and with a pixel resolution of about 230 nm, used to extract the profiles.
We make a best fit of the measured profile using a Boltzmann-like step function: Following Equation (6), c o is the intensity offset, very near to zero, A is the step function amplitude, and the X, that widens the step around the pivot point X O , assumes the role of lateral resolution [bib_ref] Supercontinuum ultra wide range confocal microscope for reflectance spectroscopy of living matter..., Selci [/bib_ref]. It is possible to compute the X parameters from the results obtained at different wavelengths.
[formula] 0 0 () 1 O pro X X X Ac Ic e .(6) [/formula]
Moreover, we know that the theoretical lateral resolution can be written as Airy X k NA . Usually, the parameter k is 0.61, in the case of a pure Airy function and, because in our case NA is nearly 0.5, the expected value is Airy X . Therefore, it is meaningful to plot the quantity exp X to check the experimental results against the theoretical values. The final result is visible in [fig_ref] Figure 8: Fit of the lateral resolution, normalized to the wavelength, and plotted as... [/fig_ref]. It is clear that the trend is very different compared to the value of 1.22, but is very near to the value of 0.8 that we have deduced from the theoretical PSF of our objective. It means that, technically, we are working in a super-resolution regime. However, the optical performances are also related to the contrast expressed by the PSF lateral lobes weight, as already explained. Nevertheless, whenever a sample is imaged with intrinsic high contrast, e.g., technological samples, reflective objectives are able to perform better than more usual objectives.
## Spectroscopy on semiconductor samples
The most obvious application of our microscope is the measure of reflectivity. The advantages of this method are various and innovative. Reflectivity is measured on a local scale, close only to the beam focus surroundings. It can be measured on an buried interface or a discontinuity, and not exclusively on the sample surface: this requires the specimen transparency but, due to the available spectral range, it is possible to reach the hidden face by choosing the right wavelengths; as an example, one can image the buried interface of evaporated layers on a semiconductor substrate simply going below the energy gap of the material and imaging from the back-side of the device [bib_ref] Supercontinuum ultra wide range confocal microscope for reflectance spectroscopy of living matter..., Selci [/bib_ref]. Moreover, the accessible spectral range is often the region of strong variation of the optical constants for the most common materials, so very precise measurements of the composition or thicknesses of a sample can be derived from experimental data. The sensitivity of the method is such that differences of few nanometers in the thickness of the observed layers can be readily observed, and an example follows.
We have used an Atomic Force Microscopy (AFM) calibration sample with periodic structures of SiO 2 on a silicon substrate (Nanosurf ® AG, Liestal, Switzerland). The silicon oxide squares have a side of 5 μm with a periodicity of 10 μm. The images of the structure in [fig_ref] Figure 9: AFM Calibration sample image at [/fig_ref] clearly show the spectroscopic effect. In fact, while in the image layer at λ = 550 nm (left) some protrusions in the black square structures are visible, in the image at λ = 650 nm (right) the squares appear as deep minima, as clarified by the profile inserts: data are dominated by spectroscopic effects, a consideration that is not taken in the proper consideration when images are obtained with standard, single wavelength, confocal microscopes. To understand this effect it is worthwhile to underline that it is quite difficult to measure the absolute reflectivity, because an absolute reference inserted in the same optical path should be available. Instead, it is possible to experimentally evaluate the ratio between two compositionally different sample zones, in this case the silicon oxide respect to the bare silicon. The experimental ratio, averaged over contiguous square zones, is shown in [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref]. The experimental data (blue line) are reproduced quite well by the theoretical calculation (red line) for a thickness of the silicon oxide hill of 108 nm, against a declared value of 100 nm and AFM measured value of 119 nm, referred to the top maximum of the structure. The theoretical lines are computed by a full multilayer calculation [bib_ref] Supercontinuum ultra wide range confocal microscope for reflectance spectroscopy of living matter..., Selci [/bib_ref] based on the Abelé s matrix method. Dielectric functions for silicon and silicon oxide are taken from Palik. The sensitivity of this methodology can be appreciated by noting that theoretical calculations for the thicknesses of 100 nm (black line) and 120 nm (dotted line) are clearly incorrect.
## Biological samples; statistical representation
The reflectance hyperspectral confocal microscope described so far has been developed in the framework of a national project aimed at performing optical spectroscopy of the skin in three dimensions in order to assess the possibility of carrying out diagnosis and classification of skin pathologies using optical spectroscopy. In this context, the investigation of biological specimens represents a primary objective. We show an example of discrimination of cellular compartments based on spectral features, achieved with a principal component analysis on hyperspectral data.
We have used M101221 melanoma cell line derived from metastasis recurrence in a patient during vemurafenib therapy treatment, kindly provided by Prof. Dummer (University Hospital, Zurich, Switzerland). Cells were grown in RPMI-1640 medium supplemented with 10% (v/v) inactivated Fetal Bovine Serum, 1% penicillin/streptomycin (medium, serum and antibiotics purchased from Lonza, Basel, Switzerland) and 1% Na pyruvate (Life technologies, Carlsbad, CA, USA). Cells were maintained at 37 °C in a humidified atmosphere of 5% CO 2 and 95% air until observation.
To perform measurements on melanoma cell culture, we used a culture chamber specially designed and 3D printed (HP Designjet 3D printer, Hewlett-Packard, Palo Alto, CA, USA) in our laboratory. Cells have been plated and cultured on 40 mm diameter round coverglass in a 60 mm Petri dish. 24 h after seeding, the coverglass was mounted as bottom of the chamber, sealed with a previously sterilized 2 mm thick silicone gasket. The chamber has then been fixed on the microscope stage to perform acquisition. Considering laser power emission, focusing optical system characteristics and culture chamber parameters, we estimated the light power on the cell layer to be less than 4 μW/nm. Extensive light exposure has been performed in different cell cultures with no evident damage. [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref] shows an image of melanoma cells (400 μm × 400 μm) at λ = 1,064 nm (left portion) and λ = 550 nm (right portion) to highlight how the availability of spectral images allows to visualize different features also in biological samples. Spectra from nuclear areas and from cytoplasmic-membrane areas of two different cells (arrows) have been acquired and averaged over a circular region of interest (23 µm diameter). As internal reference we have used a portion of glass substrate free of cells: every spectrum has been normalized to the reference spectrum. For each cell, average spectra from two different nuclear areas have been acquired. Four spectra from cytoplasmic areas in the cell labeled 1, and two spectra from cytoplasmic areas in the cell labeled 2, have been acquired as well. To visualize images and extract average spectra from the regions of interest we have used custom-developed software.
From the spectra in [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref] it is evident that both in the VIS and in the IR regions there are spectral features that may lead to the identification of different compartments. However, it must be considered that the spectral response we obtain is the combination of a morphological contribution from scattering effects and a chemical contribution from absorption effects. For this reason, especially dealing with complex samples, a pure light scattering interpretative model [bib_ref] Confocal light absorption and scattering spectroscopic microscopy monitors organelles in live cells..., Itzkan [/bib_ref] or a pure chemometric model may be reductive. 2 2 [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref]. Mean reflectance spectra of cytoplasmic areas (black) and nuclear areas (red) from melanoma cells. Each spectrum is the result of the sum of visible spectrum (500 nm-1,000 nm) data from visible sensor and infrared spectrum (1,000 nm-1,600 nm) data from IR sensor.
Recent advances in analysis instrumentation and in data collection techniques have resulted in a rapid increase in the amount of data acquired from spectral imaging. Extracting the significant information from the data produced by modern instrumentation is, in many circumstances, a nontrivial task. Here we introduce the application of statistical tools to sample analysis and modeling. These tools can be used to extract useful indications about the structure and the nature of the sample, leading to highly reliable classification.
Previously described spectra from nuclear and cytoplasmic-membrane areas of two different cells have been considered for principal component analysis (PCA). Average spectra from different areas have been smoothed with a Savitzky-Golay algorithm and compared using PCA. Normalized reflectance intensity at nine different wavelengths, spanning from 500 nm up to 1,300 nm (with 100 nm interval between wavelengths) have been considered as variables. The first two principal components have been considered (cumulative variance 87.12%) and points representative of average spectra from different areas and different cells have been plotted in a PC1/PC2 space. Then different shapes have been assigned to different compartments, and different colors have been assigned to different cells [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref].
From the plot in [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref] it is easy to recognize that spectra from different cell compartments can be separated, for they occupy different half planes of the PC1 dimension (except one point), but spectra from different cells can also be distinguished for they occupy a different half plane of the PC2 dimension. This analysis usually represents the first step of more complex classification and clusterization methods. Within this scope, the application of a multivariate approach to hyperspectral data, although not new in remote sensing and food/drug analysis [bib_ref] Study of pharmaceutical samples by NIR chemical-image and multivariate analysis, Amigo [/bib_ref] [bib_ref] Rice seed cultivar identification using near-infrared hyperspectral imaging and multivariate data analysis, Kong [/bib_ref] [bib_ref] Application of multivariate curve resolution alternating least squares (MCR-ALS) to remote sensing..., Zhang [/bib_ref] and coupled to FTIR and Raman micro-spectroscopy [bib_ref] Detection and identification of cancerous murine fibroblasts, transformed by murine sarcoma virus..., Salman [/bib_ref] , as far as we know, has not been used in reflectance confocal microscopy with cell culture. It is important to underline that PCA is an unsupervised method and that the qualitative results are independent from the specific set of sampling wavelengths, thus making it a valuable and adjustable tool to perform first analysis in different frameworks. [fig_ref] Figure 1: Microscope layout and the optical pathway [/fig_ref]. Score plot of PCA analysis on cellular compartments of melanoma cells in PC space. The first two components have been considered: points represent spectra from cell1 (black) and from cell2 (red), from nuclear areas (up triangle) or from cytoplasm (square).
# Conclusions
The new confocal microscope presented here has been realized with the idea of integrating structural and morphological information with detailed spatially resolved spectroscopic properties. Every pixel in the obtained confocal image is associated with spectral information in the wide range allowed by the combined characteristics of the laser source and the two detectors for radiation reflected by the sample (0.5 μm-2.4 μm wavelengths).
Consequently, this technique is an attractive tool for a wide range of applications in many critical areas, including bioscience and medical diagnostics, inorganic and organic material sciences, environmental sciences, forensics and archeology. Data are acquired with sub-micrometer spatial resolution and high spectral resolution. Preliminary data with clear different cells discrimination show that also if reflectance has no specific sensitivity to biochemical properties, a spectroscopic classification is possible. Optical differences between different cells can be attributed to real chemical compositional differences or to a mere variation in the density of the same constituents, contributing anyway to the index of refraction local definition and to very sensitive measurements.
[fig] Figure 1: Microscope layout and the optical pathway. Abbreviations are given in the text. [/fig]
[fig] Figure 2: Optical paths of the reflected and the refracted rays on the first beam splitter. [/fig]
[fig] Figure 3: (a) Calculated PSF profile of a 0.52 NA objective (blue line) and a 0.52 NA annular aperture objective with 17% central obscuration; (b) Calculated PSF profile of the same objective as in (a) followed by the collecting lens (coinciding with the same objective). Normalized radial units are used. [/fig]
[fig] Figure 4: PSF focus radial intensity profile of parabolic mirrors expected from construction parameters. Normalized radial units are used. [/fig]
[fig] Figure 5: Reflectance intensity profiles at different wavelengths as a function of focus axial positions. [/fig]
[fig] Figure 6: Image at 550 nm wavelength of square hole in Au coated silicon nitride membrane. The lateral side of the image is 400 μm. [/fig]
[fig] Figure 7: Reflected intensity measured on the edge of the square hole. The extraction region is indicated by the dashed line on the inset image (wavelength 550 nm, lateral size 60 μm). [/fig]
[fig] Figure 8: Fit of the lateral resolution, normalized to the wavelength, and plotted as a function of the wavelength. [/fig]
[fig] Figure 9: AFM Calibration sample image at: (a) λ = 550 nm; (b) λ = 650 nm. The lateral side of the images is 30 μm. The measured intensities profiles are indicated by the dashed lines. [/fig]
|
Comparison Between Dynamic Contour Tonometry and Goldmann Applanation Tonometry
Purpose: To compare the intraocular pressures (IOPs) measured by dynamic contour tonometry (DCT) and Goldmann applanation tonometry (GAT), and to investigate the association of IOPs on eyes of varying central corneal thickness (CCT). Methods: In this prospective study, 451 eyes of 233 subjects were enrolled. IOPs were measured by GAT and DCT. CCT was measured three times and the average was calculated. Each eye was classified into one of three groups according to CCT: low CCT (group A, CCT<520 µm, n=146); normal CCT (group B, 520 µm≤CT≤550 µm, n=163); and high CCT (group C, CCT>550 µm, n=142). In each group, we investigated the association of CCT with IOP measurement by GAT and DCT. Results: The IOPs measured by GAT and DCT were significantly associated for all eyes (R=0.853, p<0.001, Pearson correlation). CCT was related with both IOP measurement by GAT and DCT with statistical significance (mixed effect model, p<0.001). However, subgroup analysis showed that CCT affected IOP measured by GAT for groups B and C, whereas it affected IOP measured by DCT only for group C. Conclusions: IOP measured by DCT was not affected by CCT in eyes with low to normal CCT, whereas this measurement was affected in eyes of high CCT range. CCT may have less effect on IOP measurements using DCT than those obtained by GAT, within a specified range of CCT.
It is widely known that central corneal thickness (CCT) affects the measurement of intraocular pressure (IOP) by Goldmann applanation tonometry (GAT).This is because IOP measured by GAT is calculated according to the modified Imbert-Pick law with the assumption that the CCT is 550 µm.In a recent study of Zhang and colleagues, 2 the mean± SD of CCT in Chinese adults was 556±33.1 µm (median: 553 µm, range: 429-688 µm). When this wide range of CCT is considered, a tonometer minimally affected by CCT is of clinical importance.
The dynamic contour tonometer (DCT; Pascal ® ; Swiss Microtechnology AG, Port, Switzerland) is a device that may measure IOP relatively independently of CCT. The DCT sets a hypothetical corneal contour that is achieved when the pressures on the two sides of the cornea are equal. The force distribution needed to gently fit the corneal surface to that hypothetical contour counterbalances the force distribution generated by the IOP. Hence, a pressure sensor centrally and concavely embedded into the tonometer tip precisely measures the transcorneal pressure of the eye. [bib_ref] Dynamic contour tonometry: presentation of a new tonometer, Kanngiesser [/bib_ref] Although a few studies have compared IOP measured by DCT with IOP measured by GAT, it remains unclear whether either measurement is affected by CCT. Kaufmann et al. [bib_ref] Comparison of dynamic contour tonometry with Goldmann applanation tonometry, Kaufmann [/bib_ref] and Schneider et al. [bib_ref] Intraocular pressure measurement -comparison of dynamic contour tonometry and goldmann applanation tonometry, Schneider [/bib_ref] reported that IOP measured by DCT was not significantly affected by CCT. Kotecha and co-workers [bib_ref] The relative effects of corneal thickness and age on Goldmann applanation tonometry..., Kotecha [/bib_ref] concluded that DCT was less affected by CCT than was GAT. However, Doyle and Lachkar, 7 in a study which included 75 eyes, concluded that DCT permitted accurate assessment of true IOP in eyes with thin and structurally normal corneas, but had no advantage over GAT in eyes with thick corneas. Thus, this study was intended to compare IOPs measured by DCT with those obtained using GAT, and investigate the association of IOP data with CCT in a considerably large number of eyes.
# Materials and methods
This prospective single-center study assessed 451 eyes of 233 subjects (mean±SD age: 54.9±13.8 years) recruited from the glaucoma clinic at the Asan Medical Center, Seoul, Korea, from October 2005 to September 2006. All subjects were referred from primary eye care clinics upon suspicion of glaucoma.
Each subject was initially seen by one of our glaucoma specialists (J.C. or C.H.L.) and underwent IOP measurements by GAT and DCT. After one drop of 0.25% solution of sodium fluorescein in combination with Alcaine ® (proparacaine chloride; 5 mg/mL; Alcon, Fort Worth, TX) was instilled into the lower conjunctival fornix, IOP measurement was performed with GAT and DCT sequentially 5 minutes apart. GAT was performed according to the guidelines of the Eye Care Technology Forum. [bib_ref] Standardizing the measurement of intraocular pressure for clinical research. Guidelines from the..., Kass [/bib_ref] DCT was measured with a Pascal tonometer mounted on a slit-lamp. DCT shows IOP as a digital numeric outcome. Thus, prior knowledge of the GAT value would not influence the DCT result, which made it unnecessary to mask the examiner to the results of the two IOP measurements. IOP measurement with DCT was repeated if the quality score was greater than 2.
Next, CCT measurements using ultrasonic pachymetry (Pachette2 ® pachymetry; DGH Technology Inc.; Exton, PA) were performed. The examiner thus measured IOP independent of CCT. Corneal thickness was measured three times within the pupil margin, and care was taken not to dent the cornea with the pachymetry tip. Values were accepted if the standard deviation of each measurement was within 5.0 µm. An average of three measurements was obtained for data analysis.
Patients were enrolled if the IOP was below 30 mmHg, and the corneal surface was smooth and regular so that GAT could be used. As marked astigmatism of more than four diopters can result in erroneous GAT measurements, such patients were excluded. [bib_ref] Goldmann applanation tonometry in patients with regular corneal astigmatism, Holladay [/bib_ref] Patients under 20 years of age were excluded because of restlessness during IOP measurements which might lead to unreliable IOP data.
All participants gave an informed consent. All procedures conformed to the Declaration of Helsinki and the study was approved by the Ethics Committee of the Asan Medical Center at the University of Ulsan, Korea.
IOPs measured by GAT, and DCT, were tested for association using the Pearson correlation. Next, subjects were divided into three groups according to CCT measurements using 520 µm and 550 µm as cut-off values, and similar number of eyes were allocated to each group: a thin cornea group (group A, CCT<520 µm, n=146), a normal cornea group (group B, 520 µm≤CT≤550 µm, n=163), and a thick cornea group (group C, CCT>550 µm, n=142).
The effect of CCT on IOP measurement was tested by mixed effect model accounting for clustering of eyes within subject, gender and age. One eye (left eye) was incorporated into multivariate regression model for validation of those findings determined by mixed effect model. After CCT was conformed to be a statistically significant parameter based on multivariate regression model, Pearson correlation analysis was further performed to obtain coefficient (R). SAS ® (SAS Institute Inc., Cary, NC) version 9.1 and SPSS ® (SPSS Inc., Chicago, IL) version 11.5 was employed for statistical analysis.
# Results
Data on patient demographic and background variables for all subjects and for each group classified by corneal thickness are shown in [fig_ref] Table 1: Patient demographics and background variables [/fig_ref]. All patients were Korean, and 184 (40.8%) eyes were of men whereas 267 (59.2%) were of women. The mean (±SD) of age was 54.9±13.8 years (range, 21-82 years). IOP measured by DCT was significantly higher than IOP measured by GAT (14.8±3.7 mmHg versus 16.9± 3.4 mmHg; p<0.001; paired t-test). There were no significant differences in age, gender, spherical equivalent and visual field mean deviation for age, gender, spherical equivalent, and mean deviation, respectively; one-way ANOVA, chi-square test). depicted statistically significant correlations between IOPs measured by GAT and DCT (R=0.853, p<0.001, Pearson correlation analysis). showed that both IOP measurements had significant correlations with CCT (GAT; R=0.330, p<0.001), (DCT; R=0.271, p<0.001). IOP measured by GAT was associated with CCT in groups B and C whereas IOP measured by DCT in group C only [fig_ref] Table 2: Relationship of IOP measured by GAT and DCT with CCT assessed by... [/fig_ref]. [fig_ref] Table 3: Relationship of IOP measured by GAT and DCT with CCT in left... [/fig_ref] showed the effect of CCT on IOP measurement using multivariate regression in left eye. Pearson correlation coefficients (R) were assessed when IOP measurements were influenced by CCT.
## D iscussion
Many studies have discussed the influence of CCT on IOP as measured by GAT. No general consensus has been reached regarding the correction factor that should be used for adjusting IOP measured in this way.According to Ehlers and colleagues, the average error is 0.7 mmHg per 10 µm. [bib_ref] Applanation tonometry and central corneal thickness, Ehlers [/bib_ref] In a more recent study, however, Wolfs and co-workers suggested smaller values, as little as 0.19 mmHg per 10 µm, should be used to correct GAT values. [bib_ref] Distribution of central corneal thickness and its association with intraocular pressure: the..., Wolfs [/bib_ref] There is thus agreement that CCT influences IOP measurement, but different correction factors have been proposed. This suggested the need for a tonometry mode that would yield correct data independent of CCT, and DCT was developed for that purpose.
Theoretically, IOP measured by DCT should not be affected by CCT. In this study, however, IOP measured by DCT was significantly affected by CCT in the thick cornea group (p<0.001). Despite this finding, DCT was not significantly affected by CCT in groups with thin and normal corneal thickness. Lesser impact of CCT and a good correlation between DCT and GAT measurements increase the clinical value of DCT. These results agreed with those of other studies in smaller series. [bib_ref] Comparison of dynamic contour tonometry with Goldmann applanation tonometry, Kaufmann [/bib_ref] [bib_ref] Intraocular pressure measurement -comparison of dynamic contour tonometry and goldmann applanation tonometry, Schneider [/bib_ref] [bib_ref] The relative effects of corneal thickness and age on Goldmann applanation tonometry..., Kotecha [/bib_ref] [bib_ref] Comparison of dynamic contour tonometry with Goldmann applanation tonometry over a wide..., Doyle [/bib_ref] Doyle and Lachkar calculated that IOP measured by GAT was underestimated by 0.7 mmHg for every 10 µm of CCT below 520 µm, and overestimated by 0.2 mmHg for every 10 µm of CCT over 580 µm; these data support the idea that DCT is more accurate than GAT when used on thin or normal corneas. [bib_ref] Comparison of dynamic contour tonometry with Goldmann applanation tonometry over a wide..., Doyle [/bib_ref] The cited study used 75 eyes, but we examined 451 eyes. Doyle and Lachkar considered that there was much greater variability in data from the two tonometers when used on thick corneas, and added that this result was unexpected and difficult of explanation. [bib_ref] Comparison of dynamic contour tonometry with Goldmann applanation tonometry over a wide..., Doyle [/bib_ref] Similarly, we cannot offer a conclusive reason for our result that CCT influences IOP measured by DCT. It is possible that other unknown corneal factors other than CCT, such as hysteresis, might affect the measurement of IOP by DCT. Corneal thickness per se may not be the direct cause of high IOP. Instead, if eyes with thick corneas tend to have different corneal biomechanical properties that are difficult to quantify, CCT can be mistaken for the cause of high IOP although the real cause lies with biomechanical property such as hysteresis. However, data from this study, is not sufficient to explain the findings in the thick cornea group and this is one of limitation of our study.
Another limitation of this study is the arbitrary cut-off values of CCT used for dividing patients into groups. The three groups were intended to represent thin, normal, and thick cornea groups. However, there is no consensus on classification of corneas as thin, normal, or thick. Goldmann and Schmidt used 520 µm as an average corneal thickness in their work with the modified Imbert-Fick law. [bib_ref] Applanation tonometry, Goldmann [/bib_ref] In other studies, the mean values were 537 to 554 µm in normal subjects. [bib_ref] Distribution of central corneal thickness and its association with intraocular pressure: the..., Wolfs [/bib_ref] [bib_ref] Relationship between corneal thickness and measured intraocular pressure in a general ophthalmology..., Shah [/bib_ref] In addition, CCT varies with race; an average of 531 µm in one African-American population, a mean of 547 µm in Latinos, and an average of 552 µm in Japanese and Caucasians. [bib_ref] Central corneal thickness and intraocular pressure in a Monglolian population, Foster [/bib_ref] [bib_ref] Central corneal thickness of Caucasians and African Americans in glaucomatous and nonglaucomatous..., Rosa [/bib_ref] [bib_ref] Central corneal thickness in Latinos, Hahn [/bib_ref] [bib_ref] Central corneal thickness of normal tension glaucoma patients in Japan, Wu [/bib_ref] [bib_ref] Intraocular pressure, Goldmann applanation tension, corneal thickness and corneal curvature in Caucasians,..., Shimmyo [/bib_ref] Considering these findings, we divided our subjects into three groups using 520 µm and 550 µm as cut-off values, resulting in similar number of eyes allocated to each group.
A third limitation of this study is the lack of normal control group. We could not collect data from normal control group, thus further study which includes normal population can be of value to complement the limitation of this study.
In conclusion, we found that IOP measured by DCT correlates well with IOP measured by GAT. IOPs obtained by DCT are less dependent on, but not totally independent of CCT, compared to those measured by GAT, especially when the corneal thickness is normal or thin. These findings indicate that DCT measure can be less affected by CCT than that of GAT in patients who have normal or thin corneas.
[fig] Figure 1, Figure 2: Pearson correlation analysis of intraocular pressure (IOP) measurements obtained by Goldmann tonometry and dynamic contour tonometry (n=451, R=0.853, p<0.001). (A) Association of central corneal thickness (CCT) with intraocular pressure (IOP) measurement by Goldmann applanation tonometry (GAT) (n=451, R=0.330, p<0.001). (B) Association of CCT with IOP measurement by Pascal-dynamic contour tonometry (P-DCT) (n=451, R=0.271, p<0.001). [/fig]
[table] Table 1: Patient demographics and background variables (one-way ANOVA, chi-square test)Data are expressed as means±SDs (range) except for gender. IOP=intraocular pressure; GAT=Goldmann applanation tonometry; DCT=dynamic contour tonometry; CCT=central corneal thickness; MD=mean deviation; PSD=pattern standard deviation. [/table]
[table] Table 2: Relationship of IOP measured by GAT and DCT with CCT assessed by mixed effect model (p value) CCT<520 µm; Group B: 520 µm≤CCT≤550 µm; Group C: CCT>550 µm. IOP=intraocular pressure; GAT=Goldmann applanation tonometry; DCT=dynamic contour tonometry; CCT=central corneal thickness. Statistically significant p values are marked with asterisk (*). [/table]
[table] Table 3: Relationship of IOP measured by GAT and DCT with CCT in left eyes assessed by multivariate regression morel [/table]
|
A homology independent sequence replacement strategy in human cells using a CRISPR nuclease
Precision genomic alterations largely rely on homology directed repair (HDR), but targeting without homology using the non-homologous end-joining (NHEJ) pathway has gained attention as a promising alternative. Previous studies demonstrated precise insertions formed by the ligation of donor DNA into a targeted genomic double-strand break in both dividing and non-dividing cells. Here, we demonstrate the use of NHEJ repair to replace genomic segments with donor sequences; we name this method 'Replace' editing (Rational end-joining protocol delivering a targeted sequence exchange). Using CRISPR/Cas9, we create two genomic breaks and ligate a donor sequence in-between. This exchange of a genomic for a donor sequence uses neither microhomology nor homology arms. We target four loci in cell lines and show successful exchange of exons in 16-54% of human cells. Using linear amplification methods and deep sequencing, we quantify the diversity of outcomes following Replace editing and profile the ligated interfaces. The ability to replace exons or other genomic sequences in cells not efficiently modified by HDR holds promise for both basic research and medicine. Recently, increasing awareness of the fidelity and efficiency of NHEJ repair has led to the development of methods to produce genomic deletions and exogenous sequence insertions using this pathway. As NHEJ is highly active in all phases of the cell cycle, it has allowed precise edits in muscle cells and neurons[15][16][17].
# Introduction
RNA-guided nucleases [bib_ref] Cas9-crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria, Gasiunas [/bib_ref] [bib_ref] A programmable dual-RNAguided DNA endonuclease in adaptive bacterial immunity, Jinek [/bib_ref] [bib_ref] RNA-guided human genome engineering via Cas9, Mali [/bib_ref] have rapidly become foundational tools in facilitating genomic manipulations [bib_ref] The next generation of CRISPR-Cas technologies and applications, Pickar-Oliver [/bib_ref] [bib_ref] A new class of medicines through DNA editing, Porteus [/bib_ref]. These nucleases target specific genomic loci and form a double-strand break (DSB). DNA repair processes are then leveraged to produce the desired outcome of the gene editing. Conventionally, specific genomic changes are made using homology directed repair (HDR) [bib_ref] CRISPR/Cas9 β-globin gene targeting in human haematopoietic stem cells, Dever [/bib_ref] [bib_ref] Reprogramming human T cell function and specificity with non-viral genome targeting, Roth [/bib_ref] with exogenously introduced DNA containing flanking sequences homologous to the targeted locus. One limitation of HDR-mediated genome editing is its restriction to the S/G2 phase, reducing or abolishing efficacy in slowly or non-dividing cells [bib_ref] A mechanism for the suppression of homologous recombination in G1 cells, Orthwein [/bib_ref]. HDR, when used for gene editing, can be precise, but recent reports demonstrate greater error than often assumed, as incomplete or extraneous portions of the delivery vector can be copied into the genome [bib_ref] Capture of retrotransposon DNA at the sites of chromosomal double-strand breaks, Moore [/bib_ref] [bib_ref] Systematic gene tagging using CRISPR/Cas9 in human stem cells to illuminate cell..., Roberts [/bib_ref] [bib_ref] Repair of site-specific double-strand breaks in a mammalian chromosome by homologous and..., Sargent [/bib_ref] [bib_ref] Retrotransposon reverse-transcriptase-mediated repair of chromosomal breaks, Teng [/bib_ref]. On the other hand, the canonical non-homologous end-joining (NHEJ) pathway is traditionally viewed as error prone and relegated to disrupting gene function by inducing small insertions and deletions (InDels) during DSB repair. However, the highfidelity aspects of NHEJ repair are often underappreciated as mutant InDels are easily observed, whereas non-mutagenic repair is indistinguishable from the original allele [bib_ref] Is nonhomologous end-joining really an inherently errorroyalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 prone..., Bétermier [/bib_ref]. Furthermore, non-mutagenic repair by NHEJ reforms the Cas9 target site allowing for continued DSB formation. This may result in a final genomic population containing majority InDels despite NHEJ repair being predominately error-free.
Targeted deletions are produced by forming two DSBs with loss of the intervening sequence during repair. The ubiquitous nature of the NHEJ pathway allows for deletions in zygotes, as well as in adult tissue such as in vivo exon deletion in a mouse muscular dystrophy model [bib_ref] Long-term evaluation of AAV-CRISPR genome editing for Duchenne muscular dystrophy, Nelson [/bib_ref] [bib_ref] Creation of targeted genomic deletions using TALEN or CRISPR/Cas nuclease pairs in..., Brandl [/bib_ref]. Additionally, exogenously introduced dsDNA donor sequences can efficiently ligate into a single DSB by NHEJ (herein referred to as Insert targeting) [bib_ref] Plug-and-play protein modification using homology-independent universal genome engineering, Gao [/bib_ref] [bib_ref] In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration, Suzuki [/bib_ref] [bib_ref] Highly efficient CRISPR/Cas9-mediated knock-in in zebrafish by homology-independent DNA repair, Auer [/bib_ref] [bib_ref] In vivo cleavage of transgene donors promotes nuclease-mediated targeted integration, Cristea [/bib_ref] [bib_ref] Knock-in of large reporter genes in human cells via CRISPR/Cas9-induced homologydependent and..., He [/bib_ref] [bib_ref] A generic strategy for CRISPR-Cas9-mediated gene tagging, Lackner [/bib_ref] [bib_ref] Obligate Ligation-Gated Recombination (ObLiGaRe): custom designed nucleases mediated targeted integration through non-homologous..., Maresca [/bib_ref] [bib_ref] Zinc-finger nuclease-driven targeted integration into mammalian genomes using donors with limited chromosomal..., Orlando [/bib_ref] [bib_ref] CRISPaint allows modular base-specific gene tagging using a ligase-4-dependent mechanism, Schmid-Burgk [/bib_ref]. With the NHEJ pathway conserved broadly, Insert targeting has been shown in plants, fish [bib_ref] Highly efficient CRISPR/Cas9-mediated knock-in in zebrafish by homology-independent DNA repair, Auer [/bib_ref] , cell lines [bib_ref] In vivo cleavage of transgene donors promotes nuclease-mediated targeted integration, Cristea [/bib_ref] [bib_ref] Knock-in of large reporter genes in human cells via CRISPR/Cas9-induced homologydependent and..., He [/bib_ref] [bib_ref] A generic strategy for CRISPR-Cas9-mediated gene tagging, Lackner [/bib_ref] [bib_ref] Obligate Ligation-Gated Recombination (ObLiGaRe): custom designed nucleases mediated targeted integration through non-homologous..., Maresca [/bib_ref] [bib_ref] Zinc-finger nuclease-driven targeted integration into mammalian genomes using donors with limited chromosomal..., Orlando [/bib_ref] [bib_ref] CRISPaint allows modular base-specific gene tagging using a ligase-4-dependent mechanism, Schmid-Burgk [/bib_ref] , nondividing neurons and in vivo mouse tissues [bib_ref] Plug-and-play protein modification using homology-independent universal genome engineering, Gao [/bib_ref] [bib_ref] In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration, Suzuki [/bib_ref]. The ability to effectively integrate DNA across cell types has been used to tag genes with fluorophores [bib_ref] Plug-and-play protein modification using homology-independent universal genome engineering, Gao [/bib_ref] [bib_ref] Zinc-finger nuclease-driven targeted integration into mammalian genomes using donors with limited chromosomal..., Orlando [/bib_ref] [bib_ref] CRISPaint allows modular base-specific gene tagging using a ligase-4-dependent mechanism, Schmid-Burgk [/bib_ref] , identify off-target CRISPR cleavage sites [bib_ref] GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases, Tsai [/bib_ref] and as a strategy for gene therapy by inserting functional coding sequences upstream of a disease causing exon [bib_ref] In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration, Suzuki [/bib_ref].
Leveraging NHEJ repair to create large deletions and insert exogenous DNA posits the possibility of NHEJ-based sequence replacement; two DSBs are produced and a donor sequence without homology is ligated between the two breaks. This approach would enable the replacement of defective exons or regulatory sequences in a wide range of resting or dividing cells. NHEJ-based replacement has been demonstrated in plants, where HDR is often infeasible [bib_ref] Gene replacement by intron targeting with CRISPR-Cas9, Li [/bib_ref] [bib_ref] Nonhomologous end joining-mediated gene replacement in plant cells, Weinthal [/bib_ref]. In order for NHEJ-based replacement to be considered a viable approach in human cells, demonstration of its efficiency and a thorough understanding of the editing outcomes is required. Here, we demonstrate efficient replacement of genomic sequences and exons with a donor sequence in human cells using NHEJ repair; we call this method Replace (Rational end-joining protocol delivering a targeted sequence exchange). Analysis of single-cell-derived clones provides conclusive evidence of Replace editing and efficiency. We further introduce sequencing pipelines for the precise quantification of the structural variants produced during Replace targeting and the InDels at the ligated interfaces. Together, our results and analysis strategies lay the groundwork for future applications of NHEJ-based Replace editing in gene therapy and research.
# Results
Replace targeting (figure 1a) aims to exchange a genomic sequence with a double-stranded donor sequence without the use of homology. In this strategy, undesired products such as deletions or inverted donor sequences reform the Cas9 gRNA target sites and can be further targeted by Cas9, while the desired integration is captured. For initial validation, we used a fluorescence-based reporter system (figure 1b). The synthetic reporter system was created and integrated into two AAVS1 loci in a HeLa cell line. The reporter system contains a CAG promoter upstream of a BFP fluorophore. The BFP prevents the expression of a downstream Venus-pA. The cells initially are BFP + . Replace targeting exchanges the BFP cassette with a mCherry donor. Reporter HeLa cells were lipofected to deliver the donor sequence and Cas9 plasmid containing a puromycin resistance gene. Cells were selected for 48 h to ensure construct delivery and analysed after two weeks. Replacement targeting cleaved both sides of the BFP-pA cassette, with the excised sequence exchanged with the linearized mCherry donor sequence. Correct ligation of mCherry resulted in the loci expressing only mCherry. Deletion of the BFP cassette without replacement resulted in expression of the downstream Venus. Some alleles lost expression due to mutations or incorrect donor ligation.
Replace targeting of the reporter locus resulted in 34% mCherry + cells (figure 1c). We compared the effect of delivering donor sequences within a plasmid or in the form of minicircles as a previous report showed minicircles to increase Insert efficiency [bib_ref] In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration, Suzuki [/bib_ref] (figure 1d). Minicircles are minimal plasmids and contain only the donor sequence and require only a single Cas9 DSB for linearization, whereas plasmids require two DSBs to excise the donor. Donor sequences delivered as minicircles resulted in a sixfold increase in cells with mCherry expression compared to plasmid delivery. We therefore used minicircles for Replace targeting in the remainder of this work. To address if mCherry expression was driven in part by off-target integration of the donor sequence, we Replace targeted, in an otherwise identical manner, wild-type HeLa cells. As these cells do not contain the AAVS1 integrated promoter and target site, only off-target integration could result in mCherry expression (figure 1d). Wild-type HeLa cells showed no mCherry expression indicating that the 34% mCherry + cells in our original experiment are the result of correct integration at the target loci. mCherry + cells were singlecell sorted, expanded and genotyped to check for correct sequence replacement. Twenty-four out of 25 analysed clones (i.e. 32% of all cells) contained the anticipated exchange of BFP with mCherry, while one clone contained an allele with mCherry insertion upstream of BFP (figure 1e). As HeLa reporter cells contained two copies of the reporter locus, we quantify the frequency of homozygous knock-in by simultaneously transfecting two donor sequences (mCherry and miRFP670) (electronic supplementary material, figure S1). By measuring the mCherry + , RFP + and dual-positive populations, we calculated an average of 5% homozygous knock-in. Taken together, Replace targeting in our reporter system occurs as a major outcome, with a successful sequence exchange of at least one allele in 32% of cells.
During the ligation of the donor sequence into the genome, InDels may occur at the interface. To quantify short InDels, the gDNA of targeted and unsorted HeLa reporter cells was PCR amplified using primers flanking the ligated interface. The deconvolution of the Sanger traces of these amplicons provides an InDel estimate of the bulk population of Replace targeted cells [fig_ref] Figure 1: Replace targeting using HeLa reporter cells [/fig_ref]. This analysis shows that short resection occurs in a minor (less than 16%) fraction of these small amplicons. The majority contained no InDel or a small, non-random insertion. Sanger sequencing of cloned individual alleles supports the bulk analysis (electronic supplementary material, [fig_ref] Figure 2: Exon Replace editing in K562 cells [/fig_ref]. The one or two nucleotide insertions were striking in that they matched the protospacer sequence downstream of the break site. It is known that SpyCas9 does not always form a canonical blunt end break three nucleotides downstream of the PAM, but can, at some frequency, form a staggered cut [bib_ref] Precise and predictable CRISPR chromosomal rearrangements reveal principles of Cas9-mediated nucleotide insertion, Shou [/bib_ref] [bib_ref] Bidirectional degradation of DNA cleavage products catalyzed by CRISPR/Cas9, Stephenson [/bib_ref] [bib_ref] Decoding nonrandom mutational signatures at Cas9 targeted sites, Taheri-Ghahfarokhi [/bib_ref]. These non-random insertion InDels are probably caused by NHEJ acting on a Cas9-formed staggered cut (electronic supplementary material, [fig_ref] Figure 3: Long-read deep sequencing of Replace targeted cells [/fig_ref]. In this model, the sticky end cutting causes the PAM side of the break to contain extra nucleotides. These overhangs are filled during repair and appear as insertions when the two PAM sides are ligated in Replace targeting (figure 1f ). This produces insertions in the interfaces of the PAM sides and not in ligated interfaces of two Protospacer sides of the break (electronic supplementary material, [fig_ref] Figure 3: Long-read deep sequencing of Replace targeted cells [/fig_ref].
royalsocietypublishing.org/journal/rsob Open Biol. To test Replace targeting of an endogenous gene, we targeted three ubiquitously expressed loci in K562 cells: Polymerase Beta (POLB) exon 5, CCNA1 exon 2 and LMNA exon 2. We replaced exons with a splice acceptor-2A-mCherry-pA donor sequence [fig_ref] Figure 2: Exon Replace editing in K562 cells [/fig_ref]. Replace targeting resulted in reporter expression stable over weeks (figure 2c). Genotyping of mCherry + single-cell derived colonies showed mCherry integration into the targeted locus in 100% of colonies. Correct replacement ranged from 60% to 93% of the colonies, but in some cells, the donor mCherry sequence inserted next to the original exon without replacing it (figure 2d). Sanger sequencing of the genome-donor sequence interface of individual PCR amplified alleles showed modest InDel formation in the correctly exchanged alleles (figure 2e; electronic supplementary material, [fig_ref] Figure 2: Exon Replace editing in K562 cells [/fig_ref]. Replicate targeting experiments gave an average of 58%, 39% and 19% mCherry + cells for POLB exon 5, CCNA1 exon 2 and LMNA exon 2 respectively (figure 2f ). All three targeted loci are triploid in K562 [bib_ref] Comprehensive, integrated, and phased whole-genome analysis of the primary ENCODE cell line..., Zhou [/bib_ref] , assuming independence in the editing events, we can estimate the corresponding diploid cells would measure 44%, 28% and 13% mCherry + for POLB, CCNA1 and LMNA, respectively. Combining FACS and single-cell genotyping data allowed an It is known that large-scale deletions may follow a single Cas9-driven DSB [bib_ref] Repair of double-strand breaks induced by CRISPR-Cas9 leads to large deletions and..., Kosicki [/bib_ref] , and Replace targeting further complicates analysis due to the structural variants formed by the two genomic breaks and donor sequence integration. In order to quantify large deletions and the directionality of donor integration, we performed long-read deep-sequencing on amplicons of the targeted loci from unsorted Replace targeted HeLa cells and Replace targeted K562 cells (figure 3). We used primers 800-2000 bp away from the DSBs to generate long amplicons that were sequenced with PacBio technology. A bioinformatics pipeline was built to analyse large deletion and structural outcome frequencies (figure 3a) [fig_ref] Figure 4: Linear amplification analysis of POLB exon 5 [/fig_ref]. While a donor sequence with no homology is expected to integrate equally in both directions, inspired by the work of Suzuki et al. [bib_ref] In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration, Suzuki [/bib_ref] , we designed a preferred orientation into our donor sequence without the use of homology (electronic supplementary material, . When the donor integrated in the undesired direction the ligated interface reform the Cas9 target site, whereas the desired orientation is unable to be further cut. Long-read deep sequencing measured the desired orientation of mCherry in 79% of reads where BFP was replaced in HeLa and 89% of alleles with POLB exon 5 replacement in K562 (figure 3b). Even alleles containing unintended donor insertion of mCherry into a DSB flanking the targeted sequence integrated preferentially in the designed orientation.
[formula] (a) (b) (e) ( f ) (c) ( d)(a) (d) (e) ( f ) (b) (c) [/formula]
Alignment of the reads showed alleles with large-scale deletions (greater than 500 bp) occurred (figure 3c). Notably, individual reads showed that large-scale resection was frequently asymmetric with one side of the break undergoing dramatically larger resection. Viewing the frequency of a deletion at each base along the amplicon creates an averaged deletion profile and shows that the majority of loci experienced small-scale resection (figure 3d). Specifically, in successfully Replace targeted alleles, deletion mutations at the ligated junctions was smaller than 30 bp in greater than 90% of HeLa reporter reads, and smaller than 30 bp in greater than 95% of the POLB exon 5 reads. The ligated interface containing the protospacers were InDel-free in 79% of the correctly targeted reads of the HeLa reporter (electronic supplementary material, S1) and 63% InDel-free in the reads of the correctly targeted POLB K562 alleles, as measured by collapsing the long-read data (electronic supplementary material, S2).
Linear PCR methods requiring only one gene-specific primer, such as UDiTaS [bib_ref] UDiTaS, a genome editing detection method for indels and genome rearrangements, Giannoukos [/bib_ref] and LAM-HTGTS [bib_ref] Detecting DNA double-stranded breaks in mammalian genomes by linear amplification-mediated high-throughput genomewide..., Hu [/bib_ref] , offer more complete and quantitative measurements of DNA repair outcomes following a DSB. A gene-specific primer binds upstream of the targeted break site and a universal primer binding sequence is integrated downstream. Subsequently, the PCR amplifies the region across the break regardless of the structural variant, deletion size or translocation (figure 4a; electronic supplementary material, [fig_ref] Figure 4: Linear amplification analysis of POLB exon 5 [/fig_ref] , D). The UDiTaS method also contains a robust computational pipeline for CRISPR analysis. We modified this pipeline to extend the capabilities for Replace targeting with two pipelines (electronic supplementary material, [fig_ref] Figure 4: Linear amplification analysis of POLB exon 5 [/fig_ref]. Pipeline 1 closely follows the published UDiTaS pipeline; it aligns reads to the in silico reconstructed expected outcomes, performs InDel analysis and quantifies these measurements. The results of Pipeline 1 showed that at the targeted POLB locus donor sequence integrated in the preferred orientation at a 5 : 1 ratio to an inverted orientation [fig_ref] Figure 4: Linear amplification analysis of POLB exon 5 [/fig_ref]. At 39% of all POLB alleles, the integration of the donor sequence in the desired orientation is the single most frequent outcome measured. Strikingly, more than one-third of these donors were integrated without an InDel formed at the ligated interface. This highlights both the efficiency and fidelity of Replace targeting for exon replacement. As exogenously introduced DNA is known to integrate randomly into the genome [bib_ref] Inactivation of Pol θ and C-NHEJ eliminates off-target integration of exogenous DNA, Zelensky [/bib_ref] , we developed Pipeline 2 to quantify and map the integration location of the donor sequence (electronic supplementary material, [fig_ref] Figure 4: Linear amplification analysis of POLB exon 5 [/fig_ref]. Using a primer that binds the donor sequence and points towards the ligated interface, we generated amplicons that contain the flanking genomic sequence. These amplicons were Illumina sequenced, and the genomic sequences beyond the end of the donor sequence were aligned to the human genome (figure 4c; electronic supplementary material, figures S4C and S5). Sequence alignment showed 55% on-target integration into the POLB locus. Thirty-four per cent of all measured donor sequences had formed concatenations; it remains to be determined where these concatenated sequences are integrating within the genome, but concatenation of exogenous dsDNA itself is a known phenomenon [bib_ref] Patterns of integration of DNA microinjected into cultured mammalian cells: evidence for..., Folger [/bib_ref] [bib_ref] Adenoviral vector DNA for accurate genome editing with engineered nucleases, Holkers [/bib_ref] [bib_ref] 2020 DNA barcoding reveals that injected transgenes are predominantly processed by homologous..., Smirnov [/bib_ref]. The donor sequences were shown to be integrated into the genome at more than 28 loci (figure 4d). Interestingly, none of the off-target integration mapped to any of the 293 predictedSpyCas9 off-target sites.
# Discussion
This work demonstrates that NHEJ-based genomic sequence exchanges are feasible and efficient in human cells. In the four loci tested, replacement was successful in 16-54% of cells; in one case, the desired product was the major outcome. We furthermore demonstrated targeted exon replacement via NHEJ in three widely expressed human genes. Based on the comprehensive analysis of our targeted alleles, we arrive at three design principles to guide future Replace work.
The first design aspect ensures the correct orientation of the donor sequence in the genome. Linearizing the donor sequence with the same gRNA that cuts the target locus allows incorrectly ligated donors to be re-cut and excised (electronic supplementary material, . It is crucial to add a gRNA targeting the sequence formed during a deletion. This gRNA re-opens alleles that form a deletion and also excises out incorrectly ligated donor sequences. The minimal requirement for this design is two gRNAs (electronic supplementary material, . Long-read sequencing confirmed 89% of the donor sequences integrated in the designed orientation after POLB exon 5 Replace editing.
The second design principle is to avoid gRNAs that are involved in non-canonical SpyCas9 sticky end cutting. The frequency of 'InDel free' ligated interfaces measured in this work supports the idea that NHEJ repair is often not mutagenic [bib_ref] Is nonhomologous end-joining really an inherently errorroyalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 prone..., Bétermier [/bib_ref].
We believe breaks introduced by Cas9 are often re-ligated to reform the original sequence, which can then be cleaved again-forming a break ligation cycle. This cycle continues until the Cas9 is no longer active or the target site forms an InDel during repair and disrupts Cas9 binding. For efficient Replace targeting, prolongation of this cycle provides more time to acquire and ligate the donor sequence in the correct orientation. InDel mutations remove alleles from the ligation cycle and thus decrease efficiency. One avoidable driver of InDel formation is non-canonical SpyCas9 cutting in which a staggered cut is formed [bib_ref] Precise and predictable CRISPR chromosomal rearrangements reveal principles of Cas9-mediated nucleotide insertion, Shou [/bib_ref] [bib_ref] Bidirectional degradation of DNA cleavage products catalyzed by CRISPR/Cas9, Stephenson [/bib_ref] [bib_ref] Decoding nonrandom mutational signatures at Cas9 targeted sites, Taheri-Ghahfarokhi [/bib_ref]. The staggered cut is filled in and then ligated, duplicating the staggered nucleotide(s). The resulting small insertions are easily identifiable as they match the nucleotides of the protospacer sequence beyond the expected break site (electronic supplementary material, [fig_ref] Figure 3: Long-read deep sequencing of Replace targeted cells [/fig_ref]. Data from large gRNA screens suggest this mechanism as the predominant driver of +1 insertions [bib_ref] Predictable and precise template-free CRISPR editing of pathogenic variants, Shen [/bib_ref]. The non-canonical cutting of SpyCas9 may be sequence or loci-dependent. Empirical testing of a gRNA by measuring InDel outcomes, therefore, allows us to avoid sites that incur staggered cuts.
The third concept is to design sacrificial sequences around the ligated regions to buffer possible resection and sequence deletions. While the overall rate of InDels and large-scale deletion is low, detrimental effects can be further reduced. During exon Replace targeting, we cut in intronic regions outside the splice site as short intronic InDels are less likely to be detrimental to gene function. Long-range deep sequencing showed that in our systems, the vast majority of the InDels are less than 30 bp long. Considering this, we recommend a sacrificial buffer 30 bp or greater be included on the flanks of the Replace construct to protect the splicing donor/acceptor and coding sequence. We currently use minicircles but also recommend such buffers on AAV delivered donor sequences too.
NHEJ-based sequence replacement has previously been explored using PCR fragments as donor sequence [bib_ref] In vivo blunt-end cloning through CRISPR/Cas9-facilitated non-homologous endjoining, Geisinger [/bib_ref]. However, the genetic analysis in that study was not sufficient to distinguish successful replacement from other possible editing outcomes, such as unintended Insert targeting, structural rearrangements and off-target integration. Therefore, it remains to be confirmed and quantified in future work, if Replace editing with PCR donor templates is a viable strategy.
Measuring the outcomes of Replace targeting is complicated by the various structural rearrangements formed. Additionally, a growing body of literature documents complex outcomes following even simple Cas9-formed DSBs. These can include large-scale resection [bib_ref] Repair of double-strand breaks induced by CRISPR-Cas9 leads to large deletions and..., Kosicki [/bib_ref] , chromosomal fusions [bib_ref] UDiTaS, a genome editing detection method for indels and genome rearrangements, Giannoukos [/bib_ref] , mis-spliced mRNA [bib_ref] In vivo genome editing improves muscle function in a mouse model of..., Nelson [/bib_ref] and unintended vector integration into the break site [bib_ref] Long-term evaluation of AAV-CRISPR genome editing for Duchenne muscular dystrophy, Nelson [/bib_ref]. In working towards a full understanding of the outcomes of Replace targeting, we developed multiple deep sequencing pipelines. Long-read sequencing of PCR amplicons of the targeted loci proved useful in illuminating resection profiles and gives insight into the orientation of the structural variants produced. However, samples prepared for long-read sequencing used two gene-specific primers and so suffered from PCR bias, over-representing the shorter amplicons, making quantitative comparisons of alleles of different lengths impossible. Traditional two primer PCR also requires both intact binding sites, and unable to amplify more complex repair products. To address these shortcomings, we turned to single primer amplification methods such as UDiTaS and LAM for quantitative analysis, as they amplify all outcomes approximately equally and measure more complex repair events. This allowed us to measure the frequency of deletions royalsocietypublishing.org/journal/rsob Open Biol. in POLB editing to be 26% of all alleles and only 16% of alleles maintained their wild-type allele. A total of 39% of alleles show correct integration of the donor, and the rest would not produce functional protein (structural inversions or deletions). This ability to measure knock-in and knock-out rates concurrently is helpful in understanding the function at the cellular level. In contrast with other studies measuring repair outcomes of a Cas9 DSB [bib_ref] UDiTaS, a genome editing detection method for indels and genome rearrangements, Giannoukos [/bib_ref] , we did not detect chromosomal fusions at our break points. However, this may be due to our analysis time point three weeks post-targeting, where alleles could have been selected out of the population. Beyond the utility for quantitative measurements on-target, these single-gene primer protocols are powerful for measuring unintended integration of introduced DNA sequences. For example, in treating a mouse model of muscular dystrophy, linear amplification measurements showed the therapeutic AAV unintentionally integrated into the Cas9 break site and throughout the genome [bib_ref] Long-term evaluation of AAV-CRISPR genome editing for Duchenne muscular dystrophy, Nelson [/bib_ref]. These unintended integration of AAV in human cells may have a carcinogenic potential [bib_ref] AAV vector integration sites in mouse hepatocellular carcinoma, Donsante [/bib_ref]. Others have recently demonstrated high rates of unintended on-and offtarget integration of AAVs using single primer amplification [bib_ref] High levels of AAV vector integration into CRISPR-induced DNA breaks, Hanlon [/bib_ref]. Replace donor sequences have the potential to integrate into the target site or off-target into the genome. To our knowledge, this is the first work to map and quantify off-target integration or concatenation of donor sequences following NHEJ Insert or Replace targeting. Using a primer on the donor sequence, we detected substantial off-target integration of the donor. Strikingly, none of these off-target integration loci were within 5000 bases of the top 293 predicted Cas9 off-target sites. Rates of off-target integration may be similar for doublestranded HDR templates, but to our knowledge, off-target integration mapping by linear amplification has not been done after an HDR editing making comparison difficult. Singlestranded donor templates are known to integrate off-target less frequently [bib_ref] Reprogramming human T cell function and specificity with non-viral genome targeting, Roth [/bib_ref] [bib_ref] Highly efficient and marker-free genome editing of human pluripotent stem cells by..., Martin [/bib_ref] , but off-target quantification has mainly relied on integration of large fluorescent cassettes and could benefit from using single primer amplification approaches.
There are currently over 3800 genes known to cause monogenic diseases with mutations often spread across multiple exons. Gene editing holds great potential for the treatment of such diseases, but reversing the genetic defects in terminally differentiated or resting cells remains a major challenge [bib_ref] Strategies for in vivo genome editing in nondividing cells, Nami [/bib_ref]. HDR is unable to target non-dividing cells [bib_ref] A mechanism for the suppression of homologous recombination in G1 cells, Orthwein [/bib_ref] , but the NHEJ pathway is known to be preserved across cells types and cycle [bib_ref] Playing the end game: DNA double-strand break repair pathway choice, Chapman [/bib_ref]. NHEJ-based Insert targeting had previously been shown efficient in a wide variety of nondividing and dividing cells in vivo and in vitro. The use of such NHEJ Replace editing holds the most potential for therapies looking to correct mutations in non-dividing cells by the replacement of exons. However, it was not clear if the NHEJ repair would allow for effective genetic replacement, but instead result in majority deletions, inserts or InDels. Additionally, the size variation between possible repair outcomes (i.e. deletions, insertions, replacements) makes their quantitative analysis challenging. In this work, we have demonstrated that the kinetics and fidelity of the NHEJ pathway allows for efficient replace targeting in human cells, and that a thorough understanding of the edited population can be achieved based on single primer PCRs, long PCRs and tailored analysis pipelines. While many questions, such as optimal donor delivery, remain to be addressed, our work provides the foundation for future applications of Replace editing for genome engineering.
# Material and methods
## Data and methods availability
Sequencing data are available. Sequence Read Archive (SRA) accession: PRJNA622521. Extended protocols are available at https://www.protocols.io/researchers/eric-danner/ publications. Plasmids were submitted to Addgene (https://www.addgene.org/Ralf_Kuehn, #149344-#149354) and a folder of annotated genebank (.gb) files is added as electronic supplementary material, S1. All code used is available on Github: https://github.com/ericdanner. This includes scripts, Jupyter notebooks and Conda environments.
## Dna constructs
Cas9-2A-puro targeting plasmid is Addgene ID 62988 with F1 sequence removed. The AAVS1 targeting fluorescent reporter system was modified from Addgene ID 60431. The neomycinR sequence was modified to a more robust form [bib_ref] A mutant neomycin phosphotransferase II gene reduces the resistance of transformants to..., Yenofsky [/bib_ref]. The RTTA3 gene was replaced by a BFP-pA-Venus-pA where the BFP is flanked by Rosa26 sequences constructed by Gibson Assembly. Guide RNA target sequences were ligated into BbsI cleaved plasmids using synthetic oligonucleotides [fig_ref] Table 1: Primer table [/fig_ref]. When more than one guide was necessary, the plasmids were combined using Gibson Assembly.
Minicircles are produced in engineered bacteria using arabinose-induced recombination to remove the plasmid backbone [bib_ref] Efficient non-viral T-cell engineering by sleeping beauty minicircles diminishing DNA toxicity and..., Clauss [/bib_ref] [bib_ref] A robust system for production of minicircle DNA vectors, Kay [/bib_ref]. The ZYCY10P32T E. coli strain and the minicircle backbone were purchased from System Bioscience. After cloning in the sequence into the specific minicircle backbone, the plasmid is transformed into the ZYCY strain. The 200 ml culture was grown in TB media for 16 h. Then 200 µl of 20% L-arabinose was added and adjusted to pH 7 and 200 ml LB were added. The culture was then shaken at 32°C for 4 h to induce minicircle formation and slow cell division. An endotoxin-free purification kit (Macherey Nagel) was used following the protocol for low copy number plasmids. The resulting product contained plasmid and gDNA contamination. Restriction enzymes cutting the backbone and gDNA were added for 2 h. Then the resulting fragmented DNA was digested with PlasmidSafe DNase for 16 h (Epicure).
## Cell culture and targeting
HeLa cells were cultured in DMEM, 10% FBS, 1% Penicillin/ Streptomycin and passaged with trypsin every 3-4 days. To generate the fluorescent reporter line, plasmid #208 was cloned. Successful integration into the AAVS1 loci generated neomycin resistance. Cells were selected with 0.6 mg ml −1 G418 for one week. Single cells were FACS sorted into a 96-well plate and expanded. Colonies were checked for correct integration by genotyping and a clone with inserts on both alleles was expanded and used. Targeting of Reporter HeLa: 50 000 cells were reverse-transfected with 1.5 µg of Cas9_2A_puro/guide plasmid + 1.5 µg of MC or plasmid complexed with Lipofectamine 3000. The next morning 1.5 µg ml −1 puromycin was added for 48 h. Cells were then FACS analysed. mCherry + cells were single cell sorted into a 96-well plate and expanded for genotyping. For the HDR targeting experiment, the guide RNA targeting the Insert site was used together with the donor plasmid.
royalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 K-565, a leukaemia cell line, were kept in IMDM, 10% FBS, 1% penicillin/streptomycin and split every 3 days. For targeting, cells were nucleofected using the Lonza 4D strips. A total of 5 × 10 5 cells were resuspended in nucleofection bufferwith 1 µg Cas9/guide plasmid and 3 µg of minicircle and nucleofected using program FF-120. The following day puromycin (4 µg ml −1 ) was added for 48 h.
## Genotyping
For single-cell clones or bulk sequencing, genomic DNA (gDNA) was extracted by quick extract (Lucigen). PCR amplification was performed with LongAmp Polymerase (NEB) or PrimerStar GXL (Takara). Primer pairs flanking the upstream cut site or downstream cut site were used. Amplicons were verified by gel extraction and Sanger sequencing. Amplicons from bulk sequencing were cloned into the TOPO vector (Invitrogen) before Sanger sequencing.
The frequency of homozygous and heterozygous integration in HeLa cells was determined by knocking-in mCherry and miRFP670 simultaneously. By measuring mCherry + , miRFP670 + and double positive cells, the homozygous knock-in could be calculated [bib_ref] Reprogramming human T cell function and specificity with non-viral genome targeting, Roth [/bib_ref].
We used modified ICE analysis for deconvolution of amplicon Sanger trace data derived from unsorted Replace targeted cells. The amplicons were made using a primer on the donor sequence and a primer on the genomic sequence flanking the ligated site. The amplicon was cloned into the TOPO vector and individual cloned alleles were Sanger sequenced along with the mixed PCR product. A cloned colony with Replace inserts without any InDels was identified, and these Sanger trace data were used as the 'wild-type' reference in ICE analysis.
## Long-read deep sequencing and analysis
Bulk gDNA of targeted and control cells was amplified by PrimeStar GXL for Polb targeting. The HeLa synthetic reporter system required PCR with OneTaq (NEB) using the high GC content additive to amplify through the very GC-rich CAG sequence. Five-minute elongation steps were used to reduce PCR bias. Amplicons were cleaned by SPRI beads and quantified by Qubit. The Libraries were pooled and prepared for PacBio sequencing following company protocol. Data analysis was done using 'Pipeline Longread'. This pipeline uses custom Python scripts for preprocessing and bins the reads into different structural variants: original exon, replacement, insertion or deletion. Alignments were done with BBmap or MiniMap2 [bib_ref] Minimap2: pairwise alignment for nucleotide sequences, Li [/bib_ref] and visualized with IGV (Interactive Genome Viewer). Analysis of alignments was done in R using a modified script from (Github/pigX). Plotting was done in R or Python with a number of the plots included in the Jupyter Notebooks.
## Uni-directional targeted sequencing sample preparation
Wild-type and treated cells having had the POLB exon 5 targeted showing 50% mCherry expression were used. Samples were prepared either as described in LAM-HTGTS [bib_ref] Detecting DNA double-stranded breaks in mammalian genomes by linear amplification-mediated high-throughput genomewide..., Hu [/bib_ref] beginning with 500 ng of gDNA or based on the Tn5-Uditas protocol [bib_ref] UDiTaS, a genome editing detection method for indels and genome rearrangements, Giannoukos [/bib_ref] beginning with 50 ng gDNA. LAM-HTGTS was done generally as published with a few modifications. A single biotinylated gene-specific primer was used to amplify 500 ng sonicated gDNA (1 kb peak) 80× rounds. Streptavidin Dynabeads were found to inhibit PCR so the concentration was reduced to 1/10th and used to capture the amplified sequence. Capture bead-DNA was washed and then the universal primer was ligated on the end. This adapter-ligated sequence was PCR amplified with a universal primer and a nested gene-specific primer 30×. We added Nextera adapters by 10× rounds of amplification. Gel extract 300-500 bp smear 300-500 bp, quantified by Qubit and Bioanalyzer, then sequenced with Illumina MiniSeq. For Tn5 sample preparation, we modified the UDiTaS protocol, 50 ng gDNA was washed 2× with SPRI beads. Tagmentation used hyperactive Tn5 produced by the Max Delbrueck Center protein production facility following published protocols [bib_ref] Tn5 transposase and tagmentation procedures for massively scaled sequencing projects, Picelli [/bib_ref]. Samples were tagmented to add the universal primer binding site. Sample was amplified with gene-specific primer and universal primer 15×. A nested primer with Illumina adapter sequences was added and followed by PCR 15×. Then Illumina adapters were added with 10× PCR. Amplicons 300-500 bp were gel extracted, quantified by Qubit and Bioanalzyer, then sequenced with Illumina MiniSeq.
## Analysis of uni-directional targeted sequencing
All scripts and notebooks are on github.com/ericdanner/ REPlacE_Analysis. The analysis of the linear amplified sequences was based on the Uditas software (https://github. com/editasmedicine/uditas). De-multiplexed samples are run through pipeline 1 or pipeline 2. Pipeline 1 generates amplicons of the various expected outputs and does a global alignment using Bowtie2 [bib_ref] Fast gapped-read alignment with Bowtie 2, Langmead [/bib_ref]. Reads that align well and cover the ligated junctions are analysed for InDels. If the samples were prepared with Tn5, they contained UMIs. Unique UMIs are tallied and editing outcomes are quantified. LAM samples do not contain UMIs. In Pipeline 2, the reads are checked for correct on-target priming. The samples are then trimmed using Cutadapt [bib_ref] Cutadapt removes adapter sequences from high-throughput sequencing reads, Martin [/bib_ref] up to the expected break site leaving only the sequence downstream of the break site. This sequence is aligned globally using Bowtie2 to an index file containing hg38 and the targeting vector.
Data accessibility. Sequencing data are available. Sequence Read Archive
[fig] Figure 1: Replace targeting using HeLa reporter cells. (a) Overview of replace targeting concept: correct donor integration disrupts the Cas9 targeting sequence and is captured, while undesired products (deletion or inverted integrants) reform the gRNA target and can be further cut. (b) HeLa cells containing a fluorescent reporter system integrated into two AAVS1 loci. Targeting used gRNA-1 shown in (a) with two identical sites flanking BFP. Lipofection of Cas9-2A-puro/gRNA-1 and the donor sequence was followed by 48 h puromycin selection to ensure complete delivery. (c) Results of Replace targeting with a minicircle donor measured by FACS. BFP, original allele; Venus, deletion; mCherry, donor integration; colourless, allelic damage. Total sums to greater than 100% as cells can express two fluorophores. (d ) Replace targeting using minicircles and plasmid sequence donors. WT cells are HeLa cells without the integrated reporter system. WT targeted identically. p-value was calculated using Student's t-test (****p-value <0.0001). (e) Single-cell sorted Replace targeted, mCherry + cells were expanded and genotyped. Correct exchange of BFP by mCherry (Replacement) and mCherry integration flanking the BFP (Insertion) were both measured. ( f ) InDel frequencies at the interface of the integrated donor sequence in Replace targeted cells quantified by ICE deconvolution. Arrows represent primers for sequenced amplicons with sizes in grey.royalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 estimate of 54% of POLB, 23% of CCNA1 and 16% of LMNA Replace targeted cells with successful replacement in K562. [/fig]
[fig] Figure 2: Exon Replace editing in K562 cells. (a) Replacement of POLB Exon 5, CCNA Exon 2 or LMNA Exon 2 with a cherry reporter including a splice acceptor (SA), T2A self-cleaving peptide and a polyadenylation site (bpA). (b) K562 cells cotransfected with Cas9-2A-puro/guides plasmid and mCherry Minicircles were puromycin selected for 48 h and followed by FACS analysis over five weeks. (c) mCherry + count of transfected cell cultures was measured weekly by FACS. (d ) Colonies of mCherry + single cells were expanded and genotyped to check and quantify replacement of the exon with the donor sequence. (e) Sanger sequencing of amplicons showing InDels at the interface of the integrated donor sequence. ( f ) Biological replicates of Replace targeting of POLB, CCNA1 and LMNA loci. [/fig]
[fig] Figure 3: Long-read deep sequencing of Replace targeted cells. (a) Unsorted HeLa reporter cells or K562 POLB exon 5 targeted cells. Loci were amplified using the primers more than 800 bp from the Cas9 breaksites; amplicons were sequenced using PacBio technology (b). Directionality of structural variants formed after Replace targeting. BFP/exon 5 pie charts are alleles containing the original targeted allele and no donor. mCherry pie charts show alleles where mCherry replaced targeted sequence. Insertion is both donor sequence and targeted sequence in the loci. (c) Forty representative alignments of three major type of alleles: allele with original sequence, sequence exchanged with mCherry, or a deletion allele. (d) An average deletion profile of reads in which Replace editing using the mCherry donor was successful. Alignments were analysed for the fraction of reads containing a deletion at each base position. Asterisk denotes Cas9 cleavage site. royalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 (electronic supplementary material, [/fig]
[fig] Figure 4: Linear amplification analysis of POLB exon 5; quantifying on-target Replace outcomes and mapping donor sequence integration. (a) Replace overview showing primer binding used for linear amplification. Unsorted gDNA was amplified with primer 1 or primer 2 and a universal primer, sequenced using Illumina technology, and analysed. (b) Pipeline 1 analysis of Uditas prepared samples quantifies outcomes at the targeted site with the corresponding InDels quantification. (c) Pipeline 2 analysis of UDiTaS prepared samples shows mCherry integration with an overall quantification of donor sequence integration. (d ) Donor sequence integration mapped across the genome with read counts plotted logarithmically. Chromosomes with no integration sites were removed. Top 10 predicted Cas9 offtarget sites shown with red triangles. The POLB locus on chromosome (chr) 8 is marked by an asterisk. royalsocietypublishing.org/journal/rsob Open Biol. 11: 200283 [/fig]
[table] Table 1: Primer table.royalsocietypublishing.org/journal/rsob Open Biol.11: 200283 [/table]
|
Low versus standard dose intravenous alteplase in the treatment of acute ischemic stroke in Egyptian patients
وسالمتها التيباز من منخفضة جرعة نتيجة لتقييم األهداف: في احلادة الدماغية السكتة لعالج القياسية اجلرعة نظام مع باملقارنة املصريني. املرضى منفردة عشوائية غير استطالعية رصدية دراسة أجريت املنهجية: قسمنا 8102م. ديسمبر إلى 7102م نوفمبر من الفترة خالل عمياء احلادة، اإلقفارية الدماغية السكتة مرضى من ا ً مصري ا مريضً 80 مجموعتني إلى الوريد، طريق عن التيبالز على للحصول مؤهلون 0.6 بجرعة للمرضى تخثر ظهر مجموعة). كل في ا مريضً 40( املجموعة في مجم/كجم 0.9 و األولى املجموعة في مجم/كجم والنتيجة. بالسالمة يتعلق فيما املجموعتني كال قارنا الثانية. اجلمجمة داخل النزف معدل خالل من السالمة عن التعبير مت والتعبير أشهر، 3 فترة خالل والوفيات )SICH( بأعراض املصحوب (مقياس أشهر ثالثة عند اإليجابية النتائج خالل من النتيجة عن .)2 إلى 0 من املعدل[]mRS رانكني نتائج حققوا (العدد=72) 69.2% األولى املجموعة في النتائج: املجموعة في (العدد=52) 64.1% بـ مقارنة ا ً يوم 90 في إيجابية (العدد=2) 5% ا ً يوم 90 ملدة الوفيات معدل كان .)p=0.631( الثانية الثانية املجموعة في (العدد=1) 2.5% مقابل األولى املجموعة في في مرضى 3 في اجلمجمة داخل نزيف أعراض وحظ ٌ ل .)p=0.556( .)p=0.077( األولى املجموعة في وصفر الثانية املجموعة ا ً عملي ً بديال تكون أن ميكن التيبالز من منخفضة جرعة اخلالصة: فترة في خاصة احلادة الدماغية بالسكتة املصابني املصريني للسكان ساعات. 4.5 إلى 3 من تتراوح Objectives: To assess low dose altepase outcome and safety in comparison with a standard-dose regimen for acute ischemic stroke treatment in Egyptian patients. Materials: An observational prospective cohort nonrandomized single blinded study was carried out during the period from November 2017 to December 2018. Eighty Egyptian acute ischemic stroke patients, all eligible for intravenous alteplase, were subdivided into 2 groups )40 patients in each group(. Patients were thrombolysed at a dose of 0.6 mg/kg in the first groupOriginal Article and 0.9 mg/kg in the second group. Both groups were compared in regard to safety and outcome. Safety was expressed by the rate of symptomatic intracranial hemorrhage )SICH( and 3 months mortality, while outcome was expressed by favorable outcomes at three months )modified Rankin Scale ]mRS[ of 0 to 2(.Results:In the first group, 69.2% )n=27( achieved favorable outcomes at 90 days compared with 64.1% )n=25( in the second group )p=0.631(. Ninety-day mortality was 5% )n=2( in the first group versus 2.5% )n=1( in the second group )p=0.556(. Symptomatic intracranial hemorrhage was noted in 3 patients in the second group and zero patients in the first group )p=0.077(.Conclusion:Low-dose alteplase could be a practical alternative for Egyptian populations with acute ischemic stroke especially in 3 to 4.5 hours window.
C erebrovascular stroke is the second death and the seventh disability leading cause worldwide.Tissue-type plasminogen activator )tPA( alteplase was the first medication approved by the Food and Drug Administration )FDA( for the acute ischemic stroke )AIS( treatment on June 1996, within 3 hours of stroke onset with a recommended dose of 0.9 mg/kg )maximum 90mgIn 2008, the safety of using alteplase within 3 to 4.5 hours of stroke onset was approved by the Safe Implementation of Treatments in Stroke International Stroke Thrombolysis Registry )SITS -ISTR( 3 and the European Cooperative Acute Stroke Study )ECASS IIIHowever, thrombolytic therapy use has not been widely adopted, especially in developing countries. The restricted time window )3 to 4.5 hours(, intracerebral hemorrhage )ICH( risk and the drug high cost are major obstacles preventing its broad application.Coagulation and fibrinolysis responses differ among different races, which increase symptomatic intracerebral hemorrhage )SICH( risk with standard-dose alteplasein Asian populations, many Asian neurologists considered alteplase low dose to be a better alternative for ischemic stroke treatment. Many studies had been conducted in order to prove the efficacy and safety of Alteplase low dose.One of these studies was the Japan Alteplase Clinical Trial )J-ACT( conducted by Yamaguchi et alAccording to this study, using a 0.6 mg/kg dose of intravenous recombinant tissue plasminogen activator )rtPA( in Japanese patients was safe and effective. Despite the relatively stroke high rate among Egyptian populations, 963/100,000 inhabitants, only less than 1% of stroke patients receive intravenous thrombolysis. A major reason for this is the drug cost.Low-dose regimens )0.6 mg/kg( use will lower the economic burden of thrombolytic therapy in the community and will greatly promote the implementation of this therapy in Egypt. Our study aim was to assess the outcome and safety of alteplase low dose in comparison to the standard-dose regimen in AIS treatment in Egypt.
Methods. This study was an observational prospective cohort non-randomized single blinded study done during the period from November 2017 to December 2018. Eighty ischemic stroke patients within four and a half hours from symptom onset were recruited from both El-Mataria Teaching Hospital and Menoufia University Hospital. This study was approvad by the Research and Ethics Committee of the Faculty of Medicine, Menoufia University. All patients were considered eligible for intravenous thrombolysis according to the American Heart Association )AHA( and American Stroke Association )ASA( guidelines with inclusion criteria which are: age ≥18 years old, time window ≤4.5 hours, patients presented with moderate to severe symptoms and demonstrate early improvement before starting alteplase, patients with seizures at time of onset if evidence suggests that residual impairment is secondary to stroke and not a postictal phenomenon, patients with blood pressure ≤185/ 110 or those who presented with high blood pressure and responded successfully to intravenous antihypertensive and patients with early ischemic changes )other than obvious hypodensity( as demonstrated on initial noncontrast CT brain.Informed written consent forms for alteplase were signed by patients or first-degree relatives. All patients were subjected to complete history taking, including assessment of vascular risk factors, previous treatment, time window, pre-stroke modified Rankin Scale )mRS(, and initial stroke severity measured as a National Institutes of Health Stroke Study )NIHSS( score. Non-contrast computed tomography )CT( brain scans were done in all patients upon presentation, and these scans were analyzed for the presence of intracranial hemorrhage or any findings preventing thrombolysis. Alberta Stroke Program Early CT )ASPECT( scores were used to assess early signs of ischemia. 14 Trial of ORG 10172 in acute stroke treatment )TOAST( 15 and Oxfordshire community stroke project )OCSP( 16 classifications to clarify stroke subtypes. The patients were assigned to treatment groups non-randomly. The first group was given a single dose of intravenous alteplase as a low dose of 0.6 mg/kg )not exceeding 60 mg(, with 10% given as a bolus followed by a continuous infusion of the remainder over one hour. We give the low dose to patients who are more risky to haemorragic transformation as old age, when we suspect large infarction and those with time window more than 3h and less than 4.5 h. The second group was given a standard dose of 0.9 mg/kg )not exceeding 90 mg(, with 10% given as a bolus followed by a continuous infusion of the remainder over one hour. Alteplase infusion and monitoring for complications over the next 24 hours were done following AHA/ASA guidelines.Follow-up non-contrast CT brain scans 24 hours post treatment were assessed for ICH. The NIHSS was assessed 24 hours after treatment and at day seven of admission or at discharge. In-hospital mortality was recorded. All patients were followed-up with 90 days from admission by telephone calls or face-to-face interviews to assess mRS and mortality.
Primary objectives. 1( Assessment safety of low-dose and standard-dose alteplase by assessing the incidence of symptomatic intracranial hemorrhage )SICH( within 24 hours of starting treatment and assessing mortality )in-hospital mortality and 90-day mortality(. We defined SICH according to the ECASS study definition )any hemorrhage plus a neurological deterioration of ≥ 4 points from the NIHSS baseline or from the lowest Disclosure. Authors have no conflict of interests, and the work was not supported or funded by any drug company.
Low dose of altepase in acute ischemic stroke ... Salem et al www.nsj.org.sa NIHSS value after baseline for seven days, or leading to death2( Assessment of outcome of low-dose and standarddose alteplase by assessing the proportion of patients with favorable outcomes )mRS scores of 0 to 2( and unfavorable outcome )mRS scores of 3 to 6( at 90 days.
Secondary objectives. An assessment was made regarding predictors of favorable outcomes )mRS scores of 0 to 2(. Statistical methods. Results were collected, tabulated, and statistically analyzed by IBM personal computer and the Statistical Package for the Social Sciences )SPSS( version 22 )Armonk, NY: IBM Corp, 2013(. Data were presented as descriptive statistics including percentage )%(, mean )x(, standard deviation )SD(, and range. Statistical tests used included the Chi-square test )χ2(, the Student's t-test, the Mann-Whitney test, the Wilcoxon Signed Rank test, and binary -Demographic and baseline characteristics of patients treated with IV tissue-type plasminogen activator )tPA( in low dose and standard dose groups. Thirty-six patients )90%( in the low-dose group received alteplase between 3 and 4.5 hours of stroke onset, while 97.5% )n=39( of the standard-dose group received alteplase within 3 hours of stroke onset )p<0.001(. Pre-stroke modified Rankin Scale )mRS( scores ranged from 0 to 2, with most patients of both groups having pre-stroke mRS scores of 0 )86%(. Regarding stroke subtypes, there was a significantly higher percentage of patients with posterior circulation infarction )POCI( in the low-dose group )20%(, while no cases of POCI were present in the standard-dose group )p=0.003(. There was a significantly higher mean value for the Alberta Stroke Program Early CT )ASPECT( scores in the low-dose group than in the standard-dose group )6.83±3.09 versus 5.55±2.59, p=0.049(.
## Items
Regarding outcome, favorable outcomes )mRS scores of 0 to 2( at 3 months were recorded in 67.5% of the low-dose group versus 62.5% of the standard-dose group )p=0.631(. Alteplase dose was not a significant predictor of favorable outcome )OR=0.402, p=0.504, 95% CI 0.028-5.824(. The only significant predictor of a favorable outcome at 3 months was a NIHSS score ≤13. For each one-degree decrease in NIHSS, there was a 2.338 degree increase in favorable outcomes )OR=0.097, p=0.018, 95% CI )0.014-0.673((. Regarding safety, no cases of SICH were found in the low-dose group. In comparison, 3 cases )7.5%( were found in the standard-dose group )p=0.077(. The overall rate of hemorrhagic transformation was 2.5% in the low-dose group versus 12.5% in the standard-dose group. Three cases were parenchymal hemorrhage type 1 )lesion occupying <30% with no mass effect( and 3 cases were parenchymal hemorrhage type 2 )lesion occupying > 30% with definite mass effect( )(. Mortality at 3 months was 5% in the low-dose group versus 2.5% in the standard-dose group )p=0.556(. Only 2 patients died in the hospital )one from each group( )(.
# Discussion
. Low-dose alteplase use in treating acute ischemic stroke is not common in Egypt. This study demonstrated the results of thrombolysis with a low dose in a sample of the Egyptian population. The time from stroke onset to thrombolysis was strikingly different between the low and standard-dose groups. In the low-dose group, 90% of patients were thrombolysed between 3 and 4.5 hours of stroke onset, while 97.5% of the standard-dose group patients were thrombolysed within 3 hours of stroke onset. The extended time window )3 to 4 and a half hours( in the low-dose group was a strong rationale for using low-dose alteplase, as the increased time could increase hemorrhagic transformation risk. This was evidenced by Wahlgren et al 3 and Hacke et al,who recorded the standard dose efficacy in the extended window )3 and 4.5 hours( but found SICH higher risk in comparison to a 3-hour window )2.2%-2.4% versus 1.7%(.
In most of the studies comparing low and standard alteplase doses, favorable outcome was considered with a modified Rankin Scale score of 0 to 1 and a score of 0 to 2 represented functional independence. In the current study, a favorable outcome was considered with a score of 0 to 2 at 3 months. Our justification was that with scores of 0 to 2, the patient achieves a state of functional independence, which is considered a good outcome.
In the current study, 67.5% of patients in the low-dose group had favorable outcomes at 3 months in comparison to 62.5% of the standard-dose group patients, with no significant difference between both groups. This finding is consistent with the results of Kim et al,who found that 45.5% )n=450( of patients who received low-dose alteplase achieved mRS scores of 0 to 2 at 3 months in comparison to 49% )n=1,076( of the standard-dose group. In addition, Chao et al,who studied different doses of alteplase for acute stroke in Chinese patients )0.6, 0.7, 0.8, and 0.9 mg/kg(, demonstrated that patients with mRS scores of 0 to 2 at 3 months included 53% )n=146( of the low-dose )0.6 mg/kg( group and 47% )n=367( of the standard-dose group )0.9 mg kg(.
In the current study, a binary logistic regression analysis for predictors of favorable outcomes at 3 months revealed that an NIHSS score of ≤13 was the only significant predictor of a favorable outcome at 3 months, and that for each one-degree decrease in NIHSS scores, there was a 2.338-degree increase in favorable outcomes )OR=0.097, p=0.018, 95% CI: 0.014 to 0.673( . This result goes with Chao et al 19 who stated that an NIHSS score ≤8 was an independent predictor of a good outcome after intravenous alteplase )OR=0.37, p=0.0001, 95% CI(. The different values of NIHSS scores could be explained by the different scores of favorable outcomes considered -an mRS score of 0 to 1 in Chao et al 19 study versus a score of 0 to 2 in the current study.
This analysis revealed that the dose of alteplase was a non-significant predictor of stroke outcome at 3 months )OR=0.402, p=0.504, 95% CI: 0.028 to 5.824(. No patients developed SICH in the low-dose group, in comparison to 3 cases developed )7.5%( in the standard-dose group. This result agrees with Anderson et al,in which fewer major SICH occurred in the low-dose group than in the standard-dose group )1% versus 2.1%(. The difference in patients percentage that developed SICH in this study versus the study of Anderson et al 20 might be the result of SITS-MOST criteria use for SICH in the latter study. In addition, Chao et alshowed that there was no difference between their 2 treatment groups regarding SICH according to the ECASS definition )5% in both(. Nguyen et al,in their comparative study between low-dose and standard-dose alteplase in Vietnam, also documented no difference in the occurrence of SICH )per the ECASS definition( in their standard-dose group compared with their low-dose group )5% versus 2%(.
Kim et al 18 also reported that no statistically significant differences in the incidence of SICH between their standard-dose and low-dose groups )6% versus 8%(. There were no significant differences between low-dose and standard-dose groups regarding mortality at 3 months: there were 2 cases in the low dose group versus one case in the standard-dose group )5% versus 2.5%, p=0.556(. Most of the studies that compared low and standard rt-PA doses have been compatible with our present study result showing no significant difference between the 2 doses regarding the 3 months mortality. Kim et al,found that there was no significant difference between the low dose and standard dose )13% versus 14%(. Chao et aldocumented a 3 months mortality of 8% in both 0.6 and 0.9 mg/kg groups. Sharma et al,in Singapore documented that there was no significant difference in the 3 months mortality between low and standard doses )10% versus 13%(. Anderson et al 20 also showed that mortality at 90 days did not differ significantly between the 2 groups )8.5% and 10.3%(. Yang et al 23 studied a comparison between the 2 groups where the mortality in the 0.6 mg/kg dose group had a non-significantly higher incidence than that of the standard-dose group )4.3% versus 1.6% in the 0.9 mg/kg dose group, p=0.792(. The mortality in the 0.6 mg/kg dose group was higher than that in the 0.9 mg/ kg dose group )4.3% versus 1.6%, respectively; p>0.05( Nguyen et alwas the only study to show a significant difference in the mortality rate at 3 months, mortality rates of 2% in the low-dose group and 12% in the standard-dose group.
Study limitations. First, the sample size was small because of the following factors: public awareness of stroke symptoms and the relevant time window is still deficient in our region, so many stroke patients arrive at the hospital after the thrombolytic window. In addition, thrombolytic therapy was recently introduced to our center in 2017 at a high cost )which is still not covered by health insurance(, so many patients cannot afford it. Some other studies recruited small numbers of patients, like Nguyen et al 21 who included 48 in the low-dose group versus 73 standard-dose patients; Sharma et al,who included 48 low-dose patients versus 32 patients receiving the standard dose; and Zhou et al 24 who included 23 patients receiving a low dose(. Second, the alteplase dose selection was not randomized. Also, both groups were not balanced regarding to ASPECTS score and POCI. Finally, the stroke team that conducted the outcome assessments was not blinded to the treatments, raising the possibility of observer bias.
In conclusion, the present study showed that there was no statistically significant difference between low and standard alteplase doses regarding either safety or outcome. Taking into consideration the economic burden of the treatment, low-dose alteplase )0.6 mg/kg( could be a practical alternative for Egyptian populations with acute ischemic stroke especially in 3-4.5 hours window. This study compares between different doses of alteplase in acute ischemic stroke treatment in a big country like Egypt where utilization of reperfusion therapies for stroke remains <1% Zakaria et alTherefore, additional confirmation with randomized controlled trials in Egyptian patients is necessary.
|
In vivo tau staging in Alzheimer’s disease
Disease staging systems allow for the measurement of disease severity based on the identification of important milestones in the natural history of a disease. Staging systems can be used to help guide clinical care, serve as inclusion criteria for therapeutic trials, and furthermore provide a consistent framework for interpreting findings from different studies in different populations. Disease staging systems can also be used to model the progression of clinical and biomarker changes in relation to important disease landmarks. In Alzheimer's disease (AD), which is defined by the presence of amyloid-β plaques and tau neurofibrillary tangles [1,2], staging systems have been devised based on the anatomical distribution of neuropathology at autopsy. Staging of AD is already conducted routinely: postmortem neuropathological assessments of AD, considered the gold standard in AD diagnosis, rely on histopathological staging systems [1]. Although most in vivo research has focused on dichotomous classification of amyloid-PET and tau-PET imaging into +/-groups, the spatial resolution of PET provides an opportunity for staging based on the anatomical distribution of pathology, similar to postmortem staging systems[3]. Due to established relationships between tau, neurodegeneration and cognitive dysfunction, the Braak neurofibrillary tangle staging system [4] provides a useful framework for staging tau pathology, as well as AD severity. Recent work from our group has applied the Braak staging system to [ 18 F]MK6240 tau-PET data[5]. After assigning a Braak stage to all subjects based on the topography of tau-PET abnormality(Figure 1), individual-level Braak stages were employed to model the trajectory of other AD biomarkers, as well as disease Editorial
symptoms. Overall, we observed that most fluid p-tau biomarkers became subtly elevated at Braak stage II. At stages III and IV, this abnormality increased in magnitude, and subsequently plateaued at stages V and VI. Very subtle memory decline could be identified by stage II, with mild cognitive symptoms detectable by stages III and VI. Braak stages V and VI were only observed in individuals with overt cognitive impairment (mild cognitive impairment or dementia). A unique feature of in vivo disease staging, as opposed to neuropathology, is the ability to track the evolution of disease in the individual over time. Using longitudinal imaging, we observed that amyloid-PET abnormality at baseline was associated with different rates of Braak stage progression. Progression from early Braak stages (0-II) took place in the presence or absence of abnormal amyloid. However, progression from Braak stage III and above during two-year follow-up occurred exclusively in the presence of abnormal amyloid-β. Taken together, these results showcase how applying disease staging systems to living humans can provide new insights into disease progression. Another important advantage of in vivo disease staging is the ability to monitor the severity of AD specifically. This is in contrast to clinical dysfunction, which is often the result of multiple neuropathological processes. Furthermore, cognitive reserve, which can influence the extent to which individuals can carry brain pathology without symptoms, also complicates clinical staging. PET-based Braak staging can also identify preclinical disease (detectable tau pathology in cognitively unimpaired individuals). Therefore, in vivo staging of Alzheimer's disease has greater sensitivity and specificity over clinical assessment of symptoms. www.aging-us.com
## In vivo tau staging in alzheimer's disease
PET-based Braak staging may also prove useful for clinical trial enrolment. Identifying individuals at a specific stage of tau abnormality will allow for a more homogeneous study population, potentially reducing the variance in longitudinal cognitive decline. Furthermore, studies aimed at halting the progression of tau pathology may seek to recruit patients at a specific early stage. Finally, tau-PET can be used to monitor disease progression in clinical trials, with lack of tau progression as a possible signal of effectiveness. This is similar to the recent phase II donanemab trials which recruited amyloid-positive individuals with intermediate levels of tau-PET uptake, which met primary endpoints in phase II. The recent approval of aducanumab (Aduhelm) highlights the need for biomarkers to assess the severity of AD. If evidence supports the optimal utility of anti-amyloid therapies (such as aducanumab or other monoclonal antibodies) for specific stages of AD, biomarker-driven staging systems may be wellpositioned to identify who may optimally respond to a specific therapeutic option. Furthermore, if and when other therapeutic options are available, biomarker-driver staging systems will be able to determine the optimal therapeutic window for specific therapeutic modalities (i.e. anti-amyloid, anti-tau, combination therapies). Several unanswered questions remain. The first is the degree of correspondence between PET-based Braak stage and Braak stage at autopsy. Because of limitations of PET's sensitivity, some discordance between antemortem and postmortem Braak stage is expected. Furthermore, it must be highlighted that the methodology of in vivo PET staging and ex vivo neuropathological staging are very different: in vivo PET-based Braak staging relies on surpassing predefined thresholds of abnormality with specific brain ROIs representing specific Braak stages. In contrast, histopathological Braak staging relies on the detection of neurofibrillary tangles using staining techniques. Therefore, identification of even small numbers of neurofibrillary tangles using histopathological techniques can place an individual at a more advanced Braak stage, as a higher concentration of neurofibrillary tangles would be needed for the same individual to exceed a predefined threshold for abnormality. Finally, it is important to emphasize that staging systems involve dimensionality reduction. Despite the fact that atypical clinical variants of AD largely conform to the Braak system, some information about the spatial distribution of tau in atypical ADis lost when only considering an individual's tau stage.
[fig] Figure 1: Progression of PET-based Braak stages. [/fig]
|
A Dive Into Oliceridine and Its Novel Mechanism of Action
# Introduction and background
Pain is the unpleasant sensory and emotional experience associated with actual or potential tissue damage. Pain can be classified as acute or chronic, each with different management options. Acute pain is characterized as being self-limiting, lasting less than three to six months with treatment focused on the use of analgesics. Postoperative pain is a type of acute pain occurring in patients after undergoing a procedure. There is a preconceived notion that acute pain will be relatively short-lived after a successful operation. However, evidence suggests that less than half of patients who undergo surgery report adequate postoperative pain relief. Inadequate acute pain management can transition to chronic pain through neuronal plasticity and increase mortality, delay recovery, and increase hospital costs. This paper will review oliceridine (Olinvyk®, Trevena Inc., Chesterbrook, USA), its novel mechanism of action, and its role in pain management, especially with the current state of the ongoing opioid epidemic.
In 2012, about 313 million surgical operations were performed globally, while about 28 million inpatient surgical procedures and 48 million ambulatory surgeries were reported in 2006 and 2010, respectively. Postoperative pain is preventable and treatable, but it is often inadequately managed as pain response and sensitivity are different in all patients and management requires an individualized approach. More than 80% of patients who undergo surgical procedures experience acute postoperative pain and approximately 75% of those with postoperative pain report the severity as moderate, severe, or extreme.
The economic burden of postoperative pain is not easy to estimate, but with the number of patients undergoing surgery, the burden is large due to the direct costs in healthcare resources and the indirect costs from the reduction in patient function and productivity. This is furthered through the inadequate management of postoperative pain-causing extended hospitalization, increased morbidity, and mortality associated with chronic pain. The cost of transitioning from acute to chronic pain in one patient can be up to $1 million and overall the burden of cost for chronic pain in the United States is $560-$635 billion, higher than the individual costs of heart disease, cancer, and diabetes. Our knowledge about the increased risks associated with inadequate management is linked with increased utilization of potent analgesics, such as opioids. However, the prescribing of opioids to help manage postoperative pain is not appropriate for everyone. On the other side of the spectrum, the opioid crisis has led to increased healthcare costs and loss of productivity by $100 billion per year. Finding new strategies to properly control pain will drastically decrease the cost burden on the healthcare system and efficiently allow us to utilize healthcare resources.
Conventional parenteral opioids have a narrow therapeutic window and are associated with dose-limiting opioid-related adverse events (ORAEs) such as vomiting, nausea, and opioid-induced respiratory depression (OIRD). As expected, the risk of these ORAEs frequently increases in the elderly and patients with comorbidities, but studies have shown that the risk also increases in children and adolescents between the ages of 12 to 17 years. Health care providers must determine the fine line between overuse and underuse of these drugs.MOA: bind to receptors in the central nervous system and peripheral tissues and modulate the effect of the nociceptorsSide effects: a) Significant: respiratory depression b) Common: nausea, vomiting, pruritus, reduction in bowel motility Rapid onset of action with peak effect occurring in 1-2 hoursRequires special consideration of use in morbidly obese patients or chronic pain patientsPotentNeed regular monitoring of respiration and oxygen saturation in postoperative patientsIV Acetaminophen MOA: not fully understood but shown to prevent prostaglandin production in the CNS and PNS to inhibit pain impulsesSide effects: infection, phlebitis, and local irritationRapid and high plasma concentration achieved within five minutesContraindications: hypersensitivity, severe hepatic impairment or severe active hepatic diseasePain relief occurs within few minutesEasily passes through blood brain barrier and preferred in multimodal analgesiaIV NSAIDs (i.e.
ibuprofen, naproxen, ketorolac)
MOA: inhibit prostaglandin production through COX-1 and COX-2Side effects: increased risk of gastrointestinal bleeding, cardiotoxicity, hepatotoxicity, renal dysfunction, and drug induced asthmaPlasma half life of ketorolac: about 5.5 hoursContraindications/warnings: a) Avoid during pregnancy b) Patients with salicylate hypersensitivity or allergic reaction (urticaria, asthma, etc.)Ketorolac is the most commonly used IV NSAID and reduces opioid consumption by 25-45%Ketorolac cannot be used more than five daysIV Ketamine MOA: noncompetitive NMDA receptor antagonist believed to have antihyperalgesic effects and reverse opioid tolerance; may also suppress pain transmission by limiting astrocyte and microglial activationSide effects: hallucinations, agitation, euphoria, dysphoria, anxiety, nausea, tachycardia, sedation, and dizzinessHalf life: 10-15 minutesContraindications: pregnancy, psychosis, uncontrolled hypertension, severe liver dysfunctionCan be used in patients in opioid-dependent or opioid tolerant patients
## Oliceridine's mechanism of action
The United States Food and Drug Administration (FDA) granted approval to Travena Inc. for oliceridine on August 7, 2020 after following a rigorous review to evaluate the risk versus benefit as addressing the opioid epidemic is the FDA's top priority. Oliceridine is a full opioid agonist with selectivity towards the G protein pathway and is indicated in adults for the management of severe acute pain that requires an intravenous (IV) opioid analgesic after inadequate management with alternative treatments. Conventional opioids, such as morphine, produce analgesia on the central μ-opioid receptors and lead to a cascade of post-receptor signaling events through the G protein pathways, leading to the potent analgesia and the β-arrestin pathway, which results in ORAEs. Due to its selectivity, oliceridine has shown decreased recruitment of β-arrestin and is expected to cause fewer ORAEs. Oliceridine should be reserved for use in a controlled clinical setting in patients for whom alternative treatment options have not been tolerated, are not expected to be tolerated, have not provided adequate analgesia, or are not expected to provide adequate analgesia.
# Methods
A search of oliceridine on clinicaltrials.gov, a public registry listing clinical trials, was conducted to obtain relevant clinical trials and review the safety and efficacy of the drug's novel mechanism of action. The search was limited to "interventional studies (clinical trials)" and "studies with results." Search results were restricted to 2013 to the current time frame to focus on the most current pain management research. Six relevant clinical trials were found. The summary of these trials is outlined in. The relevance of a trial in this review is based on the current indication of oliceridine.
## Study title trial
## Review
## Athena-1
The ATHENA-1 trialwas a phase 3, multi-center, open-label study between December 2015 to May 2017 that evaluated the safety and effectiveness of oliceridine in patients with moderate to severe acute pain, warranting the use of a parenteral opioid. A total of 768 adult post-operative surgical and non-surgical patients with painful medical conditions reporting a pain score of ≥ 4 on an 11-point numerical rating scale (NRS) received oliceridine and were included in the safety and efficacy analysis.
Enrolled patients were treated with IV oliceridine via clinician-administered bolus dosing and/or patientcontrolled analgesia (PCA). Patients treated with IV oliceridine were given a loading dose followed by a supplemental dose if needed. Subsequent doses were administered on an as-needed basis. In patients that were also given PCA, they were provided with a loading dose and a demand dose using a six-minute lockout interval. If clinically indicated, supplemental doses were allowed as needed as early as 15 minutes after the initial dose. The duration of treatment was determined by the clinical need for parenteral opioid therapy, but the maximal duration of oliceridine treatment was limited to 14 days. There was no restriction on the prior use of opioids or non-opioid analgesics, perioperative use of local anesthetics, and epidural and intrathecal opioids, prior or concomitant use of anxiolytics, sedatives, and hypnotics, or concomitant nonopioid analgesics as part of a multimodal analgesic approach. However, patients were not allowed to use other parenteral or oral opioids during treatment with oliceridine.
Oliceridine was found to be associated with a potent analgesic effect and rapid onset of action. Lack of efficacy leading to discontinuation was reported in less than 5% of patients. The mean NRS pain score at baseline was 6.3 ± 2.1. Oliceridine showed a rapid reduction in pain intensity. The mean change from baseline was -2.2 ± 2.3 at 30 minutes after the first dose. Oliceridine showed maintenance of pain reduction. The mean change from baseline at the end of treatment was -3.1 ± 3.1.
Oliceridine demonstrated a favorable safety and tolerability profile, especially for those at risk for opioidrelated complications. Sixty-four percent of all patients reported at least one adverse event (AE) during the study. Most AEs were mild to moderate, but severe AEs occurred in 2% of patients. The intensity of the AEs was similar across all cumulative dose groups. The most commonly reported AEs were nausea (31%), constipation (11%), and vomiting (11%) and incidence was dose-dependent. There were no differences in AEs in patients receiving pain management as bolus or PCA. Serious AEs were observed in 3% of patients with most of these AEs being attributed to complications of surgery, secondary to an underlying medical condition, or secondary to opioid therapy. Three patients experienced serious AEs "possibly" related to oliceridine which were postoperative ileus, respiratory depression, and hepatic/renal failure. There was decreased sedation in the oliceridine dosing groups. A total of 97% of patients reported "none to mild" withdrawal symptoms.
## Apollo-1
The APOLLO-1 trialwas a phase 3, randomized, double-blind, placebo-and active-controlled study between May 2016 to December 2016 that measured safety, effectiveness, and tolerability of oliceridine compared to the placebo and morphine regimens. A total of 418 adults were eligible and scheduled to undergo primary, unilateral, first-metatarsal bunionectomy with osteotomy and internal fixation. Postsurgery, the patients were enrolled into the clinical trial and only were given study medication if they reported at least moderate pain, which was measured on a categorical scale (none, mild, moderate, and severe) and on an 11-point NRS (NRS>4) within nine hours after discontinuation of regional perineural anesthesia.
The enrolled patients were randomized into one of the five treatment regimens, placebo, oliceridine 0.1 mg, oliceridine 0.35 mg, oliceridine 0.5 mg, or morphine 1 mg. The patients received a clinician-administered fixed IV loading dose followed by demand doses administered through a PCA. The PCA doses were administered 10 minutes after the loading dose. Patients could receive protocol-specified open-label rescue pain medication with etodolac 200 mg every six hours as needed. These patients continued with the study medication unless both were inadequate. Additionally, prophylactic antiemetics and prophylactic supplemental oxygen were not permitted during the randomized treatment period.
The proportion of treatment responders through 48 hours was statistically superior for all of the oliceridine treatment regimens compared to placebo. Based on the incidence of AEs and comparable efficacy to morphine, oliceridine regimen 0.35 mg appears to be the optimal dose to achieve the adequate analgesic effect.
Amongst the three oliceridine treatment groups there were no statistical differences on the composite measure of respiratory safety burden (RSB) compared to the morphine group. However, when looking at these treatment groups individually, patients receiving 0.1 mg and 0.35 mg oliceridine experienced significantly lower respiratory safety compared to morphine whereas patients receiving 0.5 mg did not show a difference compared to morphine. Amongst the oliceridine treatment groups, the gastrointestinal (GI) related AEs increased in a dose-dependent manner. The most common AEs were comparable between oliceridine and conventional opioids; these included nausea, vomiting, headache, dizziness, constipation, somnolence or sedation, pruritus, and dry mouth. No patient in the oliceridine treatment groups experienced a serious AE, and only a few reported experiencing an AE that was classified as severe. Compared to morphine-treated patients (7.9%), there were fewer oliceridine patients (2.6%) who discontinued treatment due to an AE. No deaths were reported during the study.
## Apollo-2
The APOLLO-2 trialwas a phase 3, multicenter, randomized, double-blind, placebo-and activecontrolled study conducted at five sites between May 2016 to December 2016 in the United States among 407 patients recruited for an elective abdominoplasty surgery to determine whether oliceridine would provide rapid and superior acute postoperative analgesia compared to placebo with a more favorable safety and tolerability profile than morphine. The goal of this study was to confirm the previous oliceridine phase 2 findings where patients underwent either a bunionectomy or abdominoplasty. Similar analgesic efficacy was observed between the oliceridine treatment groups and morphine. Patients between the ages of 18-75 years were recruited to undergo an abdominoplasty procedure with no additional secondary procedures in the APOLLO-2 trial. Patients were eligible for randomization if they reported moderate to severe pain based on a categorical scale within four hours after surgery and if their pain was a score ≥5 on an NRS.
Patients were randomized in equal ratios to double-blind IV treatment demand dose regimens of placebo, oliceridine 0.1 mg, oliceridine 0.35 mg, oliceridine 0.5 mg, or morphine 1 mg. They began administration with a clinician-administered IV fixed loading dose followed by the demand doses administered through a PCA device or clinician-administered blinded supplemental dose. The use of concomitant analgesics was prohibited for the most part. However, patients could receive protocol-specified open-label rescue pain medication with etodolac 200 mg every six hours as needed. These patients continued with the study medication unless both were inadequate. Prophylactic antiemetics and prophylactic supplemental oxygen were not permitted during the randomized treatment period.
The proportion of treatment responders at the 24-hour NRS assessment was significantly higher in all oliceridine cohorts (0. The cumulative duration of events was also not statistically significant. Therefore, no formal inferiority analyses were conducted, but exploratory studies did indicate that the 0.35 mg and 0.5 mg demand dose regimens were noninferior to morphine with the magnitude of pain relief comparable to morphine. The "perceptible pain relief" and "meaningful pain relief" did not show a statistical difference among the three oliceridine treatment groups and morphine. The active treatment regimens of oliceridine and morphine showed a notable decrease in rescue medication use. In fact, the use of these rescue medications was similar between the 0.35 mg and 0.5 mg oliceridine treatment groups and the morphine group.
While the majority of the AEs reported were mild to moderate in intensity, five patients reported serious AEs. These were post-procedural hemorrhage, abdominal wall hematomas, syncope, and lethargy with the last two relating to oliceridine. The overall proportion of patients experiencing at least one AE was lowest with placebo at 78.3% and increased in a dose-dependent manner across the oliceridine groups (0.1 mg: 93.7%, 0.35 mg: 93.7%, 0.5 mg: 95%). The biggest proportion of patients experiencing at least one AE occurred in the morphine group at 97.6%. The most common AEs included nausea, vomiting, headache, and hypoxia. This study showed that fewer patients treated with oliceridine experienced nausea and vomiting compared to morphine. Previous studies have shown a reduced suppression of respiratory drive with oliceridine. The reduction in respiratory burden in this study was consistent with previous findings, but the differences between treatments using the composite outcome measure did not reach statistical significance. This may be due to the exclusion of patients at risk for these respiratory events. No deaths were reported in the trial.
The clinical trials mentioned above are summarized infor reference.
## Athena-1
Author
# Discussion
Surgeries such as bunionectomies and abdominoplasties may require management of postoperative complications including pain. The management of preoperative, intraoperative, and postoperative pain is continuously evolving. According to the clinical practice guidelines from the American Pain Society, the treatment of postoperative pain should be individually tailored to a patient's specific needs on the basis of adequate pain relief and the presence of AEs. With the high percentage of patients experiencing moderate to severe postoperative pain, opioids are used especially during the first 24 hours after surgery. Intravenous opioids have not shown superiority to oral administration and therefore, in the initial pain management, short-acting oral opioids are preferred for moderate to severe analgesia. While IV morphine may not seem an appropriate choice for the treatment of moderate to severe pain due to the high risk of serious AEs, it is still widely used and is the main opioid assessed in many studies of immediate postoperative pain management due to its high efficacy and potency. Patient-related risk factors may limit their ability to handle these AEs. OIRD is the most serious AE and the biggest driver for the increased deaths associated with opioid use. It can cause life-threatening respiratory and cardiopulmonary arrest.
While exploratory studies showed that RSB was more favorable in the oliceridine group compared to morphine, it was never directly studied but solely based on clinicians' subjectivity. According to a survey of 501 physicians treating postoperative pain, the highest reported unmet need for postoperative pain management is an IV opioid with fewer side effects. Oliceridine demonstrates the potential to address this gap in treatment as seen through the phase 3 clinical trials discussed earlier.
The approval of oliceridine highlights the impact of the opioid epidemic in America. This crisis has impacted the lives of many throughout the decades. Since 1999, there have been three waves of opioid overdose deaths each involving different opioids, including prescription or illicit opioids. The only way to combat this opioid epidemic is through the collaboration of many different groups, including medical personnel. This can improve the ways opioids are prescribed and reduce the number of people who misuse or overdose from these drugs. Since 2011, opioids have been prescribed less as a whole nationally in response to the opioid epidemic. However, the need for surgeries and anesthesia has increased and will continue to increase over the next decade. This trend has correlated with the increase in opioid prescribing to treat patients postoperatively and there has been an increase of 70% in the average total of morphine equivalents prescribed for postoperative pain leading to increased opioid-related morbidity and mortality.
The opioid epidemic has shed light on the importance of alleviating acute pain. Factors, such as age, gender, income, education, and type of surgery impact one's pain vulnerability. The United States Centers for Disease Control and Prevention (CDC) has developed guidelines to work towards combating the overuse of opioids in chronic pain, but pain management guidelines have done very little to address the issue of the increased prescribing in the acute setting. The answer to the opioid epidemic is not to completely abandon the use of opioids as they play a major role in the management of moderate to severe pain; it is to control the use and monitor for the addiction.
# Conclusions
## Future outlook
Through the ATHENA-1, APOLLO-1, and APOLLO-2 trials, oliceridine has demonstrated potential as an opioid with a new mechanism of action showing superior efficacy results compared to placebo, comparable efficacy results to morphine, and a favorable safety profile, but head to head studies comparing oliceridine and morphine are lacking. The exploratory studies were mainly focused on the impact of oliceridine on specific ORAEs, but as a whole, the phase 3 clinical trials conducted for oliceridine did not account for specific imbalances. For example, they predominantly enrolled females, failing to represent the findings across both genders. Additionally, prophylactic antiemetics were prohibited in APOLLO-1 and APOLLO-2. As a result, it would be difficult to predict the occurrence of postoperative nausea and vomiting with oliceridine compared to conventional opioids. As stated earlier, these ORAEs can be prevalent in younger patients and, therefore, future studies with oliceridine should be conducted in this population.
Ultimately, oliceridine provides a new alternative for clinicians due to its rapid onset of action, potent pain relief, and favorable safety profile. While a major aspect of the opioid epidemic is the addiction potential of these opioids associated with long-term use, the potential harm of ORAEs associated with short-term use is a growing area of concern. The approval of oliceridine is a stepping stone in the fight against the opioid epidemic because it paved the path towards finding the balance between adequate management and minimized AEs. The reported safety profile of oliceridine is similar to that of the current pain management options in terms of common side effects, but the selective pathway of this G protein-biased μ-opioid receptor agonist presents a new alternative to clinicians requiring IV analgesia in patients at higher risk for these ORAEs.
# Additional information disclosures
## Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.
|
Strategies for Dealing with Missing Data in Clinical Trials: From Design to Analysis
observation carried forward; NRC, National Research Council; ITT, intent-to-treat principle; MOP, manual of operations.
Randomized clinical trials are the gold standard for evaluating interventions as randomized assignment equalizes known and unknown characteristics between intervention groups. However, when participants miss visits, the ability to conduct an intent-to-treat analysis and draw conclusions about a causal link is compromised. As guidance to those performing clinical trials, this review is a non-technical overview of the consequences of missing data and a prescription for its treatment beyond the typical analytic approaches to the entire research process. Examples of bias from incorrect analysis with missing data and discussion of the advantages/disadvantages of analytic methods are given. As no single analysis is definitive when missing data occurs, strategies for its prevention throughout the course of a trial are presented. We aim to convey an appreciation for how missing data influences results and an understanding of the need for careful consideration of missing data during the design, planning, conduct, and analytic stages.
# Introduction
Scientists require the best available clinical evidence, which can come from systematic reviews, experimental trials, and observational research . Missing data are ubiquitous throughout various medical research designs, even randomized controlled trials (RCT †) that are considered the gold standard for evaluating a direct causal link between an intervention and outcome. The randomized assignment protects against selection bias by equalizing known and unknown characteristics between intervention groups. Nevertheless, randomization alone is not sufficient to provide an unbiased intervention comparison. Two additional requirements for an unbiased study are: 1) missing data from randomized patients do not bias the comparison of interventionsand 2) outcome assessments are obtained in a similar and unbiased manner for all patients. For the latter, standardized participant assessments and blinding of the intervention are commonly employed to assure this balancebut are outside the scope of this review.
Despite efforts to minimize missing data through design, it is likely to occur in the majority of RCTs.
The intent-to-treat principle (ITT) requires the complete inclusion of all data from all randomized patients in the analysis and is considered the most appropriate criteria for assessment of the utility of a new therapy. In an ITT analysis, all randomized participants have outcomes assessed and are analyzed in the group in which they were randomized (regardless of the actual intervention received). When participants drop out or miss visits, thus producing missing data, the ability to conduct an intent-totreat analysis and draw conclusions about a causal link is compromised.
For nearly a century, scientists have been dealing with missing data by deleting or arbitrarily filling in missing cases posthoc. These techniques are prone to bias to the extent the study results are meaningless, yet they continue to be utilized. Over the past 2 1/2 decades, great strides have been made in the development of analytic techniques to estimate causal effects in the presence of missing data. The increased utilization of methods such as inverse probability weighting, multiple imputation, and likelihood-based analysisvastly improved rigor over the adhoc methods (e.g., last observation carried forward, complete case analysis) that previously dominated the RCT landscape. Still, it is important to understand that these methods are tools rather than solutions. When data are missing, the result of any statistical analysis relies on the unverifiable assumptions concerning the relation between the unobserved data and the reasons they are missing. In other words, conclusions drawn from clinical trials with missing data can vary depending on the assumptions made and the analytic method chosen.
Given there is no universal method to analyze missing data, the National Research Council (NRC) released guidelines on the Handling of Missing Data in Clinical Trials. They advocate a more principled approach to design and analysis focusing on two critical elements: 1) careful design and conduct to limit the amount of missing data and 2) analysis that makes full use of information on all randomized participants and is based on careful attention to the assumptions about the nature of the missing data underlying estimates of treatment effects.
As guidance for researchers interested in performing clinical trials, this review is a non-technical overview of the consequences of missing data and a prescription for its treatment extending beyond the typically used analytic strategies to the entire research process including the stages of design, planning, and conduct. We provide an exampledriven demonstration of the potential bias from incorrectly analyzing data with missing observations and explain the advantages/disadvantages of available analytic strategies. Most notably, we discuss the need for a paradigm shift in the way trials are managed, focusing on the prevention of missing data. Many of the recommendations are applicable to research studies beyond the RCT.
## Types of missing data and analytic examples
When planning a study, conducting an analysis or critically reviewing trial results, it is necessary to contemplate how the missing data are generated. Are certain groups more likely to having missing data? Are certain responses more likely to be missing? To assist in our approach to designing and analyzing studies in the presence of missing data, Little and Rubin.
## Missing completely at random
The definition of MCAR is that the likelihood of missing data is unrelated to any observed or unobserved variables. That is, the chance of missing data is the same for individuals in different treatment groups and those who have differential disease severity or treatment response. For example, a dropped test tube in a lab or an equipment failure may lead to missing data. As this is equally likely to occur in any subject in the study (i.e., regardless of treatment received, disease severity, etc.), it represents a completely random process. Subsequently, the average effect of the treatment will be the Missing data caused by features of the study design such as participants being removed from the trial if their conditions are not controlled sufficiently well according to protocol criteria. Dropout based on recorded side-effects. Dropout based on known baseline characteristics.
Dropout based on the unobserved response (e.g., a person not responding to treatment is more likely not to provide an observation). Participants miss a visit because they've had an outcome. outcome Total measure of the relative effect of two treatments is the risk ratio (RR; the proportion of those in treatment A with an improvement divided by the proportion in treatment B with an improvement). In the study population, improvement in body fat was 1.5 times more likely for treatment A compared to B (i.e., RR = 1.5). The 95 percent confidence interval (1.05, 2.15) and the p-value (0.02) generated from the chi-square test show a statistically significant difference indicating that treatment A is superior to B.
To illustrate the MCAR mechanism, suppose that during the study the bioimpedence scale used to measure body fat was unreliable and failed in 30 percent of the participants. We also assume that this failure rate was not related to their treatment or whether or not they had an improvement in body fat (i.e., an MCAR mechanism).shows the results of the study in those without missing data (i.e., completers). The risk ratio in the completers remains unchanged at 1.5. Thus, the estimate of the treatment effect was not biased by MCAR missing data. However, because of the reduced sample size due to missing data, the standard error is larger, resulting in a wider confidence interval (0.98, 2.30) that includes the null value of 1.0 and a chi-square statistic that is no longer significant at the 0.05 threshold (p = 0.053). This example illustrates that complete case analysis does not result in a biased estimate of the treatment difference in the absence of data arising from an MCAR mechanism; however, there is a loss of precision in the estimation of the treatment difference (i.e., larger standard errors and wider confidence limits), as well as a loss of power in the test of significance (i.e., more conservative test statistics yielding higher p-values).
## Missing at random
When the likelihood of missing data is related to observed variables but not to unobserved variables, the missing data mechanism is referred to as missing at random (MAR). If in a clinical trial dropout is more likely for men compared to women, but all men have the same chance of dropout and all women have the same chance of dropout, the missing data mechanism is MAR. Other examples of the MAR mechanism are missing data caused by features of the study design (e.g., providing rescue therapy when conditions are not sufficiently controlled according to protocol criteria), dropout based on recorded side effects or lack of efficacy, or dropouts based on known baseline characteristics.
To illustrate MAR, the results of another hypothetical weight loss trial are shown in . Although the overall likelihood of improvement was higher in men (62.5 percent = 375/600) than in women (25 percent = 150/600), risk ratios were identical (RR = 1.5). Pooling of the gender specific data also results in the same crude and gender adjusted risk ratio (i.e., Mantel Haenszel RR = 1.5, 95 percent CI = 1.33, 1.70). Suppose the likelihood of having missing data was dependent on the combination of the treatment and gender (i.e., an MAR mechanism). For example, at the end of the study, 20 percent of men in treatment A, 20 percent of men in treatment B, 10 percent of women in treatment A, and 55 percent of women in treatment B had missing data. Note that the missing rates are dependent on the observed data but not on the unobserved outcome. demonstrates the cross tabulations with the missing data rates applied to the whole study population. Despite the differential missing data, gender specific risk ratios remain the same, 1.5. However, crude pooling of the gender specific completers data results in a risk ratio of 1.31 (95 percent CI = 1.12, 1.52) that is smaller than the true risk ratio of 1.5. After adjustment for gender using the Mantel-Haenszel method, the true risk ratio of 1.5 is recovered in the completers, albeit with reduced precision and slightly larger confidence limits (1.30, 1.73) compared to the whole study population.
## Missing not at random
When the likelihood of missing data depends on the unobserved data, the missing data is termed missing not at random (MNAR). For example, in substance abuse trials with abstinence as an outcome, it is usual that dropout is higher for those who have relapsed. The problem is in those who drop out, relapse status is not typically obtained. In this case, the probability of having missing data is dependent on their unobserved data -relapse status. In another example, consider a study evaluating treatments to reduce cocaine use in which the outcome is drug level from a urine drug test measured every Monday morning. Participants who use cocaine over the weekend and do not show up for their urine test would be expected to have higher levels of cocaine metabolites. Thus, the likelihood of the data being missing is directly related to the unobserved cocaine level.
Continuing with our hypothetical clinical trial evaluating the effect of treatment on improvement in body fat, demonstrates the consequences of missing data arising from a MNAR mechanism. The proportion of missing data was set at 40 percent in those who received treatment A and had an improvement in body fat and 40 percent in those who received treatment B and did not have any improvement. Therefore, miss-ing data were related to the unobserved outcome (i.e., MNAR). The net result was a complete reversal of the risk ratio when examining the whole study population , RR = 1.5) compared to the completers , RR = 0.84), both reaching statistical significance. If the complete case analysis was used, conclusions would have been the opposite from the true effect.
An alternative taxonomy refers to missing data arising from an MNAR mechanism as non-ignorable because ignoring the process that leads to the missing data will lead to biased results. In contrast, the probability of missing data for ignorable missingness (MCAR or MAR) on a particular variable does not depend on the values of that variable given other observed variables. Such data may still produce unbiased estimates without the need for a model to explain the missing mechanism. To obtain unbiased treatment differences when missing data are MNAR requires modeling the relation between the outcome of interest and the probability of non-response. Determining this relation is a difficult task that highlights the importance of obtaining outcome data on every randomized patient and collecting auxiliary data that may be predictive of dropout.
## Analytic approaches to
MiSSing dAtA: eFFective treAtMent, But not the cure As demonstrated above, failure to appropriately account for the missing data mechanism during analysis can lead to erroneous conclusions. The problem that exists at the end of a study is that there is no definitive way to demonstrate which mechanism has led to the missing data. Addressing missing data during the analysis is not an acceptable comprehensive approach to managing missing data.
Statistical tests are available to demonstrate that missing data are not MCAR. However, exclusion of MCAR is insufficient to ensure validity of a particular analysis. For instance, in a study evaluating two interventions on smoking cessation, exploration might uncover that baseline level of smoking was associated with dropout. While this observation precludes the validity of an analysis that excluded those with missing data (i.e., analysis that assumes MCAR) and suggest the need for an analytic method that is valid under an MAR mechanism, it would not rule out an MNAR mechanism. That is, unless the missing data occur as a function of the study design, such as administrative censoring, a single technique for analysis does not exist. Rather a series of missing data analytic techniques are necessary to examine missing data bias. The following is an overview of analytic approaches, their assumptions about the missing data mechanisms, and their advantages and disadvantages. The approaches fall under four general strategies for coping with missing data: 1) use only data from participants completing the trial with no missing data; 2) use all available data; 3) impute (either single or multiple) values for missing data and analyze with complete case methods; or 4) develop a model for the data that includes a model for the missing data process. The utility of each method under the different
## Predefining approach: during the design of the study
It is important to note the need to specify the plan to address the inevitable missing data during the study design phase. Pre-specification of the statistical analysis plan is important to avoid data-driven selection of results.
## Common approaches: computationally simple, but rarely acceptable
Complete Case Analysis is performed on only those subjects with a complete set of outcome data observed. Subjects with any missing data are excluded from analysis and typical statistical methods are used (e.g., chisquare analysis, t-tests, ANOVA, regression) on the reduced set of observations. For complete case analysis to provide an unbiased assessment of the intervention effect, the assumption that completers are a random sample of the full study sample (i.e., MCAR) is required. In addition to its computational simplicity whereby common statistical tests are conducted, an advantage of this method is that estimates of treatment differences are unbiased when missing data are MCAR. Disadvantages include loss of power and precision in the estimation of the treatment effect from the reduced sample size and the reliance on the strong MCAR assumption. When missing data are not MCAR, the estimate of the intervention effect will be biased.
Single Imputation creates a complete set of data for all randomized subjects by using a rule to set missing responses to a value. These approaches are computationally simple. Once created, simple statistical methods are used on the full set of data. There are many forms of single imputation, including last observation carried forward (LOCF), worst observation carried forward, and simple and conditional mean (or regression) imputation.
Due to its simplicity and its perceived (but often not true) tendency to provide conservative estimates of the treatment effect, LOCF has been a popular imputation technique. In LOCF, missing data for a subject is replaced with the last observed value for that subject. This analysis is not valid under MCAR but rather under the very specific and unrealistic assumption that the missing outcomes are equal to the last observed response. The argument that LOCF provides a conservative estimate of an intervention is not universally true. Take, for example, a trial comparing two interventions for maintaining cognitive function in patients with Alzheimer's. The typical downward trajectory for these patients may lead to an upwardly biased estimate of function in those who drop out. A higher proportion of dropouts in one of the intervention groups would favorably bias that intervention in an LOCF analysis. Imputing identical values for an individual can also lead to underestimation of variability and a low p-value.
Two other single imputation approaches are simple and conditional mean (or regression) imputation. In simple mean imputation, missing values are replaced by the mean for that variable. This approach ignores information from other variables that may be relevant if the data are MAR. While it can lead to valid estimates of treatment differences under MCAR, it will result in an underestimation of variability from the unseen data because a constant is imputed for all missing data regardless of differing participant characteristics. In contrast, conditional mean imputation accommodates associations with other observed variables by regressing the outcome on other observed variables in the completers. The missing outcomes are imputed using the regression equation that includes a random error component. Given the use of the observed data in the imputation and provided all covariates that sufficiently explain the missing data are included, valid estimates of treatment differences can be achieved under MAR. Nevertheless, the use of a single value to replace missing data does not fully capture the uncertainty that this value is correct. As with other single imputation procedures, underestimates in the variability of treatment ef-fects will lead to inappropriately low p-values.
The NRC guidelinesfor missing data state that single imputation methods like "last observation carried forward" should not be used as the primary analytic approach unless the assumptions that underlie these methods are scientifically justified.
## Acceptable approaches under mar
Inverse Probability Weighting. This method has its origins in sample survey research in which responses by survey participants are weighted to accommodate unequal probabilities of selection. The survey weight for each participant is the inverse of their probability of selection, and, thus, those with lower probabilities of selection have a greater weight in the analysis. In missing data, probabilities of being observed are estimated and analyses are weighted by their inverse. Consequently, those with a low probability of observation will be weighted more highly in the analysis. Weights may be obtained from a model such as a logistic regression that includes intervention group, previous values of the outcome of interest, and other covariates that may be predictive of being observed.
Inverse probability weighting is a natural extension of a familiar technique that provides unbiased estimates of the treatment difference under the MAR assumption. Unlike the methods described below, it does not directly make use of all available data as only subjects with complete data are included in the final weighted model. Thus, power may be attenuated. Furthermore, to obtain valid standard errors and confidence intervals, specialized software is required to appropriately accommodate the weighted analysis. Packages such as SUDAAN, SAS, and STATA are equipped to do this.
Multiple Imputation (MI). While conditional mean imputation produces unbiased estimates of the treatment effect, it consistently underestimates the variability of this effect (i.e., artificially lowering p-values). This underestimation results from treating the observed and imputed values the same despite uncertainty in the imputed values.
The method of multiple imputation corrects for this by generating m completed data sets (typically m ranges from, each with the observed values and different plausible values imputed for the missing observations. After creating the m complete data sets, each is analyzed using usual methods (e.g., AN-COVA, regression), and the results are then combined across the analyses. It is important to note that imputations are not meant to be actual observations for that individual, but rather a statistically plausible set of values based on other information for that individual. Thus, analysis from "filled-in" data sets provide statistically plausible results that would have been obtained if there were no missing data. Refer tofor examples of trials using MI.
Under the MAR assumption, MI produces unbiased estimates of the intervention effect and correct p-values. Other advantages of MI include the ability to handle not only missing outcome but missing covariate information, its relatively easy implementation, and the flexibility it provides by separating the imputation from the analytic model. The latter allows for increased complexity of the imputation model to make the MAR assumption more plausible. It also provides a simple and attractive framework for exploring sensitivity to non-random missing data. Drawbacks to MI include the inability to produce a unique estimate of the treatment effect (provides a different result each time you use MI) and the requirement for compatibility between the imputation and analysis models (e.g., the analysis model cannot contain variables, non-linearities or interactions that are not in the imputation model). The imputation model can be more complex than the analysis model, but the latter cannot contain variables that are not in the imputation model.
Likelihood-based Analysis. Maximum likelihood estimation (MLE) is a common estimation method in statistics contingent upon finding estimates of the treatment differences that maximize the probability of the observed data. To illustrate the MLE approach, suppose we do an experiment in N people where the probability of a success for an individual is p and the probability of a failure is 1-p. If n people succeed and N-n people fail, the likelihood is proportional to the product of the probabilities of successes and failures or p n (1-p) N-n . The value of p that maximizes the likelihood is n/N or the overall proportion of success. In the weight loss trial with no missing data, maximum likelihood will produce the best estimate of the difference in body fat between intervention groups that maximizes the probability of observing the data. The problem when missing data occur is that we only observe a subset of the data yet we would like to draw conclusions based on the full data. Under MAR, the likelihood-based analysis allows us to accomplish this by averaging out the missing data from the joint likelihood of the observed and missing data. This is possible under MAR, because the future statistical behavior for a subject, conditional upon observed data, is the same whether or not that subject drops out. Seefor examples of trials using MLE.
Under the MAR assumption, MLE produces unbiased estimates of the intervention effect and correct p-values and software is readily available in statistical packages (SAS, STATA, SPSS) for both continuous and discrete outcomes. Unlike MI, MLE provides a unique estimate of the treatment difference. MLE also requires fewer decisions than MI and is not reliant on compatibility of the imputation (as there is no imputation in MLE) and analysis model. Disadvantages of MLE include its reliance on parametric assumptions (e.g., normality) and that it is only appropriate for missing outcome data (i.e., it is unable to accommodate missing covariate data).
ITT or Per Protocol. ITT analyses determine the effect of intervention assignment (i.e., intention to give treatment) on the outcome regardless of the actual intervention received. It evaluates the random assignment. Per-protocol analyses evaluate the effect of the intervention for participants who adhered to the protocol. IPW, MI, and MLE methods are consistent with the ITT principle in that they do not exclude data from participants with incomplete responses and participants are analyzed within the group to which they were randomized. However, in that they assume people with missing data would have had the same outcome experience if they had completed the study as similar people without missing data (i.e., those who adhered to the protocol) the MAR methods deviate from the ITT toward the per-protocol analysis. Nevertheless, when data are missing, there is no unequivocal ITT analysis. As such, the MAR methods often provide a sensible approach for the primary analysis, but sensitivity analyses are recommended to understand the robustness to departures from the MAR assumption.
## Acceptable approaches under mnar
The basic approach to handling ignorable missing data (i.e., from an MCAR or MAR mechanism) is to adjust for all observable differences between missing and non-missing cases and assume that all remaining differences are unsystematic, thereby ignoring the process by which missing data happen. When missing data occur from an MNAR process, appropriate analysis requires the joint modeling of the outcome along with the missing data mechanism. This can be very complicated, given that under MNAR the model thereby the missing data process is rarely known for creating unverifiable assumptions for the analysis. For instance, we suspect relapse to lead to missing data in a study of treatment for substance abuse, but it is unlikely that all missing data is the result of relapse. To conduct an MNAR analysis, it is necessary to specify the strength of this relation, i.e., what's the probability of having missing data given relapse or similarly (but not equal) what's the probability of relapse in those with missing data?
Paralleling these related questions are two approaches to analysis under MNAR, selection, and pattern-mixture models. Selection models require the specification of the relation between the outcome and the probability of being missing. For instance, in our weight loss study, the probability of a non-missing observation might be lower in those who have a recent unob-served increase in body fat. On the other hand, pattern-mixture models specify the outcome distribution across the observed missing data patterns. For the weight loss example, this could correspond to indicating the probability of various weight loss profiles for those who drop out after their first visit compared to those who drop out after their second visit or those who don't ever drop out. While arbitrary, the chosen weight loss profiles could be further informed by recorded data such as the reason for dropout, where different profiles could be adopted for those dropping out because of migration compared to lack of efficacy.
While we can conceive of an endless number of MNAR analyses, in practice, a few reasonable scenarios are chosen in sensitivity analyses. Mallinckrodt et al. discuss a structure for conducting MNAR sensitivity analyses. Ranges of likely values for the non-ignorable missing data can be solicited from experts (see examples of this process). The robustness of the conclusion with regard to the treatment difference is evaluated across these scenarios. For example, in the pattern mixture model described above, a range of weight loss profiles could be examined for their impact on study conclusions. See Hedeker et al. for an applied example.
## Exploring missing data patterns
In the presence of missing data, it is important to understand how much is missing and in what patterns. Exploratory analyses include examining proportions and time to drop out, differential reasons for dropout, and characteristics of those who do and do not drop out. Kaplan Meier curves are helpful to compare time to dropout between intervention groups or other important study variables. Similar rates and reasons for dropouts between intervention groups increase the confidence in validity of an ignorable analysis. Plots of outcomes at each time point for those who do and do not drop out at the next visit determine whether dropout is conditional on observed out-comes. Logistic regression can identify factors most strongly associated with dropout. While a hypothesis test is available for MCAR, its use is limited as it only provides evidence that data are not MCAR and cannot be used as confirmation that data are MCAR. No test can provide definitive proof of the mechanism that produced missing data.
how Much MiSSing dAtA iS too Much?
The proportion of missing data alone is not sufficient to indicate study validity, but studies with minimal missing data are more likely to produce valid conclusions. Schulz and Grimessuggest that losses to follow-up less than 5 percent usually have little impact whereas losses greater than 20 percent raise serious flags about study validity. In-between levels lead to problems somewhere in the middle. To support this, Kristman et al.demonstrated through simulation that substantial bias in the estimation of odds ratios under MNAR conditions may arise in cohort studies with loss to follow-up of 20 percent. However, the 5-20 general rule of thumb has no statistical justification and oversimplifies the problem as the bias resulting from missing data also depends on the missing data mechanism and the analytic method. It is certainly possible that the use of a complete case analysis when 5-20 percent of the data are missing from an MAR mechanism can lead to a biased treatment effect.
## Practical approaches to minimize missing data
Given the uncertainties involved with the identification of the missing data mechanism and thus conclusions from analyses that rely on this identification, the best method for dealing with missing data is to prevent it. O'Neill described the need for a cultural shift, focusing on strategies to prevent missing data during the conduct and management of clinical trials rather than relying on imperfect analytic methods.
The National Research Council (NRC) has provided recent guidance, listing several recommendations to prevent missing data during the design and conduct of clinical trials. Interpretations of these guidelines and additional approaches (seeto prevent missing data have been describedas well as the potential dynamic between statisticians and clinical investigators to achieve them. The overarching Develop detailed study documentation in the form of manual of operations addressing all aspects of the study including screening procedures, training requirements, methods of communication, delivery of treatment, schedule and windows for assessments, and data collection/entry/editing procedures. 8. Develop an informed consent that distinguishes the difference between withdrawing from the treatment and withdrawing from the study. 9. Select study sites with strong track records for enrolling, following, and completing participants.. Adopt a reimbursement mechanism that encourages study completion. 11. Train/certify study personnel for participant enrollment, data collection, data entry, delivery of treatment, etc. prior to enrollment with re-certification throughout trial if necessary. 12. Highlight the continued collection of data in participants that are not adherent to treatment but remain in the study.. Test operational aspects of the trial (e.g., enrollment, retention, clarity of study manuals and data collection instruments, study burden on participants, randomization, treatment delivery).. utilize approaches to keep the study participants engaged in the study including incentives, visit reminders, newsletters, and intermittent phone calls to monitor status. 17. Outline procedures for contacting individuals with missed visits in manual of operations. Identify and intervene in participants that are likely to drop out. 18. Timely data entry allows earlier detection of problems with missing data.. Implement a verification process requiring fields to be checked for accuracy and all discrepancies resolved before data entry. 20. Devise an efficient method of communication with study personnel for identifying and resolving unanticipated issues that arise during the study.
## Analytic stage
Explore Missing Data use All Available Data Perform sensitivity analysis 21. The amount of missing data, missing data patterns and variables associated with missingness will help to inform the primary and sensitivity analyses.. For primary analysis, use methods that make use of all available data such as multiple imputation or likelihood-based approaches. These methods make weaker assumptions about the missing data compared to complete case analysis.. For primary analysis, avoid the use of ad-hoc solutions (e.g., last observation carried forward) as they make unreasonable assumptions about the mechanism that produced the missing data.. use methods such as pattern mixture or selection models to examine robustness of conclusions to reasonable MNAR mechanisms.
goal is the creation of procedures and a climate and culture that will maximize the collection of complete data. Collaboration with a data coordinating center with a history of conducting clinical trials is strongly encouraged to improve implementation.
## Approaches in study design
While it won't necessarily attenuate missing data, predicting the expected proportions of missing data is recommended during the design phase as it 1) can impact the variability and estimate of the effect size thus influencing the required sample size and 2) be helpful in directing the range of sensitivity analyses required. Of note, inflation of the sample size estimate for dropouts helps to preserve power, but it is not a comprehensive strategy for dealing with missing data as it does not preclude the opportunity for missing data-related bias.
Other design techniques are aimed at reducing the number of participants with a missing primary outcome, usually from dropout. Although collecting an abundance of data to answer secondary, tertiary, and exploratory questions is tempting, the focus of a trial is to answer a primary question. Therefore, the benefits of collection of data beyond what is required to answer the primary question and crucial supportive evidence for it must be carefully weighed against the threat of missing data. Limiting participant burden by reducing the number of visits and the amount of data collected at each visit are universally applicable study design approaches.
The applicability of other design strategies is study-specific. Run-in periods are beneficial to minimize missing data even though the original intent was to maximize adherence and exclude those intolerant to treatment. Addressing target populations with incentive to remain in the study, using flexible treatment regimens that increase adherence, using outcomes that can be ascertained in a high proportion of participants, and shortening the follow-up period for the primary outcome also are design approaches to minimize missing data.
The randomized withdrawal design was initially proposed to evaluate long-term ef-fectiveness. Participants deemed "responders" to a treatment are randomized to either maintain their current treatment group or receive placebo. For example, to minimize the long-term impact of a weight loss therapy, a short-term endpoint (e.g., 3 months), where the propensity for missing data is low, could be used to define those responding to the treatment. Responders could then be randomized to continue or receive standard care and then be evaluated at 12 months to examine maintenance of effectiveness.
## Approaches in study planning and conduct
Several common-sense opportunities should be considered during the planning and conduct of a study to minimize missing data. Approaches are directed at the participant, the data collection process, and the study team.
In addition to limiting participant burden, enhancing participant engagement is an effective approach to promote study completion. Updating contact information regularly and providing incentives for participation decrease the likelihood of dropout. Monetary incentives can be increased for subsequent visits beyond the index visit to encourage study completion. Informed consent procedures should clearly distinguish between discontinuation of study treatment and discontinuation of follow-up so participants are clearly aware of their ability to complete the study despite non-adherence. Feelings of investment and appreciation for participation may be reinforced through newsletters, study websites and social networking, access to study findings, and frequent expressions of gratitude from the study team. It's particularly important to engage participants at the greatest risk of dropout. Those at highest risk may be identified by asking at each visit about their likelihood for attendance at the next visit. If the chances of attending are low, reasons or barriers along with potential solutions can be explored.
When a participant drops out, it's helpful to determine the reason for withdrawal as well as ancillary factors that may have been associated with their decision. This information can be utilized to meet the MAR assumption (e.g., by including them as adjustment covariates in the analysis) or to inform sensitivity analyses.
In addition to participant cooperation, complete data collection is contingent upon adequate study documentation, appropriate training of study personnel, a robust clinical management procedure, and a structure for efficient communication among the study team. A manual of operations (MOP) provides documentation to guide the conduct of the study including timing windows for assessments, instructions for data collection, and training requirements. Many NIH institutes such as the National Institute on Aging offer online guidelines and templates for the MOP. Prior to enrollment, training and certification of study personnel is essential for proper execution of the protocol. Highlighting the need for complete data collection regardless of participant treatment adherence is a key component of this training. Pilot testing the protocol can identify possible sources of missing data. Development of participant calendars for visit reminders and automated participant contact can be made possible by an electronic clinical data management system. Visual data completeness checks may be performed prior to the participant leaving or electronic data capture systems programmed to require data completeness.
Constant exchange of information among the study team is essential so a structure for efficient communication is required. The process for dissemination of revised operations must be delineated. Regular study team meetings (phone or face to face) or web-based discussion boards permit the identification and an opportunity to resolve potential missing data issues.
Finally, data completeness should be evaluated through regular monitoring reports made available to the entire study team. These reports require the automization of data collection or entry of data in a timely manner (i.e., as close to real-time as possible) in order to be utilized to improve study conduct. The reports should include site-specific summaries of subject retention and data completeness, particularly for the essential study outcomes. A priori targets for unacceptable rates of missing data will aid in interpretation and possible remediation. Underperforming sites can undergo additional training or be closed. To avoid this, the track record of the site with regard to enrolling, following and collecting complete data should be an important criterion for study-site selection.
concluSion Missing data are common in clinical trials. Given that no single analysis will be definitive in the presence of missing data, limiting missing data through the creation of a climate and culture that maximizes the collection of complete data is necessary. Our proposed recommendations operationalize this by providing specific guidance for each stage of the trial. In the design stage, researchers should anticipate missing data patterns and causes and consider methods/designs that encourage participant retention. Developing detailed study documentation, training study personnel and testing operational aspects of the trial are important during the planning stage. Regular monitoring of missing data and enhanced participant contact is recommended for the conduct stage. While easy to implement, ad hoc methods such as complete case analysis and last observation carried forward are not advocated as primary analytic strategies. As a primary strategy, the use of all available data is recommended through methods such as multiple imputation and likelihood-based analysis. Sensitivity analyses under MNAR are appropriate to evaluate robustness of conclusions to a range of sensible conditions.
|
An Uncommon Case of Incessant Tachycardia-induced Cardiomyopathy in a Child
An AXIOM Sensis XP system (Siemens AG, Munich, Germany) was also used during the procedure.Narrow complex tachycardia with a cycle length (CL) of 360 ms was ongoing. The earliest atrial activations were recorded in the left posterolateral wall (ie, at 4 o'clock according to the clinical standard nomenclature by Cosio) with the shortest ventriculoatrial (VA) interval equal to 104 ms. Few sinus complexes were recorded with concentric ventricular activation(Figure 2A). The atrium to His (AH) and His to ventricle intervals were 94 ms and 50 ms, respectively (Figure 2B). The post-pacing interval was 492 ms with a ventricle-atrium-ventricle (VAV) response during overdrive pacing and entrainment of tachycardia from the right ventricle (Figure 2C). Ventricular pacing within 40 ms of the His potential advanced the atrium and the atrial activation sequence to the same as that seen during tachycardia (Figure 2D). The ventricular pacing given earlier terminated the tachycardia without advancing to the atrium (Figure 2E). The VA interval was not decremental, although the tachycardia CL fluctuated from 310 ms to 360 ms. Because of the incessant tachycardia, we could not pace the atrium in sinus rhythm; however, the short attempts showed that the AH intervals during the sinus complexes and the atrial pacing in tachycardia CL were similar. Orthodromic atrioventricular reentrant tachycardia (AVRT) with concealed slow conductive accessory pathway (AP) was diagnosed.Case presentationA nine-year-old girl was admitted to our clinic having presented with palpitation, dyspnea, and heart failure (New York Heart Association functional classification class II). A 12-lead resting electrocardiogram (ECG) revealed regular narrow complex tachycardia(Figure 1). Echocardiography revealed reduced left ventricular systolic function (ejection fraction was 35%) and normal biatrial diameter. Tachycardia was incessant, lasting more than 50% of monitoring time prior to drug administration. Intravenous adenosine and b-blocker medication were ineffective. The patient was referred for electrophysiology (EP) study and catheter ablation. The procedure was performed with the patient under sedation with intubation. A steerable decapolar catheter (Abbott Laboratories, Chicago, IL, USA) was inserted into the coronary sinus via the subclavian vein, a quadripolar catheter (Abbott Laboratories, Chicago, IL, USA) was positioned in the right ventricle, and an ablation catheter (Marinr ® MC; Medtronic, Minneapolis, MN, USA) was placed in the His position via the right femoral vein during the EP study.ABSTRACT. The case of a pediatric patient with a history of incessant narrow complex tachycardia is presented. The patient underwent successful catheter ablation for an uncommon concealed slow accessory pathway. The mechanism and ablation location are discussed.
Narrow complex tachycardia with a cycle length (CL) of 360 ms was ongoing. The earliest atrial activations were recorded in the left posterolateral wall (ie, at 4 o'clock according to the clinical standard nomenclature by Cosio) with the shortest ventriculoatrial (VA) interval equal to 104 ms. Few sinus complexes were recorded with concentric ventricular activation . The atrium to His (AH) and His to ventricle intervals were 94 ms and 50 ms, respectively . The post-pacing interval was 492 ms with a ventricle-atrium-ventricle (VAV) response during overdrive pacing and entrainment of tachycardia from the right ventricle . Ventricular pacing within 40 ms of the His potential advanced the atrium and the atrial activation sequence to the same as that seen during tachycardia . The ventricular pacing given earlier terminated the tachycardia without advancing to the atrium . The VA interval was not decremental, although the tachycardia CL fluctuated from 310 ms to 360 ms. Because of the incessant tachycardia, we could not pace the atrium in sinus rhythm; however, the short attempts showed that the AH intervals during the sinus complexes and the atrial pacing in tachycardia CL were similar. Orthodromic atrioventricular reentrant tachycardia (AVRT) with concealed slow conductive accessory pathway (AP) was diagnosed.
## Case presentation
A nine-year-old girl was admitted to our clinic having presented with palpitation, dyspnea, and heart failure (New York Heart Association functional classification class II). A 12-lead resting electrocardiogram (ECG) revealed regular narrow complex tachycardia . Echocardiography revealed reduced left ventricular systolic function (ejection fraction was 35%) and normal biatrial diameter. Tachycardia was incessant, lasting more than 50% of monitoring time prior to drug administration. Intravenous adenosine and b-blocker medication were ineffective. The patient was referred for electrophysiology (EP) study and catheter ablation. The procedure was performed with the patient under sedation with intubation. A steerable decapolar catheter (Abbott Laboratories, Chicago, IL, USA) was inserted into the coronary sinus via the subclavian vein, a quadripolar catheter (Abbott Laboratories, Chicago, IL, USA) was positioned in the right ventricle, and an ablation catheter (Marinr ® MC; Medtronic, Minneapolis, MN, USA) was placed in the His position via the right femoral vein during the EP study.
ABSTRACT. The case of a pediatric patient with a history of incessant narrow complex tachycardia is presented. The patient underwent successful catheter ablation for an uncommon concealed slow accessory pathway. The mechanism and ablation location are discussed. Trans-septal puncture was performed using a Brockenbrough curved needle (Abbott Laboratories, Chicago, IL, USA). Radiofrequency ablation was performed in the posterolateral wall of the mitral annulus. Supraventricular tachycardia (SVT) terminated at the fifth second of ablation and did not recur. Stable VA dissociation was recorded . Tachycardia was no longer inducible after a 30-minute waiting period. No other coarrhythmias were included. The procedure was completed without complication. Fluoroscopy time was 10 minutes [fig_ref] Figure 3: A [/fig_ref]. Local activation at the successful ablation site is shown in [fig_ref] Figure 3: A [/fig_ref].
No SVT recurrence was detected during the three-month follow-up period, and left ventricular ejection fraction was 58% soon after the procedure.
# Discussion
Limited literature is available on the association between incessant AVRT and tachycardia-induced cardiomyopathy in pediatric patients. Generally, the incessant nature of tachycardia favors permanent junctional reciprocating tachycardia (PJRT). [bib_ref] Variable location of accessory pathways associated with the permanent form of junctional..., Ticho [/bib_ref] The AP in PJRT usually has retrograde and anterograde decremental conduction An RF catheter is positioned in the tricuspid annulus. The first beat (1) was sinus and conducted to the ventricle through the AV node. The second beat (2) was a premature atrial conduction conducted to the ventricles by the AV node (long AV interval), which initiated (3) SVT with a cycle length (CL) of 360 ms and a constant ventriculoatrial (VA) interval of 104 ms. Aberrancy in the first tachycardic beats was recorded. B: His potential during tachycardia. The atrium to His interval was 94 ms; the His to ventricle interval was 50 ms. C: Overdrive pacing from the right ventricle was performed. Entrainment was confirmed by measurement of the AA interval during pacing (300 ms) and tachycardia (360 ms), which was also the tachycardia CL. VAV response and 492-ms PPI are illustrated. PPI-TCL = 132 ms. The earliest atrial activations were detected in CS2, positioned in the left posteroinferior wall. D: A PVC placed at the time of His bundle refractoriness advanced the atrium without changing the atrial activation sequence and reset the tachycardia, which was diagnostic for a retrograde pathway that participates in tachycardia. E: The ventricular pacing given earlier terminated tachycardia without advancing to the atrium. F: Retrograde ventriculoatrial dissociation was recorded during right ventricular pacing after ablation.
properties and is typically identified in the posteroseptal location. [bib_ref] Variable location of accessory pathways associated with the permanent form of junctional..., Ticho [/bib_ref] Critelli et al. described the anatomical substrate of PJRT as an AP with a tortuous course through the AV annulus fenestration, which probably causes its decremental properties. [bib_ref] Anatomic and electrophysiologic substrate of the permanent form of junctional reciprocating tachycardia, Critelli [/bib_ref] The surface ECG criteria for PJRT are well reported and include narrow QRS tachycardia with negative P-waves in inferior leads, a PR interval shorter than the RP interval, an atrioventricular (AV) ratio of 1:1, no evidence of a delta wave during sinus rhythm, and no episode of functional AV block during tachycardia. [bib_ref] Permanent junctional reciprocating tachycardia in children: a multicentre study on clinical profile..., Vaksmann [/bib_ref] The patient's ECG met all these criteria, but the P-waves were considered positive in inferior leads. AVRT, atrial tachycardia, and atypical AV nodal reentrant tachycardia (AVNRT) must be excluded in such cases .
A previous study reported that only 2.9% (34 patients) of 1,163 consecutive individuals with Wolff-Parkinson-White syndrome had VA conduction times greater than 80 ms. [bib_ref] Electrophysiological characteristics and radiofrequency ablation of accessory pathways with slow conductive properties, Manita [/bib_ref] Only one of these patients was younger than 18 years. [bib_ref] Electrophysiological characteristics and radiofrequency ablation of accessory pathways with slow conductive properties, Manita [/bib_ref] The authors reported long ablation times while treating this type of pathway.
## Incessant tachycardia-induced cardiomyopathy in a child
The difference between the present case and patients with PJRT is the retrograde circuit, which is capable of slow retrograde conduction over an AP without decremental properties and is located in the posterolateral zone. In our case, procedural time was also longer than usual (ie, 100 minutes).
Notably, we observed a VAV response during ventricular overdrive pacing , although it is known that a pseudo-ventricle-atrium-atrium-ventricle (pseudo-VAAV) response may occur in such a situation. [bib_ref] Establishing the mechanism of supraventricular tachycardia in the electrophysiology laboratory, Kanjwal [/bib_ref] As reported previously, a pseudo-VAAV response occurs when retrograde conduction is slow either in an atypical AVNRT or a slow AP with retrograde conduction. [bib_ref] Establishing the mechanism of supraventricular tachycardia in the electrophysiology laboratory, Kanjwal [/bib_ref] A postpacing interval-tachycardia CL (PPI-TCL) > 115 ms is usually consistent with AVNRT, while a PPI-TCL difference < 115 ms is consistent with AVRT. [bib_ref] Differentiation of atypical atrioventricular node re-entrant tachycardia from orthodromic reciprocating tachycardia using..., Michaud [/bib_ref] Nevertheless, a PPI-TCL > 110 ms can occur with AVRT circuited by a left-sided AP, because the right ventricular pacing site is far from such a circuit. During a long RP interval SVT, a corrected PPI-TCL > 110 ms should also prompt the consideration of orthodromic AVRT employing an AP with delayed retrograde conduction. [bib_ref] First postpacing interval after tachycardia entrainment with correction for atrioventricular node delay:..., Gonzalez-Torrecilla [/bib_ref] The most common maneuver is to deliver a His-synchronous premature ventricular contraction (PVC) on time or within 30 ms of the His potential. Unfortunately, figures of His refractory pacing, in which the His potential was recorded, are not presentable in the current case due to noise. In this case, atrial activation is advanced without a change in the atrial activation sequence. The fact that a His refractory PVC can affect atrial timing indicates that an AP is present. As the atrial activation sequence is unaltered, we conclude, though not with certainty, that the AP is participating in the SVT circuit, establishing a diagnosis of AVRT . The earlier PVC terminates tachycardia without conduction to the atrium. This indicates that the AP is present and is part of the circuit .
The AH interval also has an important diagnostic role in the EP study. During PJRT, this interval is a true interval, similar to the AH interval, when pacing at the tachycardia CL (difference of < 20 ms). [bib_ref] Long RP interval tachycardia. What is the mechanism?, Jiménez-Diaz [/bib_ref] tachycardia involving concealed nodofascicular APs includes a shorter AH interval during tachycardia compared with during sinus rhythm. [bib_ref] Long RP interval tachycardia. What is the mechanism?, Jiménez-Diaz [/bib_ref] In our case, because of the incessant tachycardia, we could not pace the atrium in sinus rhythm. However, the AH intervals during SVT and atrial pacing were similar to that during sinus rhythm. A short AH interval is also reportedly an important factor in the development of incessant tachycardia. [bib_ref] Electrophysiologic comparison between incessant and paroxysmal tachycardia in patients with permanent form..., Yagi [/bib_ref] Previous studies have shown that local VA at the successful ablation site is usually 25 ms to 50 ms and that a VA interval of less than 50 ms is an independent predictor for successful AP ablation. [bib_ref] Successful radiofrequency ablation of accessory pathways with the first energy delivery: the..., Xie [/bib_ref] Lin et al. described the EP characteristics of an AP with a long VA interval (arbitrarily defined as ≥ 50 ms with an absence of continuous electrical activity) and no retrograde decremental property. [bib_ref] Electrophysiological characteristics of accessory pathways with prolonged retrograde conduction, Lin [/bib_ref] Fifteen patients with the aforementioned characteristics were compared with 171 study participants with normal VA conduction; the mean VA conduction time was 77 ms ± 24 ms. Notably, the experimental group included significantly older patients with longer retrograde AP block CLs and retrograde AP effective refractory periods. In these patients, adenosine and verapamil were ineffective for terminating tachycardia. There was also a positive correlation between AP VA interval and patient age. In this current case, we observed this rare pathway in a pediatric patient, and there was no VA fusion at the successful ablation site.
# Conclusions
Slow retrograde conducted AP may be located in an unusual left posterolateral site. In this case, it induced incessant tachycardia with cardiomyopathy in a child. The EP features of this pathway are an absence of AV fusion in the ablation site and long PPI during entrainment from the right ventricle.
[fig] Figure 3: A: Fluoroscopic imaging of catheter positions during the successful ablation (right: RAO, left: LAO). The decapolar catheter was in the coronary sinus. The Marinrs MC ablation catheter (Medtronic, Minneapolis, MN, USA) was in the mitral groove. The quadripolar catheter was in the right ventricle. B: Local activation at the successful ablation site. No VA fusion was present at the ablation site. H.R. Poghosyan, A.B. Danoyan, T.B. Hovakimyan, et al. [/fig]
|
Formal consensus to identify clinically important changes in management resulting from the use of cardiovascular magnetic resonance (CMR) in patients who activate the primary percutaneous coronary intervention (PPCI) pathway
## General comments
Cardiovascular magnetic resonance (CMR) is currently playing a pivotal role in clinical cardiology. CMR is increasingly used for a subgroup of ACS patients who activate the primary percutaneous coronary intervention pathway, but it seems unclear how CMR influences clinical management in this population.This consensus study is well designed to answer this question and the message is simple, clear but could be very useful to the society.
The study is well presented in general with minor issues. I would therefore recommend for a publication with minor revision.
1. Page 1. The title needs to be reworded The objective is to determine important changes in patient management arising from the use of cardiovascular magnetic resonance (CMR) imaging; but the original title can be simplified as "consensus to identify management change resulting from CMR patients...". This is incomplete or confusing. It could be '... use of CMR in patients...' from my understanding.
## General comments
This is an interesting concept on generating consensus on patients "activating the PPCI pathway". Unfortunately, the paper is neither as study, nor describing a novel method, nor a proper consensus statement.
1.) The aims and goals of the "study" are not well described:
- After reading the title and the abstract I was still unclear what this is about. What patients exactly are looked at? What is a "formal consensus study". The paper is not a study, it is a consensus statement generated by an interactive process of (random) participants.
- The reason for the study was "to determine the feasibility of setting up a prospective registry", later it is stated, that "the key objective … was to define a primary composite outcome".
- Neither of the two reasons is addressed in the "study" 2.) The authors need to decide, whether this is a study assessing the recognition of CMR in patients after PPCI within the larger cardiology community of the UK or whether this is a consensus statement prepared by iterative majority consensus vote of some UK cardiologists. The method used has been described before and this should be introduced in the introduction and discussed in the methods. (It partially is). It would also be good to compare this approach to the ACC/AHA approach for Appropriateness Criteria.
3.) There is redundancy of the addenda and the main paper. It would help readability of the main paper if the statements receiving final support would be ordered as 1 -X and the ones without support X -12. The process, the first version of statements and the results from the first round could then be placed solely in the addendum.
4.) Some of the evidence is questionable and needs to be used with more care. E.g. the meta-analysis on CMR being superior to echo for the detection of thrombus includes several papers, which use CMR LGE as the reference standard. Such evidence should not be used.
Obtaining consensus and defining the utility of CMR has merits and the paper adds knowledge. I would suggest to focus on the consensus, do not call it a study and place the pathway for obtaining the consensus in the addendum
## Version 1 -author response
Reviewer 1
1. Page 1. The title needs to be reworded. The objective is to determine important changes in patient management arising from the use of cardiovascular magnetic resonance (CMR) imaging; but the original title can be simplified as "consensus to identify management change resulting from CMR patients...". This is incomplete or confusing. It could be '... use of CMR in patients...' from my understanding.
We have changed the title to "Formal consensus to identify clinically important changes in management resulting from the use of cardiovascular magnetic resonance imaging (CMR) patients who activate the primary percutaneous coronary intervention (PPCI) pathway".
2. Page 5. Strengths and limitations of this study. Please consider remove the last sentence. This is not really a limitation to the current study as no single study can address all patient groups that may benefit from CMR.
This sentence has been removed.
3. Page 6. consider rewriting: '...for patients with acute coronary syndrome (ACS) who activate the primary percutaneous coronary intervention (PPCI) pathway...'. To note, patients but not ACS activate PPCI pathway.
We have removed the term "acute coronary syndrome (ACS)" to clarify the sentence.
4. The term cardiovascular magnetic resonance (CMR) is the convention in the society. Please check the whole manuscript as CMR was abbreviated for both 'cardiovascular magnetic resonance' and 'cardiovascular magnetic resonance imaging'.
We have amended all relevant text to state cardiovascular magnetic resonance (CMR) or CMR.
## Reviewer 2
This is an interesting concept on generating consensus on patients "activating the PPCI pathway". Unfortunately, the paper is neither as study, nor describing a novel method, nor a proper consensus statement.
We disagree with this comment. Our research represents a study; the Nominal Group Technique is one of the four well-established methodologies for formal consensus (organizing subjective judgments and synthesising them with the available evidence); the other three are the Delphi method, RAND/UCLA Appropriateness Method (RAM), and National Institutes of Health (NIH) consensus development conference methodology. Our study is novel in that it is the first to attempt to identify potentially important changes in management arising from the use of cardiac magnetic resonance (CMR) in a specific patient group. We are not sure what the reviewer means by a "proper consensus statement" but we met our study aim of identifying important changes in management (and the specific patient subgroups these changes in management relate to) resulting from the use of CMR in a specified patient population. We identified five subgroups of ACS patients who activate the PPCI pathway for whom there was consensus that CMR changes patient management in a clinically important way (i.e. expected to prevent adverse clinical outcomes in the long term).
1. The aims and goals of the "study" are not well described:
-After reading the title and the abstract I was still unclear what this is about. What patients exactly are looked at?
We have made it clear in the title and the abstract that our study population is "patients who activate the primary percutaneous coronary intervention (PPCI) pathway".
-What is a "formal consensus study".
See response to general comments from Reviewer 2 above.
-The paper is not a study, it is a consensus statement generated by an interactive process of (random) participants.
See response to general comments from Reviewer 2 above.
-The reason for the study was "to determine the feasibility of setting up a prospective registry", later it is stated, that "the key objective … was to define a primary composite outcome". Neither of the two reasons is addressed in the "study".
We have explained clearly in the introduction that a main objective of the feasibility study was to define a primary composite outcome based on clinically important changes in management, and that we used a formal consensus method to identify these clinically important changes in management.
We have made it clear in the conclusions and future research section that the "clinically important changes in management will now be used to design a composite primary outcome for an evaluation of the effectiveness and cost-effectiveness of CMR".
2. The authors need to decide, whether this is a study assessing the recognition of CMR in patients after PPCI within the larger cardiology community of the UK or whether this is a consensus statement prepared by iterative majority consensus vote of some UK cardiologists. The method used has been described before and this should be introduced in the introduction and discussed in the methods. (It partially is). It would also be good to compare this approach to the ACC/AHA approach for Appropriateness Criteria.
The study is a consensus study to identify clinically important changes in management resulting from CMR in a specific patient population. The formal consensus method based on the modified nominal group technique used in our study is presented and referenced in the introduction (page 6, second line paragraph 3). The separate components of the nominal group technique are fully explained in the introduction (page 6 & 7, end of paragraph 3). We have added a statement about the different consensus development methods (page 24, paragraph 1), but it is beyond the scope of this paper to compare the different approaches.
3. There is redundancy of the addenda and the main paper. It would help readability of the main paper if the statements receiving final support would be ordered as 1 -X and the ones without support X -12. The process, the first version of statements and the results from the first round could then be placed solely in the addendum.
We have removed the results of the first survey from the main paper and placed these in appendices to improve readability. However, we would prefer not to change the order of the statements (i.e. place those that received support first) as suggested by the reviewer; the number order relates specifically to how the statements were developed and refined from the beginning to the end of the consensus process. We have also chosen to keep comments relating to the statements from the cardiologists who participated in the formal consensus meeting in the paper. Although these comments were part of the first round of the survey and the face-to-face meeting, they add important contextual information that will help readers interpret the results.
. Some of the evidence is questionable and needs to be used with more care. E.g. the meta-analysis on CMR being superior to echo for the detection of thrombus includes several papers, which use
CMR LGE as the reference standard. Such evidence should not be used.We have included a sentence stating that "most of the included studies in this review did not use a pathological or surgical gold standard for the detection of LV thrombus" (see page 23, paragraph 2).Obtaining consensus and defining the utility of CMR has merits and the paper adds knowledge. I would suggest to focus on the consensus, do not call it a study and place the pathway for obtaining the consensus in the addendumWe have taken all comments on board, except for the comment about referring to this paper as a study (see response to general comments from Reviewer 2 above).Please let us know if you need further information or clarification. I look forward to hearing from you.VERSION 2 -REVIEWREVIEWERTaigang He St George's, University of London REVIEW RETURNED 02-Apr-2017GENERAL COMMENTSThe authors have responded to reviewers comments. I feel the manuscript is much improved after the revision and I would recommend for publication in your journal.
|
Neuroimaging risk factors for participation restriction after acute ischemic stroke: 1-year follow-up study
The aim of the present study was to determine the neuroimaging predictors of poor participation after acute ischemic stroke. A total of 443 patients who had acute ischemic stroke were assessed. At 1-year recovery, the Reintegration to Normal Living Index was used to assess participation restriction. We also assessed the Activities of Daily Living Scale and modified Rankin Scale (mRS) score. Brain MRI measurement included acute infarcts and preexisting abnormalities such as enlarged perivascular spaces, white matter lesions, ventricular-brain ratio, and medial temporal lobe atrophy (MTLA). The study included 324 men (73.1%) and 119 women (26.9%). In the univariate analysis, patients with poor participation after 1 year were older, more likely to be men, had higher National Institutes of Health Stroke Scale (NIHSS) score on admission, with more histories of hypertension and atrial fibrillation, larger infarct volume, more severely enlarged perivascular spaces and MTLA, and more severe periventricular hyperintensities and deep white matter hyperintensities. Patients with participation restriction also had poor activities of daily living (ADL) and mRS score. Multiple logistic regression showed that, in model 1, age, male gender, NIHSS score on admission, and ADL on follow-up were significant predictors of poor participation, accounting for 60.2% of the variance. In model 2, which included both clinical and MRI variables, male gender, NIHSS score on admission, ADL on follow-up, and MTLA were significant predictors of poor participation, accounting for 61.2% of the variance. Participation restriction was common after acute ischemic stroke despite good mRS score. Male gender, stroke severity, severity of ADL on follow-up, and MTLA may be predictors of poor participation. Trial registration number ChiCTR1800016665.
# Introduction
The lifetime risk of stroke in the Chinese population is the highest in the world (approximately 39.3%). [bib_ref] Global, regional, and country-specific lifetime risks of stroke, Feigin [/bib_ref] According to the International Classification of Functioning, Disability and Health, the consequences of a disease can be categorized into three different dimensions: body function impairments, activity limitations, and participation restrictions. Participation can be defined as 'the person's involvement in a life situation' and includes daily activities and social roles.Stroke survivors often experience participation restrictions in the chronic phase, despite having
## Significance of this study
What is already known about this subject? ► Stroke survivors often experience participation restriction in the chronic phase. ► Several factors contribute to participation restriction after stroke, including cognitive impairment, emotional deficits, stroke severity, functional dependency, and older age. ► The relationship between clinical neuroimaging factors and poor participation in patients who had acute ischemic stroke has rarely been studied.
What are the new findings? ► The present study determined the relationship between neuroimaging factors and participation restriction after acute ischemic stroke. ► Participation restriction was common in patients who had stroke despite having good modified Rankin Scale score. ► Male gender, stroke severity, medial temporal lobe atrophy, and severity of activities of daily living on follow-up were important predictors of poor participation at 1 year after the index stroke.
How might these results change the focus of research or clinical practice? ► The advantage of our study was that the relationship between participation and comprehensive MRI variables, which included acute infarct and pre-existing brain abnormalities, was assessed. ► Given the very few studies examining the relationship between clinical neuroimaging factors and poor participation, our findings provide important new knowledge on participation restriction after ischemic stroke.
## Original research
favorable basic activities of daily living (ADL). [bib_ref] Participation in the chronic phase of stroke, Van Der Zee [/bib_ref] A Swedish cohort study found that a patient who had a stroke who experienced participation restriction did not want to do everyday occupations, 4 while a Netherlandish study further found that decline in participation might induce incomplete social activities even at 3 years poststroke. [bib_ref] Association between satisfaction and participation in everyday occupations after stroke, Bergström [/bib_ref] Social participation, equally important to cognition, was a strong determinant of quality of life among older adults who had a stroke. [bib_ref] Social activity one and three years post-stroke, Jansen [/bib_ref] Several factors, including cognitive impairment, 7 emotional deficits, 8 stroke severity, 9 functional dependency, [bib_ref] Level and predictors of participation in patients with stroke undergoing inpatient rehabilitation, Yang [/bib_ref] and older age, 11 might contribute to participation restrictions after stroke. However, previous studies largely focused on the relationship of demographic clinical factors with participation restriction. By contrast, the relationship between MRI variables in acute ischemic stroke and participation restriction remains unclear. Thus, the aim of the present study was to determine the relationship between neuroimaging factors assessed by MRI and participation restriction.
# Materials and methods
The study was registered at http://www.chictr.org.cn/ index.aspx. Valid written consent was obtained from all participants.
## Participants and setting
Patients who had acute ischemic stroke admitted to Division I, Department of Neurology, Dongguan People's Hospital from January 1, 2017 to December 30, 2018 were screened. The inclusion criteria were as follows: (1) age >18 years; (2) first or recurrent acute ischemic stroke confirmed by MRI 12 and admitted within 7 days after onset;
(3) had complete brain MRI examination; and (4) modified Rankin Scale (mRS) score <3 points at discharge. The exclusion criteria were as follows: (1) transient ischemic attack or hemorrhagic stroke; (2) incomplete or no brain MRI data; (3) mRS score ≥3 points at discharge, including death (mRS=6); (4) patients with severe comorbidities (eg, malignant tumor, etc); (5) patients who were unable to complete the assessment at discharge or follow-up due to severe hearing disabilities, visual disabilities, language disorders, or cognitive impairment; and (6) patients who refused to provide signed consent.
## Demographic data collection
Information on demographic and clinical variables included age, sex, history of stroke, and vascular risk factors. Severity of stroke was assessed with the National Institutes of Health Stroke Scale (NIHSS). [bib_ref] Measurements of acute cerebral infarction: a clinical examination scale, Brott [/bib_ref] Subtype of ischemic stroke was judged according to the Trial of Org 10172 in Acute Stroke Treatment subtype system. 14
## Follow-up of patients
All assessments at follow-up were completed by the patients. Participation restriction was measured by the Chinese version of the Reintegration to Normal Living Index (RNLI). [bib_ref] The psychometric properties of the Chinese versionreintegration to normal living index (C-RNLI)..., Liu [/bib_ref] The Chinese version of the RNLI was translated from the RNLI and is an easy-to-understand version with simple words and structures, which ensure patients can finish it, even for those with little or no education. [bib_ref] Assessment of global function: the reintegration to normal living index, Wood-Dauphinee [/bib_ref] The RNLI is a self-report questionnaire used to quantitatively assess reintegration to normal functioning after stroke. [bib_ref] Good outcome" isn't good enough: cognitive impairment, depressive symptoms, and social restrictions..., Kapoor [/bib_ref] We assessed participation status at 1 year after the index stroke with a good mRS score at discharge. The components of RNLI include 11 questions measuring the different levels of ability for mobility, self-care ability, daily activities, recreational activities, general coping skills, family roles, social activities, personal relationships, and presentation of self to others. [bib_ref] Assessment of global function: the reintegration to normal living index, Wood-Dauphinee [/bib_ref] Each item is scored from 1 (minimal reintegration) to 10 (complete reintegration). The total score ranges from 11 to 110. A lower RNLI score indicates more severe participation restriction in normal living. We defined poor participation as a score <P25 of the IQR of the RNLI.
We also assessed functional status and disability at 1 year according to the Lawton Activities of Daily Living Scale [bib_ref] Assessment of older people: self-maintaining and instrumental activities of daily living, Lawton [/bib_ref] and mRS.
## Neuroimaging data
Neuroimaging data were generated by an MRI examination. All patients were scanned on a 3.0 T system (Sonata; Siemens Medical, Erlangen, Germany) 19 within 7 days after admission. The following sequences were included: T1-weighted imaging, T2-weighted imaging, and diffusionweighted imaging. A neurologist (H-HZ) who was blinded to patients' clinical information reviewed the MRI data. We assessed both the acute infarcts and pre-existing abnormalities, which included enlarged perivascular spaces (EPVS), white matter lesions (WMLs), global brain atrophy, and medial temporal lobe atrophy (MTLA). 19
# Statistical analysis
Statistical analyses were performed using SPSS for Windows (V.24.00). Descriptive data are presented as proportion, mean, or median, as appropriate. A univariate analysis comparing the putative risk factors between patients with poor and favorable participation was performed. Variables with p<0.05 in the univariate analysis were included in further binary multivariate logistic regressions. The significance level was set at p<0.05 (two-sided).
# Results
One thousand and fifty-four patients were admitted during the study. Patient selection is described in the flow chart (figure 1). Finally, 443 patients were included in the final analysis. Intrarater reliability (kappa) tests were performed on 10 patients who had a stroke by the same MRI rater. The intrarater agreement for the MRI measurements was good to excellent: volume of infarction intraclass coefficient (ICC) =0.88, EPVS=0.81, WML=0.83, ventricular-brain ratio ICC=0.85, and MTLA=0.86.
There were no differences in gender between the excluded and included patients (men, 72.7% vs 69.1%, respectively; p=0.201), while there were significant differences in age (59.61±11.84 years vs 62.59±14.3 years, respectively; p<0.001) and NIHSS score on admission (2 (1-3) vs 4 (2-10), respectively; p<0.001).
The study sample consisted of 324 men (73.1%) and 119 women (26.9%), with a mean age of 59.61 years (range 26-88). The median NIHSS score at the time of first screening was 2 (range 0-17). The median RNLI score on follow-up was 103, ranging from 11 to 110 [fig_ref] Table 1: Demographic and clinical characteristics of the study patients activities of daily living [/fig_ref].
## Univariate analysis of poor participation
In the univariate analysis, patients with poor participation at 1 year were older, more likely to be men, with higher NIHSS score at admission, and with more histories of hypertension and atrial fibrillation. Further, these patients had a significant larger infarct volume, more severe EPVS and MTLA, and more severe periventricular hyperintensities (PVH) and deep white matter hyperintensities (DWMH). Patients with participation restriction also had poor ADL and mRS score [fig_ref] Table 2: Risk factors for poor participation at 1-year follow-up in the univariate analysis [/fig_ref].
## Multiple regressions for poor participation
We conducted two multiple stepwise regression models and the results are presented in table 3. In model 1, after using poor participation as the dependent variable, age, sex, NIHSS score on admission, hypertension, atrial fibrillation, stroke subtype, and ADL at follow-up were entered into the model. mRS score on follow-up was not included because it was significantly correlated with ADL (r=0.725). Age, male gender, NIHSS score on admission, and ADL at follow-up were significant predictors of poor participation at 1 year, accounting for 60.2% of the variance.
Model 2 included both the clinical and neuroimaging variables. PVH and EPVS in the centrum semiovale were not included in the model as they were highly correlated with DWMH score (r=0.695) and EPVS in the basal ganglia (r=0.608), respectively. The results showed male gender, NIHSS score at admission, ADL at follow-up, and MTLA were significant predictors of poor participation at 1 year, accounting for 61.2% of the variance. We further analyzed the interaction of age with MTLA; the OR and 95% CI of the interaction analysis of age and MTLA was 0.996 (0.981 to 1.059), showing that it was not a significant predictor of poor participation.
# Discussion
As well as reduced functional outcomes and physical disability, stroke affects multiple other levels of function. In this prospective observational study, we found that many
## Original research
patients who had a stroke were unable to socially reintegrate, despite having a good mRS score. Further, male gender, stroke severity, MTLA, and severity of ADL were important predictors of poor participation at 1 year after the index stroke. Given the very few studies examining the relationship between clinical neuroimaging factors and poor participation, these data provide important new knowledge on participation restriction after ischemic stroke. The effect of sex on social participation is controversial. Our findings suggest that men show a greater propensity for participation restriction poststroke than women. In support, women were reported to have a higher subjective well-being than men after stroke. [bib_ref] Correlates of subjective well-being in stroke patients, Wyller [/bib_ref] Further, non-white men were shown to have lower participation after stroke. [bib_ref] Quality of life during and after inpatient stroke rehabilitation, Hopman [/bib_ref] Thus, discussions on the roles of men at home and the difficulties that they may encounter in maintaining these roles after discharge should be considered as an important factor in the part of male groups.
Our data suggest that stroke severity at admission is an important risk factor for participation restriction poststroke, as previously reported. However, those studies recruited patients with a wide range of stroke-related disabilities, resulting in a heterogeneous study sample. By contrast, our study included a more homogenous population of patients who had a stroke with mild, residual disabilities. We found that the more severe the neurological deficits, the worse the participation, despite a favorable recovery. Thus, although patients may recover in terms of physical disability, they may also experience dysfunction in reintegration to normal living.
Comparing with global atrophy, MTLA may be a significant predictor of poor participation in the present study. A previous study showed that the medial temporal lobe, but not global atrophy, is a region that is more susceptible to ischemia and may predict cognitive decline in stroke survivors. [bib_ref] Medial temporal atrophy rather than white matter hyperintensities predict cognitive decline in..., Firbank [/bib_ref] We suggest that cognitive impairment may be an important mediator between MTLA and participation restriction.
The present study found that poorer ADL correlated with poor participation to normal life. In a comprehensive study by Mayo et al [bib_ref] Activity, participation, and quality of life 6 months poststroke, Mayo [/bib_ref] , restriction in ADL was a risk for social isolation, with a further negative implication on patients' health. Complete ADL requires not only dependent basal ADL, but also memory function and satisfactory executive function. [bib_ref] Visuospatial function is a significant contributor to functional status in patients with..., Fukui [/bib_ref] Impairment in ADL may induce multiple aspects of dysfunction, including memory and executive function, which mediate participation restriction. Surprisingly, after adjusted MTLA, our study found a conflicting result compared with previous studies, which have reported an increased risk associated with older age. Our contrasting findings may relate to the strong relationship between age and MTLA. In model 1, older age was an independent risk factor for poor participation, while in model 2, which included both the clinical and neuroimaging variables, age was not significantly correlated with poor participation. The OR and 95% CI of the interaction analysis of age and MTLA was 0.996 (0.981 to 1.059), which suggested that the association of age and participation restriction reported in other studies may be partly caused by more severe cerebral degeneration. Thus, our data suggest that cerebral atrophy should be considered when studying participation in stroke.
There were several advantages to our study. First, all patients had mild ischemic stroke. In addition, we assessed both acute infarct and comprehensive pre-existing brain abnormalities. There were also some limitations to our study. First, we did not evaluate the cognitive status of our patients at admission or follow-up, which may be correlated with participation restriction. Second, one of the major inclusion criteria is that patients should have complete MRI data. Of the 1054 consecutive patients, 186 were excluded due to lack of MRI or incomplete MRI data. Additionally, our finding is limited to those who had mild-moderate ischemic stroke, who could cooperate to finish the MRI. Third, there were 33 patients who had previous stroke included in the study. We did not assess their prestroke mRS score. However, we had excluded those with significant disability at discharge (mRS ≧3 points), implicating those with apparent disability before the index stroke were not included in this study.
In conclusion, we found that participation restriction was common after the index mild ischemic stroke. Male gender, stroke severity, MTLA, and severity of ADL may be predictors of poor participation.
[fig] Figure 1: Flow chart of participants. mRS, modified Rankin Scale. [/fig]
[table] Table 1: Demographic and clinical characteristics of the study patients activities of daily living; BG-EPVS, enlarged perivascular spaces in the basal ganglia; CS-EPVS, enlarged perivascular spaces in the centrum semiovale; DWMH, deep white matter hyperintensities; mRS, modified Rankin Scale; MTLA, medial temporal lobe atrophy; NIHSS, National Institutes of Health Stroke Scale; PVH, periventricular hyperintensities; RNLI, Reintegration to Normal Living Index; VBR, ventricular-brain ratio. [/table]
[table] Table 2: Risk factors for poor participation at 1-year follow-up in the univariate analysis [/table]
[table] Table 3: Multivariate logistic regression of risk factors for poor participation ADL, activities of daily living; BG-EPVS, enlarged perivascular spaces in the basal ganglia; DWMH, deep white matter hyperintensities; MTLA, medial temporal lobe atrophy; NIHSS, National Institutes of Health Stroke Scale. [/table]
|
Urothelial carcinoma of the prostate with raised β-hCG levels: a case report
Background: Trophoblastic differentiation in primary urothelial carcinoma of the prostate is extremely rare. An increased level of β-subunit human chorionic gonadotropin in serum in urothelial carcinoma is detected in approximately 30% of cases. To our knowledge, increased concentration of β-subunit human chorionic gonadotropin in serum in prostatic urothelial carcinoma has never been reported and its clinical significance is not evaluated yet.Case report: Here we present the case of a 67-year-old European patient who was admitted to the hospital with hematuria, dysuria, and enlarged painful testis. Ultrasonographic examination of the testis did not reveal any focal lesion. Magnetic resonance imaging of the pelvis showed a tumor of 62 mm diameter mainly located in the posterior part of the prostatic gland. A pathological examination from cystoscopy biopsy allowed us to set the diagnosis of high-grade invasive urothelial carcinoma with trophoblastic differentiation. The patient received neoadjuvant treatment. Nonetheless, after a short period of disease stabilization, he developed progression and brain metastasis. He died 9 months after diagnosis. During the disease course, his β-human chorionic gonadotropin level was measured repeatedly and analyzed in relation to disease progression. The level of serum β-human chorionic gonadotropin corresponded with the therapy response; it was at its lowest during stabilization and the highest in the metastatic stage.Conclusion:Our case study provides the first report of urothelial cancer of the prostate, with a concomitant increase of β-subunit human chorionic gonadotropin level with testis enlargement. Besides its rarity, it constitutes an interesting observation of increasing β-subunit human chorionic gonadotropin concentration with concomitant disease progression.
# Background
Primary urothelial carcinoma of the prostate is a rare malignancy (incidence 1-5%) with an aggressive course and poor prognosis [bib_ref] Primary urothelial carcinoma of the prostate: a rare case report, Zhou [/bib_ref]. Overall survival does not exceed 2 years as it tends to recur and metastasize quickly [bib_ref] Primary urothelial carcinoma of the prostate: a rare case report, Zhou [/bib_ref]. As trophoblastic transformation in prostatic urothelial carcinoma is extremely rare, the prognostic significance of β-subunit human chorionic gonadotropin (β-hCG) has not been investigated yet [bib_ref] Urothelial carcinomas with trophoblastic differentiation, including choriocarcinoma: clinicopathologic series of 16 cases, Przybycin [/bib_ref]. Trophoblastic differentiation results in the presence of syncytiotrophoblastic cells and in some cases in increased β-hCG levels in serum [bib_ref] Beta-hCG expression by bladder cancers, Iles [/bib_ref]. The presence of β-hCG in serum or urine has been already reported in many other malignancies, including lung, breast, stomach, bladder, and prostate adenocarcinoma [bib_ref] Beta hCG as a prognostic marker in adenocarcinoma of the prostate, Sheaff [/bib_ref] [bib_ref] Expression of human chorionic gonadotropin beta in gastric carcinoma: a retrospective immunohistochemical..., Murhekar [/bib_ref] [bib_ref] Secretion of hCG/beta-hCG by squamous cell carcinoma of the lung in a..., Yoshida [/bib_ref] [bib_ref] Immunohistochemical expression of subunit beta HCG in breast cancer, Agnantis [/bib_ref] [bib_ref] Human chorionic gonadotropin expression in lung, breast, and renal carcinomas, Kuida [/bib_ref]. It is perceived as a marker of progression as it promotes growth and invasion of tumor cells while correlating with high-grade and advanced-stage disease [bib_ref] Serum total hCGβ level is an independent prognostic factor in transitional cell..., Douglas [/bib_ref] [bib_ref] The increase in bladder carcinoma cell population induced by the free beta..., Butler [/bib_ref] [bib_ref] Human chorionic gonadotropin in cancer, Stenman [/bib_ref] [bib_ref] Human urothelial carcinomas-a typical disease of the aged: the clinical utility of..., Iles [/bib_ref]. Owing to insufficient evidence of β-hCG utility in urothelial cancer, it is nowadays not widely used in clinical practice [bib_ref] Human urothelial carcinomas-a typical disease of the aged: the clinical utility of..., Iles [/bib_ref]. Here we present the case of a 67-year-old male patient with urothelial carcinoma of the prostate, non-neoplastic testis enlargement, and increased concentration of β-hCG with fluctuations during the course of the disease.
## Case presentation
A 67-year-old European man was admitted to the Department of Urology in December 2019 with major complaints regarding recurrent hematuria, dysuria, pain, and enlargement of the left testicle, pain in the lumbar and sacral section of the spine, and hyperhidrosis. The patient also reported a recent loss of weight: around 3 kg down from 97 kg within the last month. A medical interview revealed occupational exposure to asbestos since the patient had been working in a factory producing asbestos seals for several years. His body mass index was 37.8 kg/ m 2 , and body surface area was 2.09 m 2 . Physical examination revealed lower limb edema as well as painful and enlarged left testis of regular shape and firm consistency, without any palpable focal lesion. Owing to enlarged testis, ultrasonography imaging (USG) and blood tests were performed. USG revealed diffuse testis enlargement but did not reveal any focal malignancies originating from the testis. Results of the blood test showed an increased level of β-hCG (644.4 lU/l). The other tumor markers, that is, lactate dehydrogenase (LDH) α-fetoprotein (AFP), and prostate-specific antigen (PSA) remained within the normal range (230 U/ml, 1.58 U/ml, and 0.874 ng/ml, respectively). Surprisingly, the following magnetic resonance imaging (MRI) revealed a mass up to 62 mm in the posterior part of the prostatic gland, infiltrating seminal vesicles and anterior rectum wall together with bilaterally enlarged iliac lymph nodes. Additionally, computed tomography (CT) confirmed metastasis in pelvic lymph nodes. Cystoscopy was performed, and a biopsy specimen from tumor mass from prostate and bladder was taken. Histopathological examination of the specimens obtained in the cystoscopy revealed prostate and muscle of the bladder infiltration consisting of highgrade invasive urothelial carcinoma with trophoblastic differentiation and the presence of large syncytiotrophoblastic cells. Urothelial carcinoma cells were positive for cytokeratin 20 (CK20), GATA3, p63, and CKHMW but negative for PSA. Some of these large syncytiotrophoblastic cells inside the tumor showed positivity for β-hCG [fig_ref] Figure 1: Histopathological image of specimen from cystoscopy [/fig_ref]. Keeping in mind that urothelial cancer most often originates in the bladder, the patient was diagnosed with urothelial bladder cancer. However, reassessment of the radiographic imaging led to the conclusion that the tumor was predominantly located in the prostate, which suggested the prostate as the primary origin and established the diagnosis of urothelial prostate cancer with secondary bladder involvement. However, the treatment of urothelial cancer remains the same regardless of primary origin. As subsequent radiography imaging of the chest did not reveal any focal changes, the presence of metastasis in the chest was excluded. In March 2020, the patient started neoadjuvant chemotherapy with intravenous cisplatin 75 mg/m 2 on the first day and intravenous gemcitabine 1000 mg/m 2 on the 1st, 8th, and 15th repeating cycle every 4 weeks (cisplatin-gemcitabine protocol, PG). His symptoms improved significantly, and the β-hCG concentration decreased (205.2 mIU/ml) within the first 2 months of therapy. However, after two cycles, β-hCG level slightly increased (504.8 mIU/ml) and hyperhidrosis recurred. After four cycles, computed tomography (CT) revealed that local treatment could not be implemented owing to rectum infiltration. PG chemotherapy was continued up to six cycles. Four months later, after completion of PG chemotherapy, MRI showed disease progression. Pelvic tumor enlarged to 90 mm in diameter, and pelvic lymph nodes enlarged to maximum 21 mm in short-axis diameter [fig_ref] Figure 2: Imaging studies of patient's urothelial tumor after progression on computed tomography [/fig_ref]. Simultaneously, β-hCG concentration increased to 9446 mIU/ml. In October 2020, palliative chemotherapy was implemented (paclitaxel 80 mg/m 2 once a week). After the fifth cycle of paclitaxel, the patient presented with polyneuropathy, so gabapentin was administered but withdrawn because of dizziness. However, the dizziness aggravated, so a CT scan of the brain was performed. It revealed small, diffuse metastases up to 8 mm in the pons. At that time, the β-hCG concentration reached the highest value of 31,163 mIU/ml. Palliative radiotherapy was planned, but owing to rapid deterioration of general condition (PS ¾), the patient did not manage to start the treatment. Eventually, the patient died in November 2020, 2 weeks after diagnosis of brain metastasis, because of disease progression.
# Discussion and conclusions
β-hCG is routinely used in testicular cancer in multiple stages of the disease. It supports diagnostics, correlates with stage and patients' risk, and helps to monitor response to the treatment or predicts disease relapse [bib_ref] EAU guidelines on testicular cancer: 2011 update, Albers [/bib_ref] [bib_ref] American Society of Clinical Oncology Clinical Practice Guideline on uses of serum..., Gilligan [/bib_ref]. In urothelial cancer of the bladder, it showed to have a prognostic value. Douglas et al. utilized β-hCG serum levels in patients with urothelial transitional cell carcinoma to predict patients' prognosis. They showed that total serum level of β-hCG is an independent prognostic factor in patients receiving chemotherapy for urothelial transitional cell carcinoma [bib_ref] Serum total hCGβ level is an independent prognostic factor in transitional cell..., Douglas [/bib_ref]. This observation was further confirmed in other studies, which showed that expression of β-hCG in bladder urothelial cancer is associated with radioresistance, high grade, and muscle invasion [bib_ref] Urothelial carcinoma of the bladder with trophoblastic differentiation: a case report, Tuna [/bib_ref] [bib_ref] The expression of beta human chorionic gonadotrophin (β-HCG) in human urothelial carcinoma, Venyo [/bib_ref] [bib_ref] Is beta-human chorionic gonadotrophin production by transitional cell carcinoma of the bladder..., Moutzouris [/bib_ref]. Moreover, Malkhasyan et al. reported a case where the level of β-hCG corresponded with tumor progression and treatment response [bib_ref] The use of serum hCG as a marker of tumor progression and..., Malkhasyan [/bib_ref]. High levels of β-hCG were shown to be associated with high grade, resistance to chemotherapy, and poor survival in other malignancies as well, such as in lung cancer and colorectal cancer [bib_ref] Frequency and clinical significance of beta-subunit human chorionic gonadotropin expression in non-small..., Szturmowicz [/bib_ref] [bib_ref] Serum HCG beta, CA 72-4 and CEA are independent prognostic factors in..., Louhimo [/bib_ref]. It should be remembered that an elevated level of β-hCG may also affect the secondary hyperplasia of organs. It is reported to influence the growth of breasts causing gynecomastia and enlargement of testicles and penis [bib_ref] Penile growth in response to human chorionic gonadotropin (HCG) treatment in patients..., Kim [/bib_ref] [bib_ref] Gynecomastia attributable to human chorionic gonadotropin-secreting giant cell carcinoma of lung, Yaturu [/bib_ref] [bib_ref] HCG-secreting malignancies-diagnostic pitfalls, Michalski [/bib_ref]. Here we present a unique case of testis enlargement caused by β-hCG produced by urothelial prostate cancer. Increased levels of β-hCG and testicle enlargement lead to the suspicion of a germ cell tumor malignancy, which was excluded on USG imaging. Presumably, the β-hCG elevation caused a secondary growth of the testicles, which could mislead the diagnosis. Moreover, to our knowledge, the β-hCG concentration of 31,163 mUl/ml is the highest recorded so far. Owing to multiple measurements of serum β-hCG concentration during disease, our case presents the dynamics of β-hCG level and its further association with disease progression. At the beginning of treatment, the patient presented a good response with the stabilization and decrease of β-hCG to 205.2 mIU/ml. One month after stabilization, the level increased to 504.8 mIU/ml, and 5 months later, disease progression was observed. Enlargement of the tumor to the size of 90 mm and brain metastasis resulted in the increase of β-hCG to 9446 mIU/ml and 31,163 mIU/ml, respectively. Fluctuation of β-hCG with disease progression is shown in [fig_ref] Figure 3: Relation between concentration of β-hCG and course of the disease [/fig_ref].
This case report is in our opinion a valuable example of a nonobvious presentation of high-grade invasive urothelial carcinoma of the prostate with trophoblastic differentiation and increased β-hCG serum concentration and secondary testis enlargement. Moreover, repeatedly measured β-hCG provided an occasion to observe how its serum concentration correlates with gradual disease progression. To our knowledge, urothelial prostate cancer has been reported in the literature only once, but this is the first report of such clinical presentation with increased concentration of β-hCG and testis enlargement [bib_ref] Urothelial carcinomas with trophoblastic differentiation, including choriocarcinoma: clinicopathologic series of 16 cases, Przybycin [/bib_ref]. We hope that our case will lead to the widening of the clinical perception of urothelial prostate cancer and the extension of the differential diagnosis of neoplasms with a high β-hCG level to include urothelial prostate cancer with trophoblastic differentiation.
# Acknowledgments
We are grateful to Antoni Siejka and Ireneusz Sołek for proofreading and grammar check.
[fig] Figure 1: Histopathological image of specimen from cystoscopy. A Gross picture of invasive high-grade urothelial carcinoma [hematoxylin and eosin staining (H&E); 10× magnification]. B Syncytiotrophoblastic cells in the urothelial carcinoma (H&E; 400× magnification). C Positive immunohistochemical (IHC) staining for cytokeratin 20 (CK20, DAKO) in nest of carcinoma cells (50× magnification). D Positive IHC reaction for β-human chorionic gonadotropin in large syncytiotrophoblastic cells inside the urothelial carcinoma (anti-β-hCG, DAKO, 100× magnification) [/fig]
[fig] Figure 2: Imaging studies of patient's urothelial tumor after progression on computed tomography (CT) (A-C) and magnetic resonance imaging (MRI) (D). A Longitudinal dimension of the tumor. B Short-axis diameter of pelvic lymph node. C Transverse dimension of tumor. D MRI of patient's tumor in pelvis [/fig]
[fig] Figure 3: Relation between concentration of β-hCG and course of the disease [/fig]
|
Metabolic Signaling into Chromatin Modifications in the Regulation of Gene Expression
The regulation of cellular metabolism is coordinated through a tissue cross-talk by hormonal control. This leads to the establishment of specific transcriptional gene programs which adapt to environmental stimuli. On the other hand, recent advances suggest that metabolic pathways could directly signal into chromatin modifications and impact on specific gene programs. The key metabolites acetyl-CoA or S-adenosyl-methionine (SAM) are examples of important metabolic hubs which play in addition a role in chromatin acetylation and methylation. In this review, we will discuss how intermediary metabolism impacts on transcription regulation and the epigenome with a particular focus in metabolic disorders. Int. J. Mol. Sci. 2018, 19, 4108 2 of 15 the TCA cycle for energy production. The slicing of fatty acids through beta-oxidation produces acetyl-CoA which also enters TCA cycle. In addition, protein degradation can also feed TCA cycle through acetyl-CoA production from ketogenic amino acids. Acetyl-CoA has versatile functions in different crucial pathways. For example, lipids, cholesterol and steroid synthesis are derived from acetyl-CoA utilization as a carbon source. Moreover, ketone bodies are derived from acetyl-CoA. It also plays an important role in processes such as protein glycosylation which will not be discussed in detail in this review.One of the most abundant chromatin modifications that directly regulates transcriptional activity is histone acetylation[2]. The fact that acetyl-CoA is the only donor for histone acetylation connects the cellular metabolic status with gene control[3]. Generally, when energy status is high, acetyl-CoA levels are elevated and this correlates with global histone acetylation[4,5]. On the contrary, when acetyl-CoA drops, histone acetylation decreases[4,5]. It is not fully understood whether histone acetylation uses particular subcellular pools of acetyl-CoA. Mitochondria contain a large pool of acetyl-CoA, however, its membrane is impermeable for acetyl-CoA. Mitochondrial-produced acetyl-CoA can contribute to cytosolic acetyl-CoA thanks to a shuttle system mediated by ATP Citrate Lyase (ACLY) in the cytoplasm. After acetyl-CoA enters the TCA cycle, citrate is made and transported out to the cytoplasm from the mitochondria(Figure 1). ACLY then catalyzes the conversion of citrate to oxaloacetate by transferring 2 carbons to Coenzyme A leading to acetyl-CoA formation. Wellen et al. elegantly showed that the increase of histone acetylation in response to growth factors and glucose availability is dependent on ACLY[6].Although the correlation between high energetic status (reflected by high levels of cellular acetyl-CoA) and histone acetylation is well accepted [7], its relevance in specific biological contexts or tissues is not well understood. Whether for example fluctuations of acetyl-CoA levels could lead to differential acetylation on specific loci is not known. In this respect, the activity of Histone Acetyltransferases (HATs) could play a major role. Since the discovery of the first HAT, Gcn5 in yeast, [8] a large number of studies in different fields have identified multiple regulatory functions of HATs not only for histone but also for non-histone protein acetylation[9]. HATs play an essential role in transcriptional activation by forming protein complexes at gene promoters and catalyzing lysine histone acetylation. However, contrary to phosphorylation, the kinetics of histone acetylation largely depends on the acetyl-CoA concentration, particularly acetyl-CoA versus free CoA ratio[3]. While kinases function with saturating ATP concentrations, HATs activity depends on acetyl-CoA concentration. This concentration ranges between 3-30 µM in both yeast and mammalian cells[7]and most HATs have a dissociation constant (Kd) in the low micromolar or high nanomolar range. Therefore, the abundance of acetyl-CoA influences HAT activity, further suggesting a key connection between metabolic status and gene regulation. In addition to concentration, the subcellular localization of acetyl-CoA plays a regulatory role since it is not permeable to mitochondria. This fact leads to the existence of differentially regulated nuclear versus cytosolic acetyl-CoA pools that may indirectly impact HAT activity. It is not known whether additional differential compartmentalization or fluxes of acetyl-CoA could exist within cytosolic and nuclear subcellular compartments. Although acetyl-CoA is freely diffusible between cytoplasm and nucleus, there might be particular physical constraints due to for example high level of chromatin compaction in some nuclear areas which could lead to acetyl-CoA subnuclear compartmentalization[7]. A similar phenomenon has been recently identified for NAD + , which has been found to be synthesized differently in subnuclear compartments and concomitantly regulate gene transcription[10]. ACLY is also localized in the nucleus, however, it is not known how its nuclear import is regulated. Recently, DNA damage has been shown to induce ACLY phosphorylation which leads to the recruitment of BRCA1 for homologous recombination in DNA repair[11]. Hence, the translocalization of ACLY may be regulated through post-translational modifications.
# Introduction
An essential adaptation of most living organisms is to sense and respond to nutrient availability. A classic example of this interaction is how bacteria regulate expression of metabolic genes in function of the presence of specific nutrients in the environment. These processes were discovered by Nobel laureates Jacob and Monod in the sixties [bib_ref] Genetic regulatory mechanisms in the synthesis of proteins, Jacob [/bib_ref]. Mammals have evolved sophisticated ways to respond to nutrients which involve multiple processes and adaptive mechanisms including hormonal inter-organ communication, fasting/feeding and circadian cycles, storage of energy and the possibility to alternate fuel usage. Despite the profound differences between unicellular and pluricellular life in respect to metabolic regulation, it is tempting to speculate that some basic processes could have been conserved or converged during evolution. At the cellular level, multiple metabolic pathways and metabolites are intricately connected to the regulation of gene expression. A paradigmatic example of such regulation is the link of acetyl-CoA between multiple metabolic pathways and its function as donor of acetyl groups for histone acetylation, a crucial chromatin modification involved in active gene transcription.
In this review, we aim at studying how different metabolites impact on gene regulation by directly playing a donor or cofactor function of key chromatin modifying enzymes. In addition, we focus on how these processes could be regulated in different physiological contexts particularly in metabolic disorders. The interaction with diets and potential therapeutic opportunities will also be discussed.
## Interaction of metabolites with chromatin modifications
## Acetyl-coa and regulation of acetylation
Acetyl-CoA is a central two-carbon carrier which is produced by different fuel and intermediary sources and connects different metabolic pathways. Carbohydrate catabolism can lead to acetyl-CoA production by the pyruvate dehydrogenase complex (PDH) in the mitochondria, which then feeds produced acetyl-CoA can contribute to cytosolic acetyl-CoA thanks to a shuttle system mediated by ATP Citrate Lyase (ACLY) in the cytoplasm. After acetyl-CoA enters the TCA cycle, citrate is made and transported out to the cytoplasm from the mitochondria [fig_ref] Figure 1: Figure 1 [/fig_ref]. ACLY then catalyzes the conversion of citrate to oxaloacetate by transferring 2 carbons to Coenzyme A leading to acetyl-CoA formation. showed that the increase of histone acetylation in response to growth factors and glucose availability is dependent on ACLY [bib_ref] ATP-citrate lyase links cellular metabolism to histone acetylation, Wellen [/bib_ref]. Interaction between metabolism and histone acetylation and DNA/histone methylation. Different nutrient substrates including glucose, fatty acids, amino acids and acetate lead to production of intermediary metabolites which play a role in protein acetylation. Acetyl-CoA derived from glucose, fatty acid or amino acid metabolism is the substrate for histone acetylation after conversion into citrate by TCA cycle and back to Acetyl-CoA in the cytoplasm by ACLY. Acetate is also a source of acetyl-CoA which leads to histone acetylation. Histone and DNA methylation depends on the dietary methionine which enters a cycle for conversion into SAM which is used as a donor of the methyl group. This leads to formation of SAH which is recycled back to methionine through [fig_ref] Figure 1: Figure 1 [/fig_ref]. Interaction between metabolism and histone acetylation and DNA/histone methylation. Different nutrient substrates including glucose, fatty acids, amino acids and acetate lead to production of intermediary metabolites which play a role in protein acetylation. Acetyl-CoA derived from glucose, fatty acid or amino acid metabolism is the substrate for histone acetylation after conversion into citrate by TCA cycle and back to Acetyl-CoA in the cytoplasm by ACLY. Acetate is also a source of acetyl-CoA which leads to histone acetylation. Histone and DNA methylation depends on the dietary methionine which enters a cycle for conversion into SAM which is used as a donor of the methyl group. This leads to formation of SAH which is recycled back to methionine through methylation of homocysteine. PDC: Pyruvate Dehydrogenase Complex; ACLY: ATP-dependent Citrate Lyase; SAM: S-Adenosylmethionine; SAH: S-Adenosyl-Homocysteine; DNMTs, DNA N-Methyltransferase; MATs: Methionine Adenosyltransferase. Dashed arrows: multiple-step metabolic pathway; solid arrows: one-step metabolic reaction.
## Nad + and regulation of sirtuin-dependent deacetylation
The counterpart of histone acetylation is the removal of acetyl groups by lysine deacetylases (KDACs). These enzymes are divided in two large families: the zinc-dependent deacetylases classically known as histone deacetylases (HDACs), and the family of nicotinamide dinucleotide (NAD + ) dependent deacetylases; the sirtuins [bib_ref] NAD + and sirtuins in aging and disease, Imai [/bib_ref] [bib_ref] The many roles of histone deacetylases in development and physiology: Implications for..., Haberland [/bib_ref] [bib_ref] Regulation of chromatin and gene expression by metabolic enzymes and metabolites, Li [/bib_ref] [bib_ref] Erasers of histone acetylation: The histone deacetylase enzymes, Seto [/bib_ref]. HDACs usually associate with co-repressors complexes including SIN3A, SMRT or NCOR [bib_ref] Erasers of histone acetylation: The histone deacetylase enzymes, Seto [/bib_ref]. Among their wide variety of functions, they also regulate systemic metabolism by the deacetylation of key metabolic transcription factors [bib_ref] Protein acetylation in metabolism-Metabolites and cofactors, Menzies [/bib_ref]. There are 7 sirtuins (Sirtuin 1 to 7) that also have non-histone targets. These include transcription factors and metabolic enzymes. The discovery of Sir2 in yeast as a regulator of lifespan extension during caloric restriction conditions exemplified the role of sirtuins in metabolic homeostasis [bib_ref] NAD + and sirtuins in aging and disease, Imai [/bib_ref]. The key connection of sirtuin's function with intermediary metabolism is its dependence on NAD + . The catalytic deacetylation of substrates by sirtuins uses NAD + as the acceptor of the acetyl group. This reaction leads to the dissociation of NAD + into nicotinamide (NAM) and 2 OAADPr. NAD + is a small molecule ubiquitously present in energy metabolism that serves as a coenzyme of oxidoreductase enzymes in the transfer of electrons between metabolites and NADH. Because the TCA cycle and the electron transport chain (ETC) require both NAD + and NADH, an optimal NAD + /NADH ratio is required for an efficient metabolism. The activity of sirtuins consumes NAD + which reduces its concentration and needs to be replenished for basal metabolic functions. Another family of enzymes which consume NAD + are poly-ADP-ribose polymerases (PARPs), which are involved in DNA repair [bib_ref] Histone ADP-ribosylation in DNA repair, replication and transcription, Messner [/bib_ref]. NAD + in mammals is synthesized from NAM, tryptophan and nicotinic acid. Diet can influence NAD + levels through the uptake of the essential amino acid tryptophan (followed by de novo biosynthesis of NAD + ), nicotinic acid (Preiss-Handler pathway) or nicotinamide riboside (another precursor of NAM) [bib_ref] NAD + Metabolism and the Control of Energy Homeostasis: A Balancing Act..., Cantó [/bib_ref]. The major source of NAD + is NAM through the salvage pathway, however, this view has been recently challenged by recent work showing that tryptophan contributes largely to NAD + synthesis in the liver [bib_ref] Quantitative Analysis of NAD Synthesis-Breakdown Fluxes, Liu [/bib_ref] [bib_ref] De novo NAD + synthesis enhances mitochondrial function and improves health, Katsyuba [/bib_ref]. In the salvage pathway, nicotinamide phosphoribosyltransferase (NAMPT) catalyzes the conversion of NAM into NMN, which is then converted into NAD + by NMNAT. Three different enzymes NMANT1, 2 and 3 catalyze this reaction in nucleus, cytoplasm and mitochondria respectively, where these enzymes are located [bib_ref] Metabolic regulation of transcription through compartmentalized NAD + biosynthesis, Ryu [/bib_ref].
## Methionine and regulation of methylation
Methylation reactions including histone and DNA methylation require methyl groups contributed by dietary methyl donors and by 1-carbon methyl cofactors. The metabolic cycle of methionine is essential to feed most of methylation reactions. Methionine is a sulfur-containing essential amino acid [bib_ref] Methionine: A metabolically unique amino acid, Brosnan [/bib_ref] which serves as the substrate for the production of other amino acids such as cysteine and homocysteine [bib_ref] Dealing with methionine/homocysteine sulfur: Cysteine metabolism to taurine and inorganic sulfur, Stipanuk [/bib_ref]. In the methionine metabolic cycle, methionine is firstly converted into S-adenosylmethionine (SAM) by methionine adenosyltransferase (MAT) with the addition of adenosine that is donated by adenosine triphosphate (ATP) [bib_ref] The impact of metabolism on DNA methylation, Ulrey [/bib_ref]. As a universal cellular methyl donor, SAM can be used in different forms of methylation such as DNA, RNA and protein methylation, catalyzed by various methyltransferases. The transfer of the methyl group to the respective substrates leads to the formation of S-adenosylhomocysteine (SAH), which then undergoes hydrolysis into homocysteine by SAH hydrolase with the removal of the adenosine group. Homocysteine can subsequently enter different pathways including remethylation to form back methionine or irreversible transsulfuration for cysteine or α-ketobutyrate production. The catalytic activity of methyltransferases depends interestingly on a high SAM/SAH ratio [bib_ref] Regulation of chromatin and gene expression by metabolic enzymes and metabolites, Li [/bib_ref].
In the context of DNA methylation, which typically acts to repress gene transcription, the palindromic CpG dinucleotides in the genome are often the targets for DNA methyltransferases (DNMTs). DNMTs catalyze the transfer of a methyl group from SAM to covalently bind to the carbon-5 position of the cytosine residues in the CpG. This added methyl group can then block transcription factors from binding to the target genome sites, resulting in differential gene regulation [bib_ref] CpG methylation inhibits proenkephalin gene expression and binding of the transcription factor..., Comb [/bib_ref]. Since methyl groups involved in DNA methylation derived from methionine, an essential amino acid, it suggests that nutrition and metabolism could have an indirect effect on DNA methylation. In the methionine metabolic cycle, it was found that the accumulation of adenosine and homocysteine favors the biosynthesis of SAH rather than hydrolysis [bib_ref] Methionine metabolism in mammals, Finkelstein [/bib_ref]. In a study which tested rat liver, it was found that high levels of adenosine and homocysteine could lead to drastic decrease in the ratio of SAM/SAH, resulting in the inhibition of SAM-dependent methyltransferases and a global decrease in DNA methylation [bib_ref] S-Adenosylmethionine and S-adenosylhomocystein metabolism in isolated rat liver. Effects of L-methionine, L-homocystein,..., Hoffman [/bib_ref] [bib_ref] Influence of threonine metabolism on S-adenosylmethionine and histone methylation, Shyh-Chang [/bib_ref]. Moreover, anti-diabetic drug metformin has recently been found to boost DNA methylation by simultaneously promoting the accumulation of SAM and reducing SAH levels, further suggesting that high SAM:SAH ratio favors DNA methylation [bib_ref] Metformin regulates global DNA methylation via mitochondrial one-carbon metabolism, Cuyàs [/bib_ref].
## Α-ketoglutarate (αkg) and regulation of demethylation
DNA methylation is thought to be a relatively more stable epigenetic marker in comparison to histone modification [bib_ref] TET enzymes, TDG and the dynamics of DNA demethylation, Kohli [/bib_ref]. Yet, in the past decades, studies have shown that DNA methylation is a dynamic process in different biological contexts [bib_ref] Active DNA demethylation: Many roads lead to Rome, Wu [/bib_ref] [bib_ref] Epigenetic Reprogramming of Mouse Germ Cells toward Totipotency, Surani [/bib_ref]. There are in general two types of DNA demethylation: active or passive. Active DNA methylation is believed to be an enzyme-catalyzed process which leads to the removal or modification of the methyl group from 5-methylcytosine (5mC) which is mostly present in the CpG dinucleotides. Passive DNA demethylation on the other hand, refers to the loss of methyl group from 5mC during rounds of replication when DNMTs activities are reduced or inhibited. Although the passive mechanism is generally well studied, the active DNA demethylation mechanism still remains to be elucidated [bib_ref] Active DNA demethylation: Many roads lead to Rome, Wu [/bib_ref] [bib_ref] The colorful history of active DNA demethylation, Ooi [/bib_ref]. There are currently several proposed mechanisms for active DNA demethylation, one of them is oxidative demethylation which links cellular metabolism with DNA demethylation.
Oxidative demethylation is dependent on a group of enzymes called ten eleven translocation (TET) proteins that belong to a family of three AlkB-like Fe(II)/α-ketoglutarate-dependent dioxygenases: TET1, TET2 and TET3 [bib_ref] Role of TET enzymes in DNA methylation, development, and cancer, Rasmussen [/bib_ref]. Several studies discovered that TET proteins can catalyze the successive oxidation of 5mC to 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) [bib_ref] Conversion of 5-Methylcytosine to 5-Hydroxymethylcytosine in Mammalian DNA by MLL Partner TET1, Tahiliani [/bib_ref] [bib_ref] Tet-Mediated Formation of 5-Carboxylcytosine and Its Excision by TDG in Mammalian DNA, He [/bib_ref] [bib_ref] Tet Proteins Can Convert 5-Methylcytosine to 5-Formylcytosine and 5-Carboxylcytosine, Ito [/bib_ref]. These modified 5mC products were believed to serve as intermediates for the subsequent conversion back to unmodified cytosine through further reactions [bib_ref] TET enzymes, TDG and the dynamics of DNA demethylation, Kohli [/bib_ref]. This TET-mediated reaction is found to be dependent on α-ketoglutarate (αKG) which is a key metabolite in TCA Cycle. TETs use oxygen as the substrate for TETs to catalyze the oxidative carboxylation of αKG, producing CO 2 , enzyme-bound succinate and a reactive high-valent Fe(IV-oxo) intermediates [bib_ref] Reversing DNA Methylation: Mechanisms, Genomics, and Biological Functions, Wu [/bib_ref]. This intermediate then reacts with 5mC/5hmC/5fC for the production of 5hmC/5fC/5caC. Since αKG is produced mainly from oxidative decarboxylation of isocitrate mediated by isocitrate dehydrogenase (IDH) in the TCA cycle, the need of αKG in this process has linked metabolism with active DNA demethylation [bib_ref] Influence of threonine metabolism on S-adenosylmethionine and histone methylation, Shyh-Chang [/bib_ref].
## Metabolic-chromatin signaling in different physiological contexts
A key question regarding the metabolic connection of gene control is whether changes in metabolite fluxes and concentrations could influence specific gene programs and ultimately cell function or differentiation. Examples of such mechanisms have been shown in different contexts including stem cells differentiation, regulation of inflammation and tumorigenesis. The aim of this review is not to cover extensively these areas but to illustrate examples where metabolic signaling plays a role. The interface of metabolism and gene regulation in the context of metabolic disorders will be discussed.
## Metabolic reprogramming in stem cells
Stem cells undergo a metabolic reprogramming to allow cell differentiation [bib_ref] The metabolic programming of stem cells, Ng [/bib_ref]. Recent hypothesis suggests that metabolic changes could drive epigenetic modulation of the earliest steps of development. Pluripotent stem cells are able to self-renew and differentiate into all adult tissues in mammals [bib_ref] The metabolic programming of stem cells, Ng [/bib_ref]. There are two defined states of pluripotency, "naïve" and "primed" which have higher and reduced developmental potential respectively [bib_ref] Open chromatin in pluripotency and reprogramming, Alexandre [/bib_ref]. Naïve versus primed states have distinct epigenetic states, whereby primed state is more associated with repressive histone methylating marks. Interestingly, the laboratory culture conditions of naïve stem cells require a GSK3β and a MEK inhibitor (known as 2i) which leads to high levels of αKG [bib_ref] The ground state of embryonic stem cell self-renewal, Ying [/bib_ref]. The family of DNA and histone demethylating enzymes harbor a dioxygenase catalytic activity, which requires αKG/Fe 2+ . It has been shown that high αKG levels are associated with a demethylating activity which erases DNA and chromatin repressing marks and maintains stem cell pluripotency [bib_ref] Intracellular α-ketoglutarate maintains the pluripotency of embryonic stem cells, Carey [/bib_ref].
Recent work shows that metabolome regulates naïve to primed stated embryonic stem cell transition through the regulation of global H3K27me3 [bib_ref] The metabolome regulates the epigenetic landscape during naive-to-primed human embryonic stem cell..., Sperber [/bib_ref]. The authors found that H3K27me3 are indirectly regulated by the activity of nicotinamide N-methyltransferase (NNMT), which consumes SAM. NNMT was found particularly elevated during the naïve state which results in decreased SAM levels leading to reduced H3K27me3 [bib_ref] The metabolome regulates the epigenetic landscape during naive-to-primed human embryonic stem cell..., Sperber [/bib_ref]. These results contrast however with other findings showing that H3K27me3 is less sensitive to SAM compared to other histone methylations such as H3K4me3 [bib_ref] Histone Methylation Dynamics and Gene Regulation Occur through the Sensing of One-Carbon..., Mentch [/bib_ref] [bib_ref] Methionine metabolism influences genomic architecture and gene expression through H3K4me3 peak width, Dai [/bib_ref]. On the other hand, the naïve state requires high levels of histone acetylation, which leads to an open chromatin conformation [bib_ref] Open chromatin in pluripotency and reprogramming, Alexandre [/bib_ref]. In addition, the transition into a primed state is accompanied by a metabolic shift to a high glycolytic status and reduced fatty acid oxidation [bib_ref] The metabolome regulates the epigenetic landscape during naive-to-primed human embryonic stem cell..., Sperber [/bib_ref] [bib_ref] Metabolic exit from naive pluripotency, Wu [/bib_ref]. This leads to high levels of acetyl-CoA production which is used for histone acetylation necessary for pluripotency [bib_ref] Glycolysis-Mediated Changes in Acetyl-CoA and Histone Acetylation Control the Early Differentiation of..., Moussaieff [/bib_ref]. On the other hand, differentiation of adipocytes require high levels of acetyl-CoA [bib_ref] ATP-citrate lyase links cellular metabolism to histone acetylation, Wellen [/bib_ref] , although this could be also due to the requirements acetyl-CoA for fatty acid synthesis during adipogenesis.
A glycolytic switch also takes places in satellite muscle cells during its activation which leads to proliferation and differentiation into myoblasts [bib_ref] Metabolic Reprogramming of Stem Cell Epigenetics, Ryall [/bib_ref]. shown that this glycolytic shift decreases intracellular NAD + , reducing the activity of SIRT1 deacetylase activity [bib_ref] The NAD + -Dependent SIRT1 Deacetylase Translates a Metabolic Switch into Regulatory..., Ryall [/bib_ref]. This is associated with an increase in H4K16 acetylation and an activation of the myogenic differentiation program of muscle stem cells [bib_ref] The NAD + -Dependent SIRT1 Deacetylase Translates a Metabolic Switch into Regulatory..., Ryall [/bib_ref].
## Metabolic reprogramming in immune cells
Recent studies highlight the relevance of metabolic inputs into inflammatory functions by modulating cell differentiation [bib_ref] Metabolic Regulation of Immune Responses, Ganeshan [/bib_ref]. It has been shown that some metabolites play a signalling function during immune stimulation in both adaptive and innate immunity. This has been studied in the metabolic role of lymphocyte T cell activation and macrophage polarization.
Macrophages are the first line of defense against pathogens. The activation of monocytes into distinct and specialized macrophage polarization states combats bacterial or parasitic infections, namely pro-inflammatory versus anti-inflammatory respectively. Macrophages with anti-inflammatory activation also participate in processes such as wound healing. Both anti-versus pro-inflammatory have profound metabolic differences [bib_ref] The Many Alternative Faces of Macrophage Activation, Hume [/bib_ref]. The first evidence of a metabolic change during macrophage activation (pro-inflammatory state) dates from the 70s, where a switch from oxidative phosphorylation to glycolysis was identified [bib_ref] Some biochemical aspects of the immune macrophage, Hard [/bib_ref]. The Pro-Inflammatory Toll-like receptor 4 (TLR4) activation by lipopolysaccharide (LPS) results in a shift to glycolytic metabolism and impaired mitochondrial respiration. Furthermore, TLR signaling results in marked shifts in NAD + /NADH ratios, which influence the activities of the sirtuins, potentially modifying histone acetylation status [bib_ref] NAD + -dependent sirtuin 1 and 6 proteins coordinate a switch from..., Liu [/bib_ref]. In addition, some evidence suggests that certain metabolites could directly influence macrophage polarization. Liu et al. showed that αKG production via glutaminolysis promotes anti-inflammatory polarization by modulating the histone demethylase JMJD3, which mediates epigenetic changes. In addition, αKG inhibits pro-inflammatory induction through inhibition of NF-κβ [bib_ref] α-ketoglutarate orchestrates macrophage activation through metabolic and epigenetic reprogramming, Liu [/bib_ref].
Adaptive immunity seems also to be subjected to metabolic regulation. Activation of lymphocyte T-cells is initiated by the presentation of antigens through the T cell receptor (TCR) by antigen presenting cells (APCs). Naïve CD4 + T-cells then undergo differentiation into an activated "effector" state which mediates an immune response. Interestingly, this transition is associated with a metabolic switch from high rates of fatty acid oxidation to aerobic glycolysis [bib_ref] The CD28 signaling pathway regulates glucose metabolism, Frauwirth [/bib_ref]. Overall, the metabolic switch sustains the rapid cell proliferation necessary to activate inflammation. Aerobic glycolysis, mediated by increased lactate dehydrogenase activity, induces T helper 1 cell differentiation through elevation of acetyl-CoA levels resulting in histone acetylation of the IFN-γ promoter [bib_ref] Aerobic glycolysis promotes T helper 1 cell differentiation through an epigenetic mechanism, Peng [/bib_ref].
Similar to macrophages, αKG metabolism also determines T cell activation through its regulatory role of TET enzymes. A competitive inhibitor of αKG, 2-Hydroxyglutarate (HG) has been found to control histone and DNA methylation by affecting TET enzymes. HG accumulates in CD8 + T cells followed T-cell receptor activation and determines T-lymphocyte fate through an epigenetic mechanism [bib_ref] S-2-hydroxyglutarate regulates CD8 + T-lymphocyte fate, Tyrakis [/bib_ref]. One-carbon metabolism pathway is also important for T cell differentiation, as a dietary restriction of serine and glycine was associated with deficient effector T cell activation and IFN-γ production [bib_ref] The CD28 signaling pathway regulates glucose metabolism, Frauwirth [/bib_ref] [bib_ref] Serine Is an Essential Metabolite for Effector T Cell Expansion, Ma [/bib_ref].
## Metabolic signaling in the regulation of gene expression of metabolic disorders
Modern lifestyle, particularly in wealthier countries is associated with an increase of sedentarism and excessive food intake. This has led to a triplication of obesity rates within the last 20 years and the trend continues as childhood obesity has also risen [bib_ref] Current pharmacotherapy for obesity, Srivastava [/bib_ref]. Moreover, obesity is the cause of several associated pathologies including mainly type 2 diabetes (T2D), cardiovascular disease (CVD) and some types of cancer [bib_ref] Adiposity and gastrointestinal cancers: Epidemiology, mechanisms and future directions, Murphy [/bib_ref] [bib_ref] Insulin action and resistance in obesity and type 2 diabetes, Czech [/bib_ref]. Besides the socio-economic and environmental causes of obesity, the biological mechanisms underlying the regulation of energy balance are still intriguing. Although progress has been made in the understanding of the physiological, endocrine, cellular and molecular functions of different metabolic tissues and the central nervous system, there are still no current effective therapies to treat obesity [bib_ref] Current pharmacotherapy for obesity, Srivastava [/bib_ref]. Novel alternative therapeutic strategies are needed and advances have been made in particular in the identification of the molecular mechanisms that control energy expenditure through adipose tissue thermogenesis [bib_ref] Basic and Applied Thermogenesis Research' Bridging the Gap, Carobbio [/bib_ref]. However, how nutrients and derived intermediary metabolites control energy balance through gene expression regulation is not fully understood.
Given the connection between key intermediary metabolites and chromatin or DNA modifications, a relevant question is how an obesogenic nutrient load would impact the metabolite-chromatin regulatory axis. Because glucose is a major source of acetyl-CoA, it is tempting to speculate that high glucose levels induced by obesity would impact into histone acetylation through modulation of acetyl-CoA metabolism. In relation to this hypothesis, a comprehensive targeted metabolomic approach to quantify acyl-CoAs thioester compounds using the liver of high-fat diet fed mice did not show significant differences on the levels of acetyl-CoA in response to diet [bib_ref] High-Resolution Metabolomics with Acyl-CoA Profiling Reveals Widespread Remodeling in Response to Diet, Liu [/bib_ref]. However, Carrer et al. showed a decreased acetyl-CoA and acetyl-CoA/CoA ratio in liver, adipose tissue, and pancreas of mice fed a high-fat diet [bib_ref] Impact of a High-fat Diet on Tissue Acyl-CoA and Histone Acetylation Levels, Carrer [/bib_ref]. This was in addition associated with a global decrease in histone acetylation only in white adipose tissues but not in the liver. This reduction could be resulted from a decrease in the expression of ACLY induced by high-fat diet [bib_ref] Impact of a High-fat Diet on Tissue Acyl-CoA and Histone Acetylation Levels, Carrer [/bib_ref]. It was shown that acetylation of many histone lysines were correlated with acetyl-CoA levels [bib_ref] Impact of a High-fat Diet on Tissue Acyl-CoA and Histone Acetylation Levels, Carrer [/bib_ref]. On the other hand, it is not known whether other histone acylation modifications could be induced upon high-fat diet or exacerbated fatty acid metabolism. Acyl-CoAs can be generated by various intermediate metabolic pathways and it has been shown that short acyl-chains including propionyl-CoA, butyryl-CoA or crotonyl-CoA can post-translationally modify histones [bib_ref] Metabolic regulation of gene expression through histone acylations, Sabari [/bib_ref]. It has been postulated by histone acylation mark actively transcribed genes [bib_ref] Metabolic regulation of gene expression through histone acylations, Sabari [/bib_ref]. In relation to fatty acid metabolism, a recent study using isotope tracing has shown that lipids are also a source of histone acetylation through acetyl-CoA [bib_ref] Lipids Reprogram Metabolism to Become a Major Carbon Source for Histone Acetylation, Eoin [/bib_ref]. The authors showed that at least, the short chain lipid octanoate leads to a specific lipid gene program regulation in hepatocytes [bib_ref] Lipids Reprogram Metabolism to Become a Major Carbon Source for Histone Acetylation, Eoin [/bib_ref]. However, one of the technical limitations is to unveil whether the acetyl-CoA derived from lipid oxidation is the source of the histone acetylation of the genes whose expression was regulated.
Another interesting question is how different metabolic fluxes would impact on cell differentiation of metabolic tissues, particularly adipocytes, which can be very relevant in the context of obesity. Pioneering work by Wellen et al. showed that in the presence of glucose, ACLY, which is responsible for the conversion of citrate into acetyl-CoA, promotes histone acetylation, including the promoter of Slc2a4 encoding the glucose transporter GLUT4 [bib_ref] ATP-citrate lyase links cellular metabolism to histone acetylation, Wellen [/bib_ref]. Moreover, ACLY function was also required for adipocyte differentiation, suggesting that acetyl-CoA and high levels of histone acetylation is needed for adipocyte differentiation as we previously described. However, further research is needed to elucidate the dependence of metabolite signaling into specific adipogenic differentiation programs.
Several transcriptional regulators control white and brown adipocyte differentiation, including some brown adipose-specific regulators such as PRDM16, PGC-1α or EBF2 [bib_ref] Transcriptional and epigenetic control of brown and beige adipose cell fate and..., Inagaki [/bib_ref]. By means of genetic loss of function, Yang et al. recently showed that AMPK activation leads to the elevation of αKG which promotes the demethylation of the promoter of Prdm16 by TET enzymes [bib_ref] AMPK/α-Ketoglutarate Axis Dynamically Mediates DNA Demethylation in the Prdm16 Promoter and Brown..., Yang [/bib_ref]. The authors showed that the specific Prdm16 demethylation committed pre-adipocytes precursors into brown adipogenic differentiation. Interestingly, pharmacological activation of AMPK through metformin (the main drug used in T2D) or AICAR rescued the obesity-induced suppression of brown adipogenesis and thermogenesis [bib_ref] AMPK/α-Ketoglutarate Axis Dynamically Mediates DNA Demethylation in the Prdm16 Promoter and Brown..., Yang [/bib_ref].
Histone demethylases have also been involved in the formation of obese phenotypes. For example, the knockout mouse model of the H3K9-specific demethylase KDM3A (also known as JHDM2A or JMJD1) leads to adipose tissue accumulation and insulin resistance [bib_ref] Role of Jhdm2a in regulating metabolic gene expression and obesity resistance, Tateishi [/bib_ref]. KDM3A was shown to deacetylate H3K9me2 at the peroxisome proliferator activator receptor response element (PPRE) which controls the expression of the thermogenic gene Uncoupling Protein 1 (Ucp1) in brown adipose tissue. The expression of KDM3A was in addition induced by environmental cold exposure and recruited PPARγ-RXRα and PGC-1α to the Ucp1 promoter which enhanced Ucp1 expression [bib_ref] Role of Jhdm2a in regulating metabolic gene expression and obesity resistance, Tateishi [/bib_ref]. Recent work has shown that KDM3A is phosphorylated in response to cold-induced adrenergic signaling at S265 [bib_ref] JMJD1A is a signal-sensing scaffold that regulates acute chromatin dynamics via SWI/SNF..., Abe [/bib_ref] [bib_ref] Histone demethylase JMJD1A coordinates acute and chronic adaptation to cold stress via..., Abe [/bib_ref]. Using KDM3A-S265A knock-in mice, the authors showed that S265 phosphorylation is required to the demethylate the promoter of Ucp1 in beige adipocytes [bib_ref] JMJD1A is a signal-sensing scaffold that regulates acute chromatin dynamics via SWI/SNF..., Abe [/bib_ref] [bib_ref] Histone demethylase JMJD1A coordinates acute and chronic adaptation to cold stress via..., Abe [/bib_ref]. This promoter is highly methylated (H3K9me2) in white adipocytes unless there is a cold-induced chromatin reprogramming leading to KDM3A-dependent demethylation. KDM3A was shown in addition to interact with PRDM16 and recruit the PPARγ-PGC-1α complex [bib_ref] JMJD1A is a signal-sensing scaffold that regulates acute chromatin dynamics via SWI/SNF..., Abe [/bib_ref] [bib_ref] Histone demethylase JMJD1A coordinates acute and chronic adaptation to cold stress via..., Abe [/bib_ref]. Another demethylase, LSD1 was recently showed to mediate repression of white adipose selective targets in brown adipose tissue through the demethylation of the activating mark H3K4me3 [bib_ref] Lysine-specific demethylase 1 promotes brown adipose tissue thermogenesis via repressing glucocorticoid activation, Zeng [/bib_ref]. Adipose-specific ablation of LSD1 reduced whole-body energy expenditure through a reduced mitochondrial fatty acid oxidation of the brown adipose tissue which lead to increased fat deposition [bib_ref] Lysine-specific demethylase 1 promotes brown adipose tissue thermogenesis via repressing glucocorticoid activation, Zeng [/bib_ref]. Several other chromatin modifying enzymes, particularly HDACs, control multiple metabolic processes. This topic has been extensively reviewed previously [bib_ref] Protein acetylation in metabolism-Metabolites and cofactors, Menzies [/bib_ref].
Besides the direct function played by chromatin regulators on metabolic processes, how metabolic pathways connect directly the activity of transcription or chromatin factors remains less understood. One example of such regulation is the activation of the transcriptional co-activator PGC-1α. PGC-1α was identified as a cold response inducible factor in brown adipose tissue where it activates thermogenesis through the co-activation of nuclear hormone receptors PPARγ (Peroxisome Proliferator Activator Receptor) [bib_ref] A cold-inducible coactivator of nuclear receptors linked to adaptive thermogenesis, Puigserver [/bib_ref]. PGC-1α has been shown to connect multiple energy homeostasis pathways through the co-activation of several transcription factors including PPARα, RAR (retinoic acid receptor), and TR (thyroid receptor), ERR (estrogen-related receptor) or YY1 (Yin Yang 1) [bib_ref] Peroxisome Proliferator-Activated Receptor-γ Coactivator 1α (PGC-1α): Transcriptional Coactivator and Metabolic Regulator, Puigserver [/bib_ref]. The activity of PGC-1α is regulated by its acetylation, PGC-1α is active when it is deacetylated by SIRT1 and deactivated when it is acetylated by GCN5 [bib_ref] Nutrient-dependent regulation of PGC-1α's acetylation state and metabolic function through the enzymatic..., Dominy [/bib_ref]. Therefore, the acetylation and its dependence on SIRT1 and GCN5 links PGC-1α activity to the nutritional status. A low nutritional status leads to an increase in NAD + which activates SIRT1-dependent PGC-1α deacetylation to activate the transcription factors of energy generating pathways [bib_ref] Nutrient-dependent regulation of PGC-1α's acetylation state and metabolic function through the enzymatic..., Dominy [/bib_ref]. On the other hand, upon high energy status GCN5 acetylates and represses PGC-1α activity [bib_ref] Nutrient-dependent regulation of PGC-1α's acetylation state and metabolic function through the enzymatic..., Dominy [/bib_ref]. This has implications in metabolic disorders such as diabetes, since PGC-1α target genes in oxidative phosphorylation were found downregulated in human diabetes [bib_ref] PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes, Mootha [/bib_ref]. More recently, HDAC3 was shown to also deacetylate PGC-1α in brown adipose tissue and predispose brown adipose tissue to acute cold exposure [bib_ref] Histone deacetylase 3 prepares brown adipose tissue for acute thermogenic challenge, Emmett [/bib_ref].
In addition to chromatin modifying enzymes which sense metabolic status, it has been shown that an increasing number of cytosolic metabolic enzymes is found in the nucleus [bib_ref] Metabolic Enzymes Moonlighting in the Nucleus: Metabolic Regulation of Gene Transcription, Boukouris [/bib_ref]. They sometimes play additional enzymatic functions which are referred as a moonlight role [bib_ref] Metabolic Enzymes Moonlighting in the Nucleus: Metabolic Regulation of Gene Transcription, Boukouris [/bib_ref]. Most of cytosolic glycolytic and TCA cycle enzymes are surprisingly also located in the nucleus where they seem to perform functions in transcription or DNA replication and repair or even unknown roles [bib_ref] Metabolic Enzymes Moonlighting in the Nucleus: Metabolic Regulation of Gene Transcription, Boukouris [/bib_ref]. It is not known how the nuclear translocation of metabolic enzymes is regulated or whether it differs according to the tissue specificity. Post-translational modifications may control the localization of metabolic enzymes in the nucleus, but evidence is lacking. A recent study identified the interaction of the subunits E1b and E2 of the pyruvate dehydrogenase complex (PDH) with the transcription factor STAT5 in the nucleus of adipocytes [bib_ref] Pyruvate dehydrogenase complex (PDC) subunits moonlight as interaction partners of phosphorylated STAT5..., Richard [/bib_ref]. Moreover, E1b and E2 were found to be associated with chromatin through STAT5 interaction. Since STAT5 is involved in adipocyte differentiation, the authors suggested that the interaction with PDH subunits could be involved in histone acetylation of STAT5 targets [bib_ref] Pyruvate dehydrogenase complex (PDC) subunits moonlight as interaction partners of phosphorylated STAT5..., Richard [/bib_ref].
In summary, future research may reveal additional metabolite-dependent orchestration of complex cellular outputs through chromatin modifications in the context of metabolic disorders.
## Therapeutic diet interventions targeting metabolic-chromatin axis
During the last 200 years, humanity has experienced a doubling of the life expectancy in most developed countries [bib_ref] Facing up to the global challenges of ageing, Partridge [/bib_ref]. Scientific advances (immunization against infectious diseases, and antibiotics) together with social (food and water quality, housing and lifestyle) changes are among the main causes [bib_ref] Facing up to the global challenges of ageing, Partridge [/bib_ref]. However, we are today facing a burden of non-communicable diseases, usually appearing at late stages in life, including cancer, metabolic disorders and neurodegenerative diseases which reduce the health-span expectancy [bib_ref] Facing up to the global challenges of ageing, Partridge [/bib_ref]. Cardiovascular disease and diabetes (mainly due to obesity) are the main causes of global morbidity in both men and women [bib_ref] Facing up to the global challenges of ageing, Partridge [/bib_ref].
Diet plays a major role in the development or prevention of diseases, particularly CVD and T2D, however, there are still many open questions partly due to the individual variation and different response to interventions [bib_ref] The Hunger Genes: Pathways to Obesity, Van Der Klaauw [/bib_ref] [bib_ref] A systematic review and mixed treatment comparison of pharmacological interventions for the..., Gray [/bib_ref]. Drug therapies are usually not efficacious, accompanied with side-effects and in most of the cases require a chronic treatment. There is, therefore, an urge to identify novel biological mechanisms and pathways that would lay the ground for novel effective therapeutic approaches or diet interventions.
From a metabolic standpoint, it is tempting to ask how would diet influence on specific gene programs. Particularly, a recurrent question is the effect of nutrition/diet with stable chromatin/DNA modifications. A paradigmatic biological example of the effect of nutrition in epigenetics is the feeding of honeybees (Apis mellifera) with "royal jelly". The feeding of this unique diet to bee larva induces profound phenotypic changes including fertility through its transformation into a "queen" bee. One of the intriguing questions was the fact that, despite the massive phenotypic transformation of the fertile queen, compared to the sterile worker bees, all larvae are genetically identical. The biological mechanism of this transformation, lies on the differential epigenetic DNA methylation, proving a direct link of diet and epigenetics [bib_ref] Nutritional control of reproductive status in honeybees via DNA methylation, Kucharski [/bib_ref]. Potentially, analogous molecular mechanisms may exist in mammals including humans.
Today, ketogenic diet, intermittent fasting or caloric restriction are interventions which have proved some success in some contexts including neurodegenerative diseases, cancer, aging, metabolic disorders or exercise performance [bib_ref] Dietary Protein and Energy Balance in Relation to Obesity and Co-morbidities, Drummen [/bib_ref]. We will address here the type of therapies or interventions which directly interact with chromatin function . effects and in most of the cases require a chronic treatment. There is, therefore, an urge to identify novel biological mechanisms and pathways that would lay the ground for novel effective therapeutic approaches or diet interventions. From a metabolic standpoint, it is tempting to ask how would diet influence on specific gene programs. Particularly, a recurrent question is the effect of nutrition/diet with stable chromatin/DNA modifications. A paradigmatic biological example of the effect of nutrition in epigenetics is the feeding of honeybees (Apis mellifera) with "royal jelly". The feeding of this unique diet to bee larva induces profound phenotypic changes including fertility through its transformation into a "queen" bee. One of the intriguing questions was the fact that, despite the massive phenotypic transformation of the fertile queen, compared to the sterile worker bees, all larvae are genetically identical. The biological mechanism of this transformation, lies on the differential epigenetic DNA methylation, proving a direct link of diet and epigenetics [bib_ref] Nutritional control of reproductive status in honeybees via DNA methylation, Kucharski [/bib_ref]. Potentially, analogous molecular mechanisms may exist in mammals including humans.
Today, ketogenic diet, intermittent fasting or caloric restriction are interventions which have proved some success in some contexts including neurodegenerative diseases, cancer, aging, metabolic disorders or exercise performance [bib_ref] Dietary Protein and Energy Balance in Relation to Obesity and Co-morbidities, Drummen [/bib_ref]. We will address here the type of therapies or interventions which directly interact with chromatin function .
## Figure 2.
Influence of different diets in chromatin function. Ketogenic diet promotes increased fatty acid oxidation rates which elevates acetyl-CoA production and therefore ketone bodies. The ketone body β-hydroxybutyrate inhibits class I HDACs leading to increased H3K9ac and H3K14ac. Calorie restriction leads to increased NAD+ levels and activation of SIRT1 and SIRT6, which promote histone deacetylation and delays aging. Nutrient overload leads to obesity, it is not fully understood how acetyl-CoA pools may affect specific gene programs in the context of obesity. HDACS: Histone Deacetylases; HATS: Histone Acetyltransferases.
## Ketogenesis
A high-fat, adequate protein and very low carbohydrate diet known as ketogenic diet (KD) induce a switch into fatty acid oxidation as fuel usage, resulting in an excessive acetyl-CoA production which leads to ketone bodies formation. In recent years, KD has been used in therapy of epilepsy, is considered metabolically healthy and promotes weight loss. The end product of ketogenesis is β-hydroxybutyrate (βOHB) and its circulating levels have been shown to increase by KD [bib_ref] Ketone bodies as signaling metabolites, Newman [/bib_ref]. βOHB has been shown to inhibit class I HDACs and concomitantly increase histone acetylation (H3K9ac and H3K14ac) [bib_ref] Suppression of oxidative stress by β-hydroxybutyrate, an endogenous histone deacetylase inhibitor, Shimazu [/bib_ref]. Therefore, KD could potentially directly modify chromatin . Influence of different diets in chromatin function. Ketogenic diet promotes increased fatty acid oxidation rates which elevates acetyl-CoA production and therefore ketone bodies. The ketone body β-hydroxybutyrate inhibits class I HDACs leading to increased H3K9ac and H3K14ac. Calorie restriction leads to increased NAD+ levels and activation of SIRT1 and SIRT6, which promote histone deacetylation and delays aging. Nutrient overload leads to obesity, it is not fully understood how acetyl-CoA pools may affect specific gene programs in the context of obesity. HDACS: Histone Deacetylases; HATS: Histone Acetyltransferases.
## Ketogenesis
A high-fat, adequate protein and very low carbohydrate diet known as ketogenic diet (KD) induce a switch into fatty acid oxidation as fuel usage, resulting in an excessive acetyl-CoA production which leads to ketone bodies formation. In recent years, KD has been used in therapy of epilepsy, is considered metabolically healthy and promotes weight loss. The end product of ketogenesis is β-hydroxybutyrate (βOHB) and its circulating levels have been shown to increase by KD [bib_ref] Ketone bodies as signaling metabolites, Newman [/bib_ref]. βOHB has been shown to inhibit class I HDACs and concomitantly increase histone acetylation (H3K9ac and H3K14ac) [bib_ref] Suppression of oxidative stress by β-hydroxybutyrate, an endogenous histone deacetylase inhibitor, Shimazu [/bib_ref]. Therefore, KD could potentially directly modify chromatin function as some studies suggest.
Daily metabolic fluctuations and circadian rhythms are interlocked, as such; ketogenesis is induced during the daily fasting periods coupled with the circadian clock. KD has been shown to induce circadian-like changes in the gut and liver which were distinctly controlled in both tissues [bib_ref] Distinct Circadian Signatures in Liver and Gut Clocks Revealed by Ketogenic Diet, Tognini [/bib_ref]. Particularly, the observed circadian oscillation of serum βOHB induced a coupled cyclic HDAC inhibition and H3 acetylation [bib_ref] Distinct Circadian Signatures in Liver and Gut Clocks Revealed by Ketogenic Diet, Tognini [/bib_ref]. Another recent study, found β-hydroxybutyrylation as a novel type of histone modification which was induced in the liver of long-fasted mice [bib_ref] Metabolic Regulation of Gene Expression by Histone Lysine β-Hydroxybutyrylation, Xie [/bib_ref]. The genes marked by this chromatin mark correlated with the upregulated genes during prolonged fasting [bib_ref] Metabolic Regulation of Gene Expression by Histone Lysine β-Hydroxybutyrylation, Xie [/bib_ref].
## Calorie restriction
There is a vast literature, including controversies, regarding the potentially beneficial health effects and lifespan extension induced by calorie restriction (CR) in different organisms [bib_ref] Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan, Howitz [/bib_ref] [bib_ref] Calorie restriction, SIRT1 and metabolism: Understanding longevity, Bordone [/bib_ref]. It is however well supported that the physiological effects of CR involve the activation of sirtuins (NAD + dependent deacetylases) [bib_ref] NAD + and sirtuins in aging and disease, Imai [/bib_ref] [bib_ref] Calorie restriction and sirtuins revisited, Guarente [/bib_ref]. SIRT1 to SIRT6 target different tissues relevant to CR in a systemic coordinated response. These include hypothalamus, skeletal muscle, vasculature, liver, kidney, pancreatic β-cells and adipose tissue [bib_ref] Calorie restriction and sirtuins revisited, Guarente [/bib_ref]. Many of the sirtuin actions in these tissues involve the regulation of key transcriptional regulators. There is a direct link between dietary inputs and sirtuins which play a nutrient-sensing regulatory role. The pharmacological intervention to activate sirtuins has been also intensely pursued as a promising strategy not absent of difficulties. On the other hand, manipulation of NAD + levels was found to ameliorate metabolic and aging dependent disorders [bib_ref] The NAD + precursor nicotinamide riboside enhances oxidative metabolism and protects against..., Cantó [/bib_ref] [bib_ref] Nicotinamide mononucleotide, a key NAD + intermediate, treats the pathophysiology of diet-and..., Yoshino [/bib_ref] [bib_ref] The Biology and Therapeutic Potential of NMN and NR, Yoshino [/bib_ref]. Recently, cellular senescence, a hallmark of aging, has been shown to be delayed by restoring the normal age decline of NAD + through nicotinamide riboside nutrition in mice [bib_ref] NAD + repletion improves mitochondrial and stem cell function and enhances life..., Zhang [/bib_ref].
## Concluding remarks
The identification of nutrient regulation of gene expression in the 1960 s set the ground for more complex understanding of how genes are activated or repressed. Then, the era of modern molecular biology has led to a deep knowledge in how cells coordinate transcriptional programs. Nevertheless, how metabolic networks directly confluence with gene expression regulation is still not fully understood. In mammalian systems, hormonal control followed by intracellular signaling cascades are used to sense and transmit environmental signals to regulate specific gene expression programs. Increasing evidence suggests that an additional layer of regulation could be supported by an interface between metabolic intermediates and chromatin modifying enzymes. Metabolites provide therefore a direct nutritional information into gene expression control. Several examples in stem cell differentiation, macrophage or T cell activation and tumorigenesis highlight how specific cell decisions are controlled by the concentration of some key metabolites. Yet, additional examples arise in different fields including metabolic homeostasis where nutritional status plays even a more crucial role in regulating transcriptional outputs. Future research will unveil novel players in the metabolic-chromatin axis which may provide new potential therapeutic targets or specific diet interventions to combat the still expanding rates of obesity and metabolic diseases.
[fig] Figure 1: Figure 1. Interaction between metabolism and histone acetylation and DNA/histone methylation. Different nutrient substrates including glucose, fatty acids, amino acids and acetate lead to production of intermediary metabolites which play a role in protein acetylation. Acetyl-CoA derived from glucose, fatty acid or amino acid metabolism is the substrate for histone acetylation after conversion into citrate by TCA cycle and back to Acetyl-CoA in the cytoplasm by ACLY. Acetate is also a source of acetyl-CoA which leads to histone acetylation. Histone and DNA methylation depends on the dietary methionine which enters a cycle for conversion into SAM which is used as a donor of the methyl group. This leads to formation of SAH which is recycled back to methionine through [/fig]
[fig] Funding: This research was funded by Swiss National Foundation (31003A_172871). [/fig]
|
Comprehensive DNA methylation analysis of tissue of origin of plasma cell-free DNA by methylated CpG tandem amplification and sequencing (MCTA-Seq)
Background: Comprehensive analysis of the tissue of origin of plasma cell-free DNA (cfDNA) remains insufficient. A genome-scale DNA methylation method for this analysis is of both biological and clinical interest.Methods: We used the methylated CpG tandem amplification and sequencing (MCTA-Seq), which is a genomescale DNA methylation method, for analyzing cfDNA. We performed MCTA-Seq to pair plasma cfDNA and white blood cell genomic DNA from 14 healthy individuals for comparative analysis, with eight tissues being analyzed for identifying tissue-specific markers. The relative contributions of multiple tissues to cfDNA were calculated for plasma cfDNA obtained from healthy adults (n = 25), cholelithiasis patients (n = 13), liver cirrhosis patients (n = 17), hepatocellular carcinoma patients (n = 30), and acute pancreatitis patients (n = 8).Results: We identified a total of 146 tissue-specific hypermethylation markers. Simulation analysis showed that MCTA-Seq can accurately measure DNA fractions contributed by multiple tissues to cfDNA. We demonstrated that the liver is the major non-hematopoietic tissue contributing to plasma cfDNA in healthy adults. The method also detected increases in the liver-derived DNA in the blood from patients with liver diseases, which correlate with an increase in the liver enzyme level. Furthermore, the results indicated that blood cells make a major contribution to the elevation of cfDNA levels in acute pancreatitis, liver cirrhosis, and hepatocellular carcinoma patients. Finally, we characterized a novel set of tissue-specific hypermethylation markers for cfDNA detection, which are located within the intragenic regions of tissue-specific highly expressed genes. Conclusions: We have used MCTA-Seq for simultaneously measuring cfDNA fractions contributed by multiple tissues. Applying this approach to healthy adults and liver and pancreas disease patients revealed the tissue of origin of cfDNA. The approach and the identified markers should facilitate assessing the cfDNA dynamics in a variety of human diseases.
# Background
The biological and diagnostic applications of circulating cell-free DNA (cfDNA) have attracted great interest in recent years. The non-invasive prenatal testing of fetal chromosomal aneuploidy by cfDNA sequencing has been widely used in the clinic. Also, "liquid biopsy" of tumor-specific mutations in cfDNA has great promise for cancer diagnosis and monitoring.
While genetic-based approaches have been successfully used, DNA methylation has long been studied as a promising epigenetic cfDNA biomarker because of its stable and informative nature. A unique characteristic of DNA methylation vs genetic-based markers is tissue specificity. Different cell types have distinct DNA methylation patterns, which can be used to distinguish the tissue of origin of cfDNA. Indeed, individual tissue-specific DNA methylation markers have recently been shown to be sensitive for blood-based detection of tissue cell death in several diseases. Genomewide and targeted approaches have also recently been used to inspect the fractions of cfDNA contributed by multiple tissues.
Despite these progresses, DNA methylation analysis of the tissue of origin of plasma cfDNA is still insufficient. For example, a number of previous studies have shown that cfDNA levels are elevated in many clinical disorders including cancer, autoimmune diseases, liver diseases, intensive exercise, and acute medical emergencies such as acute pancreatitis, trauma, stroke, myocardial infarction, and sepsis. The cellular origin of the increased cfDNA level remains incompletely clarified. Additionally, complementary approaches are needed for identifying novel tissue-specific methylation markers for cfDNA detection.
We recently developed methylated CpG tandem amplification and sequencing (MCTA-Seq) for cfDNA analysis. As a genome-scale DNA methylation method that enriches methylated CpG islands, MCTA-Seq is informative, sensitive, and cost-effective and is suitable for efficient screening of novel cfDNA methylation markers. Here, we extended this method to assess the relative contributions of multiple non-hematopoietic tissues to plasma cfDNA. We have analyzed 60 tissue samples and 85 plasma samples, identifying a total of 146 tissuespecific methylation markers and investigating the tissue of origin of plasma cfDNA in healthy individuals and liver and pancreas disease patients.
# Methods
## Subject
All subjects were recruited from the Department of Surgery, Beijing Shijitan Hospital, Capital Medical University of China (which is also the Ninth School of Clinical Medicine, Peking University). The study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University. Written informed consents were obtained from all subjects before inclusion in the study.
## Dna extraction and mcta-seq library preparation
Commercialized genomic DNAs for normal tissues, including the lung (n = 2), stomach (n = 2), colon (n = 2), kidney (n = 2), pancreas (n = 2), muscle (n = 2), and skin (n = 2) were purchased from BioChain. The DNAs were extracted from white blood cells (WBC) using the DNeasy Blood & Tissue Kit (Qiagen) according to the manufacturer's protocol. Plasma cfDNA was obtained as described previously. The cfDNA concentration was quantified using the Qubit dsDNA HS Kit (Invitrogen, Q32854). The MCTA-Seq library was prepared and sequenced as described previously with small modifications. Briefly, cfDNA obtained from 2 mL plasma (up to 24 ng) or 400 ng tissue gDNA were treated by bisulfite and purified using the MethyCode bisulfite conversion kit (Invitrogen). All bisulfite-converted cfDNA or 60 ng bisulfite-converted gDNA (quantified using the Qubit ssDNA Kit, Q10212) were then amplified using MCTA-Seq primers A and B as described previously. The full amplicon was amplified in a 50-μL reaction by adding a 30-μL solution containing 1× Ex Taq Buffer, 250 μM each dNTP, 2 μM primer C (5′-AATGATACGGCGAC CACCGAGATCTACACTCTTTCCCTACACGACGCTC TTCCGATCT-3′) and 2 μM primer D (5′-CAAGCAGA AGACGGCATACGAGATCTGATCGTGACTGGAGTT-CAGACGTGTGCT-3′), and the reaction was subjected to 14 cycles of 95°C for 30 s, 65°C for 30 s, 72°C for 1 min, and a final cycle of 72°C for 5 min. Then, instead of processing individual samples, we pooled six samples with different Illumina index sequences by taking 30-μL reactions for each sample and then purified. The pooled product was resolved on a 3% agarose gel (Takara, Agarose LM SIEVE, D614), and the fraction between 180 and 250 bp was excised and then purified. For the tissue sample, the product is usually ready for serving as the library for sequencing. For the plasma sample, one to two additional rounds of amplification (using primers QP1 (5′-AATG ATACGGCGACCACCGA-3′) and QP2 (5′-CAAGCAG AAGACGGCATACGA-3′)) and gel purification are usually needed to clean up the primer dimers and obtain enough materials for sequencing.
## Data processing
FASTQ format R2 reads were first processed and filtered as described previously (step i to step v), using hg19 as the reference sequence. Instead of considering all alleles within a CGI as the unit for calculation as in our previous study, we focused on the fully methylated molecules (FMMs) amplified from a CGCGCGG as the unit for calculation in the present study. Specifically, for a certain CpG site (from the 1st to the 10th) downstream a CGCGCGG, the methylation value is calculated as the number of FMMs, i.e., all CpGs between this CpG site and the CGCGCGG site were methylated, normalized by the total number of reads uniquely mapped to the whole human genome, and expressed as methylated alleles per million mapped reads (MePM). We calculated the MePM value for both the plasma and tissue samples.
## Identification of tissue-specific methylation markers
Tissue-specific CGCGCGG methylation markers among 8 normal tissue types, including the liver, lung, stomach, colon, kidney, pancreas, muscle, and skin, were identified using the following criteria: (i) a tissue-specific index (τ) was defined, and a threshold τ > 0.9 was used to define tissue-specific methylation; (ii) MePM > 10 in the most hypermethylated tissue; (iii) 90% percentile MePM = 0 in a training set of 29 WBC samples; (iv) average MePM < 1 in the liver if the most hypermethylated tissue is not the liver; and (v) the distance between the CpG site and the CGCGCGG site is less than 60 bp. These criteria yielded 59, 9, 20, 13, 18, and 8 markers for the liver, the stomach, the colon, the kidney, the pancreas, and the skin, respectively. Since no markers for the lung and muscle were identified using these criteria, we omitted the tissue-specific index restriction and identified 8 lung markers and 11 muscle markers.
## Deconvolution analysis for mcta-seq cfdna tissue mapping
The mathematical relationship between the methylation values (MePM) of plasma and the corresponding MePM in each tissue of marker i can be expressed by the following formula:
[formula] MP i ¼ X k MT ik à P k ; [/formula]
in which MP i represents the methylation value of the tissue-specific CGCGCGG marker i in plasma cfDNA; MT ik represents the methylation value of marker i in tissue k; and P k represents the proportional contribution of tissue k to plasma cfDNA. The particle swarm optimization programwas used to solve the simultaneous equations. In practice, we added up the markers except the maximum of each tissue type as one tissue-specific methylation marker. We also omitted the contribution of WBC since the markers were selected as having very low methylation values in WBC (90 percentile MePM = 0). Thus, a total of 8 simultaneous equations representing 8 nonhematopoietic tissue types were generated to be solved. The median value of ten runs was used as the readout. The range of P k should fulfill the expectation of 0 to 1. To further eliminate any effect from non-specific methylation in WBC, the average tissue fraction values in fourteen paired WBC samples (0.015%, 0, 0.36%, 0, 0, 0, 0.066%, and 0 for the liver, lung, stomach, colon, kidney, pancreas, muscle, and skin, respectively) were subtracted from the measured tissue fractions. In addition, the measured tissue fractions less than the average values plus three standard deviations of WBC samples (0.11%, 0, 2.0%, 0.082%, 0.036%, 0.023%, 0.57%, and 0 for the liver, lung, stomach, colon, kidney, pancreas, muscle, and skin, respectively) were set to zero. Combining the sensitivities calculated in the simulation analysis, we estimate that the low detection limits of the MCTA-Seq cfDNA tissue mapping method are 0.25% for the liver and pancreas; 0.5% for the colon, kidney, and skin; and 2% for the lung, stomach, and muscle.
For the simulation analysis, MCTA-Seq data of the nonhematopoietic tissue types and WBC (from 14 healthy individuals) were mixed in serial ratios with tissue fractions ranging from 0.25 to 16%, in a total of approximately 3 million uniquely mapping paired-end reads. We then plotted the expected and measured tissue fractions.
For analyzing the plasma of HCC patients, we only used 85 markers that are not hypermethylated in HCC tissues (liver-specific markers: liver tumor tissue vs the adjacent non-cancerous liver tissues, one-tailed MWW test, P > 0.05; other tissue specific markers: 90 percentile of liver tumor tissues < 10) to exclude the effects of the cancer-released DNA on the tissue mapping.
The relationship between the tissue-specific hypermethylation markers and the tissue-specific genes RNA-Seq data of normal human tissues were downloaded from the Human Protein Atlas. A threshold z-score > 1.5 was used to define tissue-specific genes. The intragenic region was defined as starting at 300-bp downstream of a RefGene transcription start site (TSS) and ending at 300-bp downstream of a RefGene transcription end site.
## Bioinformatics and statistical analysis
Custom R scripts and R packages were used to construct boxplots, heatmaps, and bar plots and to perform statistical analysis. The MePM of plasma and paired WBC from normal individuals were normalized to the same duplication rate (plasma, 1.4 ± 0.3; WBC, 1.3 ± 0.1 (mean ± SD)).
# Results
## Genome-wide dna methylation comparison between plasma cfdna and wbc gdna
We designed a systematic and unbiased strategy to investigate the non-hematopoietic tissue of origin of plasma cfDNA in healthy individuals. This strategy first identified the regions that are significantly hypermethylated in plasma cfDNA vs white blood cell (WBC) genomic DNA (gDNA) extracted from matched buffy coats using the MCTA-Seq method. Then, the tissue-specific methylation signatures of these regions were resolved to elucidate which non-hematopoietic tissues contribute to plasma cfDNA .
We performed MCTA-Seq on matched plasma and WBC samples obtained from 14 healthy individuals. The plasma cfDNA data were reported in our previous study. A schematic diagram of the sample collection is shown in Additional file 1: . The WBC samples were sequenced to a median depth of 6 million raw reads pair similar to the plasma samples (see Additional file 2: for sequencing information). A total of A B D C E Genome-wide DNA methylation comparison between plasma and WBC. a Schematic diagram of the strategy for a systematic and unbiased examination of the non-hematopoietic tissue of origin of plasma cfDNA. b Volcano plot showing differentially methylated CGCGCGGs (dmCGCGCGGs) between plasma and WBC. The x-axis shows the fold change of the average methylation value between plasma cfDNA and WBC gDNA, and the y-axis shows the q value as the FDR analog of the P value (−Log 10 (q value)) for a two-tailed MWW test of differences between two groups. The horizontal gray line indicates statistical significance (FDR < 0.05). c A heatmap showing methylation of the dmCGCGCGGs in different tissues and 14 paired WBC and plasma samples. d, e Boxplots of the dmCGCGCGGs showing d the methylation values (methylated alleles per million mapped reads, MePM) of MCTA-Seq or e the methylation levels in the published DNA methylome data 77 CGCGCGGs were found to be significantly differentially methylated (dmCGCGCGGs) in the plasma cfDNA (average methylation (alleles per million mapped reads, MePM) fold changes of cfDNA vs WBC > 5, false discovery rate (FDR) < 0.05, two-tailed Mann-Whitney-Wilcoxon (MWW) test, . In contrast, only 8 dmCGCGCGGs were identified in WBC gDNA.
To obtain the tissue-specific methylation pattern, we performed MCTA-Seq on seven major tissues, including the lung (n = 2), stomach (n = 2), colon (n = 2), kidney (n = 2), pancreas (n = 2), muscle (n = 2), and skin (n = 2). The data of the normal liver tissues (n = 3) were reported in our previous study. Notably, we found that most plasma dmCGCGCGGs were hypermethylated in the liver . The methylation values of these dmCGCGCGGs in the liver were substantially higher than those in other tissues (MePM: median 33.2 in the liver vs 0.4 in the lung, 2.1 in the stomach, 0.6 in the colon, 1.0 in the kidney, 1.1 in the pancreas, 0 in the muscle, 0.4 in the skin, and 0.1 in WBC, P < 0.01, two-tailed MWW test, . We also validated the methylation values of these regions by using published DNA methylome data. The results showed that these dmCGCGCGGs in cfDNA have a median methylation level of 42% in the liver, which is considerably higher than that in the seven other tissues (P < 0.01, two-tailed MWW test, . In addition, these dmCGCGCGGs have low methylation levels in other tissues, including brain and heart (0% in the brain, 0% in the heart).
## Identification of tissue-specific methylation markers
To verify the results and to determine whether any other tissues also contribute detectably to cfDNA, we used a complementary approach that first identified the tissuespecific methylation CGCGCGG markers and then examined whether they are differentially hypermethylated in cfDNA. MCTA-Seq was performed on an additional set of 29 WBC samples as a training group for selecting markers with very low methylation levels in WBC (90th percentile = 0, MePM).
We identified a total of 146 tissue-specific methylated CGCGCGGs, among which 59,, and 11 markers are for the liver, the stomach, the colon, the kidney, the pancreas, the skin, the lung, and the muscle, respectively (see the "Methods" section). Detailed information and data of these markers are provided in Additional file 2: Tables S2 and S3. Genomic views of representative markers validated by published DNA methylome data were shown in Additional file 1: . We then compared the methylation values of these markers between 14 pairs of plasma and WBC samples. To exclude the effects of sequencing depth, we also normalized the data of each pair of matched cfDNA and WBC gDNA to the same depth (Additional file 2: . The results showed that most liver-specific markers (80%, 47 out of 59) were significantly differentially methylated in cfDNA (P < 0.01, two-tailed MWW test, , confirming that the liver prominently releases DNA into the plasma under physiological conditions. None of the other tissues showed clear evidence of release based on the analysis of individual CGCGCGG markers.
To maximize the detection sensitivity, we then combined the CGCGCGG makers for each tissue by adding up their methylation values. As expected, combining the liver markers resulted in more significant hypermethylation differences in the cfDNA vs WBC (P = 4.5e−6, twotailed MWW test, . Interestingly, we found that combining 8 lung markers revealed significantly differential hypermethylation in plasma cfDNA vs WBC (P = 1.7e−3, two-tailed MWW test, . The difference was still significant when combining any 7 of 8 markers, indicating that it was not due to any individual marker (P < 0.01). In contrast, no significant release was found for any other tissues (P > 0.1).
Together, the systematic comparison and the tissuespecific marker analysis indicated that the liver is the major examined non-hematopoietic tissue that contributes to plasma cfDNA in healthy adults.
## Deconvolution analysis for plasma of healthy individuals
Next, we sought to quantify the non-hematopoietic tissue DNA fractions in plasma cfDNA using deconvolution analysis. All tissue-specific CGCGCGG markers except for one stomach marker, which showed high methylation levels in 14 WBC samples, were used to set up the deconvolution algorithm (see the "Methods" section). To test the accuracy and sensitivity of the approach, we performed simulation analysis to generate a series of synthetic datasets with the tissue fraction ranging from 0.25 to 16%. The results showed linear regression lines between the simulated fractions and the estimated fractions with R 2 greater than 0.95 for most tissue types and Additional file 1: . The sensitivities, which were determined by the lowest distinguishable tissue fractions from WBC, were 0.25% for the liver and pancreas; 0.5% for the colon, kidney, and skin; and 2% for the lung, stomach, and muscle (P < 0.05, two-tailed t test, see the "Methods" section). Analyzing the synthetic data generated by mixing two tissue types, e.g., the liver and the pancreas, into WBC showed that DNA fractions of each tissue type were accurately estimated .
We next calculated the percentage of DNA derived from the liver and other tissues in the plasma cfDNA of healthy individuals. We analyzed all 25 plasma samples obtained from healthy adults with measured cfDNA concentrations in our previous study. The deconvolution analysis showed that the liver contributed a median fraction value of 1.3% (interquartile range (IQR) 0.48-2.0%) of plasma cfDNAand Additional file 2: . All other tissues yielded a median fraction of 0. We further calculated the absolution copies of the tissue DNAs and determined that the median absolute copy of liverderived DNA is in these healthy individuals was 34 haploid genomic equivalents (GE) per milliliter plasma (IQR, 15-5;, Additional file 2: . The liver-derived DNA fraction was not related to gender or age (Additional file 1:.
## Deconvolution analysis for plasma from patients of liver diseases
Then, we examined plasma cfDNA obtained from the liver disease patients including cholelithiasis patients (n = 13), . The cholelithiasis patients include those suffering from cholecystolithiasis (n = 8), choledocholithiasis (n = 3), and hepatolithiasis (n = 2) (Additional file 2: . The data of the cirrhosis and HCC patients were reported in our previous study.. Notably, two hepatolithiasis patients (BLD5 and BLD6), who also showed evidence of liver damage with elevation of the liver enzyme alanine aminotransferase (ALT; normal range 9-50 U/L; 154 and 110 U/L for BLD5 and BLD6, respectively, Additional file 2: , displayed prominently higher levels of liver-derived DNA (817 and 649 GE/mL, or 30% and 25% for BLD5 and BLD6, respectively) comparing with patients suffering cholecystolithiasis and choledocholithiasis patients.
The methylation values of the liver-specific markers uniformly increased in these patients, indicating authentic signals. Consistent with previous reports, we found that the total cfDNA concentrations were significantly higher in both liver cirrhosis and HCC patients in comparison with healthy individuals (median 7.2, 13.3, 29.9, 39.0, 20.1 ng/ mL for the cholelithiasis patients, the liver cirrhosis patients, and the HCC patients with small, middle, and large tumors, respectively, vs 6.5 ng/mL for healthy individuals, two-tailed MWW test, Additional file 1:and Additional file 2: . We performed a deconvolution analysis to investigate the tissue of origin of the elevated cfDNA. The results interestingly indicated that the blood cells were the major contributor to the cfDNA increase of both the liver cirrhosis and HCC patients. For the cirrhosis patients, we consistently did not detect an increase of the liver-derived DNA. For the HCC patients, though both the absolute and fractional liver-derived DNA prominently increased, the blood cells still made a major contribution to the elevated cfDNA level in most cases.
## Deconvolution analysis for plasma from acute pancreatitis patients
Elevated total cfDNA concentrations in acute pancreatitis (AP) patients have been reported to correlate with disease severity. However, the tissue of origin of the increased cfDNA is unknown. We therefore performed MCTA-Seq on plasma cfDNA obtained from eight AP patients, with six patients suffering from severe AP and two patients with mild AP (Additional file 2: Tables S6 and S8). Consistent with previous reports, we found that total cfDNA concentration was notably higher in the AP patients in comparison with the healthy individual (AP vs healthy individual: median 15 vs 6.5 ng/mL, P < 0.01, two-tailed MWW tests;. Those patients with severe AP had a prominently higher cfDNA level than the patients with mild AP (PA5 and PA8;, c; Additional file 2: , which is consistent with the previous report. Unexpectedly, we only detected a clear increase in pancreas-derived DNA in patient PA4 representing 652 GE/mL or 2.8% cfDNA. The deconvolution analysis indicated that the blood cells made a major contribution to the elevated cfDNA levels. We also detected the increase in the liver-derived DNA in acute biliary pancreatitis patients (Additional file 2: Tables S5 and S8). The results of all pancreatitis and cholelithiasis patients showed that the absolute but not fractional liver-derived DNAs correlated well with ALT levels, suggesting that the absolute value is a better indication of the liver cell death than the fractional value due to the extra release from the blood cell (Spearman correlation 0.94 vs 0.58;and Additional file 1:. HCC(>5cm) patients P 5 8 P 1 1 3 P 5 9 P 7 4 P 8 0 P 4 4 P 8 1 P 3 6 P 5 1 P 1 1 4 P 5 7 P 1 1 5 P 1 0 P 6 0 P 7 7 P 6 2 P 3 7 P 8 5 P 1 6 6 P 1 3 4 P 3 9 P 4 1 P 5 0 P 1 0 9 P 9 2 P 1 1 9 P 3 P 1 0 1 P 2 9 P 1 5 8 Healthy individuals Tissue-derived DNA (GE/mL) P 7 0 P 7 1 P 8 8 P 8 9 P 9 6 P 9 8 P 6 9 P 6 6 P 5 5 P 6 8 P 6 7 P 9 4 P 6 5 P 6 1 P 5 4 P 7 3 P Together, we have detected liver-or pancreas-derived DNA in the plasma from patients of corresponding diseases. Our results indicated that the blood cells made a major contribution to the elevated cfDNA levels in the liver cirrhosis, HCC, and AP patients.
## A novel set of tissue-specific hypermethylation markers for cfdna detection
Recent studies have identified tissue-specific hypomethylation cfDNA markers at the promoter regions of tissuespecific genes. Interestingly, we found that 15 of 146 tissue-specific hypermethylation markers were located within tissue-specific highly expressed genes and showed a positive correlation between the methylation value and gene expression. These genes included many typical cell type-specific genes such as the hepatocyte-specific F7, F12, and TFR2 genes, which encode the coagulation factor VII, coagulation factor XII, and transferrin receptor 2, respectively. Instead of being located at the promoter region as the hypomethylation markers, most of these hypermethylation markers were intragenic CpG islands. Genomic views of these markers validated using published DNA methylome data were shown inand Additional file 1:. Hence, our results indicated that the intragenic CGIs of tissuespecific genes can be candidate target regions for identifying tissue-specific hypermethylation markers suitable for cfDNA detection.
# Discussion
In the present study, we have used MCTA-Seq to establish a deconvolution algorithm for analyzing the tissue of origin of plasma cfDNA from healthy individuals and patients of liver and pancreatic diseases. We have also identified novel tissue-specific methylation markers useful for cfDNA detection.
For healthy individuals, our results showed that, among all eight examined non-hematopoietic tissues, the liver is the major one contributing to plasma cfDNA corresponding to a median percentage of 1.3% or 33 GE/mL in healthy adults; in contrast, other tissues give minor contributions. The result is consistent with recent studies using individual liver-specific DNA methylation markers and the DNA methylation array. During the. Their results interestingly indicated that vascular endothelial cells and hepatocytes are the two major non-hematopoietic sources of plasma cfDNA from healthy individuals. The MCTA-Seq should be performed to vascular endothelial cells in the future.
These findings clearly demonstrate that the liver releases a substantial amount of DNA into the blood in healthy adults. This reflects the fact that the liver is the largest organ in the body, with abundant blood circulation. These results suggest a daily turnover of hepatocytes in healthy individuals. Assuming that the volume of distribution of DNA is 60-70 mL/kg, the concentration of cfDNA is 6 ng/mL and corresponds to a genomic DNA content of 1000 diploid cells per milliliter, with a half-life of approximately 1 h. The contribution of 1~2% of total plasma cfDNA of a 70-kg person means that approximately 1.2 × 10 6 hepatocytes undergo apoptosis or necrosis with the DNA releasing to the blood per day, accounting for 0.0005% of all~2.4 × 10 11 hepatocytes in the liver. This number is also largely consistent with the estimation that approximately 7.5 × 10 6 liver stem cells divide per day. Even though organs such as the colon and skin have high daily cellular turnover, we and others did not detect their DNA in the plasma, given the fact that dead cells in the mucosa of the intestinal tract and in the epidermis of the skin generally fall off of the body. We also detected potential lung-derived cfDNA in healthy individuals. However, since our method has relatively low sensitivity for detecting lung-derived DNA, we are not able to draw a conclusion in the present study.
We have detected increases in liver or pancreas-derived DNA in patients of corresponding diseases, which revealed potential clinical applications of the method. The correlation between the levels of liver-derived DNA and the liver enzyme ALT is consistent with the report of Moss et al.. Comparing with the liver enzymes, the cfDNA-based assays may be useful for clinical detection of the liver damage with its distinct characteristic including a definite indication of hepatocyte death and quick clearance.
Our results interestingly indicated that the blood cells make a major contribution to the elevated cfDNA levels in the liver cirrhosis, HCC, and AP patients. Recently, by using an array-based DNA methylation deconvolution method, Moss et al. showed that the elevated cfDNA level in transplantation and sepsis patients was mostly derived from the blood cells. By studying the sexmismatched hematopoietic stem cell transplantation and liver transplantation patients, Suzan et al. also showed that exercise-induced increases in plasma cfDNA mainly originate from cells of the hematopoietic lineage. These results indicate that the blood cell is an important source of increased cfDNA levels in human diseases. There should be different causes for different diseases. While strong immune reactions may lead to the increase of leukocyte-derived cfDNA in the AP patients, destruction of blood cells by hypersplenism is likely the cause of elevated hematopoietic-derived cfDNA the liver cirrhosis patients. Cancer-related immune response and hypersplenism may jointly result in the increase of hematopoietic-derived cfDNA in the HCC patients, as most HCCs develop in cirrhotic livers.
Finally, we have identified 146 tissue-specific DNA methylation markers suitable for cfDNA detection. Since MCTA-Seq detected plenty of loci that are included in the commonly used DNA methylation array, it serves as a method for efficiently screening novel cfDNA methylation markers. Recent studies have demonstrated the utility of individual tissue-specific hypomethylation markers for detecting tissue cell death using cfDNA. Typical tissue-specific hypomethylation markers are present in the promoter regions of highly expressed tissue-specific genes, which are specifically unmethylated in the corresponding tissue, e.g., the insulin gene promoter, which is unmethylated in insulin-producing pancreatic β cells. In the present study, we identify a group of tissue-specific DNA hypermethylation markers that are suitable for cfDNA detection. In contrast to hypomethylation markers, these hypermethylation markers are located within the intragenic CpG islands of highly expressed cell type-specific genes, such as the hepatocytespecific F7, F12, and TFR2 genes. Combined with the information of gene expression, these loci are more explicit cell type-specific markers. The tissue-specific DNA methylation pattern of intragenic CGIs has been previously reported, and a mechanism of transcription-mediated DNA methylation has been shown. The work done by ourselves and others provides a rational basis for identifying new individual DNA methylation markers for the noninvasive detection of tissue-specific cell death.
# Conclusions
We have developed a MCTA-Seq deconvolution approach for simultaneously assessing the proportions of plasma cfDNA derived from multiple non-hematopoietic tissues. Applying this approach to healthy individuals and liver and pancreas disease patients revealed tissue of origin of plasma cfDNA. We have also identified novel tissue-specific cfDNA hypermethylation markers. The approach and the identified markers have many research and diagnostics applications in a broad spectrum of human diseases.
## Additional files
Additional file 1: Genome-scale DNA methylation analysis of tissue of origin of plasma cell-free DNA. . Schematic diagram of the study design and sample collection. . Genomic views of representative tissue-specific methylation markers. The red triangle indicates the CGCGCGG sequence. . Detection sensitivity of the MCTA-Seq deconvolution analysis. The DNA percentages of different tissues estimated using the MCTA-Seq deconvolution analysis were plotted against the varying percentages of A) the lung, stomach, colon, kidney, muscle, and skin. A linear fit was observed. The error bars represent means ± SD.. The relationship between liver-derived DNA fraction and A) gender or B) age.. MCTA-Seq deconvolution analysis of healthy individuals, liver disease patients, and acute pancreatitis patients. A) Boxplots showing cfDNA concentration (ng/mL) in the plasma of healthy individuals and liver disease patients. ***P < 0.01; *P < 0.1; nd, no difference. B) Correlation between liver-derived DNA fraction concentration and ALT in cholelithiasis and AP patients.
|
Ultrasound Elastography for the Evaluation of Lymph Nodes
The differential diagnosis of lymphadenopathy is important for predicting prognosis, staging, and monitoring the treatment, especially for cancer patients. Conventional computed tomography and magnetic resonance imaging characterize lymph node (LN) with disappointing sensitivity and specificity. Conventional ultrasound with the advantage of high resolution has been widely used for the LN evaluation. Ultrasound elastography (UE) using color map or shear wave velocity can non-invasively demonstrate the stiffness and homogeneity of both the cortex and medulla of LNs and can detect early circumscribed malignant infiltration. There is a need of a review to comprehensively discuss the current knowledge of the applications of various UE techniques in the evaluation of LNs. In this review, we discussed the principles of strain elastography and shear wave-based elastography, and their advantages and limitations in the evaluation of LNs. In addition, we comprehensively introduced the applications of various UE techniques in the differential diagnosis of reactive LNs, lymphoma, metastatic LNs, and other lymphadenopathy. Moreover, the applications of endoscopic UE and endobronchial UE are also discussed, including their use for improving the positive rate of diagnosis of fine-needle aspiration biopsy.
# Introduction
Various benign and malignant disorders can result in lymphadenopathy; the differential diagnosis of lymph node (LN) is important for predicting prognosis, staging, and monitoring the treatment. Conventional computed tomography (CT) and magnetic resonance imaging (MRI) characterize LN relying on size and topographic distribution, but with disappointing sensitivity and specificity, since it is not rare that malignant LN infiltration occurs in normal-sized LN.
Conventional ultrasound (US) with the advantage of high resolution has been widely used for imaging superficial organs, particularly for the LN evaluation. Compared with conventional CT and MRI, B-mode US can provide more detailed information on shape, contour, inner texture, maximum short axis diameter, long to short axis ratio, absence of hilus, and presence of necrosis. Color Doppler US and spectral Doppler US can image the hemodynamic characters of LN and add values for the differentiation of malignant from benign LNs. Benign LNs often show hilar predominant vessel architecture and have lower resistive index (RI), while malignant LNs usually show peripheral or mixed vascularity and disappearance of hilar vascularization and have higher RI. However, Doppler techniques have limitations in small LN since the vascularity is often undetectable.
US elastography (UE) is a new technique that uses color map or shear wave velocity (SWV) to non-invasively demonstrate stiffness and homogeneity. It has rapidly become one of the most popular US-based techniques. Clinically, it can be used in the early detection and differential diagnosis of focal diseases; in improving the accuracy for diagnosing diffuse diseases, such as fibrosis and atherosclerosis; and in the assessment of response to treatments, such as thermal ablation and chemotherapy.
UE is able to demonstrate the stiffness of both the cortex and medulla of LNs and to detect early circumscribed malignant infiltration. Studies have been published on the evaluation of LNs by strain elastography (SE) or shear wave elastography (SWE). This review aims to comprehensively discuss the current knowledge of the applications of various UE techniques in the evaluation of LNs.
## Principles and techniques of ultrasound elastography
UE is a technique in which the stiffness of the tissue can be imaged as color map or SWV. The principle of UE is based upon tissue reactions, such as changes in displacement, strain, or speed, by applying an external or internal static (quasi-static) or dynamic excitation. Differences in tissue reactions are calculated, identified, and reflected by computers.
Depending on the type of excitation applied, UE is classified into two categories, i.e., 1) SE, which is composed of static or quasi-static strain imaging and acoustic radiation force impulse (ARFI) imaging; and 2), SWE which is composed of SWV measurement and SWV imaging .
## Strain elastography
Technique SE includes static/quasi-static imaging and ARFI imaging. It is based upon the fact that hard tissue is more difficultly compressed than soft tissue. SE is a technique that measures tissue deformation generated by compression, which may be applied with a probe on the body surface for static/quasi-static imaging and may also be applied with acoustic radiation force for ARFI imaging. The tissue deformation is measured by US system and displayed as a color or gray map. On the screen of the US system, both the B-mode image and corresponding elastography image could be simultaneously displayed.
The parameters commonly used to indicate tissue hardness include elasticity score and strain ratio (SR). The elasticity score indicates the strain (with color or brightness) distribution within a selected area. The SR refers to the ratio of strain between area A (usually a mass) and area B (usually a normal surrounding tissue, fat, or muscle tissue) within the region of interest (ROI).
# Advantages and limitations
SE, especially static/quasi-static imaging, is suitable for superficial organs and thus is the most commonly used method for the evaluation of superficial LNs. The operation method of SE is simple, and the operation skills can be mastered in a short time of training. However, SE is a qualitative analysis technique, and it is not able to analyze tissue hardness quantitatively. The performance of static/quasi-static imaging FIGURE 4 | Ultrasound elastography (UE) techniques. UE techniques can be classified by the type of excitation applied: 1) strain elastography (top) and 2) shear wave elastography (bottom). Excitation methods of strain elastography include constant force-induced displacement (static/quasi-static imaging) or acoustic energyinduced physiologic motion (ARFI). Excitation method of shear wave elastography where the shear waves are produced by a transducer. Shear wave elastography is classified as transient elastography (TE), point shear wave elastography (pSWE), two-dimensional shear wave elastography (2D-SWE), and three-dimensional shear wave elastography (3D-SWE), according to different measurement and imaging methods. is not good at analyzing the deep LNs. Moreover, SE is user dependent and subjective.
## Shear wave speed measurement
Techniques Shear wave speed measurement technology is a method to generate shear waves and measure SWV. Based on the principle of fast propagation of shear wave speed in hard tissue and slow propagation in soft tissue, the hardness of tissue is indirectly reflected by measuring shear wave speed. Shear wave speed can be converted to Young's modulus by Young's model formula:
[formula] E = 3rC 2 [/formula]
where E represents stiffness (Young's modulus [kPa]), r is the density (kg/m 3 , approximately equal to 1), and C is the shear wave speed (m/s).
The shear wave speed measurement techniques mainly include transient elastography (TE) and point SWE (pSWE). TE is the first shear wave speed measurement technology applied in clinical practice, but it is only used in the liver so far; therefore, this technique is not discussed in this review. The principle of pSWE is similar with ARFI: the probe applies an acoustic radiation force to the ROI of the tissue and generates transverse vibration shear waves. The receiver can detect the speed of shear wave in ROI, which is expressed by speed or by kPa value through Young's model formula.
# Advantages and limitations
pSWE can detect both deep organs (the liver, etc.) and superficial organs (the thyroid, etc.), and therefore, this technique is suitable for both superficial and deep LNs. However, the ROI is with fixed size; it can only measure one part of a LN but may be too large if the LN is very small.
## Shear wave speed imaging
## Techniques
The principles of shear wave speed imaging are that the US probe sends out the multipoint focused acoustic radiation force pulse, which makes the tissues at different depths along with the acoustic axis shift at almost the same time, producing plane shear wave, and then the image processing technology detects the SWV, forms color image, and calculates Young's modulus (elasticity index (EI)). So compared with that of pSWE, the size of ROI in 2D-SWE can be adjustable. Some US diagnostic instruments are equipped with 3D probes with high-speed acquisition capability of mechanical scanning 2D-sensor sequences, which can conduct 3D reconstruction of tissue hardness.
# Advantages and limitations
The diagnosis of shear wave speed imaging is less influenced by the operator's experience and operation than SE, because it does not rely on freehand compression. It can display the conventional US images and elastic US images synchronously and measure SWV in real time. However, multicenter studies have shown that the repeatability of shear wave elastic imaging is affected by the size, location, depth, and other factors.
## Clinical applications of elastography
## Reactive lymph nodes (inflammation)
Acute or chronic inflammation is the prime cause of LN enlargement. The elastographic architecture of LNs is kept in most inflammatory processes. Therefore, like in normal LNs, the cortex is also stiffer than the hilum in inflammatory LNs.
## Strain elastography
Both elasticity score and SR have been studied to evaluate the stiffness of reactive LN. Firstly, due to the lack of a unified classification method for US elastograms, different researchers classified US elastograms of LNs into a 4-point, 5-point, 6-point, 7-point, or 8-point rating scale. Secondly, some researchers compared the strain in target region with adjacent reference region to differentiate benign from malignant LNs.
Lyshchik classified US elastograms of LNs with a 4-point rating scale according to visibility, brightness compared with surrounding neck muscles, regularity, and definition of outline. Several studies classified elastograms of the LNs into five patterns according to relative distribution and ratio of soft or hard regions of the LN: pattern 1, absent or very small blue (hard) areas; pattern 2, total blue areas of less than 45%; pattern 3, total blue areas of greater than 45%; pattern 4, peripheral blue area and central green (soft) area; and pattern 5, blue area with or without a green rim. Tan et al. found that 87.9% of benign LNs manifest pattern 1 or pattern 2 (3).
Besides, Lyshchik defined the surrounding neck muscles to LN SR as strain index; using strain index value of <1.5 in benign LN classification, SE showed 79% accuracy, 85% sensitivity, and 98% specificity. Acu et al. calculated each LN with mean strain index. With the use of strain index value of <1.7, SE differentiates benign LNs from malignant ones with 75% accuracy, 71.6% sensitivity, and 76.5% specificity. Özel et al. reported that elastography SRs were lower in benign LNs than malignant LNs. Many studies have shown that SE has potential diagnostic value in lymphadenopathy; however, high user dependence is the limitation, especially using SR. Adjacent reference region was selected differently for SR measurement of LNs in different regions; in general, muscles as adjacent reference tissues were usable in cervical region, and fat tissue as an adjacent reference region may be a good choice in the axilla.
## Shear wave-based elastography
Compared with SE, shear wave-based elastography is regarded as potentially more objective. In most published researches, virtual touch tissue imaging (VTI) grade and SWV of ARFI imaging were used to evaluate reactive LNs, and the diagnostic performance of VTI is higher than that of SWV. In a study including 263 pediatric LNs, Bayramoglu et al. found that median elasticity and velocity values were higher in reactive LNs compared with normal LNs; with the use of the cutoff median elasticity and velocity values of >15 kPa and 2.24 m sn −1 for differentiating reactive LNs from normal LNs, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy were 27%, 96%, 82%, 74%, and 74% and 25%, 97%, 82%, 73%, and 74%, respectively. Many studies should be conducted on the evaluation of reactive LNs by shear wave-based elastography to explore the significance of SWE in the evaluation of reactive LNs and to analyze the potential factors affecting SWE imaging.
## Malignant lymph nodes (carcinoma)
Malignant tumor cells proliferate rapidly, causing internal pressure and increasing tissue stiffness in LNs. Therefore, the elastographic architecture of LNs changed compared with reactive LNs. Typically, the well-differentiated carcinoma initially infiltrates LNs in a circumscribed manner (focally stiffer and harder), whereas the undifferentiated carcinoma leads to a diffuse (mostly or completely stiffer and harder) infiltration.
## Strain elastography
Several pilot studies have evaluated the ability of SE to detect LN metastases in the cervical or axillary LNs. Both elasticity score and SR have been studied, which showed that SE and conventional US may play complementary roles in differentiating malignant LNs and assessing the risk of metastatic LNs. Firstly, suspected cervical LN metastases from hypopharyngeal and thyroid carcinomas have been recently investigated using SE (real-time elastography (RTE)). An EI has been created by comparing the elasticity of the LN with the surrounding head and neck muscle tissue (muscle to LN SR). With the use of a ratio of >1.5 as an indicator of malignant infiltration, the sensitivity was 85% and the specificity was 98%, which are superior to the best B-mode criteria. These data have been reproduced by Tan et al. Moreover, inter-observer agreement with SE was very high (kappa 0.88-0.946) (3).
Secondly, some researchers qualitatively classified US elastograms of LNs into a 4-point, 5-point, 6-point, 7-point, or 8-point rating scale. Metastatic LNs were mostly evaluated to 3-4 points in a 4-point rating scale. Suzanfound that the sensitivity, specificity, and accuracy of RTE in differentiating benign LN from squamous cell carcinoma and malignant melanoma group were 91%, 70%, and 86%, respectively.
In a 5-point rating scale, Tan et al. reported that 50 of malignant and 74.5% of metastatic LNs manifested pattern 3 or 4, while all primary malignant LNs manifested pattern 2 (3). In another study including 97 axillary LNs, using the criteria of score 1 and 2 as benign and scores 3, 4, and 5 as metastatic, the sensitivity, specificity, PPV, NPV, and accuracy were 78%, 93%, 93%, 79%, and 86%, respectively. Although qualitative strain methods based on elasticity score and SR have been widely studied all over the world for axillary and cervical LNs (17), SR >1.5 or hard composition over 50% can be a good indicator of malignancy. However, as compared with SWE, its dependence on operators cannot be overcome, and absolute quantitative elastic measurement cannot be provided; and for LNs with deep vertical distance and small volume, the judgment of RTE on LNs hardness is prone to false-positive results, which affects the accuracy of SE.
## Shear wave-based elastography
Clinically and theoretically, SWE seems to be an effective, quantitative tool for differential diagnosis of malignant and benign LNs in many researches, especially in small LNs. Based on previous researches, the maximum SWV (2.93 m/s) (23) and elastic value ratiocan be used as reliable indices to predict benign and malignant lymphatic nodes. KılıçA et al. conducted a prospective study comparing conventional US with VTI quantification (VTIQ), and when using a cutoff value of 3.03 m/s, VTIQ differentiates malignant LNs from benign ones with 75% accuracy, 93% sensitivity, and 59% specificity.
Some researchersqualitatively classified SWE images of axillary LN (ALN) into 4-point patterns, which was similar to SWE patterns of breast lesions (31): color pattern 1, homogeneous pattern; color pattern 2, filling defect within LN; color pattern 3, homogeneous within LN with a localized colored area at the margin; and color pattern 4, filling defect within LN with a localized colored area at the margin. The benign ALNs usually manifest color pattern 1, while ALN metastases (ALNMs) usually manifest color patterns 2-4, and the sensitivity, specificity, PPV, NPV, and area under the receiver operating characteristic (ROC) curve (AUC) were 96.7%, 100%, 100%, 96.8%, and 98.3%, respectively. In addition, Luo et al.and Lin et al.directly compared the diagnostic performance of qualitative and quantitative SWE, and they found that qualitative SWE had better diagnostic performance than quantitative SWE in detecting ALNM.
However, a meta-analysis compared the diagnostic performance of qualitative elastography with quantitative elastography for ALNM in breast cancer and found that quantitative and qualitative elastography had similar diagnostic performance and good clinical utility. More studies with SWE should be conducted to get more reliable cutoff values of SWV and elastic value ratio in different sites.
## Lymphoma
Lymphomas are a primary malignant tumor of LNs, lymphoid tissues outside LNs, and mononuclear macrophage system. Because of a highly heterogeneous group of lymphoproliferative malignancies, the biological behavior and pathological types of lymphomas are different, especially for non-Hodgkin's lymphoma. But the incidence of lymphomas represents approximately 4%, and newly diagnosed cancers increases each year; moreover, lymphomas are more commonly seen in developed countries, which may seriously endanger people's health. Knowledge of elastography in lymphoma is very limited. So far, different lymphomas cannot be differentiated. Initial experience suggests that focal LN infiltrationis indicative of low-grade follicular lymphoma, whereas diffuse and homogenous LN infiltration is typically found in high-grade lymphoma.
## Strain elastography
Few studies have reported on the evaluation of lymphoma with SE. With a 5-point rating scale of US elastograms of LNs, Acu et al. reported that most lymphoma manifested patterns 1 and 2. Clinically and theoretically, the stiffness degree of lymphoma is different from that of metastatic and benign LNs. In most studies, the hardness of the lymphoma was low. Thus, when metastasis and lymphoma were considered as positive, reactive LNs were considered as negative in the differential diagnosis; and the sensitivity, specificity, and accuracy with a point rating scale of US elastograms were affected. With quantitative analysis of SE, elasticity parameter strain index showed high diagnostic accuracy for distinguishing lymphoma from lymphadenitis; the cutoff value of the strain index of the cervical LNs compared with sternocleidomastoid muscle has been reported to be 1.18 in a recent study. Though it is difficult to differentiate different lymphomas, the treatment effect evaluation with SE in Hodgkin's disease may be useful.
In the study on the efficacy of refractory and recurrent Hodgkin's disease, it has been shown that the hardness of some lymphoma nodules changes with the treatment effect. It indicates that SE could be reliable for therapy response monitoring of Hodgkin's lymphoma.
## Shear wave-based elastography
Currently, there are few studies on the evaluation of lymphoma by SWE. The number of enrolled lymphoma in these studies was small, including several case reports. Soo et al. qualitatively categorized shear speed map in a of total five SWE patterns in cervical LNs: pattern 1, absent or very small red (stiff) area; pattern 2, small scattered red areas, which mean total red area less than 45%; pattern 3, large red area, equal or more than 45%; pattern 4, peripheral red area and central green (soft) area, suggesting central necrosis; and pattern 5, almost red area with or without a green rim. None of lymphoma manifested pattern 4 and pattern 5; and absolute values and ratio of both elasticity and speed were significantly lower in lymphomas than metastatic LNs. Based on a recent study in pediatric LN with quantitative evaluation of SWE, elasticity values higher than 17 kPa and velocity values higher than 2.45 m/s would be considered as lymphoma rather than lymphadenitis in an enlarged LN with at least a 91% diagnostic accuracy. Several case reports have evaluated uncommon different lymphomas with SWE. A report used SWE to evaluate primary B-cell lymphoma of the breast. The study showed that the mass of primary B-cell lymphoma on SWE was considerably stiff but softer than typical invasive ductal cancers. In the future, a prospective study with large-scale samples should be conducted to investigate quantitative or qualitative SWE features of primary B-cell lymphoma.
## Other lymphadenopathy
In the preliminary study of reactive and metastatic LNs, the AUC for combined evaluation is 0.97, which is much higher than that for B-mode US or elastography alone. The analysis of parameters can be used to quantitatively evaluate the characteristics of different LN diseases; it shows that LNs of tuberculosis (TB) are softer than metastatic LNs but harder than benign LNs. However, LNs of TB have a wide range of stiffness; the stiffness is related with internal structures, increased fibrous tissue and calcification can account for high stiffness, and liquefaction necrosis can decrease the stiffness. Cheng et al. found that only 50% LNs of TB can be correctly diagnosed by elastography.
In further studies, the combination of B-mode US and elastography may have important clinical value in differential diagnosis. Few researches have been done on relapsing or chronic lymphadenitis or rare benign diseases such as Kikuchi or Kimura disease (KD). The research shows that the LNs with KD show malignant signs in conventional US, but benign signs in SE; therefore, SE can help patients avoid unnecessary needle biopsy and inappropriate treatment. In a study of children's cervical LNs, the stiffness of the largest LNs in patients with bacterial cervical lymphadenitis (BCL) was significantly higher than that in patients with LN-first presentation of Kawasaki disease (NFKD) and healthy children, with a cutoff of 14.55 kPa; the sensitivity, specificity, and AUC were 89%, 76%, and 88.5%, respectively. So SWE is a potential method to differentiate early NFKD.
## Applications of endoscopic ultrasound and endobronchial ultrasound elastography
Endoscopic US (EUS) and endobronchial US (EBUS) are important tools to assess the digestive tract and surrounding organs, but the limited capacity to determine the exact pathological results is the major limitation. As a non-invasive technique, EUS and EBUS elastography have been proven to be able to provide complementary stiff information added to conventional EUS and EBUS imaging, becoming promising examination methods to differentiate benign from malignant LNs.
## Differentiation of benign and malignant lymph nodes
Recently, an increasing number of literatures focused on the use of EUS and EBUS to diagnose mediastinal LNs and peritoneal lymphadenitis.
EUS elastography was originally used for the differential diagnosis of pancreatic lesions. Studies on the difference between benign and malignant pancreatic masses and LNs by SE showed that EUS elastography had more advantages than conventional US.
Similar to superficial LNs, physiological and reactive peritoneal LNs manifest homogeneous or scattered soft pattern with delineated vascular structures of LN hilum. And the LN medulla may manifest as slightly softer than the LN cortex. Malignant LNs are the most characterized by a homogeneous hard elastographic pattern, especially in diffuse metastatic infi ltration; however, malignant LNs may display inhomogeneous but hard patterns because of incomplete metastatic infiltration and focal necrosis. More and more studies differentiated benign from malignant LNs with EUS; most of them were qualitative with elastographic histogram, using EUS-fine-needle aspiration biopsy (FNAB), histology, and/or surgical pathology as a reference standard.
Multiple studies have demonstrated that EUS and EBUS elastography can effectively identify the benign and malignant mediastinal and peritoneal LNs(47-51, 53-59). In addition, under the guidance of elastographic imaging, EUS-FNAB or EBUS-FNAB can improve the positive rate of diagnosis and avoid false-positive results.
With qualitative analysis of elastographic histogram, elasticity pattern and SR have been studied to evaluate the stiffness of LNs. Giovannini et al. firstly evaluated the ability of EUS elastography to differentiate benign from malignant LNs with elasticity pattern in 2006. In this color-coded scale of elastographic patterns, yellow means normal tissue, green fibrosis, red fat, and blue malignant tissue. They conducted a multicenter study in 2009 and found improved specificity of 82.5% compared with 50% in the previous study. What is more, the sensitivity, specificity, PPV, NPV, and global accuracy of EUS elastography were 91.8%, 82.5%, 88.8%, 86.8%, and 88.1%, respectively, which were significantly better than the respective parameters of B-mode. In a study including 40 patients with a 4-point elasticity score, using the criteria of elasticity scores 1 and 2 as benign and elasticity scores 3 and 4 as malignant, the sensitivity, specificity, PPV, NPV, and diagnostic accuracy were 87.5%, 41.7%, 83.3%, 50%, and 60%, respectively.
Besides, some researchers qualitatively classified EBUS elastograms into three patterns: pattern 1, predominantly non-blue (green, yellow, and red); pattern 2, partly blue and non-blue (green, yellow, and red); and pattern 3, predominantly blue. With the use of the criteria of pattern 1 elastogram as benign and pattern 3 as malignant for differentiating malignant and benign mediastinal LNs with EBUS elastography, the sensitivity, specificity, PPV, NPV, and diagnostic accuracy were 90.6%, 82.6%, 71.6%, 94.7%, and 85.2%, respectively. But the central necrosis within malignant LNs and the fibrotic component within benign LNs may influence the accuracy of elastographic evaluations. What is more, the definitions of elastography patterns were subjective and may be hard to repeat by other researchers.
When judging malignant LNs with SR, previous research showed that with the cutoff point of SR >2.. These studies reported the SR was more accurate than conventional EUS or EBUS, and EUS elastography combined with other sonomorphologic features is a potentially useful prognostic index differentiating malignant from benign LNs. Besides, a meta-analysis found that the sensitivity and specificity of UE in differentiating benign and malignant LNs were 88% and 85%, respectively. However, the SR was generally calculated by two selected target regions, which makes it hard to precisely represent the stiffness of the whole LN. Thus, some studies used software to semiquantitatively analyze the color distribution of LN elastogram. Nakajima and his colleagues analyzed 49 LNs with stiff area ratio; they found that the sensitivity and specificity were 81% and 85%, respectively, for predicting metastatic disease, using a cutoff value of 0.311 for stiff area ratios (57). Sun et al. used a software and transformed the elastographic image into gray scale, which varied form 0 (all red pixels) to 255 (all blue pixels). This method could calculate the mean gray value inside the target and reflect the stiffness of the targeted LN. They found that non-small cell lung cancer (NSCLC) showed a higher gray value than small cell lung cancer (SCLC) (201..
In the future, more studies should be conducted to compare the qualitative and quantitative EUS elastography in differentiating benign from malignant LNs, in order to find a more suitable, accurate method for clinical practice.
# Conclusion
UE is a promising method for measuring tissue hardness and has been widely used in differentiating reactive LNs, lymphoma, metastatic LNs, and other lymphadenopathy. Besides, EUS and EBUS elastography are non-invasive techniques and have been proven to be able to provide complementary stiff information for conventional EUS imaging; the positive rate of diagnosis of EUS-FNAB or EBUS-FNAB can be improved under the guidance of elastographic imaging. There are some studies that used elastography in cervical, axillary, mediastinal, and peritoneal LNs, but further studies with unbiased large-scale samples in different sites are still required. Also, the direct comparison between qualitative and quantitative elastography and new solutions for current elastographic limitations should be pursued. The current consensus for LNs diagnosis is that no single parameter has sufficient diagnostic performance, and the combination of UE and traditional US technology is conducive to the differential diagnosis of LNs. In conclusion, UE can aid in the differentiation of benign and malignant LNs and has immense potential clinical values.
# Author contributions
|
Extremely delayed solitary cerebral metastasis in patient with T1N0M0 renal cell carcinoma after radical nephrectomy
Rationale: Although renal cell carcinoma (RCC) is one of the common origins of brain metastasis, few cases of extremely delayed brain metastasis from RCC, more than 10 years after nephrectomy, have been reported. We present a rare case of extremely delayed brain metastasis from RCC, also performed a literature review to increase knowledge of the characteristics for extremely delayed brain metastasis from RCC.Patient concerns: A 72-year-old man presented with right-sided hemiplegia and dysarthria. The patient had a history of radical nephrectomy for RCC with stage T1N0M0 15 years earlier.Diagnosis: Magnetic resonance imaging with contrast revealed a 2-cm sized non-homogenous enhanced mass in the left frontal lobe with peritumoral edema. The pathological examination after surgery reported metastatic clear cell RCC.Interventions: A craniotomy for removal of the mass was performed at the time of diagnosis. Stereotactic radiosurgery was performed for the tumor bed 3 weeks after craniotomy, and then, chemotherapy was started 2 months after the SRS.Outcomes: Metastasis progressed to multiple organs 6 months after the craniotomy. The patient chose a hospice and no longer visited the hospital.Lessons: In cases with a history of nephrectomy for RCC, long period follow-up is necessary for monitoring RCC brain metastasis and pathologic diagnosis should be confirmed.Abbreviations: H&E = hematoxylin and eosin, MRI = magnetic resonance imaging, PET = positron emission tomography, RCC = renal cell carcinoma, SRS = stereotactic radiosurgery, TNM = tumor node metastasis.
# Introduction
Renal cell carcinoma (RCC) is the most common kidney cancer with an incidence of 2% to 3% of all malignant cancers in adults.It is observed that extremely delayed distant metastasis that occurs in other organs including the lung, bone, and liver 10 years after nephrectomy for RCC is not rare, with a prevalence of 4.7% to 11%.However, few cases of extremely delayed metastasis to the brain, occurring more than 10 years after the initial diagnosis of RCC, have been reported,and the mechanism of delayed metastasis is not clearly known. Here, we report a case of extremely delayed solitary brain metastasis of RCC with lymph node metastasis that occurred 15 years after nephrectomy.
## Case presentation
A 72-year-old man presented with right-sided hemiparesis and dysarthria. Magnetic resonance imaging (MRI) with gadolinium showed a 2 cm-sized non-homogeneous enhanced and round-shaped mass in the left frontal lobe with peritumoral edema. Fifteen years previously, he had undergone a right radical nephrectomy for a 6 cm-sized mass on the kidney, following the histopathologic diagnosis of clear cell typed RCCThe legal guardian has consented to submit of the case to the journal and to publishing his data and radiologic images. The human Investigation Committee (IRB) of our institute approved this study. The case report was written in accordance with COPE guidelines and complied with the CARE statement.
The authors report no conflict of interest concerning the materials or methods used in this study or the findings described in this paper.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. metastasis (TNM) system, which is the most commonly used staging system established by the American Joint Committee on Cancer.A whole-body positron emission tomography (PET) study performed at the time of the current presentation showed a hypometabolic lesion in the left frontal lobe due to peritumoral edema, which was considered as brain metastasis, and several hypermetabolic lymph nodes at station 4R (right lower paratracheal nodes) and 7 (subcarinal nodes).
The patient underwent brain surgery for histopathologic diagnosis and tumor removal. He underwent a frontal craniotomy and gross total resection of the tumor (Figs. 1E and 2A). The final histopathologic report revealed metastatic clear cell RCC with a Ki-67 index of 40%, which is primary in the kidney for both tumor and peritumoral tissue. Hematoxylin and eosinstained tissue showed clear cytoplasm and round-to-oval-shaped nuclei. Three weeks after the craniotomy, stereotactic radiosurgery (SRS) was performed for the tumor bed because of the possibility of tumor cells based on the reports of the biopsy. There was no evidence of recurrence on a follow-up brain MRI performed 4 months after surgery. Two months after the SRS, he complained of chest pain in the right lateral side, and multiple bone metastases and left back muscle metastases were found on follow-up whole-body PET. Although chemotherapy was started at the oncology department, multiple distant metastases including the lung, liver, spleen, and adrenal gland were found 6 months after the operation. Subsequently, the patient chose a hospice and no longer visited the hospital.
# Discussion
RCC is the most common kidney cancer with an incidence of 2% to 3% of all malignant cancers in adults.According to the TNM staging system, distant metastasis indicates difficulty in expecting a good prognosis.In fact, patients with stage IV RCC, distant metastatic RCC, had less than 10% of a 5-year survival with a median overall time of 6 to 10 months.Extremely delayed distant metastasis of RCC, over 10 years after nephrectomy, is not very rare, with a prevalence of 4.7% to 11%.The most common metastatic sites of RCC are the lungs, lymph nodes, bone, and liver.In comparison, brain metastasis occurs in 3.9% to 24% of patients with RCC and is most frequently detected within an average of 1 to 3 years after the nephrectomy.Similar to the cases of distant metastasis to other parts other than the brain, the prognosis of brain metastasis from RCC is poor; the median overall survival time after a diagnosis of brain metastasis was 10.7 months, and the 5-year survival rate was 12%.To the best of my knowledge, a total of 20 cases of extremely delayed brain metastases from RCC have been reported in English so far,making it a rare occurrence; all 21 cases including the present case are listed in . The cases summarized in were searched by various combinations of search terms such as 'brain or cerebral', 'late or delayed', 'metastasis or metastatic', 'renal or renal cell carcinoma' and 'nephrectomy' in databases such as PubMed, Scholar Google, and Embase, and also referred to the lists summarized published paper. The median interval period from nephrectomy to brain metastasis diagnosis was 15 years (range, RCC is considered to be radio-and chemo-resistant.Thus, surgical total resection is a standard treatment option for patients with brain metastases.Of the 21 patients with extremely delayed brain metastasis of RCC, 18 patients underwent total resection and 3 patients underwent subtotal resection. Although the total number of cases was not large (21 cases), the proportion of cases with a good prognosis was much higher in cases of total resection. In 2 cases, the patients expired: 1 patient underwent total resection for a solitary metastatic lesion and expired due to systemic spread,and the other underwent subtotal resection for multiple metastatic lesions and expired without systemic spread.We present a case of systemic metastasis accompanied by lymph node metastasis at the time of brain metastasis diagnosis, although total resection and SRS were performed for a single lesion. In addition, Fukushima et alreported that even in the case of multiple brain metastases, a good prognosis can be expected through total surgical resection. Total resection could be quite effective for the local control of extremely delayed brain metastatic lesions. In addition, SRS is also known to be effective in local metastatic tumor control.There are several hypotheses about the mechanism of extremely delayed metastasis after nephrectomy for RCC. First, it is probable that the dissemination of tumor cells occurred before the nephrectomy and grew slowly.Second, the microscopic metastatic lesions remain dormant for decades and begin to grow when the host immunopotency decreases. Several basic studies using rodent models have shown that single tumor cells spread to distant sites early on and have a period of dormancy.Histopathologic confirmation is essential for the diagnosis. Bademci et aland Montano et alreported a metastatic RCC mimicking meningioma, which was initially diagnosed as a meningioma on radiologic imaging tests. The histopathological characteristics of RCC, especially the clear cell type which occupy the RCC, are clear cytoplasm with a high lipid content during histological preparation.In some reports, the MiB-1 labeling index was less than 1%or 7%as a cell proliferation marker, and in this case, the cell proliferation rate was comparatively high, with a Ki-67 index of 40%. In the present case, the progression of the systemic spread of RCC after the first diagnosis of metastasis was relatively fast compared to that in previously reported cases of extremely delayed brain metastases of RCC. Cell proliferation is thought to be related to the prognosis of metastatic RCC.
# Conclusion
We report a rare case of extremely delayed brain metastasis from RCC. If there is a history of RCC, it is necessary to conduct long-
|
School Feeding as a Protective Factor against Insulin Resistance: The Study of Cardiovascular Risks in Adolescents (ERICA)
# Introduction
Adolescence, which includes persons aged 10-19, is a transitional period characterized by important biological, cognitive, emotional, and social changes. Additionally, this period is marked by an increase in autonomy and independence in relation to their families and a growing desire for new behavior and experiences. Behaviors created and established during this period can be carried into adulthood, affecting several health-related issues, such as unhealthy practices and chronic diseases .
Schools are an important social environment for adolescents, where peer interaction, emotional control, behaviors, and attitudes related to health and food consumption are promoted, which are able to determine their health in adult life [bib_ref] Adolescence: A foundation for future health, Sawyer [/bib_ref]. In addition, school is an influencing environment, with an important role in health promotion, which can help reduce health inequities [bib_ref] The World Health Organization's Health Promoting Schools framework: A Cochrane systematic review..., Langford [/bib_ref].
Nutrition programs and policies are implemented through a healthy school environment by offering adequate meals and making it a health-promoting place, thereby improving health and nutritional status and food consumption and reducing the risk of chronic diseases, such as obesity and hypertension [bib_ref] The School Food Environment and Obesity Prevention: Progress Over the Last Decade, Welker [/bib_ref]. In Brazil, the National School Feeding Program (PNAE) serves all public schools offering planned, healthy, and free meals, which must provide 70% of the nutritional needs of those who study full-time and 20% of those who study part-time; thus, this program is considered one of the most important public health programs in the world. Some studies have investigated the relationship of specific policies in the school environment with effects on metabolic risk, eating behaviors, and adiposity [bib_ref] Effectiveness of school food environment policies on children's dietary behaviors: A systematic..., Micha [/bib_ref] , and have also shown that the school environment can be both protective and risky for cardiovascular disease [bib_ref] Coexistence of risk factors for cardiovascular diseases among Brazilian adolescents: Individual characteristics..., Silva [/bib_ref]. However, few studies have directly addressed the school food environment, including planned school meal consumption, and the potential effects of insulin resistance markers in a representative sample of adolescents.
According to the American Diabetes Association, there are some methods for the assessment of insulin resistance, such as the homeostatic model assessment for insulin resistance (HOMA-IR), but screening for type 2 diabetes mellitus in individuals with altered blood glucose is also recommended. In addition, early insulin resistance detection is important as part of the cardiovascular risk factors for metabolic syndrome (MS), which is responsible for the increase in cardiovascular disease and mortality [bib_ref] The metabolic syndrome and cardiovascular risk: A systematic review and meta-analysis, Mottillo [/bib_ref]. According to data from the Study of Cardiovascular Risks of Adolescents (ERICA in Portuguese) [bib_ref] Prevalence of metabolic syndrome in Brazilian adolescents, Kuschnir [/bib_ref] , approximately 2.8% of Brazilian adolescents from public schools had MS, with insulin resistance present in 19% of the sample, evaluated by HOMA-IR [bib_ref] Cutoff values for HOMA-IR associated with metabolic syndrome in the Study of..., Chissini [/bib_ref]. Furthermore, the prevalence of overweight and obesity was 17% and 8.4%, respectively, and 14.4% presented prehypertension [bib_ref] Prevalence of metabolic syndrome in Brazilian adolescents, Kuschnir [/bib_ref] [bib_ref] The Global Epidemic of the Metabolic Syndrome, Saklayen [/bib_ref]. Altogether, these data point to the potential cardiovascular risk, which reinforces the importance of increasing attention to this population.
In this context, biochemical markers in addition to the prevalence of specific clinical conditions help to understand the health status of the adolescent population, which seems to be strongly associated with the food environment. Considering the dimensions of the country, the importance of adolescent health, and the magnitude of the PNAE, this study conducted unprecedented analysis using data from the ERICA from Brazilian public schools to investigate the role of school feeding in insulin resistance markers.
# Materials and methods
## Study design
This study analyzed national data from the ERICA, which is a cross-sectional, schoolbased study including school adolescents aged 12 to 17 years. The ERICA study aimed to estimate the frequency of cardiovascular risks in this population. The study was conducted between 2013 and 2014 and evaluated approximately 75,000 adolescents from public and private schools in urban and rural areas in 124 Brazilian municipalities. The students answered a self-administered questionnaire using the LG GM750Q personal digital assistant model about themselves or their family members. For the environmental data, the school principals participated in an interview in which a questionnaire about the school was completed by the interviewers.
## Setting and participants
The ERICA included students who lived in Brazilian municipalities with more than 100,000 inhabitants and excluded those with temporary or permanent physical or mental disabilities and pregnant adolescents. For the present study, public school students and students who had collected blood for biochemical examinations were selected.
The ERICA sample was stratified into 32 geographic strata constituting the 27 Brazilian capitals and 5 more macroregions. For each geographic stratum, schools were selected with a probability proportional to size and inversely proportional to the distance from the capital. The ERICA sample is representative of larger municipalities at the national, regional, and Brazilian capital levels. More details about the sample and study design can be found in [bib_ref] The study of cardiovascular risk in adolescents-ERICA: Rationale, design and sample characteristics..., Bloch [/bib_ref] [bib_ref] The study of cardiovascular risk in adolescents-ERICA: Rationale, design and sample characteristics..., Bloch [/bib_ref] and.
In the present study, data from 27,990 adolescents were analyzed, corresponding to a response rate of eligible individuals of 44.7% [fig_ref] Figure 1: Flow chart of adolescents included [/fig_ref].
## Variables and categories
For the present analysis, the adolescents' individual and contextual (environment) characteristics were considered. From the individual characteristics, the following variables were analyzed: age, sexual maturation, waist circumference, physical activity, mother's education, ethnicity/skin color, self-reported consumption of meals prepared in school, purchased food at school, HOMA-IR, insulin, and blood glucose levels. For the environmental characteristics, the following information from the school's questionnaire was considered: location area, sale of food at school, presence of vending machines, presence of advertising of industrialized foods, and sale of food in the vicinity of the school [fig_ref] Table 1: Individual and environmental characteristics variables [/fig_ref].
## Insulin resistance markers (outcome)
Biochemical analysis was performed only with students who studied in the morning and who provided blood samples after obtaining written permission from their parents or guardians. Adolescents were instructed to fast for 12 h before the exam, and analyses of glucose and insulin were applied. Blood samples were processed and separated into plasma and serum within 2 h after collection and stored at temperatures ranging from 4 to 10 - C.
The hexokinase method was used for glucose analysis and chemiluminescence for insulin analysis. For HOMA-IR calculation, the formula described by [bib_ref] Homeostasis model assessment: Insulin resistance and β-cell function from fasting plasma glucose..., Matthews [/bib_ref] [bib_ref] Homeostasis model assessment: Insulin resistance and β-cell function from fasting plasma glucose..., Matthews [/bib_ref] was used: insulin (mU/L) × (glucose [mg/dl] × 0.0555)/22.5.
The insulin level cutoff points used for the assessment were desirable (<15 mU/L), borderline (15-20 mU/L), and high (≥20 mU/L) [bib_ref] I Diretriz Brasileira de Diagnóstio Tratamento da Síndrome Metabólica, De Carvalho [/bib_ref]. Borderline and high insulin levels were included in the undesirable group. The HOMA-IR cutoff points used were described by [bib_ref] Cutoff values for HOMA-IR associated with metabolic syndrome in the Study of..., Chissini [/bib_ref] [bib_ref] Cutoff values for HOMA-IR associated with metabolic syndrome in the Study of..., Chissini [/bib_ref] as follows: 2.80 for general adolescents, 2.32 for female adolescents, and 2.87 for male adolescents. The following blood glucose cutoff points used were recommended by the American Diabetes Association (2020): normal fasting blood glucose < 100 mg/dL and undesirable blood glucose ≥ 100 mg/dL.
## Contextual characteristics (exposure variables related to school)
Schools were classified according to their location (rural or urban) and through the responses collected in the questionnaire or by observing the school environment. The following questions were asked: "Does the school offer meals prepared on your premises?" (yes or no); "At school, there are self-service machines that are working for the sale of food, such as soft drinks, sweets, potato chips, and others." (yes or no); "Is there a way to sell food at school?" (yes or no); "What foods are sold? (sweets, candies, lollipops, chocolates, cookies, soft drinks, natural guarana, fried or baked snacks, sandwiches, pizzas, or others);" "Is there advertising for industrialized foods at school?" (yes or no); "What kind of advertising is there at school? (sweets, candies, lollipops, chocolates, cookies, soft drinks, natural guarana, fried or baked snacks, sandwiches, pizzas, or others);" "Is there a street vendor selling food or nonalcoholic drinks at the door or around the school?" (yes or no); "What is sold? (food, candies, chocolates, lollipops, popcorn, drinks, food, and drinks)."
## Individual characteristics (exposure variable related to school)
Using the questionnaire answered by the adolescents, the answers to the following questions were used: "Do you eat the meal offered by the school?" (yes or no), and "Do you buy food in the school cafeteria?" (yes or no).
## Individual characteristics (confounders)
The demographic characteristics analyzed were sex, age (<15 years and ≥15 years), ethnicity/skin color (white, black, or brown, Asian, indigenous, and does not know/prefers not to answer), and mother's education (illiterate or elementary school incomplete, elementary or high school complete, complete or incomplete higher education, or does not know/does not remember). Sexual maturation was classified to Tanner's stages, which assessed adolescents' self-perception validated method, according to images of breasts for girls, genitalia for boys, and pubic hair for both, and classified as pubertal if the adolescent fulfilled at least one puberty characteristic (stage 4 or 5). Central obesity was defined based on the waist circumference criteria (percentile > 90: female ≥ 82.6 and male ≥ 86.2). Adolescents were in light clothes, without shoes, and maintained proper body posture. Anthropometric assessments were performed by trained investigators; those who reported engaging in physical activity ≥ 300 min/week were considered sufficiently active, and those who did not reach this value (<300 min/week) were considered insufficiently active [bib_ref] Validade e reprodutibilidade de um questionário para medida de atividade física em..., Farias Júnior [/bib_ref].
# Statistical analysis
In the descriptive stage, the prevalence and its respective confidence intervals (95%) were calculated to characterize the sample and determine the distribution of the variables of interest. The natural weights of the sample design and the use of poststratification estimators were considered. In the analytical stage, to investigate the factors associated with insulin resistance, a multilevel mixed-effects generalized linear model was performed at the contextual and individual levels with each marker (HOMA-IR, blood glucose, and insulin) in sequential steps and using school as a second level. This method of analysis was applied considering that the school is a cluster and that it can influence some individual characteristics analyzed. Mixed models aggregate fixed and random effects in the same analysis, which is indicated in the context where students are grouped in schools, but it is intended to identify the variance of the effect of the school environment for the outcome variables [bib_ref] Random-Effects Models for Longitudinal Data, Laird [/bib_ref] [bib_ref] Multilevel modelling of complex survey data, Rabe-Hesketh [/bib_ref]. For the first stage, the independent variables were analyzed individually for each outcome, and those with p < 0.20 were selected for the adjusted multivariate analysis. Then, the variables sex, sexual maturation, physical activity, and maternal education were used to adjust the analysis. Statistical significance was set at p < 0.05.
In both models (individual and contextual levels), the original weights were considered from the sample design. Descriptive analyses were performed using the command survey ("svy") in the statistical program Stata version 16. The command "meglm" was used for multilevel linear regression.
# Ethical aspects
The original ERICA project was approved by the Ethical Committee of the Federal University of Rio de Janeiro in 2009 (protocol number 45/2008) and in the other 26 states and the Federal District. Approval was also obtained from all state and local departments of education in all schools that participated in the study. This study was conducted according to the principles of the Declaration of Helsinki. The students signed a consent form, and the parents of those students who collected a blood sample provided written consent as well.
# Results
Most adolescents were 15 years of age or older and declared themselves black or brown. Most of them ate school meals, although they also bought food sold at school. Regarding insulin resistance markers, there was a prevalence of 12.2% (95% CI; 11.1, 13.5) in undesirable insulin, 24.7% (95% CI; 22.8, 26.7) in high HOMA-IR, and 4.6% (95% CI; 3. in undesirable blood glucose. There were greater levels of fasting insulin and HOMA-IR in female adolescents, but the most remarkable results were observed for blood glucose levels in males [fig_ref] Table 2: Adolescents' general characteristics by sex [/fig_ref]. Regarding school environment [fig_ref] Table 3: Characteristics of school environment [/fig_ref] , most schools were in urban areas, less than half of the students studied in schools that sold food, but more than half had access to food sold in the school's immediate vicinity. Students reported (52.3%; 95% CI; 46.6, 58.9) daily exposure to a variety of processed foods and beverages in and around the school.
School feeding was positively associated with the insulin resistance marker HOMA-IR, both in the crude analysis and in the adjusted analysis, where eating school food decreased HOMA-IR by 0.135 units (95% CI; −0.19, −0.08). Similarly, eating the meal offered by the school corroborated the findings for insulin and blood glucose levels, which decreased by 0.469 µU/L (95% CI; −0.66, −0.28) and 0.634 mg/dL (95% CI; −0.87, −0.39), respectively (p < 0.001). Finally, buying food at school was considered a negative association for increased blood glucose levels by 0.455 mg/dL (95% CI; 0.16, 0.75; p = 0.002) [fig_ref] Table 4: Multilevel linear regression of contextual and individual level and HOMA-IR, insulin, and... [/fig_ref].
# Discussion
This study, through a robust methodology, demonstrated that the consumption of school feeding provided by Brazilian public schools has a protective effect on the variables related to insulin resistance, such as HOMA-IR, insulin, and blood glucose. In turn, buying food in school cafeterias showed a significant inverse association with blood glucose levels, which suggests that this practice is a risk factor for increasing levels of this metabolic marker.
All schools analyzed in this study were public schools, which are part of the PNAE. The PNAE, in addition to carrying out food and nutrition education actions, offers planned, healthy, and free meals during the school term for children and adolescents that cover their nutritional needs at this time. However, this is not the adolescents' only option to eat while in school since they can purchase food in school cafeteria self-service machines and from places in the school's immediate vicinity, all of which are highly present. Therefore, the adolescents' option of choosing not to consume school feeding should be considered, which can have a negative impact on their health.
In the present study, behavior was a determinant in the investigated associations, and the simple fact of having other foods offered in cafeterias at school or in the surroundings and vending machines and advertising was not associated with insulin resistance markers. The behavior, whether consciously or automatically, occurs within a social and environmental context and is influenced by several factors, such as at the individual, social, economic, community, and family levels [bib_ref] Theoretical explanations for maintenance of behavior change: A systematic review of behavior..., Kwasnicka [/bib_ref] [bib_ref] Factors influencing children's eating behaviours, Scaglioni [/bib_ref]. Adolescents' choices and behaviors are quite complex and can vary according to gender, age, and parental education [bib_ref] Factors influencing children's eating behaviours, Scaglioni [/bib_ref] [bib_ref] Individual and environmental influences on adolescent eating behaviors, Story [/bib_ref] ; these variables were controlled in the present study to avoid possible biases.
Gonçalves et al. (2019)investigated the eating environment in schools participating in the ERICA study to identify individual and contextual factors associated with hypertension and obesity, and found that there was a 35% lower chance of obesity among students who received school feeding compared with those who were enrolled in schools that did not offer meals, in addition to being less likely to have high blood pressure when they ate meals. Another study by the ERICA (Recife-PE) aimed to identify schools that promote healthy eating and physical activity and their relationship with overweight, hypertension, insulin resistance, and hypercholesterolemia in adolescents; this study found that there was a higher prevalence of overweight adolescents in those schools with an unhealthy food environment. In addition, those schools had a higher prevalence of students with hypercholesterolemia [bib_ref] Health promotion initiatives at school related to overweight, insulin resistance, hypertension and..., De Assunção Bezerra [/bib_ref]. Cardiovascular risk factors associated with insulin resistance and visceral obesity are emerging at an increasingly early stage, such as during childhood and adolescence, and are associated with a high probability of chronic diseases in adulthood [bib_ref] Assessing and managing the metabolic syndrome in children and adolescents, Deboer [/bib_ref].
The demand for health services during adolescence is low, and biochemical assessments, particularly blood glucose, HOMA-IR, and insulin levels, can be important early markers of diabetes. According to data from the National Adolescent School-Based Health Survey (2015) [28], a cross-sectional study consisting of Brazilian ninth graders and adolescents aged 13 to 17 years, only 53.2% of public school students use these services. Importantly, health changes in adolescence can be perpetuated into adulthood, including the development of diseases [1]. In the present study, we verified the high prevalence of high HOMA-IR (24.7%; 95% CI; 22.8, 26.7) and insulin (12.2%; 95% CI; 11.1, 13.5), despite having a low prevalence of hyperglycemia (4.6%; 95% CI; 3.8, 5.4). Early recognition of these risks, adequate treatment, and lifestyle modifications have been associated with reductions in MS in adults [bib_ref] Assessing and managing the metabolic syndrome in children and adolescents, Deboer [/bib_ref].
The incidence of diabetes has increased worldwide, as well as associated diseases, such as increased cardiovascular risk (increasing hypertension and dyslipidemia) and obesity [bib_ref] Independent predictors of insulin resistance in Brazilian adolescents: Results of the study..., De Andrade [/bib_ref]. Diabetics can also develop microvascular complications, such as diabetic kidney disease, retinopathy, and neuropathy. In addition, it is estimated that the annual cost of diagnosed diabetes in 2017 was $327 billion, generating costs for individuals and society. Strategies for early risk identification of diabetes development and control are necessary to ensure it does not continue into adulthood.
Although the ERICA is a large school-based study with Brazilian representation, some methodological limitations should be considered. The cross-sectional study design does not allow for the inference of causality, and the environment characterization represents adolescents, not schools. Information bias may be present since the questionnaire was self-reported by adolescents, and blood collection was performed in the morning after an overnight fast, yet it is not possible to verify whether this request was followed. The lack of information about the frequency of consumption of school meals provided by the PNAE, in addition to the frequency of purchase of food at the school cafeteria, also limits our power of inference from the findings of the present study. Even so, the methodological rigor of the ERICA study is highlighted, with team training and standard laboratory analysis methods, in addition to robust statistical analyses, with data weighting and control of possible confounding variables. Further studies are needed to investigate possible changes that may have occurred in the Brazilian school environment in recent years, as well as the results of isolated intervention programs.
# Conclusions
This study presents the association of school feeding consumption as a protective effect for insulin resistance variables and demonstrates the risk of purchasing food to increase blood glucose level. The importance of providing food and nutrition education in schools, in order to increase students' knowledge is evident, and the need to implement policies to promote health in the school environment is clear, with the aim of strengthening adherence to the school feeding program.
[fig] Figure 1: Flow chart of adolescents included. Study of Cardiovascular Risks in Adolescents (ERICA), Brazil, 2013-2014. [/fig]
[fig] Author: Contributions: A.B.O. collaborated in the analysis and interpretation of the data and the initial writing of the article. V.S.S.G. and K.M.B.d.C. collaborated on the conception and design, interpretation of data, and relevant critical review of intellectual content. All authors declare to be responsible for all aspects of the work in guaranteeing the accuracy and completeness of any part of the work. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research was funded by the Department of Science and Technology of the Secretariat of Science and Technology and Strategic Inputs of the Ministry of Health (Decit/SCTIE/MS) and the Health Sector Fund (CT-Saúde) of the Ministry of Science, Technology, and Innovation (MCTI) (Protocols: FINEP-01090421 and CNPq-565037/2010-2). Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the Ethical Committee of the Federal University of Rio de Janeiro in 2009 (protocol number 45/2008) and the other 26 states and the Federal District. [/fig]
[table] Table 1: Individual and environmental characteristics variables. Study of Cardiovascular Risks in Adolescents (ERICA), Brazil, 2013-2014. [/table]
[table] Table 2: Adolescents' general characteristics by sex. Study of Cardiovascular Risks in Adolescents (ERICA), Brazil, 2013-2014. [/table]
[table] Table 3: Characteristics of school environment. Study of Cardiovascular Risks in Adolescents (ERICA), Brazil, 2013-2014. [/table]
[table] Table 4: Multilevel linear regression of contextual and individual level and HOMA-IR, insulin, and glucose markers. Study of Cardiovascular Risks in Adolescents (ERICA), Brazil, 2013-2014. [/table]
|
Radical Scavenging Potential of the Phenothiazine Scaffold: A Computational Analysis
## Table of contents
. Cartesian Coordinates (in Å), electronic energies (in Ha), Gibbs free energy (in Ha), zeropoint correction , thermal correction to energy , enthalpy and Gibbs free energy computed at SMD-M06-2X/6-311++G(d,p), cc-pVTZ
[fig] 1: Cartesian Coordinates (in Å) and electronic energies (in Ha), computed at SMD-M06-2X/6-311++G(d,p), cc-pVTZ(-PP). .............................................................................................. S1 General ...................................................................................................................................... S1 Hydrogen Atom Transfer .......................................................................................................... S4 Radical Adduct Formation ...................................................................................................... S14 Single Electron Transfer.......................................................................................................... S78 Direct Oxidation ...................................................................................................................... S80 [/fig]
|
Feature fusion network based on strip pooling
Contextual information is a key factor affecting semantic segmentation. Recently, many methods have tried to use the self-attention mechanism to capture more contextual information. However, these methods with self-attention mechanism need a huge computation. In order to solve this problem, a novel self-attention network, called FFANet, is designed to efficiently capture contextual information, which reduces the amount of calculation through strip pooling and linear layers. It proposes the feature fusion (FF) module to calculate the affinity matrix. The affinity matrix can capture the relationship between pixels. Then we multiply the affinity matrix with the feature map, which can selectively increase the weight of the region of interest. Extensive experiments on the public datasets (PASCAL VOC2012, CityScapes) and remote sensing dataset (DLRSD) have been conducted and achieved Mean Iou score 74.5%, 70.3%, and 63.9% respectively. Compared with the current typical algorithms, the proposed method has achieved excellent performance.OPEN Context information.In semantic segmentation, contextual information aggregation can significantly improve the segmentation effect. Some methods with aggregating context information have been studied. DeepLabs 15,16 uses dilated convolutions with different rates to stack the ASPP (atrous spatial pyramid pooling) structure to capture different contextual information. Zhao et al. 17 stacked pooling layers of different sizes to form a pyramid structure to capture more contextual information. GCN 18 aggregates more context information by using a larger convolution kernel and reduces the number of parameters by convolution decomposition. Some works, such as SegNet 19 , U-Net 20 , U-Net++ 21 and ENet 22 , use encoding-decoding structures to aggregate low-dimensional information. In addition, the attention mechanism is also introduced to aggregate context information. DFNet 23 and EncNet 24 are inspired by SENet 25 and assign different weights to different channels. Zhao et al.26proposed a PSA module to generate an over-completed map to adaptively aggregate context information. These methods with aggregating certain contextual information can assist the network focus on areas of interest effectively.Self-attention mechanism.Long-term dependence is a significant issue that influences the effect of semantic segmentation. One solution is to use the self-attention mechanism 27 which is first applied in the field of natural language processing to calculate the affinity matrix between each pixel. It can be expressed as: Q, K, and V are weight matrixes, and √ dk is a constant. The self-attention mechanism needs to multiply two matrices, which will cause huge computation.The non-local 28 module employs the relationship between two locations to capture long-term dependence of the network, which enable each position on the feature map to obtain information of other positions. CCNet 29 has designed a CC module to capture the horizontal and vertical direction information. By repeatedly stacking the CC modules, the similarity of any two positions on the feature map can be calculated. DANet 30 and DRANet 31 learn the correlation between space and channel respectively through position and channel attention modules. TANet 32 combines channel and spatial attention to improve the segmentation effect. OCNet 33 combines the self-attention mechanism with the ASPP structure.Different from previous self-attention modules, we design a lightweight self-attention module based on strip pooling, which can capture global information more efficiently and uses the linear layer to make up for the loss of information caused by the pooling operation.The proposed self-attention module transforms the original matrix multiplication into vector multiplication, which greatly reduces the cost of computation and memory.MethodThis section first introduces the overall architecture of the network and describes the FF module that can capture sparse local information. Then recurrent the FF module to capture the dense global information. Finally it describes how to aggregate all the module together.
Semantic segmentation, which is the fundamental and challenging problem in computer vision, is to parse the category of each pixel in the image. It has been extensively researched in a variety of fields, such as autonomous driving, remote sensing images, medical diagnosis, and so on.
With the emergence of Fully Convolutional Neural Networks (FCNs) 1 , many methods have made remarkable progress in semantic segmentation. However, due to the limitations of the network structure, the traditional FCN only obtains the local information of the image and lacks sufficient contextual information, which can easily lead to incorrect segmentation results.
In recent years, many novel networks [bib_ref] Jointly network image processing: Multi-task image semantic segmentation of indoor scene based..., Huang [/bib_ref] [bib_ref] Semantic segmentation for multiscale target based on object recognition using the improved..., Jiang [/bib_ref] [bib_ref] Robust real-time hand detection and localization for space human-robot interaction based on..., Gao [/bib_ref] [bib_ref] Gesture recognition based on surface electromyography-feature image, Cheng [/bib_ref] [bib_ref] Self-attention negative feedback network for real-time image super-resolution, Liu [/bib_ref] [bib_ref] Review of road segmentation for SAR images, Sun [/bib_ref] have tried to seek new methods to solve FCN's issues. UperNet 8 uses feature pyramid network (FPN) to capture multi-scale features and analyze different scenes. DenseASPP 9 combines dense connection with ASPP, which is composed of the dilated convolution with different rates, to generate different receptive fields. Affinity Loss 10 was proposed by to distinguish the relationship between different pixels. HRNet 11 maintains high-resolution representations by connecting high-resolution to low-resolution convolutions in parallel. LedNet 12 uses attention pyramid network (APN) to capture contextual information, and uses convolutional decomposition and channel separation to reduce network complexity. HANet 13 introduces a highly-driven attention module to improve image segmentation in urban scenes. SPNetproposes the strip pooling to solve the long-term dependence of the network.
In order to complete the semantic segmentation task more quickly and accurately, a novel semantic segmentation network is designed, which can efficiently aggregate context information. Specifically, it consists of a series of convolution branches and two FF modules. The FF module uses strip pooling and two linear layers to generate the affinity matrix, which can capture the correlation between any features. For each spatial position on the affinity matrix, it collects all the information from the local feature map. The main contributions of this study can be summarized as follows:
1. We design a new network with self-attention mechanism, to solve the long-term dependence problem in semantic segmentation tasks.
2. An FF module is proposed to reduce the computational cost of affinity matrix. It efficiently captures contextual information by converting matrix multiplication to vector multiplication.
3. The experiments show that the proposed method has better performance on three mainstream benchmarks including PASCAL VCO 2012, Cityscapes and DLRSD.
The remaining paper is organized in the following way. "Related work" examines the top-ranking related work on semantic segmentation. The proposed method is introduced in "Methods". In "Experiments", We have conducted a large number of ablation and comparative experiments to verify the effectiveness of the proposed method. "Conclusion" is the summary of this paper. Network architecture. The overall architecture of the network is shown in [fig_ref] Figure 1: The overall structure of our network [/fig_ref]. The convolutional layer in the figure represents a convolution, BN, and ReLU. CNN uses ResNet50 with dilated convolution. To retain more detailed information, dilated convolution is used in the last two blocks of ResNet50. The height and width of the output are 1/8 of the input I. The extracted feature is processed by 3 × 3 convolutional layer to get I' (the number of channels decreased from 2048 to 512). Then, the network with Q, V, and X branches is designed. The Q has two serial FF modules. The first FF module generates feature map F by extracting information in the horizontal and vertical directions. The second FF module generates the affinity matrix F' , which is the result of the weighted summation of all pixels. The V and X directly reduce the channel dimension through the 3 × 3 convolution with BN and ReLU (the number of channels decreased from 512 to 128). And the result of V branch and F' are multiplied to generate the attention matrix M. The result of X branchand M are added to enhance the www.nature.com/scientificreports/ feature representation. Finally, the fused feature map is sent to the convolutional layer and generates ultimate prediction images. [fig_ref] Figure 2: The main architecture of the FF module [/fig_ref] , given feature map F (CxHxW), which divides into two branches q and k. In the q branch, it performs 1 × 1 convolutional to reduce dimension to C' × H × W (C' is half of the C) and a column average pooling to compress height dimension to get Y (C' × 1 × W), where Y ∈ R. Then reshapes to C' × W (remove height dimension) and gets feature vector q' by using two linear layers. Among them, the function of the linear layer is to convert the strip pooling result and reduce feature loss caused by strip pooling. It is worth noting that the output size of the linear layer is C' → C'/4 → C' , and they all use linear activation functions. This process can be described as:
## Feature fusion module. as shown in
Equation (2) shows the process of column average pooling, where H represents the height of the feature map, t i represents the ith element in each column. Y is the result of column average pooling. Eq. (3) shows the process of fully connected layer, where g represents linear layer, W is the learnable weight matrix of the linear layer. It can be found that q' is generated after the input feature map is compressed and then space transformation is performed.
The k branch is similar to the q branch, and k' will be obtained after row average pooling and two linear layers. After reshaping q' and k' , matrix multiplication is performed to produce the output E. Then, use E to generate output O. Please note that O is equal to E in the first FF module, but in the second FF module, O is obtained after E passes through the Softmax function. As shown in [fig_ref] Figure 2: The main architecture of the FF module [/fig_ref] , the way of the FF module collects information is marked in red. Each position on O combines information from row and column of the feature map F. An FF module cannot collect enough global information, so we feed O to the FF module again to capture global information, and calculate the affinity matrix between pixels through the Softmax function. Note that the linear layer can only output a fixed size. So,we use 1 × 1 convolution instead of the linear layer to achieve an output of any size. Experiments have also proved that the 1 × 1 convolution and linear layer are equivalent.
## Experiments
we first introduce PASCAL VOC2012, Cityscapes, and DLRSD, then introduce the experimental environment and details, and finally compare and verify the proposed method on different datasets.
Datasets. PASCAL VOC2012 is a segmentation dataset. It has 21 categories, including airplanes, bicycles, boats, etc. The dataset has 10,582 images for training and 1449 images for verification.
CityScapes is a city segmentation dataset. It collects road landscape images of 50 cities, each image size is 2048 × 1024. The dataset contains 19 common categories in road scenes, with a total of 5000 high-quality pixel-level labels. The training set contains 2979 images, the validation set contains 500 images, and the test set contains 1525 images.
DLRSD is a dense labeling dataset that builds for remote sensing image segmentation tasks. It contains 2100 images with a pixel size of 256 × 256, covering 17 common remote sensing image scene categories. We divide the training set and validation set according to the ratio of 0.8:0.2 for each category.
Experimental settings. The implementation of our network is based on the Pytorch framework. Its version is 1.1.0, and the CUDA version is 10.0. We only use a Nvidia GTX 1080TI to complete the experiment. Like the previous method, it uses the 'Poly' strategy to update the learning rate. The decoder initial learning rates of where TP represents true positive, TN represents true negative, FP represents false positive, FN represents false negative, and k represents the number of categories.
[formula] (2) Y = 1 H H i=0 t i (3) q ′ = g(Y , W) [/formula]
# Results analysis
Ablation study. We use the same hyperparameters for experiments. As shown in [fig_ref] Table 1: Results of ablation experiments [/fig_ref] , the ablation experiments on the PASCAL VOC 2012 are performed.
In the Table1, the second row is the result of one FF module. And the fifth row is the result of two FF modules. Obviously, The FF module can significantly improve the segmentation accuracy. Compared with the base-line FCN8s (use ResNet50 as the backbone network), using an FF module can bring an 8.4% improvement on mIoU. When stacking two FF modules repeatedly, the proposed method can increase mIoU from 72.8% to 74.5%. And it can help the network better aggregate contextual information. We add the FF module to different backbone networks to verify its effectiveness. Like ResNet50, we replace the last convolutional layer of the backbone network with dilated convolutional and fine-tune it. When the FF module is combined with the lightweight backbones MobileNet v2 and EfficientNet b0, 67.7% and 70.4% mIoU can be achieved respectively. It is worth noting that when we use ResNet101, its feature extraction ability is stronger, which can bring the highest mIoU of 75.8%.
In [fig_ref] Figure 3: Visualization of PASCAL VOC 2012 [/fig_ref] , we visualize feature maps (come from ResNet50) at different positions. the images in the 4th and 5th columns are the output of the 13th and 15th channels respectively. It shows that the proposed method can get better features. The 6th and 7th columns are the output of the first FF module and the second FF module respectively. After the second FF module, the relationship between each pixel will be calculated, and important information will be given higher weight (such as the bright spot in [fig_ref] Figure 3: Visualization of PASCAL VOC 2012 [/fig_ref]. The attention map (come from attention matrix), which is generated after aggregating context information, is shown in the final image. It is not difficult to find that the attention map can make the network pay more attention to the area of interest. [fig_ref] Table 2: Performance comparison of different models in PASCAL VOC 2012 [/fig_ref].
[formula] (4) mIoU = 1 k + 1 K i=0 TP FN + FP + TP (5) PA = TP + TN TP + TN + FP + FN [/formula]
Obviously, the proposed method is better than other methods. Compared with other attention methods, such as DANet and CCNet, the proposed method achieves a higher mIoU (74.5%). In terms of model complexity, the proposed method parameter is only 279 MB, which is about 1/3 less than the most recent mainstream models, such as SPNet and DRANet. The segmentation results of each category of PASCAL VOC 2012 (val) are shown in . For categories with a small number and a small area, such as "bicycles" and "bottles", the proposed model considers rich context information, which make segmentation more delicate and better segmentation results.
# Results on cityscapes.
We conduct experiments on the CityScapes. The experimental results are shown in [fig_ref] Table 4: Segmentation results on the CityScapes [/fig_ref]. It can be found that the proposed method achieves 70.3% mIoU, which surpasses the previous mainstream methods. Compared with DANet and DRANet, which also use the self-attention mechanism, the proposed method has 2.9% and 1.1% improvements in mIoU, respectively.
As shown in [fig_ref] Figure 4: Visualization of CityScapes [/fig_ref] , we visualize the most recent mainstream methods on the CityScapes. The proposed network can obtain a global perspective and accurately segment the image based on contextual information. For example, red boxes for the "road " or "building" in the image, the proposed method can correctly judge the target around the "road " according to the context information and make the segmentation more accurate.
# Results on dlrsd.
The DLRSD dataset is taken from the sky. The background of the objects in the image is complex and the scale is changed drastically, which makes segmentation very difficult. [fig_ref] Table 5: Segmentation results on the DLRSD [/fig_ref] shows the verification results on the DLRSD, where FLOPs are measured when the input size is 3 × 248 × 248 and the number of outputs is 17. Compared with DANet, which also uses the self-attention mechanism, the parameter amount of the proposed method is 20% lower than it. Computational complexity can be measured in FLOPs. The proposed network has far fewer FLOPs than the dual-channel self-attention network DANet. Compared with lightweight network LedNet, the proposed method has higher computational complexity, but more computation brings higher segmentation accuracy. The proposed method can automatically aggregate contextual information and achieve 63.9% in the mIoU. shows the corresponding visualization results. In red boxes, for large-scale targets, such as "aircraft", the proposed method can make a more complete segmentation. For small-scale targets, such as "cars", the proposed method can perceive their existence from a global perspective, which is less missed than other methods.
# Conclusion
We propose an efficient self-attention segmentation network (FFANet). FF module that can efficiently capture contextual information is designed. It uses strip pooling to reduce the complexity of the affinity matrix. The spatial transformation is performed through the linear layer to compensate for the information ambiguity caused by the strip pooling. Experiments show that the proposed method can effectively solve long-term dependence and make the segmentation result more accurate. It achieves 74.5% mIoU on the PASCAL VOC 2012, 70.3% mIoU on the CityScapes, and 63.9% mIoU on the DLRSD. Although the use of a linear layer can reduce the information loss caused by the pooling operation, some information will still be lost. Therefore, in the future research, we will explore other feature compression methods to capture global information more effectively (Suppl. Information).
[fig] Figure 1: The overall structure of our network. Scientific Reports | (2021) 11:21270 | https://doi.org/10.1038/s41598-021-00585-z [/fig]
[fig] Figure 2: The main architecture of the FF module. Scientific Reports | (2021) 11:21270 | https://doi.org/10.1038/s41598-021-00585-z www.nature.com/scientificreports/ the PASCAL VOC 2012, DLRSD, and CityScapes datasets are 0.05, 0.008, and 0.01. During the training process, the learning rate of the encoder is 1/10 of that of the decoder. We employ the SGD optimizer, where weight decay and momentum are 0.0001 and 0.9 respectively. For the backbone network, it chooses ResNet50 34 with dilated convolution, which has been pre-trained on ImageNet. For the sake of generality, all the networks in the experiment do not use auxiliary loss functions. We use flip, rotate, zoom, random scramble, random crop, and blur operations on the dataset to augment the data. The batch sizes of the PASCAL VOC 2012, and DLRSD are 8, the CityScapes is 4. And the size of the input is randomly cropped to 384, 224, and 512 respectively. The number of epoch of the PASCAL VOC 2012 and DLRSD is 180 and the CityScapes is 120. Besides, the pixel accuracy (PA) and the mean intersection of union (mIoU) are used as the main evaluation indicators of the experiment, and their definitions are as follows: [/fig]
[fig] Figure 3: Visualization of PASCAL VOC 2012 (val). Scientific Reports | (2021) 11:21270 | https://doi.org/10.1038/s41598-021-00585-z www.nature.com/scientificreports/ Results on PASCAL VOC 2012. The segmentation results of PASCAL VOC 2012 are show in [/fig]
[fig] Figure 4: Visualization of CityScapes (val). [/fig]
[table] Table 1: Results of ablation experiments. [/table]
[table] Table 2: Performance comparison of different models in PASCAL VOC 2012 (val).Table 3. Per-class results on PASCAL VOC 2012 (val). [/table]
[table] Table 4: Segmentation results on the CityScapes (val). [/table]
[table] Table 5: Segmentation results on the DLRSD (val). [/table]
|
Systematic review of the costs and effectiveness of interventions to increase infant vaccination coverage in low- and middle-income countries
Background: In recent years, several large studies have assessed the costs of national infant immunization programs, and the results of these studies are used to support planning and budgeting in low-and middle-income countries. However, few studies have addressed the costs and cost-effectiveness of interventions to improve immunization coverage, despite this being a major focus of policy attention. Without this information, countries and international stakeholders have little objective evidence on the efficiency of competing interventions for improving coverage. Methods: We conducted a systematic literature review on the costs and cost-effectiveness of interventions to improve immunization coverage in low-and middle-income countries, including both published and unpublished reports. We evaluated the quality of included studies and extracted data on costs and incremental coverage. Where possible, we calculated incremental cost-effectiveness ratios (ICERs) to describe the efficiency of each intervention in increasing coverage. Results: A total of 14 out of 41 full text articles reviewed met criteria for inclusion in the final review. Interventions for increasing immunization coverage included demand generation, modified delivery approaches, cash transfer programs, health systems strengthening, and novel technology usage. We observed substantial heterogeneity in costing methods and incompleteness of cost and coverage reporting. Most studies reported increases in coverage following the interventions, with coverage increasing by an average of 23 percentage points post-intervention across studies. ICERs ranged from $0.66 to $161.95 per child vaccinated in 2017 USD. We did not conduct a metaanalysis given the small number of estimates and variety of interventions included. Conclusions: There is little quantitative evidence on the costs and cost-effectiveness of interventions for improving immunization coverage, despite this being a major objective for national immunization programs. Efforts to improve the level of costing evidence-such as by integrating cost analysis within implementation studies and trials of immunization scale up-could allow programs to better allocate resources for coverage improvement. Greater adoption of standardized cost reporting methods would also enable the synthesis and use of cost data.
# Background
A large body of evidence has demonstrated the effectiveness and cost-effectiveness of infant immunization for reducing the burden of vaccine-preventable diseases. Routine immunization programs, supplemented by periodic campaigns, cover the majority of the target population in most countries, yet in many settings coverage remains below program goals. Expanding coverage is a major objective of immunization programs, both to increase the magnitude of health benefits from vaccination and to reduce disparities in outcomes among underserved populations. The remaining gaps in target coverage are especially prominent in low-and middle-income countries (LMICs), and so these settings have been a focus for ongoing efforts to improve immunization efforts with both vertical and horizontal programs.
Several recent studies have been undertaken to describe the costs of providing national immunization services. This research provides precise estimates of the cost of providing services at current coverage levels in a range of countries, and describes variation in the costs and operating practices of individual immunization sites. However, these studies do not provide direct evidence on the costs or cost-effectiveness of strategies and interventions used to scale up immunization coverage. Without this information, countries and the international stakeholders have little quantitative evidence on the best ways to use scarce resources to achieve immunization objectives. One possible source of information is studies that report the costs and effects of specific coverage-enhancing interventions.
We conducted a systematic review on the costs and effects of interventions to improve immunization coverage in LMICs, including research reported in the grey literature. This review updates past initiatives to survey the evidence on interventions to improve immunization coverage in low-income, high-burden settings. Several recent and historical studies have reviewed evidence related to immunization coverage improvements, but the majority of these have limited their scope to reporting effects on coverage or related programmatic outcomes, and excluded costs. Three reviews have considered both costs and effectiveness. In 2004, Pegurri, et al. summarized evidence in the peer-reviewed literatureand Batt, et al. reviewed the relevant grey literature. Both of these reviews identified a limited number of LMIC studies meeting inclusion criteria, and concluded that heterogeneity in methods adopted by these studies prevented quantitative synthesis of results. Both reviews found that few studies reported costs, with 10 out of 60 and 15 out of 34 identified studies including costs, respectively. A recent review by Ozawa, et al. including articles through 2016, adopted a broader scope, including peer-reviewed studies from both low-and high-income country settings, and covering all age groups, but excluding grey literature. While this study was able to undertake some quantitative synthesis of results, the authors acknowledged difficulties due to the heterogeneity of methods and reporting adopted by included studies, and the majority of estimates in their final sample came from high-income settings.
Our review returns to the approach of the earlier reviews with a focus on LMIC settings. In our review, we extracted data for studies conducted after the period covered by the Pegurri, et al. and , as earlier studies were already included in these prior reviews, and following the reasoning that older research would be less relevant to contemporary planning and budgeting decisions. These earlier reviews stressed the importance of the grey literature in documenting findings in this field, and consequently we included both the peer-reviewed and grey literature in our review. The objective of this review was to describe the incremental cost and effectiveness of interventions to increase coverage of infant immunization in LMICs, as defined by World Bank income group. The primary outcomes of interest were the incremental costs and incremental changes in target population coverage associated with a coverage-improvement intervention, as compared to routine program performance.
# Methods
This systematic review was registered with PROSPERO, an international prospective register of systematic reviews (record number 69586). We included interventions directed solely at increasing infant immunization coverage, as well as interventions designed to improve multiple aspects of immunization performance including coverage. The target age group for infants was defined as age 1 year and below; therefore, measles, mumps, rubella, and varicella vaccination interventions were included. We only included studies that reported empirical data, and excluded modeled analyses to minimize the introduction of additional bias into any summary results. We excluded interventions designed to improve health service delivery generally (e.g., improvements in access to primary health care). We also excluded studies that reported changes in immunization coverage but did not describe specific interventions used to impact infant immunization coverage. Studies with no data on intervention costs were excluded, as were those targeting adult immunization and animal studies. We included studies published between January 2003 (the end date of the period covered by older reviewsand May 2019.
## Search strategy and extraction
We identified eligible studies in the published literature by searching the following electronic databases, without language limitations: CEA Registry, Cochrane Library, EconLit, Embase, PubMed, Social Science Research Network (SSRN), and Web of Science. We also reviewed the reference lists of recent reviews of coverage improvement interventionsas well as selected studies found in the initial title search.
We identified eligible studies in the grey literature by searching relevant databases and repositories, including World Health Organization (WHO) regional databases Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS) and African Index Medicus (AIM), ELDIS, the World Bank working papers, GreyNet, and Grey Literature Report. Reports found to have a matching publication in the published literature were excluded. We adopted previously developed sensitivity-and specificity-optimized search strategies for identifying healthcare cost studies and economic evaluations. We tested the robustness of our search strategy by applying it to the timeframe of the Pegurri, et al. review in PubMed and confirming that all studies included in the earlier review were identified by our search terms. Additional file 1 lists the complete search terms utilized for each of the databases. Search terms fell into four categories: (1) immunization terms; (2) coverage terms; (3) cost terms; and (4) LMIC terms. Some databases limited the number of terms that could be searched; in such cases, the number of search terms was consolidated and, when necessary, distinct searches were performed for each category. Due to these limitations, LMIC search terms could not be included in every database.
Record titles and abstracts were independently screened by one of two reviewers (CM and AP) to identify studies that met the inclusion criteria. Reports not meeting study inclusion criteria were excluded. Uncertainty about study inclusion was resolved by a third independent investigator (NM).
From each of the included studies, we extracted information on the type of vaccination coverage intervention and study design; urban vs. rural setting; campaign vs. routine delivery; vaccination delivery platform; the baseline and endline coverage or incremental coverage, where applicable; the intervention costs and intervention cost per person exposed; and the incremental costeffectiveness ratio. This information was extracted for each unique intervention or country in a study. For example, if a study analyzed two different interventions in two different countries, we extracted information related to each of the four observations.
# Analysis
## Quality evaluation
We evaluated the methodological quality and risk of bias of included studies using the Consensus on Health Economic Criteria (CHEC) list. We excluded the CHEC list items that were specific to modeling analyses, as our review focused on empirical analyses. For each of the relevant items on the checklist, we assigned studies a score of zero or one (Additional file 2). A score of "0" indicated that the selected element was not present in the article; "1" indicated the element was present.
## Calculating incremental coverage
The primary outcomes of interest were the incremental costs and incremental changes in target population coverage associated with a coverageimprovement intervention, as compared to routine program performance. We calculated incremental coverage (defined as the percentage point change in vaccination coverage for the target group) based on the reported study design and outcomes. For studies that measured coverage at baseline and at the intervention conclusion, in both control (ctl) and intervention (int) groups (pre-post with control, including randomized control trials), incremental coverage was calculated as:
[formula] Incremental coverage ¼ Intervention coverage int −Baseline coverage int À Á − Intervention coverage ctl −Baseline coverage ctl À Á : [/formula]
In studies which measured coverage only at the intervention conclusion (post-test with control), incremental coverage was calculated as:
[formula] Incremental coverage ¼ Intervention coverage int −Intervention coverage ctl À Á [/formula]
In studies that measured coverage before and after the intervention without a control group (pre-post without control), incremental coverage was calculated as:
[formula] Incremental coverage ¼ Intervention coverage int −Baseline coverage int À Á [/formula]
## Calculating incremental intervention cost
For studies that measured costs at baseline and at the intervention conclusion, incremental intervention costs were calculated as:
[formula] Incremental intervention cost ¼ Intervention cost−Baseline cost ð Þ [/formula]
If studies included only incremental intervention costs, with no baseline costs stated, the stated intervention costs were listed. All costs were converted to 2017 USD using local inflation according to the consumer price index and local currency to USD exchange rates .
## Calculating incremental cost effectiveness ratios (icers)
ICERs were calculated as: ICER ¼ Incremental intervention costs=ðIncremental coverage
## Ãpopulation exposed to interventionþ
Some studies evaluated interventions that combined multiple health services, some of which were not focused on immunization. The costs reported by these studies were not broken out by immunization activities vs. other areas. For these studies, the ICER assumes that the costs of all included interventions are attributable solely to increasing immunization coverage, while the incremental coverage benefit is specific to immunization.
In other studies, the currency years of costs were unclear. We contacted the authors of these studies to request clarification of cost currency years, where possible, and assumed a currency year according to the year of the intervention otherwise. Nevertheless, several studies remained in which the ICER could not be calculated with the cost and coverage data provided.
# Results
A total of 2325 records were identified from 15 databases, with an additional 4 records obtained from reference lists of other articles. After removing 114 duplicates, 2215 titles were screened for eligibility, of which 1629 were removed. Of the 586 records remaining, 545 were removed after abstract review, leaving 41 articles which received full-text review. Of these, 27 were excluded for not meeting study inclusion criteria. We examined the LMIC studies included in the Ozawa review and added two relevant articles that we had previously excluded. Fourteen studies were included in the final analysis.describes general characteristics of the 14 studies. One study describes interventions from two countries (Mexico and Nicaragua); these were considered separate observations when extracting cost and effectiveness data. More than half of the observations (n = 10) were conducted in Asia, with the remaining in Africa (Ethiopia, Guinea-Bissau, and Madagascar) and Central America (Mexico and Nicaragua). Observations were nearly evenly split with regard to setting: 6 were conducted in rural areas; 3 in urban; 3 in mixed settings and 3 were not specified. About half of the interventions (n = 7) were delivered in a mobile outreach format; 4 were delivered in a fixed (i.e., health facility) format, and 4 interventions utilized both a mobile and fixed format. Additionally, 10 interventions were administered within routine vaccination systems, while 5 were administered using a campaign format. One study utilized both routine and campaign formats. Intervention duration ranged from 2 weeks to 5 years.
More than half of the studies (n = 9) reported randomized controlled trials, with the remaining studies reporting non-randomized pre-post evaluations (1 with control group, 3 without control groups) or cross-sectional designs. The majority of interventions were aimed at approaches for vaccine delivery (n = 5) or approaches to encourage additional vaccine uptake (i.e., demand . The interventions targeted coverage improvements for a range of vaccines, with DPT3 (diphtheria-pertussis-tetanus vaccine third dose) and measles vaccination most commonly addressed.summarizes coverage and cost data in the reviewed studies. If relevant values were not provided directly by studies, values were derived from the study data, such as incremental coverage, intervention cost (converted to 2017 USD), intervention cost in 2017 USD per person exposed, and ICERs. Intervention and baseline costs were primarily presented as total costs, with some studies presenting cost per person exposedor cost per community health worker trained.
## Coverage and cost information
Most studies reported increases in vaccination coverage following the interventions. In our review coverage increased by an average of 23 percentage points postintervention across the studies, with rates ranging from 8to 72percentage points. Interventions aimed at improving and increasing delivery mechanisms saw the largest incremental coverage increases, at an average of 36 percentage points post-intervention. While some studies did not provide baseline coverage, there was no apparent relationship between incremental coverage and baseline coverage, even when looking across intervention types. There was no discernable pattern between low-income and middle-income country studies.
Given the small total number and methodological heterogeneity of included studies, we decided not to attempt a quantitative synthesis of results. For those studies reporting coverage improvements with the intervention, ICERs ranged from $0.66to $161.95per child vaccinated in 2017 USD. There was also variation in the intervention costs reported. Studies most commonly reported site-level immunization costs (46%, n = 6). The next most common reported costs were supply chain and management costs (38%, n = 5), other costs (36%, n = 5), and vaccine costs (31%, n = 4).
## Quality ratings
Additional file 2 lists the quality scores each study received, according to the CHEC list, out of a total Endline levels lower than baseline levels for both treatment and control areas, but the "program did lead to an equalization of vaccination rates between the treatment and control group, despite the treatment group's coverage rate being 3 percentage points lower than in the control area at baseline"c Intervention contained a package of health services (immunization and other health activities). However, intervention costs were given only at an aggregate level and therefore immunization ICERs could not be calculated d Intervention contained a package of health services (immunization and other health activities). However, intervention costs were given only at an aggregate level and therefore immunization ICERs could not be calculated e Intervention cost minus hospital cost savings (both in USD 2017) f Baseline and endline coverage estimates compiled from aggregating values for children 12-23 months in 6 intervention / control areas g Includes both vaccination and non-vaccination interventions h Size of the population exposed to the intervention not stated i Only one province (of three described in study) was scaling up pre-existing immunization services. However, costs were reported as an annual net cost to the government for the new device across all three provinces. We conservatively assumed this aggregate cost was specific to the single relevant province for the cost per person exposed and ICER calculations j Costs were only reported per community health worker. Costs per exposed child could not be determined k Cost per mother given the information intervention l Cost per additional child vaccinated with DPT3 required for the incremental analysis that we conducted, indicating a risk of bias within the calculated ICERs.
# Discussion
We reviewed the recent literature describing the incremental cost and impact of efforts to improve immunization program coverage in LMIC, and identified 14 studies that containing sufficient cost and coverage data to be included in this study. The interventions reviewed in these studies covered a wide range of geographic settings, vaccines, intervention types, delivery mechanisms and scales. About half of the studies were randomized controlled trials, and the average study score on the CHEC list was 14 out ofAlthough all reviewed studies provided costs of the intervention itself, only 2 studies also considered changes in the costs of providing routine immunization services, despite the fact that such changes are a likely consequence of efforts to increase coverage. Eleven studies (79%) provided adequate data to calculate ICERs. From these, 14 ICERs of different interventions were calculated, ranging from $0.66to $161.95per child vaccinated in 2017 USD. In several cases, calculations of ICERs relied on substantial assumptions. Given the considerable differences in settings and perspectives, as well as different methods and cost categories included, an observed difference between ICERs may well reflect an artifact of the different study designs rather than a true feature of the interventions under study. Therefore, estimates should be compared with caution, and with full knowledge of the methodological and contextual differences between two studies. Funding devoted to coverage improvements may not be utilized efficiently in the absence of better evidence on optimal approaches. Several factors prevented ICERs from being calculated in all studies. The interventions described in some studiescontained a package of health services (immunization and other health activities). However, intervention costs were given only at an aggregate level and therefore immunization ICERs could not be calculated. Additionally, in one study the size of the population exposed to the intervention was not stated. Similarly, in one study the costs were calculated at the level of health workers, so cost per exposed child could not be determined. Due to the small total number and methodological heterogeneity of the extracted studies, we were unable to conduct a quantitative synthesis with cost estimates stratified by intervention type and country category.
Despite the 15-year gap since previous systematic reviews focusing on LMICs, there is still a scarcity of evidence on the cost-effectiveness of options for improving immunization coverage. Studies rarely report estimates of the incremental change in both cost and coverage. Several of the studies that do provide such information failed to report key features of the study methods, which limits the utility of their results. In particular, future studies should include detailed discussions of the intervention type (i.e., aimed at demand generation, delivery mechanisms, etc.), target population, baseline and endline coverage, specific intervention costs, changes in costs of routine service provision as well as information on effect modifiers that could affect incremental costs and coverage. Such information should be reported at a level of granularity that other programs would be able to interpret and modify those costs to fit their individual setting. These details are necessary to navigate the heterogeneity of the studies and to directly compare and synthesize the results to produce generalizable conclusions. The biggest challenge is that efforts to scale up immunization coverage often lack provision for a costing study, which represents a huge missed opportunity to understand the best use of scarce resources for improving coverage. We challenge the immunization community at country, regional, and global levels to incorporate costing studies into scaling up efforts.
# Conclusions
This research underscores the need for broad adoption of standardized reporting methods in immunization costing studies. This will be essential to increase our knowledge of improving immunization coverage as well as child immunization intervention costs.
|
Homeobox gene expression in acute myeloid leukemia is linked to typical underlying molecular aberrations
Background: Although distinct patterns of homeobox (HOX) gene expression have been described in defined cytogenetic and molecular subsets of patients with acute myeloid leukemia (AML), it is unknown whether these patterns are the direct result of transcriptional alterations or rather represent the differentiation stage of the leukemic cell. Method: To address this question, we used qPCR to analyze mRNA expression of HOXA and HOXB genes in bone marrow (BM) samples of 46 patients with AML and sorted subpopulations of healthy BM cells. These various stages of myeloid differentiation represent matched counterparts of morphological subgroups of AML. To further study the transcriptional alterations of HOX genes in hematopoiesis, we also analyzed gene expression of epigenetic modifiers in the subpopluations of healthy BM and leukemic cells. Results: Unsupervised hierarchical clustering divided the AMLs into five clusters characterized by the presence of prevalent molecular genetic aberrations. Notably, the impact of genotype on HOX gene expression was significantly more pronounced than that of the differentiation stage of the blasts. This driving role of molecular aberrations was best exemplified by the repressive effect of the PML-RARa fusion gene on HOX gene expression, regardless of the presence of the FLT3/ITD mutation. Furthermore, HOX gene expression was positively correlated with mRNA levels of histone demethylases (JMJD3 and UTX) and negatively correlated with gene expression of DNA methyltranferases. No such relationships were observed in subpopulations of healthy BM cells. Conclusion: Our results demonstrate that specific molecular genetic aberrations, rather than differentiation per se, underlie the observed differences in HOX gene expression in AML. Moreover, the observed correlations between epigenetic modifiers and HOX ex pression that are specific to malignant hematopoiesis, suggest their potential causal relationships.
# Introduction
The clustered homeobox (HOX) genes encode a large family of transcription factors characterized by the presence of a highly conserved nucleotide sequence called the homeodomain. This 61-amino-acid helix-turn-helix domain is responsible for the binding of HOX proteins to their target DNA sequences. In humans, the 39 HOX genes are organized into four genomic regions (the HOXA, B, C and D clusters) located on four chromosomes (chromosomes 7, 17, 12 and 2, respectively). Each cluster consists of 9 -11 genes arranged in the same orientation and in paralogous groups [bib_ref] The structural and functional organization of the murine HOX gene family resembles..., Duboule [/bib_ref] [bib_ref] The structure and function of the homeodomain, Scott [/bib_ref].
HOX genes play essential roles during embryogenesis by controlling cell fate along the anterior-posterior axis and specifying segment identity [bib_ref] Hox genes and regional patterning of the vertebrate body plan, Mallo [/bib_ref] [bib_ref] HOX genes in implantation, Vitiello [/bib_ref] [bib_ref] Hox patterning of the vertebrate axial skeleton, Wellik [/bib_ref]. The characteristic expression of HOX genes can also be detected in various adult tissues [bib_ref] Hox gene expression in adult tissues with particular reference to the adrenal..., Neville [/bib_ref] [bib_ref] Expression profiles of 39 HOX genes in normal human adult organs and..., Takahashi [/bib_ref]. During hematopoiesis, the highest expression of HOX genes occurs in the stem and early hematopoietic progenitor cells. During maturation, HOX expression gradually decreases, and it is minimal in differentiated hematopoietic cells [bib_ref] Hox regulation of normal and leukemic hematopoietic stem cells, Abramovich [/bib_ref] [bib_ref] Differential expression of homeobox genes in functionally distinct CD34+ subpopulations of human..., Sauvageau [/bib_ref]. The expression of HOX genes throughout the maturation of hematopoietic cells is tightly regulated, suggesting that disruption of this regulation contributes to the process of malignant transformation.
The oncogenic potential of HOX genes in leukemia has been intensively studied for more than two decades. Several chromosomal translocations in leukemia involve HOX genes either directly (e.g., NUP98-HOX fusion) or via their upstream regulators (e.g., MLL rearrangements) [bib_ref] The t(7;11)(p15;p15) translocation in acute myeloid leukaemia fuses the genes for nucleoporin..., Borrow [/bib_ref] [bib_ref] Direct and Indirect Targets of the E2A-PBX1 Leukemia-Specific Fusion Protein, Diakos [/bib_ref] [bib_ref] NUP98-HOXD13 gene fusion in therapy-related acute myelogenous leukemia, Raza-Egilmez [/bib_ref] [bib_ref] Expression of HOX genes in acute leukemia cell lines with and without..., Quentmeier [/bib_ref]. Moreover, the overexpression of certain HOX genes and their cofactors are known as poor prognostic markers in leukemia patients [bib_ref] Quantitative HOX expression in chromosomally defined subsets of acute myelogenous leukemia, Drabkin [/bib_ref] [bib_ref] HOX gene expression in phenotypic and genotypic subgroups and low HOXA gene..., Starkova [/bib_ref] [bib_ref] HOX expression patterns identify a common signature for favorable AML, Andreeff [/bib_ref]. The overexpression of HOX genes is believed to induce myeloproliferation, which together with additional aberrations, may lead to leukemia.
The regulation of gene expression during hematopoiesis is controlled by the cooperation of transcription factors and the dynamic architecture of chromatin. The specific epigenetic landscape influences target gene accessibility. As major executors of epigenetic regulation, chromatin-modifying enzymes mediate DNA and histone modifications responsible for the unique dynamics of chromatin observed throughout hematopoiesis. The deregulation of this process likely contributes to the malignant transformation of hematopoietic cells.
In embryogenesis, spatio-temporal expression of HOX genes is regulated by the trithorax-group (TrxG) and polycomb-group (PcG) proteins. PcG genes maintain HOX gene silencing through methylation of histone 3 lysine 27 (H3K27). In contrast, TrxG genes are responsible for maintaining previously established HOX gene expression through trimethylation of histone 3 lysine 4 (H3K4) [bib_ref] Hox regulation of normal and leukemic hematopoietic stem cells, Abramovich [/bib_ref] [bib_ref] Polycomb group proteins and heritable silencing of Drosophila Hox genes, Beuchle [/bib_ref]. A similar effect of PcG and TrxG genes has been proposed in the regulation of HOX gene expression in hematopoiesis as suggested by the severe defects of hematopoietic cells that have been reported in PcG and TrxG knock-out models [bib_ref] Polycomb-group genes as regulators of mammalian lymphopoiesis, Raaphorst [/bib_ref] [bib_ref] Roles of a trithorax group gene, MLL, in hematopoiesis, Ono [/bib_ref]. In addition, the H3K4 demethylase LSD1 and JmjC-domain-containing H3K27 demethylases JMJD3 (KDM6B) and UTX (KDM6A) have been shown to contribute to HOX gene regulation in embryonic development [bib_ref] Loss of LSD1 (lysine-specific demethylase 1) suppresses growth and alters gene expression..., Jin [/bib_ref] [bib_ref] UTX and JMJD3 are histone H3K27 demethylases involved in HOX gene regulation..., Agger [/bib_ref]. LSD1 establishes an inactive chromatin configuration by H3K4 demethylation, whereas JMJD3 and UTX activate chromatin by demethylation of H3K27. Finally, DNA methylation has been shown to participate in the establishment of HOX gene expression patterns, further supporting the role of epigenetics in the regulation of these genes [bib_ref] DNA methylation and differentiation: HOX genes in muscle cells, Tsumagari [/bib_ref].
In this paper, we sought to determine whether the pattern of leukemic HOX gene expression was primarily driven by the differentiation stage of hematopoietic cells or determined de novo during the process of malignant transformation. To approach this question, the expression patterns of the HOX genes were correlated with the molecular genetics and morphological characteristics of the leukemic cells of patients with childhood acute myeloid leukemia (AML). To further study the regulation of HOX gene expression, we also examined the relationships of chromatin modifiers and HOX genes in normal and malignant myelopoiesis.
# Methods
## Bm samples of healthy donors and patients with aml
Subpopulations of healthy BM representing developmental stages of hematopoiesis were sorted from the samples of healthy volunteers or minimal residual disease (MRD)negative leukemia patients in long-term complete remission (Fluorescence Activated Cell Sorter (FACS) Aria, BD, San Jose, CA, USA). The combination of surface markers that was used to identify the particular stages of myeloid lineage differentiation is listed in [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref]. To ensure adequate analysis sensitivity, we pooled the sorted samples of each subpopulation from five control donors and processed them as described below.
In total, 46 patients with childhood AML enrolled in the study were diagnosed and treated from 1998 to 2010 at the Czech Pediatric Hematology Working Group centers (Additional file 1: [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref]. Following the University Hospital Motol ethical committee's approval number P304/12/2214 and written informed consent, mononuclear cells were isolated from the diagnostic BM samples using a density gradient medium (Ficoll-Paque Plus, GE Healthcare Life Sciences, Uppsala, Sweden) and stored at −80°C.
RNA from both the patient samples and the healthy donor BM subpopulations were isolated with RNeasy Mini Kit (Qiagen, Hilden, Germany) and transcribed to cDNA using the iScript kit (Bio-Rad, Hercules, CA, USA).
## Real-time quantitative polymerase chain reaction (qpcr)
The quantification of gene expression was performed using the iCycler iQ System (BioRad, Hercules, CA, USA). The primer design and qPCR conditions for amplification of the HOXA and HOXB genes in the sorted populations (HOXA3, A4, A5, A6, A7, A9, A10, B2, B4, B5, B6, and B7) and patient samples (HOXA1, A3, A4, A5, A6, A7, A9, A10, A11, A13, B1, B2, B4, B5, B6, B7, B8, and B9) as well as the chromatin modifier genes (PcG family: EZH2 and BMI1; Trx family: MLL; DNMTs: DNMT1, DNMT3a, and DNMT3b and histone demethylases: JMJD3, UTX and LSD1) were performed as previously described [bib_ref] Quantitative HOX expression in chromosomally defined subsets of acute myelogenous leukemia, Drabkin [/bib_ref] [bib_ref] HOX gene expression in phenotypic and genotypic subgroups and low HOXA gene..., Starkova [/bib_ref] [bib_ref] Loss of LSD1 (lysine-specific demethylase 1) suppresses growth and alters gene expression..., Jin [/bib_ref] [bib_ref] Gene networking and inflammatory pathway analysis in a JMJD3 knockdown human monocytic..., Das [/bib_ref] [bib_ref] A functional role for the histone demethylase UTX in normal and malignant..., Liu [/bib_ref]. To normalize the gene expression levels, we used the ABL1 gene, which is known to be stably expressed during the development of myeloid lineage cells.
# Mutation analysis
The mutation statuses of the NPMI, NRAS, KRAS, CEBPa, c-KIT and FLT3 genes were determined in 12 patients for whom material was available and who were negative for the presence of the four major molecular aberrations (Additional file 1: [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref]. The analysis was performed by qualitative PCR followed by the sequencing of particular amplicons with the primers and PCR conditions as described earlier [bib_ref] Cytoplasmic nucleophosmin in acute myelogenous leukemia with a normal karyotype, Falini [/bib_ref] [bib_ref] Characterization of CEBPA mutations and promoter hypermethylation in pediatric acute myeloid leukemia, Hollink [/bib_ref] [bib_ref] Prevalence and prognostic significance of KIT mutations in pediatric patients with core..., Pollard [/bib_ref] [bib_ref] RAS mutations are frequent in FAB type M4 and M5 of acute..., Sano [/bib_ref].
# Statistical analysis
Data were analyzed using the statistical software packages Prism (GraphPad, La Jolla, CA, USA), Excel (Microsoft Corporation, Redmond, WA, USA), StatView (SAS Institute, Cary, NC, USA) and R-project (Vienna, Austria). The statistical significance of the differences among the subgroups of samples was assessed using non-parametric tests (Mann-Whitney and Kruskal-Wallis tests with Dunn's multiple comparison post test). Gene expression correlations were estimated by Spearman's rank correlation.
Unsupervised hierarchical cluster analysis (HCA), performed with the Genesis software (Institute for Genomics and Bioinformatics, Graz University of Technology (IGB-TUG), Graz, Austria), was used to identify the subgroups of samples with similar gene expression patterns.
# Results
Expression patterns of HOX genes in sorted subpopulations of healthy BM cells representing different stages of myelopoiesis Using FACS, we obtained eight subpopulations of normal BM donor cells, based on characteristic surface markers, which represent distinct stages of myeloid differentiation. The subpopulations corresponding with particular maturation stages were selected based on our expertise and previously published studies [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref] [bib_ref] Immunophenotypic differentiation patterns of normal hematopoiesis in human bone marrow: reference patterns..., Van Lochem [/bib_ref] [bib_ref] Four-color flow cytometry shows strong concordance with bone marrow morphology and cytogenetics..., Kussick [/bib_ref] [bib_ref] Using 4-color flow cytometry to identify abnormal myeloid populations, Kussick [/bib_ref]. Notably, we frequently observed the asynchronous expression of antigens and the overlap of immunophenotypic maturation stages in the leukemic blasts.
To better demonstrate the dynamics of HOX gene expression throughout hematopoiesis, data from the subpopulations of the two developmental lineages of myelopoises (granulocytic and monocytic) with the parallel differentiation stage were pooled together. This resulted in the discernment of four consecutive stages of myeloid development (stage 1 = G1 + M1, stage 2 = G2 + M2, stage 3 = G3 + M3 and stage 4 = G4 + M4). In accordance with previously published data, the expression of HOXA and particular HOXB (HOXB2 and HOXB4) genes gradually decreased during myeloid maturation [fig_ref] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM... [/fig_ref] and Additional file 2: [fig_ref] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM... [/fig_ref]. As assessed by comparisons on an one-to-one basis we also observed a clear positive correlation of HOX gene expression within HOXA cluster and HOXB cluster as well as between both clusters (Additional file 3: [fig_ref] Figure 2: Differing HOX gene expression levels observed in FAB subtypes with respect to... [/fig_ref]. Main characteristics of stages of myeloid lineage differentiationsurface markers and their counterparts with morphological subtypes of AML.
## Expression patterns of hox genes in bm samples from patients with childhood aml
Distinctive HOX gene expression patterns were observed among the French-American-British classification (FAB) AML subtypes (Kruskal-Wallis test: p < 0.0001 for the HOXA and p = 0.0016 for the HOXB cluster; [fig_ref] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM... [/fig_ref] and Additional file 4: [fig_ref] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to... [/fig_ref]. The M3 FAB subtype had the lowest levels of HOXA and HOXB gene expression compared with other FAB subtypes (see Additional file 4: [fig_ref] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to... [/fig_ref]. In contrast, AML M5 had the highest HOXA gene expression levels, along with the largest number of individual HOXA genes showing significant differential expression (Additional file 4: [fig_ref] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to... [/fig_ref]. Significant differences were also found among subgroups defined according to molecular genetics (i.e., Kruskal-Wallis test: p < 0.0001 for the HOXA and p = 0.0001 for the HOXB cluster; [fig_ref] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM... [/fig_ref] and Additional file 5: . Patients with PML-RARa fusion showed the lowest levels of HOXA and HOXB gene expression, while those with MLL rearrangements expressed HOXA genes at the highest levels (the majority of individual HOX gene comparisons revealed significant differential expression in the PML-RARa and MLL+ patients, respectively). Moreover, unsupervised hierarchical clustering based on HOX expression divided the leukemias into five main clusters characterized by the presence or absence of prevalent gene rearrangements, i.e., PML-RARa, RUNX1-RUNX1T1(AML1-ETO), CBFb-MYH11 and MLL alterations (Additional file 6: . Interestingly, three patients from cluster 1 (i.e., those having the overall highest levels of HOX gene expression and absence of these translocations) harbored a mutation in the NPM1 gene, similar to what has been reported in adult AML [bib_ref] HOX expression patterns identify a common signature for favorable AML, Andreeff [/bib_ref]. High risk (HR) patients expressed HOXA genes at significantly higher levels compared with patients who were assigned to the standard risk (SR) group (p < 0.0001 for HOXA3 -A13 and p = 0.0004 for HOXA1). However, no differences were observed in HOXB expression between high and low risk patients. In addition, both HOXA and HOXB gene expression were not found to be related to the patient age or risk stratification (Additional file 7: ; risk group stratification of childhood AML (AML-BFM 98 and 2004): standard risk -FAB M1/M2 with Auer rods, M3, M4eo, Down sy, t(8;21), t(15;17), inv [bib_ref] HOX expression patterns identify a common signature for favorable AML, Andreeff [/bib_ref] , and ≤ 5% of blasts in BM at D15 (except M3); high risk -others).
## Impact of molecular aberrations on hox gene expression within the morphological subgroups of aml
The effects of genetic aberrations on HOX gene expression were even more apparent when analyzed within the morphological FAB subtypes. In AML M4, the CBFb-MYH11+ patients exhibited statistically significant levels of lower HOX expression compared with those lacking the rearrangement. For AML M2, the RUNX1-RUNX1T1+ patients tended to show reduced levels of HOX gene expression compared with patients without the rearrangement [fig_ref] Figure 2: Differing HOX gene expression levels observed in FAB subtypes with respect to... [/fig_ref] In adults with normal cytogenetic AML, NPMI1 mutations are associated with high HOX expression and those leukemias have a higher frequency of FLT3 mutations [bib_ref] Drabkin H a: Hox expression in AML identifies a distinct subset of..., Roche [/bib_ref]. In our samples, HOX gene expression in the PML-RARa+ patients remained at very low levels regardless of the presence of FLT3/ITD mutations (FLT3/ITD+ (N = 4) vs. FLT3/ITD-(N = 4); [fig_ref] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to... [/fig_ref] ; p values indicated in figure legend). In contrast, in the absence of a PML-RARa fusion, HOX levels were higher when FLT3 was mutated. These results were further emphasized by the analysis of gene expression data from a larger (N = 48) independent cohort of FLT3/ITD+ childhood AML patients [bib_ref] Zwaan CM: NUP98/NSD1 characterizes a novel poor prognostic group in acute myeloid..., Hollink [/bib_ref] , which demonstrated that HOXA and HOXB gene levels were significantly reduced in FLT3/ITD+ patients with PML-RARa (N = 12) compared to those without the fusion gene (N = 36; [fig_ref] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to... [/fig_ref].
HOX gene expression patterns in corresponding differentiation stages of normal and malignant myelopoiesis
As indicated above, we sorted the subpopulations of normal BM cells from the healthy donors according to the specific stage of myelopoiesis. Our gating strategy enabled a comparison of these subpopulations with the FAB subtypes of AML patients exhibiting the similar stage of myeloid maturation arrest. The list of AML subtypes assigned to particular stages of myelopoiesis is provided in [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref]. Differential HOX gene expression patterns were identified between the normal and malignant hematopoietic counterparts, as demonstrated by comparing AML M3 with the corresponding stage of promyelocytes (ID = G2). These differences were particularly evident for HOXA5, HOXA6, HOXA9, HOXA10 and HOXB4 . Similarly, differential expression of HOXA3, HOXA4, HOXA5, HOXA6, HOXA7, HOXA9, HOXA10, HOXB5 and HOXB6 distinguished leukemic cells of the M5a and M5b FAB subtypes from the matched normal counterparts, represented by the sorted M3 and M4 population, respectively (Additional file 8: and 7B).
## Expression patterns of chromatin modifiers and their role in hox gene regulation
In subpopulations of healthy BM cells, we analyzed the expression of chromatin modifying genes, previously shown to contribute to HOXA and HOXB regulation during embryogenesis [bib_ref] UTX and JMJD3 are histone H3K27 demethylases involved in HOX gene regulation..., Agger [/bib_ref] [bib_ref] Hox gene regulation and timing in embryogenesis, Montavon [/bib_ref] [bib_ref] Polycomb complexes repress developmental regulators in murine embryonic stem cells, Boyer L A [/bib_ref]. Based on observed expression in the sorted subpopulations, the chromatin modifiers were divided into three groups. The first group, "Modifiers 1", included genes that did not exhibit varying expression levels during differentiation (EZH2, BMI1, MLL, LSD1 and DNMT1). The second group, "Modifiers 2", consisted of genes showing increased expression during hematopoiesis (JMJD3 and UTX). Expression levels of the third group, "Modifiers 3" (DNMT3a and DNMT3b), showed a decrease concomitant with differentiation, which were statistically inversely correlated with Modifiers 2 (R = −0.922; . However, we did not observe a clear pattern of corresponding HOX gene expression changes in these cells.
In leukemic cells, the expression of Modifier 1 genes, except for BMI1, paralleled their normal counterparts, being largely unchanged among the morphological AML subgroups. The expression of Modifier 2 and 3 genes markedly differed among AML samples . In contrast to normal cells, the expression of Modifier 2 and 3 genes appeared to mirror the differences in HOX mRNA levels. In general, HOXB expression was positively correlated with Modifier 2 genes (e.g., H3K27 demethylases; R = 0.874). The correlation of HOXA gene expression levels with Modifiers 2 genes was less pronounced (R = 0.506) in all cases with the exception of the AML M3 subgroup. For the Modifier 3 genes (e.g., DNMTs), there was a substantial negative correlation with HOXB expression (R = −0.442; . However, in contrast to normal hematopoiesis, the observed inverse correlation between Modifier 2 and 3 genes was much less pronounced (except for AML M3 and M4; R = −0.178; .
For individual genes in the leukemic samples (Additional file 9: , there were several notable correlations. For instance, the expression of JMJD3 with HOXB4 and HOXB6 was strongly positively correlated (p = 0.0003 and 0.0012, respectively), while negatively correlated with DNMT3b (p = 0.03). There was also strong correlation on a one-to-one basis between genes in the HOXA cluster and for genes in the HOXB cluster. However, in contrast with the healthy samples, there was only a weak correlation between the particular HOX genes from different HOX clusters.
# Discussion
Several reports have demonstrated that HOX genes are not only potent regulators of embryonic development but also play significant roles in the regulation of many processes in adult organisms, including hematopoiesis [bib_ref] The role of HOX genes in normal hematopoiesis and acute leukemia, Alharbi [/bib_ref] [bib_ref] Homeobox genes in normal and abnormal vasculogenesis, Cantile [/bib_ref] [bib_ref] Molecular regulation of mullerian development by Hox genes, Du [/bib_ref]. The overall role of HOX clusters in addition to that of particular HOX genes in hematopoiesis have been revealed by various knock-out and overexpression studies of human hematopoietic cells or by studies using mouse models [bib_ref] Molecular interactions involved in HOXB4-induced activation of HSC self-renewal, Beslu [/bib_ref] [bib_ref] Mice bearing a targeted interruption of the homeobox gene HOXA9 have defects..., Lawrence [/bib_ref] [bib_ref] Loss of function of the homeobox gene Hoxa-9 perturbs early T-cell development..., Izon [/bib_ref] [bib_ref] Disruption of the homeobox gene Hoxb-6 in mice results in increased numbers..., Kappen [/bib_ref]. However, the degree to which HOX genes contribute to the process of leukemogenesis has not yet been elucidated.
The aberrant expression of HOX genes has been reported in the majority of leukemia patients. However, it remains unknown whether this aberrant expression represents a genuine driver of leukemogenesis or a passenger effect reflecting a differentiation block. Another possible explanation takes into consideration an impact of the molecular aberrations present in leukemic cells with further biological consequences. Here, we attempted to shed light on the expression of HOX genes in normal hematopoietic precursor cells versus their malignant counterparts with respect to their differentiation stage arrest in AML.
The crucial prerequisite for such an analysis is the appropriate identification of subpopulations of healthy BM cells representing the stages of myelopoiesis that can be matched to their respective morphological counterparts in AML. We managed to sort these subpopulations and analyzed their HOX gene expression patterns. The expression of HOX genes was higher at the initial stages of hematopoiesis and gradually decreased with the maturation of the hematopoietic cells, supporting the generally accepted hypothesis that HOX genes are strong regulators of hematopoiesis (particularly at the early stages) [bib_ref] The role of HOX genes in normal hematopoiesis and acute leukemia, Alharbi [/bib_ref].
A comparison of matched normal and malignant hematopoietic precursor cells at the same differentiation stage demonstrated the distinct expression patterns of the HOX genes in the leukemic cells. This indicates that the aberrant patterns of HOX gene expression cannot be simply explained by the differentiation statuses at which the cells have been arrested. This is similar to what we previously observed in pediatric patients with ALL, who were found to exhibit differential HOX gene expression between the subgroups and their matched normal precursors according to differentiation stage [bib_ref] HOX gene expression in phenotypic and genotypic subgroups and low HOXA gene..., Starkova [/bib_ref]. Altogether, our results support the hypothesis that the dysregulation of HOX genes is involved in the process of neoplastic transformation.
The analysis of childhood AML patients revealed a different expression profile of HOX genes among the FAB subtypes and the subgroups of patients bearing unique molecular rearrangements. The most diverse subgroup of AML was AML M3, which showed the lowest levels of HOX gene expression. This subgroup is characterized by the presence of the PML-RARa fusion gene, which generates an aberrant retinoic acid receptor unresponsive to the physiological levels of this molecule. RUNX1-RUNX1T1+, CBFb-MYH11+ and MLL-rearranged AML patients also showed unique HOX gene expression patterns. MLL rearrangements have been previously shown to have a determinant role on HOX gene expression [bib_ref] MLL: a histone methyltransferase disrupted in leukemia, Hess [/bib_ref]. Moreover, we revealed that AML patients bearing the PML-RARa fusion gene had low expression levels of the HOX genes regardless of FLT3/ITD status. This finding is even more interesting considering that FLT3/ ITD has been shown to be associated with the upregulated expression of HOX genes in leukemia patients [bib_ref] Drabkin H a: Hox expression in AML identifies a distinct subset of..., Roche [/bib_ref]. Therefore, we performed an analysis of a larger cohort of AML patients [bib_ref] Zwaan CM: NUP98/NSD1 characterizes a novel poor prognostic group in acute myeloid..., Hollink [/bib_ref] from Erasmus MC-Sophia Children's Hospital and replicated the results drawn from our cohort of pediatric AML patients. This analysis showed that despite the overall upregulation of the HOX genes in FLT3/ITD+ AML patients, HOX gene expression in FLT3/ITD+ PML-RARa+ patients was significantly lower compared to the FLT3/ITD+ patients without this fusion protein. Therefore, in this case, the PML-RARa fusion gene may be superior to FLT3/ITD with respect to its role in the process of malignant transformation. Based on these results, we suggest that AML-specific fusion oncoproteins may impact the upstream pathways that deregulate the HOX genes, thereby acting as the major underlying factors of their characteristic expression patterns observed in leukemic cells.
Our analysis of the AML patients also showed significantly lower expression levels of HOXA in the SR compared with the HR patients (in accordance with a previous study [bib_ref] Quantitative HOX expression in chromosomally defined subsets of acute myelogenous leukemia, Drabkin [/bib_ref]. These results suggest that the assessment of HOX gene expression patterns may allow for the prediction of aggressive cases of leukemia and may therefore be taken into consideration in risk stratification. However, we suggest that this observation is a consequence of the allocation of patients with different molecular aberrations to particular AML risk groups (i.e., PML-RARa+ patients with the lowest HOXA gene expression levels being assigned to the SR group and MLL+ cases with the highest expression levels of HOXA genes being allocated to the HR group) and not an independent prognostic factor.
Considering the profound contribution of chromatin modifiers to the embryonic regulation of HOX genes and the essential roles of HOX genes in hematopoiesis, the dysregulation of chromatin modifiers may deregulate the entire process of hematopoiesis and subsequently lead to malignant transformation. However, the exact roles of epigenetic modifications in the regulation of leukemic HOX gene expression remain to be elucidated. It has recently been shown that HOX genes possess unique chromatin regions called bivalent domains. These domains are characterized by the presence of both repressive (methylated H3K27) and activating (methylated H3K4) histone methylation marks and are found in genes poised to be activated according to cell-specific requirements [bib_ref] A bivalent chromatin structure marks key developmental genes in embryonic stem cells, Bernstein [/bib_ref]. To determine the role of chromatin modifiers in the regulation of HOX genes in normal hematopoiesis and leukemogenesis, we analyzed the expression patterns of DNA methyltransferases, histone H3K27/ H3K4 demethylases, and selected PcG and TrxG genes in subpopulations of healthy BM cells and BM samples of patients with AML. We found an inverse correlation of histone demethylase (Modifiers 2) and DNMT (Modifiers 3) gene expression in normal and malignant hematopoiesis. In contrast to healthy hematopoiesis, we found an interesting correlation between chromatin modifier gene expression and that of the HOX genes in the AML samples. The most pronounced correlation
## Aml patients -morphological subgroups
Healthy BM -stages of myeloid differentiation A B Expression patterns of chromatin modifiers and HOX genes in subpopulations of healthy BM and AML cells. A. Subpopulations of healthy BM; B. AML patient subgroups. Lines connect values between categories (e.g., differentiation stages) to visually enhance depiction of trend across subgroups.
was observed with the AML M3 subtype. The specific relationship of the HOX genes with the epigenetic modifiers in this morphological subgroup could be affected by the presence of the PML-RARa fusion gene. In particular, HOX gene expression was positively associated with the histone H3K27 demethylases, JMJD3 and UTX, and inversely correlated with DNMT3b. Notably, both JMJD3 and UTX have recently been suggested to play roles in hematopoiesis [bib_ref] A functional role for the histone demethylase UTX in normal and malignant..., Liu [/bib_ref] [bib_ref] Contrasting roles of histone 3 lysine 27 demethylases in acute lymphoblastic leukaemia, Ntziachristos [/bib_ref]. Moreover, UTX has been shown to directly bind to the HOXB1 locus [bib_ref] UTX and JMJD3 are histone H3K27 demethylases involved in HOX gene regulation..., Agger [/bib_ref] [bib_ref] A functional role for the histone demethylase UTX in normal and malignant..., Liu [/bib_ref]. Taken together, the results implicate chromatin modifiers in the establishment of the aberrant leukemic expression of HOX genes in pediatric AML patients.
Although the expression of BMI1 was not altered during hematopoiesis, a Spearman correlation analysis showed that this gene was positively correlated with HOX gene expression in the leukemic samples. It has been reported that BMI1 determines the proliferating abilities of the cells by inhibiting the p16 gene. HOXA9 was also shown to target p16 and impair the senescence of cells [bib_ref] Interplay between Homeobox proteins and Polycomb repressive complexes in p16INK 4 a..., Martin [/bib_ref]. Thus, the expression levels of the histone methyltransferase BMI1 are likely to reflect the proliferation statuses of leukemic cells without directly impacting HOX gene expression [bib_ref] Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells, Lessard [/bib_ref].
Interestingly, the PML-RARa and RUNX1-RUNX1T1 fusion oncogenes have been shown to cooperate with repressive complexes, leading to alterations in chromatin architecture. PML-RARa causes profound changes in the epigenetic landscape, mainly by recruiting chromatin-modifying enzymes to target sequences or by the deregulation of their functions [bib_ref] Methyltransferase recruitment and DNA hypermethylation of target promoters by an oncogenic transcription..., Di Croce [/bib_ref] [bib_ref] Fusion proteins of the retinoic acid receptor-alpha recruit histone deacetylase in promyelocytic..., Grignani [/bib_ref]. Furthermore, recent studies have shown that the degradation of the PML-RARa oncoprotein results in dramatic changes to the landscape of histone modifications [bib_ref] PML-RARalpha/RXR Alters the Epigenetic Landscape in Acute Promyelocytic Leukemia, Martens [/bib_ref]. Similarly, RUNX1-RUNX1T1 has also been shown to recruit epigenetic modifiers to target sequences [bib_ref] AML1-ETO triggers epigenetic activation of early growth response gene l, inducing apoptosis..., Fu [/bib_ref]. These findings together with our data suggest that AML-specific oncoproteins regulate HOX gene expression through epigenetic modifications. However, further studies are needed to understand the roles of epigenetic modifiers in the regulation of normal as well as leukemic HOX gene expression and their cooperation with AML fusion oncoproteins.
In summary, we found that the expression patterns of the HOX genes in leukemic cells are not solely determined by their particular differentiation stages. Conversely, we assume that the specific molecular aberrations that are typical of AML are the major determinants of the leukemic expression patterns of the HOX genes. Our results also demonstrate the differing contributions of epigenetic modifiers to HOX gene expression in healthy and malignant hematopoiesis.
## Additional files
Additional file 1: [fig_ref] Table 1: Identification of subpopulations of healthy BM cells [/fig_ref]. Patients' characteristics.
Additional file 2: [fig_ref] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM... [/fig_ref]. mRNA expression of particular HOXA and HOXB genes in subpopulations of healthy BM. Relationship of expression patterns of chromatin modifiers and HOX genes in subpopulations of healthy BM and AML cells.
Lines connect values between categories (e.g., differentiation stages) to visually enhance depiction of trend across subgroups.
[fig] Figure 1: Expression levels of HOXA and HOXB genes in subpopulations of healthy BM and samples of patients with AML. A. four consecutive stages of myeloid lineage differentiation; B. morphological subgroups of AML patients; C. subgroups of AML with typical molecular aberrations. [/fig]
[fig] Figure 2: Differing HOX gene expression levels observed in FAB subtypes with respect to the presence of typical molecular aberrations. A. CBFb-MYH11+/− patients with AML M4. Asterisks indicated statistical significance (* ≤ 0.05, ** ≤ 0.01; additional borderline significance: p = 0.0679 for HOXA7, HOXA10 and HOXB9, p = 0.0732 for HOXA13); B. RUNX1-RUNX1T1+/− patients with AML M2. Asterisks indicated statistical significance (* ≤ 0.05, ** ≤ 0.01; additional borderline significance: p = 0.1745 for HOXA5, p = 0.1745 for HOXA9 and P = 0.1172 for HOXA10). [/fig]
[fig] Figure 3: Differing HOX gene expression levels observed in PML-RARa+ patients with respect to the presence of FLT3/ITD. A. HOX gene expression in the subgroup of PML-RARa+ patients with FLT3/ITD compared to the other PML-RARa+ patients. Asterisks indicated statistical significance (* ≤ 0.05, ** ≤ 0.01; additional borderline significance: p = 0.0833 for HOXA7); B. HOX gene expression levels in FLT3/ITD+ patients with PML-RARa compared to the other FLT3/ITD patients. Median expression (log2) of HOXA (left) or HOXB genes is indicated by bullets colored according to p values (Mann-Whitney tests) of differences between PML-RARa + and other groups. Color legend indicated at right. Dashed line indicating equal expression is also provided. [/fig]
[table] Table 1: Identification of subpopulations of healthy BM cells [/table]
|
Laparoscopic hysterectomy for benign indications: clinical practice guideline
Results Ten topics were considered in this guideline, including amongst others, the different approaches for hysterectomy, advice regarding tissue extraction, pre-operative medical treatment and prevention of ureter injury. This work resulted in the development of a clinical practical guideline of LH with evidence-and expert-based recommendations. The guideline is currently being implemented in The Netherlands. Conclusion A guideline for LH was developed. It gives an overview of best clinical practice recommendations. It serves to standardize care, provides guidance for daily practice and aims to guarantee the quality of LH at an (inter) national level.
# Introduction
Since the introduction of laparoscopic hysterectomy (LH) more than 2 decades ago, a rapid implementation of this procedure has been observed in many countries [bib_ref] Ten years of progress-improved hysterectomy outcomes in Finland 1996-2006: a longitudinal observation..., Makinen [/bib_ref] [bib_ref] Nationwide trends in the performance of inpatient hysterectomy in the United States, Wright [/bib_ref]. For The Netherlands, the percentage of hysterectomies performed laparoscopically has increased from 3% in 2002 to 36% in 2012 [bib_ref] Trends in the implementation of advanced minimally invasive gynecologic surgical procedures in..., Driessen [/bib_ref] and similar increases have been observed in other parts of the world [bib_ref] Ten years of progress-improved hysterectomy outcomes in Finland 1996-2006: a longitudinal observation..., Makinen [/bib_ref] [bib_ref] Nationwide trends in the performance of inpatient hysterectomy in the United States, Wright [/bib_ref]. Such rapid implementation can potentially result in unwarranted practice variations in health care delivery [bib_ref] The challenge of variation in medical practice, James [/bib_ref]. Unexplained differences in health care delivery should be addressed as they are usually the consequence of a lack of consensus and/or available evidence [bib_ref] Medical practice variations: what the literature tells us (or does not) about..., Mercuri [/bib_ref] [bib_ref] Does practice variation matter?, Hlatky [/bib_ref]. Without a convenient standard of care, doctors are more prone to adopt medical practices that are based on personal experience [bib_ref] Medical practice variations: what the literature tells us (or does not) about..., Mercuri [/bib_ref] [bib_ref] Does practice variation matter?, Hlatky [/bib_ref]. Furthermore, studies
# Abstract
Purpose Since the introduction of minimally invasive gynecologic surgery, the percentage of advanced laparoscopic procedures has greatly increased worldwide. It seems therefore, timely to standardize laparoscopic gynecologic care according to the principles of evidence-based medicine. With this goal in mind-the Dutch Society of Gynecological Endoscopic Surgery initiated in The Netherlands the development of a national guideline for laparoscopic hysterectomy (LH). This present article provides a summary of the main recommendations of the guideline. Methods This guideline was developed following the Dutch guideline of medical specialists and in accordance with the AGREE II tool. Clinically important issues were firstly defined and translated into research questions. A literature search per topic was then conducted to identify relevant articles. The quality of the evidence of these articles was rated following the GRADE systematic. An expert panel consisting of 18 selected gynecologists was consulted to formulate best practice recommendations for each topic.
have shown that standardizing care on best practices is associated with better outcomes and reduced costs [bib_ref] Nationwide outcomes measurement in colorectal cancer surgery: improving quality and reducing costs, Govaert [/bib_ref]. As a result, it seems timely to define a standard of care for LH, according to the principles of evidence-based medicine.
With this goal in mind, the Dutch Society of Gynecological Endoscopic Surgery (WGE) initiated the development of a guideline for LH. This guideline aims to provide gynecologists with an overview of best practices, directly applicable for daily practice. The guideline should also ensure a minimum quality of care and enhance patient safety. This article provides a summary of the main recommendations of the guideline.
# Materials and methods
## Development of the guideline
The WGE, a working group of the Dutch Society of Obstetrics and Gynecology (NVOG), initiated the development of the guideline. A guideline working group was assembled and consisted of three gynecologists and one resident (WJKH, PMG, ART and EMS). The guideline was developed in accordance with the Dutch guideline of medical specialists. This document, recognized by all Dutch medical societies, provides a detailed overview of the process of developing an evidence-based guideline using the GRADE method [bib_ref] GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables, Guyatt [/bib_ref]. The Appraisal of Guidelines for Research and Evaluation instrument (AGREE II), an internationally recognized assessment tool, was used in a second stage to evaluate the methodological rigor, transparency and quality of the developed guideline. In the next subsections, the different steps undertaken to create this guideline will be briefly described.
Step 1: Key topic analysis A brainstorming session was organized by the WGE with 40 gynecologists, all performing advanced laparoscopic procedures. During that meeting, key topics for this guideline were determined and transformed into appropriate clinical research questions.
Step 2 and 3: Literature selection, data extraction and assessment of risk of bias For each research question, a literature search was set up in collaboration with a clinical librarian. PubMed, Medline and Cochrane databases were searched up to 1 st of March 2016. Each research question had its own inclusion and exclusion criteria. Overall, we first searched for systematic reviews. If none were available, we focused on randomized controlled trials (RCTs) and, if necessary, added cohort studies as well. Studies from the eligible systematic reviews were reviewed to avoid duplicate inclusions. Only LH for benign indications and/or low-grade malignancy were considered and will hereinafter be referred to as 'laparoscopic hysterectomy' (LH). Studies focusing on endometriosis sanitation with concomitant LH as well as highgrade malignancy were not included. Study reports, letters, non-published manuscripts and articles that were not published in English were also excluded. After selecting the eligible studies, these studies were summarized in evidence tables and when possible, extracted for meta-analysis using Review Manager (version 5.2 Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012). The quality of evidence was rated for the different outcomes following the GRADE method [bib_ref] GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables, Guyatt [/bib_ref]. The online GRADE program was used for this purpose (GRADEpro Guideline Development Tool [Software], McMaster University, Hamilton, ON, Canada, 2015, developed by Evidence Prime, Inc., available from gradepro.org).
## Step 4: concept guideline
From the initial group of 40 gynecologists who participated in the brainstorm session, an expert panel of 18 members was selected. The expert panel and the members of the guideline met a few times to discuss the research questions according to a standard template. The final recommendations were graded according to the classification used by the American Association of Gynecologic Laparoscopists (AAGL) which was adapted from the US Preventive Services Task Force [11]: Level A: Recommendations are based on good and consistent scientific evidence; Level B: Recommendations are based on limited or inconsistent scientific evidence; Level C: Recommendations are based primarily on consensus and expert opinion.
The experts wrote the first draft, after which the working guideline group merged the different topics into one document and finalized the guideline. All experts involved in the development of this document approved the guideline in its present form.
Step 5: Validation of the guideline Two independent reviewers, different committees within the NVOG as well as the independent Knowledge Institute of Medical Specialists (KIMS) reviewed the guideline. After approval, our guideline was published on the website of the NVOG to allow all Dutch gynecologists to give feedback. The guideline will be soon adopted in The Netherlands and is valid for 5 years, after which it will be updated. If necessary, it will be (partially) updated earlier.
## Findings overall
For each of the ten main topics raised during the first brainstorm session, a literature search was performed. In total 5233 articles were reviewed and 119 unique articles were included in the guideline. In the following section, each topic and its best practice recommendations are briefly summarized. More detailed information regarding the selected literature, the quality of evidence according to the GRADE method, the search strings of the different topics and the forest plots of the main outcomes, will be published in the fall of 2017 on the website of the NVOG (http:// www.nvog.nl).
## Topic 1: a comparison of surgical approaches for hysterectomy
According to the Cochrane review on this topic, vaginal hysterectomy (VH) should be, when technically feasible, the approach of first choice, followed by LH and finally abdominal hysterectomy (AH) [bib_ref] Surgical approach to hysterectomy for benign gynaecological disease, Aarts [/bib_ref]. However, limitations of the Cochrane review are the lack of differentiation between the various subtypes of LH (total laparoscopic hysterectomy (TLH); laparoscopic-assisted vaginal hysterectomy (LAVH) and robotic hysterectomy (RH)), and the inclusion of data from older trials performed in the implementation period. Because of these potential limitations, a new literature search was performed for this guideline, specifically comparing TLH to VH. In topic 2, the different subtypes of LH were also compared to TLH. To limit the bias of a learning curve and reflect current practice, we only focused on studies published in the last 15 years (from 1st of January 2000 up to 1st of March 2016).
## Tlh versus vh
As can be observed in [fig_ref] Table 1: Summary of outcomes comparing TLH to VHTLH total laparoscopic hysterectomy, VH vaginal... [/fig_ref] , VH was associated with a significantly shorter operative time, a lower risk of conversion and a lower risk of vaginal cuff dehiscence. Patients in the TLH group had lower postoperative pain scores and required analgesia for a shorter period. The other outcomes were similar, and notably the risk of ureter and bladder injury did not differ between the groups, in contrast to what was found in previously published studies [bib_ref] Surgical approach to hysterectomy for benign gynaecological disease, Aarts [/bib_ref]. Many factors, such as patient and surgeon characteristics, influence the choice of approach. Our results show that since the implementation of LH, the differences in clinical outcomes between VH and TLH have been minimized. However, when both approaches are feasible, VH is still associated with more relevant benefits compared to LH and should therefore be the approach of first choice.
## Recommendations
- When both approaches are feasible, VH still offers the most relevant benefits and should be the approach of first choice. (level A-C, table 1)
## Tlh versus rh
The results of the meta-analysis showed no clinically relevant differences between TLH and RH for most surgical and patient outcomes. Regarding the costs of the procedure, no meta-analysis could be performed because of incomplete data. Yet, all studies showed that LH was significantly less expensive with mean differences of 1.916 US dollars [bib_ref] Robotic hysterectomy versus conventional laparoscopic hysterectomy: outcome and cost analyses of a..., Sarlos [/bib_ref] , 3.049 US dollars [bib_ref] A systematic review and cost analysis of robot-assisted hysterectomy in malignant and..., Tapper [/bib_ref] and 11.214 US dollars [bib_ref] Costs and outcomes of abdominal, vaginal, laparoscopic and robotic hysterectomies, Wright [/bib_ref].
## Recommendations
- For LH, RH has no advantages and is associated with higher costs. (level B)
## Tlh versus supra-cervical laparoscopic hysterectomy (slh)
The results of the meta-analysis for this topic are summarized in [fig_ref] Table 2: Summary of outcomes comparing TLH to SLH [/fig_ref]. Despite the fact that most included studies were underpowered and nonrandomized, the expert panel concluded that no major differences were observed between the two procedures, except potentially for complications. In addition, it is important to realize that in the SLH group morcellation is always necessary, which could result in more (mini)laparotomies (topic 8). Finally, the pre-operative cervix cytology, the impact of follow-up screening and the increased risk of cyclic bleeding should also be considered when weighing the pros and cons of the two procedures.
## Recommendations
- No clinically relevant surgical differences were found between TLH and SLH, except potentially for complications. (level B)
- It is important to counsel a patient about the pros and cons of both approaches . (level C)
- Shared decision making is recommended. (level C)
## Topic 3: what is the added value of pre-operative treatment-gonadotropin-releasing hormone agonists (gnrha) or ulipristal-prior to lh for uterine fibroids?
This topic evaluated the effect of pre-operative medical treatment (GnRHa and/or Ulipristal) on complication risk, conversion risk, intra-operative blood loss and operative time during LH. The available evidence was limited, especially because many studies did not differentiate between the different approaches of hysterectomy (abdominal, vaginal and laparoscopic). Based on the selected literature, we concluded that there is currently no need to standard pre-operatively treat patients who desire LH for uterine fibroids as the advantages are marginal. However, substantial volume reduction has been demonstrated in some studies (2 weeks in gestational age [bib_ref] Pre-operative GnRH analogue therapy before hysterectomy or myomectomy for uterine fibroids, Lethaby [/bib_ref] , including a 47% reduction in the study of Donnez et al [bib_ref] Ulipristal acetate versus leuprolide acetate for uterine fibroids, Donnez [/bib_ref]. Therefore, for each patient a well-considered decision should be made, taking into account the expected volume reduction and hence the possibility for a minimally invasive approach, the side effects and the costs of treatment. - If prescribed, GnRHa should be given for at least three months. (level C)
## Recommendations
## Topic 4.1: which patient characteristics influence surgical outcomes during laparoscopic hysterectomy?
To answer this research question, one systematic review was selected [bib_ref] Case-mix variables and predictors for outcomes of laparoscopic hysterectomy: a systematic review, Driessen [/bib_ref]. In this review, associations between patient characteristics and surgical outcomes of LH were described based on 85 articles (four RCTs, 29 prospective cohort studies, 47 retrospective cohort studies and five case-control studies).
## Recommendations
- It is necessary to discuss with patients the fact that high BMI, large uterine weight and/or previous surgeries (e.g., intra-abdominal adhesions) influence intraoperative blood loss, operative time and complication and conversion risks (level A)
## Topic 4.2: what is the added value of bimanual examination and medical imaging (ultrasound, mri) prior to hysterectomy?
Pre-operative gynecological examination (speculum and bimanual examination) gives surgeons information on uterine mobility and an appropriate estimation of the uterine weight. These findings are relevant for determining the operability of the patient (i.e., best surgical approach). Additionally, an ultrasound is useful for detecting potential intra-abdominal pathologies. The expert panel agreed that an MRI is not necessarily superior to ultrasound for hysterectomy with benign indications. should always be performed to estimate the operability of a patient and predict the best surgical approach. (level C)
- A MRI is not a standard requirement for LH.
Ultrasound is sufficient to detect potential additional pathology. (level C)
## Topic 5: which instrument is the most appropriate: bipolar electrothermical energy or ultrasonic energy?
The aim of this topic was to compare bipolar electrothermical energy with ultrasonic energy, particularly with respect to patient safety. Electrothermical energy with monopolar instruments was not included in this topic.
Because of the rapid development of (new) instruments, studies quickly become outdated. The differences observed in surgical outcomes between instruments (bipolar electrothermical energy versus ultrasonic energy) were probably also influenced by surgeon's experience and preference as well as by the surgical task performed. As differences in clinical findings were small, the expert panel concluded that there was no preference of one instrument over the other. The expert panel emphasized that experience with a specific instrument is valuable and essential for a safe procedure.
## Recommendations
## Topic 7: which techniques prevent and/or detect ureter injuries during lh?
## Ureter stents
As limited evidence was available for benign LH, the search was extended to articles included oncological and endometriosis/DIE cases. Ureter stents do not seem to prevent ureter injury as no significant difference was observed for ureter injuries between the group with and the group without stents [OR 2.45 (0.28; 21.29)]. Standard stent placement could also result in unnecessary complications. Stents are, however, easy to insert and improve the identification of the ureters. In the Delphi study by Janssen et al., the experts did not reach consensus regarding the additional value of ureter stents during LH [bib_ref] Recommendations to prevent urinary tract injuries during laparoscopic hysterectomy: a systematic Delphi..., Janssen [/bib_ref].
## Recommendations
- Standard insertion of ureter stents during LH is not recommended. (level B)
- In case of expected distorted anatomy (e.g., oncology, DIE), stents can be considered. (level C) Cystoscopy Cystoscopy appears to be safe and results in limited extension of the operative time (mean additional time 13 min). When the overall risk of bladder and/or ureter injuries is below 2%, a standard cystoscopy is not cost-effective for LH [bib_ref] Cost-effectiveness of universal cystoscopy to identify ureteral injury at hysterectomy, Visco [/bib_ref]. The American Association of Gynecologic Laparoscopists (AAGL) have recommended the standard use of a cystoscopy after LH. The expert panel, on the other hand, concluded that based on available evidence, including incidence data and data on cost-effectiveness, there is insufficient justification to recommend routine cystoscopy after LH. However, the threshold to perform a cystoscopy should be low. When injuries are suspected intra-operatively, additional diagnostics during surgery is recommended and for this a cystoscopy can be of additional value. At last, one should be aware that a normal cystoscopy does not exclude the presence of (lateral thermal) injury, especially for ureter injuries.
## Recommendations
- A standard cystoscopy after LH is not recommended as the additional value of it has not been proven. (level B)
- When a urinary tract injury is suspected intra-operatively, a low threshold for additional diagnostics is recommended Based on the available evidence, we concluded that the incidence of unexpected sarcoma varies between 1:350 and 1:2000 [bib_ref] The prevalence of occult leiomyosarcoma at surgery for presumed uterine fibroids: a..., Pritts [/bib_ref] and increases with age [bib_ref] Options on fibroid morcellation: a literature review, Brolmann [/bib_ref]. Other risk factors associated with uterine sarcomas are the following: African race, Tamoxifen use, previous radiotherapy in the pelvic area, HLRCC syndrome and retinoblastoma in the past medical history [bib_ref] Options on fibroid morcellation: a literature review, Brolmann [/bib_ref]. The exact impact of malignant spill on overall survival is uncertain, but the risk of upstaging due to morcellation has been estimated to be between 15 and 64% [bib_ref] The prevalence of occult leiomyosarcoma at surgery for presumed uterine fibroids: a..., Pritts [/bib_ref]. One of the proposed solutions to minimize spillage of occult malignancy or parasitic myomas is the use of containment bags during morcellation. Although these bags are certainly not optimal yet, they are theoretically able to prevent spread of (malignant) tissue in the abdomen. Gynecologists performing LH should thoroughly counsel their patients and should acquire the skills of inbag morcellation so that they can offer all the options to their patients. The ESGE developed a flow chart allowing patients to be classified into a low-or high-risk category for sarcomas based on their risk factors and ultrasound results [bib_ref] Options on fibroid morcellation: a literature review, Brolmann [/bib_ref]. However, as long as the nature of the uterine mass cannot be diagnosed pre-operatively with certainty, such classifications are not entirely reliable. - When uncontained morcellation is estimated to be unsafe, perform 'contained morcellation' or a (mini)laparotomy to obtain the specimen. (level C) Topic 9: When is the best moment to remove the urinary catheter after LH?
## Recommendations
Using a urinary catheter during LH is recommendedbut the best moment to remove it is unclear. Although evidence was limited, particularly for LH, it seems safe to remove the urinary catheter immediately after hysterectomy. Insufficient evidence was available to determine if leaving the catheter for 6 h offers better outcomes than immediate removal. Leaving the catheter longer than 6 h does not seem to offer any benefits whereas it does increase the risk of urinary tract infection and prolonged hospital stay.
## Recommendations
- It is recommended to remove the urinary catheter within six hours after LH. (level C) Topic 10: What advice and/or interventions are helpful to promote postoperative recovery?
Sufficient evidence is available to state that LH is associated with a shorter hospital stay and a quicker recovery than AH [bib_ref] Surgical approach to hysterectomy for benign gynaecological disease, Aarts [/bib_ref]. However, research has shown that the time to return to normal activities after LH (i.e., time to return to work) takes overall longer than would be expected [bib_ref] Prediction of time to return to work after gynaecological surgery: a prospective..., Vonk [/bib_ref]. To maximize the benefits of minimally invasive surgery, it is important to adequately guide patients during recovery at home. The complexity of the surgery, the pre-operative expectations of the patient and their pre-operative mental status seem to directly influence the patients' risk of prolonged absence due to sickness. Therefore, it is important to pre-operatively discuss expectations with the patients. In addition, structured and specific advice results in quicker recovery and E-Health programs can be used for that purpose, Finally, specific advice is needed for each type of hysterectomy as advice is not generalizable for all approaches of hysterectomy [bib_ref] A personalised eHealth programme reduces the duration until return to work after..., Vonk [/bib_ref].
# Discussion and conclusion
This guideline serves as a summary of best practices of LH, and it should provide clinicians with relevant and evidencebased information for daily practice. In other countries such as Germany, guidelines on hysterectomy have been developed as well with similar recommendations [bib_ref] Hysterectomy for benign uterine disease, Neis [/bib_ref]. Besides the fact that such guidelines provide surgeons with an overview of the most relevant topics, studies have shown that standardization of care and subsequent guideline compliance is associated with better outcomes and reduced medical liability [bib_ref] Reduced medicolegal risk by compliance with obstetric clinical pathways: a case-control study, Ransom [/bib_ref] [bib_ref] Does standardization of care through clinical guidelines improve outcomes and reduce medical..., Kirkpatrick [/bib_ref]. Regarding the medico-legal consequences of this guideline, it is probable that in the future it may be used for litigation in The Netherlands. Deviating from this standard of care is obviously allowed, provided that the motivation is thoroughly documented.
Regarding the methodology of this guideline, we focused on systematic reviews and RCTs. If insufficient evidence was available from the RCTs, we added cohort studies to our analysis. A limitation of this approach is that it increases the methodological and clinical heterogeneity. For instance, by including cohort studies, differences in baseline characteristics might exist, which could have influenced the outcomes. On the other hand, this method can also been seen as a strength because for rare events RCTs are often not the best study design as they are often underpowered. During the development of this guideline, we realized that, although GRADE is currently a well-established instrument to assess the quality of evidence [bib_ref] GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables, Guyatt [/bib_ref] , it has its limitations as well. The main problem we encountered was that for many topics the available evidence was limited and therefore the quality of the evidence was instantly downgraded to 'low' or 'very low'. This point has been raised previously by other authors [bib_ref] The GRADE system for rating clinical guidelines, Kavanagh [/bib_ref] and the GRADE working group [bib_ref] Grading quality of evidence and strength of recommendations: a perspective, Ansari [/bib_ref] has stated that on occasion even low available evidence can lead to strong recommendations. The GRADE working group has also emphasized that clinical and cultural settings are of influence and might result in (slightly) different recommendations across countries [bib_ref] Grading quality of evidence and strength of recommendations: a perspective, Ansari [/bib_ref]. Therefore it is essential to choose an expert panel that is well-supported [bib_ref] Grading quality of evidence and strength of recommendations: a perspective, Ansari [/bib_ref]. As the development of our guideline was initiated by the Dutch medical society itself, we believe we had support from the entire country, especially since the panel was a good representation of all Dutch gynecologists.
# Conclusion
The guideline for LH serves as guidance for gynecologists performing LHs. The recommendations in this best practice review should enhance quality of care, minimize (unfavorable) practice variations at the (inter)national level and thereby increase patient safety.
[fig] •: Standard pre-operative treatment of patients with fibroids does not seem advisable as the advantages are marginal. (level B) When uterine volume reduction is desirable, especially to increase the possibility for a minimally invasive approach, pre-operative treatment with GnRHa should be considered. (level B) [/fig]
[table] Table 1: Summary of outcomes comparing TLH to VHTLH total laparoscopic hysterectomy, VH vaginal hysterectomy; RCT randomized controlled trial [/table]
[table] Table 2: Summary of outcomes comparing TLH to SLH [/table]
|
Long-Term Care and the COVID-19 Pandemic
INTRODUCTIONApproximately 74% of the deaths in the United States from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that causes COVID-19 have occurred in adults age 65 years and older. 1 Intrinsic vulnerabilities include comorbidity and frailty; residence in communal living spaces considered "hot spots" have contributed to higher risk. 2 US nursing homes (NHs) and other long-term care (LTC) communities such as assisted living and adult day care services have been disproportionally affected by COVID-19. 2-6 Nurses and health care workers provided care and services despite health concerns for themselves and family members. Nurses on the frontline were called to act with extraordinary tenacity, skill, flexibility, and creativity. 7 The purpose of this article is to provide an overview of the challenges posed by the COVID-19 pandemic in LTC settings and the strategies prioritized and implemented with interdisciplinary colleagues in NHs, assisted living, and adult day services.
THE NURSING HOME CONTEXT, STRATEGIES, AND LESSONS LEARNED The most severely affected by COVID-19 were residents in the approximately [bib_ref] Characteristics of U.S. Nursing Homes with COVID-19 Cases, Abrams [/bib_ref] NHs, also referred to as skilled nursing facilities. [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] Individual states have reported that up to 50% of cases and deaths attributed to the virus have occurred in these facilities.These care communities are certified by Medicare and/or Medicaid and include almost 1.3 million residents in the United States. [bib_ref] Characterization of COVID-19 in Assisted Living Facilities -39 States, Yi [/bib_ref] Most NHs provide both LTC for residents with high care dependency and complex health care needs, as well as short-term, postacute care for patients admitted from the hospital who require highly skilled nursing care and/or rehabilitation. [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] Short-term posthospital care is largely reimbursed by Medicaid, whereas most LTC is reimbursed by Medicaid. [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] [bib_ref] Driven to Tiers: Socioeconomic and Racial Disparities in the Quality of Nursing..., Vincent Mor [/bib_ref] NHs that are predominantly dependent on Medicaid have less resources and lower staffing levels, are located in the poorest neighborhoods, and have the most quality problems. [bib_ref] Driven to Tiers: Socioeconomic and Racial Disparities in the Quality of Nursing..., Vincent Mor [/bib_ref] Nursing home residents have many of the comorbidities that are considered risk factors for COVID-19 mortality (cardiovascular diseases, diabetes, chronic respiratory disorders, hypertension, and cancer). [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] Research on the facility characteristics associated with higher COVID-19 cases are conflicting. Some research indicated that NHs with low RN and total staffing levels have seemed to leave residents vulnerable to COVID-19 infections. [bib_ref] Nursing staffing and coronavirus infections in California NHs, Harrington [/bib_ref] [bib_ref] COVID-19 infections and deaths among Connecticut nursing home residents: facility correlates, Li [/bib_ref] Lower nursing home quality ratings (overall quality rating of 1 to 5 stars based on performance on 3 domains, each rated on 1 to 5 stars: health inspections, nurse staffing, and resident quality measures) were also found to be associated with higher COVID-19 incidence, mortality, and persistence. [bib_ref] COVID-19 infections and deaths among Connecticut nursing home residents: facility correlates, Li [/bib_ref] [bib_ref] The association of nursing home quality ratings and spread of COVID-19, Wiliams [/bib_ref] [bib_ref] Association Between CMS Quality Ratings and COVID-19 Outbreaks in Nursing Homes -West..., Bui [/bib_ref] In contrast, it was the location of a nursing home, asymptomatic spread, and availability of testing-not quality ratings, infection citations or staffing-that were found to be determining factors in COVID-19 outbreaks, according to independent analyses by leading academic and health care experts, as well as government researchers. [bib_ref] Characteristics of U.S. Nursing Homes with COVID-19 Cases, Abrams [/bib_ref] [bib_ref] Rates of COVID-19 Among Residents and Staff Members in Nursing Homes -United..., Bagchi [/bib_ref] [bib_ref] Risk Factors Associated With All-Cause 30-Day Mortality in Nursing Home Residents With..., Panagiotou [/bib_ref] Larger facility size, urban location, and a greater percentage of African American residents were also associated with higher rates of infection. [bib_ref] Characteristics of U.S. Nursing Homes with COVID-19 Cases, Abrams [/bib_ref] A study of 12,576 US NHs, conducted by Li and colleagues [bib_ref] COVID-19 infections and deaths among Connecticut nursing home residents: facility correlates, Li [/bib_ref] found that NHs with higher numbers of racial and ethnic minorities reported greater incidences of confirmed COVID-19 cases and deaths. In addition, in a cohort study of US nursing home residents with COVID-19, increased age, male sex, and impaired cognitive and physical function were independently associated with mortality. 19
## Nurse leaders influencing policy
NHs experienced a steep learning curve during the COVID-19 pandemic. [bib_ref] Long-term care facilities and the coronavirus epidemic: practical guidelines for a population..., Dosa [/bib_ref] [bib_ref] Nursing homes during the COVID-19 pandemic: a scoping review of challenges and..., Giri [/bib_ref] [bib_ref] Expert nurse response to workforce recommendations made by The Coronavirus Commission For..., Bakerjian [/bib_ref] [bib_ref] A call to the CMS: Mandate adequate professional nurse staffing in nursing..., Kolanowski [/bib_ref] NHs were required to quickly mobilize to implement the use of personal protective equipment (PPE), testing, and restrictive visitation policies. The fast spread of COVID-19 challenged federal and state agencies to provide timely guidance and NHs to implement those changes and still meet the daily health care needs of their residents.Nurse leaders assumed a critical role at the national and facility level to provide information, develop and help implement policy, and educate staff. For example, Deb Bakerjian developed and continues to update the Web site, "Coronavirus Disease 2019 (COVID-19) and Safety of Older Adults Residing in Nursing Homes" on the Agency for Healthcare Research and Quality (AHRQ) Patient Safety network. The primer is a compilation of information that has affected the safety of older adults and has been published on federal Web sites, in professional and academic literature and in the press.A critical nursing role that has been highlighted by the pandemic is the infection preventionist (IP). The Centers for Medicare and Medicaid Services (CMS) in October 2016 expanded NH infection prevention and control (IPC) requirements to include an antibiotic stewardship program and a designated individual to serve as an IP to oversee the program; this was based on research demonstrating that NHs with IPs who had specialized training were 5 to 13 times more likely to have a stronger IPC program.Before pandemic, IPs typically had multiple roles and thus are not able to dedicate their full time to IPC. [bib_ref] Infection prevention and control in nursing homes during COVID-19: An environmental scan, Rubano [/bib_ref] The pandemic illuminated the need for a full-time essential IP to coordinate interdisciplinary activity including tracking cases of infection, educating staff, overseeing testing, and monitoring facility IPC practices. [bib_ref] Changes in US nursing home infection prevention and control programs from, Agarwal [/bib_ref] [bib_ref] Infection preventionist staffing in nursing homes, Stone [/bib_ref] The California state nursing workforce center (HealthImpact) partnered with nurse leaders within the state to address macrolevel nursing workforce issues during the pandemic. Academic and clinical practice nurse leaders created guidance documents for schools of nursing and clinical agencies to support and encourage safe academicpractice partnerships during the pandemic. [bib_ref] The Importance and Impact of Nurse Leader Engagement With State Nursing Workforce..., Chan [/bib_ref] The coalition worked with the California Board of Registered Nursing to also create official guidance documents to explain the various roles nursing students can assume to contribute to the workforce during the pandemic. They developed free high-quality simulations and supported a bill that codified the governor's waiver for increased use of simulation in disaster situations. The coalition also created a toolkit to help introduce or refresh essential knowledge for retired nurse returning to the workforce and nursing students and worked with other nurse leader organizations to create resources, webinars, and podcasts to support the well-being and health of nurses. Finally, the coalition created a volunteer registration and matching system for interprofessional health care licensees and students to staff vaccination events throughout California especially in communities of color and hard-to-reach communities. [bib_ref] The Importance and Impact of Nurse Leader Engagement With State Nursing Workforce..., Chan [/bib_ref] The CMS recognized the urgency of the COVID-19 crisis and convened the Coronavirus Commission for Safety and Quality in NHs in April, 2020.The 25-member commission was composed of academicians, clinicians, NH administrators, family members, residents, industry professionals, and scientific experts. The members were charged with making recommendations to improve infection prevention and control, safety procedures, and the quality of life of residents in NHs. The final report of the Commission contained 27 recommendations and was released in September 2020.A group of geriatric nurse experts responded to this report and while confirming the committee's observations, posited that there were other policy weaknesses that have been long-standing problems and exacerbated the COVID-19 crisis. The experts identified weaknesses including chronic overall understaffing and insufficient numbers of professional nurses (that is, registered nurses [RNs]), insufficient geriatric expertise for managing complex care problems, and a culture focused on regulatory compliance, to the exclusion of quality improvement. [bib_ref] Nursing homes during the COVID-19 pandemic: a scoping review of challenges and..., Giri [/bib_ref] Consequently, the nurse experts made the following recommendations: (1) ensuring RN coverage on a daily, around the clock basis and providing adequate compensation to provide total staffing levels that are commensurate with the needs of the residents; (2) ensuring RNs have both clinical and leadership competencies in geriatric nursing, including quality improvement and supervisory skills; (3) increasing efforts to recruit and retain the NH workforce, particularly RNs; and (4) supporting care delivery models, including the professional practice model, that strengthen the role of the RN. [bib_ref] Expert nurse response to workforce recommendations made by The Coronavirus Commission For..., Bakerjian [/bib_ref] [bib_ref] A call to the CMS: Mandate adequate professional nurse staffing in nursing..., Kolanowski [/bib_ref] [bib_ref] Adapting a professional practice model, Silverstein [/bib_ref]
## Care of residents and families
In addition to enormous mortality and morbidity, nursing home residents faced the unintended consequences of restricted visitation and isolation practices that were implemented to curtail exposure to COVID-19. [bib_ref] Social isolation-the other COVID-19 threat in nursing homes, Abbasi [/bib_ref] [bib_ref] Loneliness and social isolation during the COVID-19 pandemic, Hwang [/bib_ref] [bib_ref] Competing crises: COVID-19 countermeasures and social isolation among older adults in long-term..., Chu [/bib_ref] [bib_ref] The Impact of COVID-19 on Social isolation in long-term care homes: Perspectives..., Chu [/bib_ref] [bib_ref] High depression and anxiety in people with Alzheimer's disease living in retirement..., El Haj [/bib_ref] [bib_ref] Psychosocial Impact of COVID-19 nursing home restrictions on visitors of residents with..., O' Caoimh [/bib_ref] [bib_ref] Loneliness and isolation in long-term care and the COVID-19 pandemic, Simard [/bib_ref] NHs reported that that residents had stopped eating and had "given up" without family visitation. [bib_ref] Weight Loss in COVID-19-Positive Nursing Home Residents. Research Letters, Martinchek [/bib_ref] Residents were Long-Term Care and the COVID-19 Pandemic confined to their rooms, precluding their engagement in communal meals, activities, and normal levels of physical activity, all factors associated with delirium, nutritional problems, psychological distress, functional decline, and falls. [bib_ref] Weight Loss in COVID-19-Positive Nursing Home Residents. Research Letters, Martinchek [/bib_ref] [bib_ref] Nursing home resident weight loss during coronavirus disease 2019 restrictions, Danilovich [/bib_ref] [bib_ref] Experiences of residents, family members and staff in residential care settings for..., Sweeney [/bib_ref] A scoping review of restricted visitation described significant mental health consequences for the patient, including anxiety, loneliness, depressive symptoms, agitation, aggression, reduced cognitive ability, and overall dissatisfaction. [bib_ref] Social connection in long-term care homes: A scoping review of published research..., Bethell [/bib_ref] Residents' families were also negatively affected by this practice; lack of connectivity with loved ones added an additional layer of stress to the worry and anxiety they experienced in the face of the pandemic crisis. [bib_ref] Experiences of residents, family members and staff in residential care settings for..., Sweeney [/bib_ref] Simple interventions promoted connectivity with families. For example, families were encouraged to drop off letters, drawings, or other packages and maintain contact with regular telephone check-ins. [bib_ref] Social connection in long-term care homes: A scoping review of published research..., Bethell [/bib_ref] Video calls for residents to interact virtually with their families have been shown to decrease depressive symptoms, loneliness, and increase social interaction and quality of life. [bib_ref] Practical nursing recommendations for palliative care for people with dementia living in..., Bolt [/bib_ref] [bib_ref] Videoconference program enhances social support, loneliness, and depressive status of elderly nursing..., Tsai [/bib_ref] [bib_ref] Social isolation and psychological distress among older adults related to COVID-19: a..., Gorenko [/bib_ref] However, many NHs do not have the resources to promote everyday technology use to enable residents to connect with providers and families, which worsened health disparities. [bib_ref] Telemedicine disparities during COVID-19: provider offering and individual technology availability, Jacobs [/bib_ref] [bib_ref] Excess mortality in long-term care residents with and without personal contact with..., Savage [/bib_ref] For example, a study conducted by Savage and colleagues [bib_ref] Excess mortality in long-term care residents with and without personal contact with..., Savage [/bib_ref] found that residents who did not have personal contact, including phone calls, with loved ones during COVID-19 restrictions experienced 35% greater excess mortality compared with residents who had personal contact. When tablets and laptops have been made available, nursing staff have reported benefits such as increased resident well-being and increased family engagement in care plan meetings. [bib_ref] Psychosocial Impact of COVID-19 nursing home restrictions on visitors of residents with..., O' Caoimh [/bib_ref] [bib_ref] Excess mortality in long-term care residents with and without personal contact with..., Savage [/bib_ref] [bib_ref] Telemedicine in long-term care facilities during and beyond COVID-19: challenges caused by..., Seifert [/bib_ref] In addition, staff described greater ease in educating families and connecting the residents to health care consultants. [bib_ref] Telemedicine in long-term care facilities during and beyond COVID-19: challenges caused by..., Seifert [/bib_ref] [bib_ref] Communication Technology Improved Staff, Resident, and Family Interactions in a Skilled Nursing..., Vu [/bib_ref] Resident unfamiliarity with technology and staff concern about technology being an additional burden have been barriers. However, studies demonstrated that residents were receptive to technology and technology actually reduced staff burden by keeping residents engaged and freeing up staff for other activities. [bib_ref] Videoconference program enhances social support, loneliness, and depressive status of elderly nursing..., Tsai [/bib_ref] [bib_ref] Social isolation and psychological distress among older adults related to COVID-19: a..., Gorenko [/bib_ref] Another technology known as simulated presence therapy, whereby recorded video messages from family or friends are replayed frequently, led to the enhanced well-being of residents living with dementia and decreased behavioral symptoms of distress. [bib_ref] Loneliness and isolation in long-term care and the COVID-19 pandemic, Simard [/bib_ref] The pandemic also demanded excellent basic nursing care. [bib_ref] Social connection in long-term care homes: A scoping review of published research..., Bethell [/bib_ref] Nursing staff have reported that the use of face masks that have a clear window has helped with communication with residents. Preference congruence and positive interactions were supported by interdisciplinary assessment of the residents' psychosocial needs, backgrounds, and preferences. [bib_ref] Social connection in long-term care homes: A scoping review of published research..., Bethell [/bib_ref] [bib_ref] Communication Technology Improved Staff, Resident, and Family Interactions in a Skilled Nursing..., Vu [/bib_ref] [bib_ref] Leading a long-term care facility through the COVID-19 crisis: successes, barriers and..., Havaei [/bib_ref] Improvement initiatives focused on measures such as structured, consistent rounding to provide cognitive stimulation and social interaction, snacks and fluids, physical activity. The use of exergames have been shown to improve functional capacity and increased interaction with other residents, families, and friends. In addition, simple approaches to increase physical activity, including sit-to-stand exercises, walking, and encouragement of self-care in hygiene and grooming, could be incorporated into rounds and room visitations. [bib_ref] Care homes and COVID-19 in Hong Kong: how the lessons from SARS..., Chow [/bib_ref] [bib_ref] Person-centered physical activity for nursing home residents with dementia: The perspectives of..., Chu [/bib_ref] [bib_ref] The importance of physical activity to care for frail older adults during..., Aubertin-Leheudre [/bib_ref] [bib_ref] Providing simultaneous COVID-19-sensitive and dementia-sensitive care as we transition from crisis care..., Canevelli [/bib_ref] Trauma-Informed Care
For residents, staff, and families, the COVID-19 pandemic has been marked by isolation, uncertainty, and loss, all hallmarks of trauma. [bib_ref] Trauma-Informed Care and Regulatory Expectations, Levenson [/bib_ref] The Substance Abuse and Mental Health Services Administration (SAMHSA) guidelines define trauma as an event or pattern of occurrences that is experienced as harmful or threatening and has ongoing negative effects on the person.Viewing the pandemic as traumatic underscores the need to provide support to mitigate the effects of trauma. LTC residents' response to trauma can present as increased anxiety, agitation, aggression, or withdrawal. Cornerstones of trauma-informed care include trying to understand the person (knowing the person, looking for the reason behind behavior), acting with empathy, engaging the person's social support, and providing choice whenever possible. The mainstay of treatment is nonpharmacological interventions including communicating and acting with empathy.Trauma experts have provided guidance during the pandemic to support the well-being of all members of the NH community, in the face of trauma that feels unrelenting and tests resilience. Box 1 provides a summary for increasing compassion resiliency and moral wellness. [bib_ref] Best Practices for Increasing Compassion Resiliency and Moral Wellness in Post-Acute and..., Beausoleil [/bib_ref] Nursing home staff were particularly exposed and vulnerable to COVID-19. Many of these workers struggle with grief over the suffering they have witnessed, both at work and in their communities. Some have been infected with COVID-19 and recovered physically-but not emotionally. In addition to many experiencing COVID illness, staff at all levels described dramatic levels of personal distress, including feelings of helplessness, fear, and anxiety. [bib_ref] Best Practices for Increasing Compassion Resiliency and Moral Wellness in Post-Acute and..., Beausoleil [/bib_ref] Balancing high, often concurrent, and competing demands of work and family was reported as a major source of stress. There was frustration with operational challenges including inadequate access to personal protective equipment and COVID testing and lack of information and consistent guidance. [bib_ref] Grief reactions in dementia carers: a systematic review, Chan [/bib_ref] Moreover, staff were dealing with their grief when residents died.
Attention to signs of burnout, difficulty coping, and substance misuse have required promote attention and referral to employee assistance programs and mental health professionals.
Strategies that NHs have implemented to support staff well-being include attention to the work environment including ensuring sufficient supply of PPE; consistency in assigned units; and clear, direct, and frequent communication from supervisors. Another strategy included instituting formal, scripted pauses at set times or debriefing after resident deaths or other stressful events. Moments of pause and huddles to acknowledge and verbalize difficulties and promote safe discussion have been helpful. Nurse administrators have also promoted staff morale and well-being through numerous small acts such as acknowledging team members for their work contributions, celebrating patient recoveries, and highlighting recognition from patients and families. [bib_ref] Turning the Lens of Trauma-Informed Care Toward Staff with Stress First Aid, Sjostrom [/bib_ref] [bib_ref] Face to Face with Coronavirus Disease 19: Maintaining Motivation, Psychological Safety, and..., Lateef [/bib_ref] The pandemic crisis has demanded administrators and managers that are consistently present; attentive to multiple sources of information; nimble in their planning and responses; and effective in their communication with staff, residents, and families. The Role of the Nurse Practitioner in COVID-19-Related Care
Nurse practitioners have a proven track record of providing safe and cost-effective care in LTC, improving resident outcomes, in the areas of polypharmacy, falls, restraint use, and transfers to acute care and thus are well positioned to lead and respond to the complexity of COVID-19 care. [bib_ref] A mixed methods quality improvement study to implement nurse practitioner roles and..., Kilpatrick [/bib_ref] [bib_ref] The APRN role in changing nursing home quality: the Missouri quality improvement..., Popejoy [/bib_ref] [bib_ref] Results of the Missouri quality initiative in sustaining changes in nursing home..., Vogelsmeier [/bib_ref] [bib_ref] The effect of Evercare on hospital use, Kane [/bib_ref] In addition to conducting comprehensive assessments, rapid recognition and response to clinical deterioration, symptomatic care, psychological support, and prevention of multiple potential complications, nurse practitioners provided other critical functions. Qualitative research with nurse practitioners described the foci of their work during the pandemic. They described the responsibilities associated with containing the spread of COVID-19 within the LTC homes, including working with management to implement and communicate the evolving COVID-19 recommendations from local health authorities and implement pandemic protocols such as resident cohorting and isolation plans, testing, and infection control practices. In many cases, the nurse practitioners provided the in-person clinical visits with residents when the physicians were making virtual visits only. A third critical function was providing emotional support to fearful staff and residents' families and education about the pandemic and resident care needs. Maintaining relationships between residents and their families was also a focus of the clinicians' work. Finally, the nurse practitioners acted as a liaison with other health care systems including acute care, home care, and emergency departments and worked with them to create policies, strategies, and algorithms to minimize fragmentation in care. [bib_ref] Nurse Practitioners Rising to the Challenge During the Coronavirus Disease 2019 Pandemic..., Mcgilton [/bib_ref] Box 2 Items in the check-list of best practice policies for nursing leaders In another study nurse practitioners described the complexity of their roles in promoting a dignified death for LTC residents during the COVID-19 pandemic. [bib_ref] Complexity of Nurse Practitioners' Role in Facilitating a Dignified Death for Long-Term..., Vellani [/bib_ref] They described intensive engagement with residents and more frequently their care partners, to facilitate advance care planning and goals of care conversations. The nurse practitioners promoted comfort at end of life by prescribing pharmacologic and nonpharmacologic interventions symptom management; consulting with expert clinicians where needed; addressing the psychosocial needs of residents and families, and providing education to staff on comfort measures. Nurse practitioners also facilitated compassionate visits to care partners when their residents were imminently dying, allowing them to stay as long as possible while making sure that their PPE remained safely useable. The nurse practitioners also described "care after death," which included informing the care partners of the death, upholding the pandemic-related policy of allowing only one grieving care partner at a time, and completing death certificates. They also educated and often supervised safe care of the resident's body after death. A good deal of time was spent providing emotional support to staff, with minimal time to focus on in their own self-care. [bib_ref] Complexity of Nurse Practitioners' Role in Facilitating a Dignified Death for Long-Term..., Vellani [/bib_ref] Experts in advanced practice registered nurse (APRN) practice have recommended that APRNs (both nurse practitioners and clinical specialists) use their leadership skills to provide consultation and lead quality improvement efforts beyond the COVID-19 pandemic. There is a critical need to use evidence of past successes to influence NH owners, operators, and policy makers to extend and strengthen the APRN role in LTC. [bib_ref] Impact of COVID-19 pandemic on APRN practice: Results from a national survey, Kleinpell [/bib_ref] [bib_ref] The Advanced Practice Registered Nurse Leadership Role in Nursing Homes Leading Efforts..., Bakerjian [/bib_ref] ASSISTED LIVING AND ADULT DAY SERVICE: CONTEXT, STRATEGIES, AND LESSONS LEARNED Assisted Living
More than 800,000 residents live in the 28,000 US assisted living facilities (ALFs); 52% are age 85 years and older and 30% are between the ages of 75 and 84 years. ALFs are not licensed as health care facilities; they do not provide round-the-clock skilled nursing care. There are no federal regulations for ALFs, and state regulations vary considerably. ALFs also vary widely in the array of services provided, ranging from around the clock assistance with daily living to on-call assistance. ALF care is largely private pay, although a small amount is reimbursed via Medicaid waiver programs. Some ALFs specialize in the care of people with dementia and other forms of cognitive impairment.The staffing, structure, and resources of ALFs limit the capacity of ALFs to respond to the COVID-19 outbreak.The staff are largely unlicensed direct care workers, and the number of nurses varies widely. Unlike NHs, there is no requirement for a medical director or an infection control practitioner. There are also no standard requirements for infection control or an infection control practitioner, as there are in NHs. Residents may receive care from their personal medical provider; they are not required to have regular medical visits. Residents often rely on external providers to provide home care attendants in their own apartment or rooms. Although residents could be restricted to their rooms, it would require significant staff to provide needed care and residents would need to agree to adhere to such restrictions, which makes it difficult to enforce such universal precautions. This structure is also not as conducive as NHs to cohort residents. Consequently, ALFs have relied on close collaboration with local health authorities and service agencies to guide infection control practices and monitoring. [bib_ref] Characterization of community-wide transmission of SARS-CoV-2 in congregate living settings and local..., Terebuh [/bib_ref] [bib_ref] The Unique Challenges Faced by Assisted Living Communities to Meet Federal Guidelines..., Dobbs [/bib_ref] Approximately 42% of assisted living residents have dementia, who may be particularly vulnerable to the adverse effects of social isolation and loneliness from visitor Long-Term Care and the COVID-19 Pandemic restriction and curtailing of group activities. [bib_ref] Loneliness and isolation in long-term care and the COVID-19 pandemic, Simard [/bib_ref] [bib_ref] Experiences of residents, family members and staff in residential care settings for..., Sweeney [/bib_ref] The overall lower numbers of staff and lack of a requirement for a nurse on staff predispose the residents to complications including falls, dehydration/nutritional problems, and delirium. Advanced practice RNs and home care nurses with palliative care expertise have been engaged in some ALFs and in some instances have addressed these challenges. In one ALF, in Washington, the palliative care nurse collaborated with APRNs to triage residents, manage symptoms, coordinate the functions of the interdisciplinary team, and monitor health patterns among all residents. A dedicated palliative RN was enlisted to coordinate prompt goals of care conversations with all residents, which was necessary because of the rapid deterioration of some residents once they became symptomatic. The palliative team supplemented the efforts of the remaining staff to provide an extra layer of support to meet the social, emotional, and spiritual needs of the residents who were finding themselves suddenly shut off from their usual social support in a time of crisis. [bib_ref] Ethical implications of COVID-19: Palliative care, public health, and long-term care facilities, De Campos [/bib_ref] In addition to the care provided to individual residents, the palliative care nurses and team supported the implementation of systemic approaches to support comfort and well-being in residents. The team brainstormed with staff to develop a tool to measure and track baseline function and promote activity and "new normal" activities of daily routine. Residents who qualified for a restorative plan of care were formally admitted to home care. Those who were at risk for deconditioning participated in exercises in their rooms; some followed exercise instruction via a closed-circuit television channel. The team worked with the facilities' dietary staff to provide fluids on each floor that could be offered between meals. Because residents missed the socialization and ambience of communal meals, staff members sat in the room with residents to socialize during meals and replated the meal from Styrofoam onto the resident's own dinnerware. Intake and output sheets were posted so that poor intake could be identified and addressed. Residents who were COVID-positive often deteriorated rapidly and were provided hospicelike care, including bereavement support to other residents, staff, and families. [bib_ref] Ethical implications of COVID-19: Palliative care, public health, and long-term care facilities, De Campos [/bib_ref] Even with the support of external services agencies and services, COVID-19 has presented ongoing challenges for assisted living to maintain resident quality of life. Notably, there have been more than 44,000 additional deaths due to dementia since February 1, 2020.Experts opine that this statistic underscores a need for attention to resident health and quality of life and the integration of more, consistent psychosocial and medical care into ALs. This type of progress would need to be informed by research on different models of integrated, interdisciplinary health care that engage relevant stakeholders including residents, families, staff, administrators, and regulators in their development and evaluation. [bib_ref] COVID-19 Recommendations for Assisted Living: Implications for the Future, Vipperman [/bib_ref]
## Adult day services
Adult day services provide a planned program offered in a group setting for older adults and persons with disabilities that offers social activities, meals, and health care monitoring. Services offered can vary; some provide case management, complex nursing care, rehabilitation, and caregiver training. [bib_ref] COVID-19 and the Need for Adult Day Services, Gaugler [/bib_ref] Adult day services also provide respite and relief to family caregivers so they can work or attend to other responsibilities and self-care needs. The programs are staffed by nurses, social workers, health aides, activity professionals, and other health professionals such as rehabilitation therapists. As of 2016, there were 4600 adult day programs serving approximately 286,300 older adults throughout the United States. [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] The predominant payer is Medicaid (66%). [bib_ref] Implications of the COVID-19 Pandemic on Adult Day Services and the Families..., Parker [/bib_ref] The Veterans Administration is the second largest public source of reimbursement. Medicare does not pay for adult day services. Some participants pay out of pocket for care and even fewer use LTC insurance to pay for care. [bib_ref] Long-term care providers and services users in the United States, Harris-Kojetin [/bib_ref] Participants in adult day services tend to have a high prevalence of chronic health conditions that have been associated with risk for severe illness from COVID-19 such as hypertension, diabetes, or dementia. [bib_ref] Implications of the COVID-19 Pandemic on Adult Day Services and the Families..., Parker [/bib_ref] Because of their multiple comorbidities, many clients who attend adult day services would be eligible for nursing home level care, yet because of their preferences to age-in-place they remain in the community and use adult day services. The COVID-19 pandemic forced ADS to close and abruptly end in-person services to clients. Most of them were closed due to a state mandate. [bib_ref] Understanding and addressing older adults' needs during COVID-19, Sands [/bib_ref] In a national survey, sites continued to provide included telephone support (n 5 22, 100%), delivery of food (n 5 8, 36.4%), medical check-ins (n 5 9, 40.1%), and activity via Zoom or YouTube (n 5 14, 63.6%). Most of these services were provided without reimbursement. In these cases, nurses and other staff and administrators volunteered their time demonstrating extraordinary commitment to their clients. In some states, Medicaid waivers covered reimbursement for daily telephonic wellness check-ins, online social and activities, and care-coordination services. [bib_ref] Service provision, hospitalizations, and chronic conditions in adult day services centers: Findings..., Caffrey [/bib_ref] An important lesson offered by the experience of adult day services is the valuable contribution of remote and flexible services offered during the pandemic. Future research is warranted that examines the clinical efficacy and cost-effectiveness of reimbursing these services.
## Summary
The needs of older adults admitted to LTC settings are increasingly complex. The COVID-19 pandemic has highlighted the need to not only prepare for crises but also respond with nurse-led care and services that are person-centered, familyengaged, and support interdisciplinary collaboration. Furthermore, models of care need to incorporate a comprehensive commitment to function and well-being-physical, social, and emotional-of older adults, their families, and staff.
## Clinics care points
The infection preventionist (IP) nurse is an essential role to track cases of infection, educate staff, oversee testing, and monitor facility infection prevention and control practices.
The unintended consequences of restricted visitation and isolation practices for residents include depression, anxiety, delirium, nutritional problems, symptoms of psychological distress, delirium, and functional decline.
Nursing interventions include encouraging regular family contact via telephone or video messages and recorded video messages from family/friends. Structured, consistent rounding to provide cognitive stimulation and social interaction, snacks and fluids, and physical activity help prevent delirium and functional decline. Sit-to-stand exercises, walking, and encouragement of self-care in hygiene and grooming can be incorporated into rounds and room visitations.
Trauma-informed resident care includes knowing the person, looking for the reason behind behaviors, acting with empathy, engaging the person's social support, and providing choices.
Strategies to support staff well-being include ensuring sufficient supply of equipment, consistent assignments, and clear, solid communication from supervisors, staff huddles that allow pause, acknowledging staff contributions, celebrating resident recoveries, and highlighting recognition from patients and families.
# Disclosure
The authors have nothing to disclose.
Long-Term Care and the COVID-19 Pandemic
|
Infectious virus in exhaled breath of symptomatic seasonal influenza cases from a college community
A complete list of the EMIT Consortium can be found in the Supporting Information.This article contains supporting information online at www.pnas.org/lookup/suppl/
Little is known about the amount and infectiousness of influenza virus shed into exhaled breath. This contributes to uncertainty about the importance of airborne influenza transmission. We screened 355 symptomatic volunteers with acute respiratory illness and report 142 cases with confirmed influenza infection who provided 218 paired nasopharyngeal (NP) and 30-minute breath samples (coarse >5-μm and fine ≤5-μm fractions) on days 1-3 after symptom onset. We assessed viral RNA copy number for all samples and cultured NP swabs and fine aerosols. We recovered infectious virus from 52 (39%) of the fine aerosols and 150 (89%) of the NP swabs with valid cultures. The geometric mean RNA copy numbers were 3.8 × 10 4 /30-minutes fine-, 1.2 × 10 4 /30-minutes coarse-aerosol sample, and 8.2 × 10 8 per NP swab. Fine-and coarse-aerosol viral RNA were positively associated with body mass index and number of coughs and negatively associated with increasing days since symptom onset in adjusted models. Fine-aerosol viral RNA was also positively associated with having influenza vaccination for both the current and prior season. NP swab viral RNA was positively associated with upper respiratory symptoms and negatively associated with age but was not significantly associated with fine-or coarse-aerosol viral RNA or their predictors. Sneezing was rare, and sneezing and coughing were not necessary for infectious aerosol generation. Our observations suggest that influenza infection in the upper and lower airways are compartmentalized and independent.
influenza virus | aerosol | airborne infection | vaccination effects | viral shedding T he nature of infectious contacts and the relative importance of contact, large-droplet spray, and aerosol (droplet nuclei) transmission remain controversial [bib_ref] Questioning aerosol transmission of influenza, Lemieux [/bib_ref] [bib_ref] Review of aerosol transmission of influenza A virus, Tellier [/bib_ref] [bib_ref] Aerosol transmission of influenza A virus: A review of new studies, Tellier [/bib_ref] [bib_ref] Transmission of influenza: Implications for control in health care settings, Bridges [/bib_ref] [bib_ref] Routes of influenza transmission, Killingley [/bib_ref]. Nonpharmaceutical interventions have been employed to control and reduce the impact of influenza epidemics and pandemics [bib_ref] Research findings from nonpharmaceutical intervention studies for pandemic influenza and current gaps..., Aiello [/bib_ref]. However, to design effective nonpharmaceutical interventions, it is necessary to accurately define the relative and absolute contribution of each route of transmission [bib_ref] Quantifying the routes of transmission for pandemic influenza, Atkinson [/bib_ref] and implement interventions that impede those of principal importance.
Mathematical models that have been used to understand and estimate the contribution of each mode are very sensitive to estimates of unmeasured parameters [bib_ref] Informing optimal environmental influenza interventions: How the host, agent, and environment alter..., Spicknall [/bib_ref] , such as the viral load in exhaled breath and coughs and the frequency of sneezing by influenza cases [bib_ref] Quantifying the routes of transmission for pandemic influenza, Atkinson [/bib_ref]. However, due to limitations inherent to sampling virus shedding via various routes from infected individuals, and the difficulty of distinguishing routes of transmission in observational studies, the quantitative dynamics and relative contributions of each route remain elusive [bib_ref] Quantifying the routes of transmission for pandemic influenza, Atkinson [/bib_ref]. Recent reports have shown that infectious influenza virus can be recovered from exhaled aerosols [bib_ref] Viable influenza A virus in airborne particles expelled during coughs versus exhalations, Lindsley [/bib_ref] [bib_ref] Influenza virus aerosols in human exhaled breath: Particle size, culturability, and effect..., Milton [/bib_ref] [bib_ref] Measurements of airborne influenza virus in aerosol particles from human coughs, Lindsley [/bib_ref]. These studies, based on small numbers of cases or artificial breathing maneuvers, do not provide sufficient data to quantify the extent of aerosol shedding during natural breathing, nor do they identify the contributions of spontaneous coughs and sneezes commonly thought to be the most important mechanism for viral shedding, or identify other factors that may impact viral aerosol shedding. We address these key knowledge gaps by characterizing influenza virus in exhaled breath from community-acquired influenza cases during natural breathing, prompted speech, coughing, and sneezing, and assess the infectivity of naturally occurring influenza aerosols.
# Results
We screened 355 volunteers with acute respiratory illness; the 178 volunteers who met enrollment criteria provided 278 visits for sample collection. We confirmed influenza infection in 156 (88%) of the enrolled participants using qRT-PCR; 152 had at least one positive nasopharyngeal (NP) swab and 4 (3%) were confirmed based on positive aerosol samples alone. NP swab analysis was positive for 8 (33%) of 24 randomly selected volunteers from among the 177 screened who did not meet enrollment criteria; thus, sensitivity and specificity of our enrollment criteria, during the 2012-2013 season, were ∼73% [95% confidence interval (CI) 62-84%] and 84% (95% CI 80-88%), respectively. In the reported analyses, we excluded 8 visits made on the day of symptom onset, 10 made >3 d after onset, 7 with missing data for cough, and 3 with Significance Lack of human data on influenza virus aerosol shedding fuels debate over the importance of airborne transmission. We provide overwhelming evidence that humans generate infectious aerosols and quantitative data to improve mathematical models of transmission and public health interventions. We show that sneezing is rare and not important for-and that coughing is not required for-influenza virus aerosolization. Our findings, that upper and lower airway infection are independent and that fine-particle exhaled aerosols reflect infection in the lung, opened a pathway for a deeper understanding of the human biology of influenza infection and transmission. Our observation of an association between repeated vaccination and increased viral aerosol generation demonstrated the power of our method, but needs confirmation.
incomplete qRT-PCR data [fig_ref] Figure 1: Histograms of symptom scores [/fig_ref] and [fig_ref] Table 1: Characteristics of study population [/fig_ref]. The resulting dataset for confirmed cases with complete data on RNA copies, cough, and symptoms included 218 visits by 142 cases: 89 influenza A (83 H3, 3 pdmH1, 3 unsubtypable), 50 influenza B, and 3 dual influenza infection cases.
Our study population [fig_ref] Table 1: Characteristics of study population [/fig_ref] consisted mostly of young adults (19-21 y) with a high asthma prevalence (21%), normal body mass index (BMI, median = 22.7; 7% underweight, 20% overweight, and 8% obese) [fig_ref] Table 2: Viral shedding [/fig_ref] , and a low self-reported influenza vaccination rate (22%). We observed at least one cough during 195 (89%) and one or more sneezes during 11 (5%) of the 218 visits. Cough frequency varied considerably, from 5 per 30 min at the 25th percentile to 39 per 30 min at the 75th. Most volunteers rated their upper respiratory symptoms as mild to moderate, systemic symptoms as moderate to severe, and lower respiratory symptoms as mild [fig_ref] Figure 1: Histograms of symptom scores [/fig_ref].
Infectious virus was recovered from 52 (39%) fine-aerosol samples and 150 (89%) NP swabs [fig_ref] Table 2: Viral shedding [/fig_ref]. Quantitative cultures were positive for 30% of the fine-aerosol samples, with a geometric mean (GM) for positive samples of 37 fluorescent focus units (FFU) per 30-min sample [fig_ref] Figure 2: Viral shedding [/fig_ref] and for 62% of NP swabs with GM for positive samples of 2,500. Using Tobit analysis to adjust the estimate of the GM for the presence of samples below the limit of detection, we obtained a GM 1.6 (95% CI 0.7-3.5) for fine aerosols and a GM 60.6 (95% CI 22.7-1.6 × 10 2 ) for NP swabs.
Influenza virus RNA was detected in 76% of the fine-aerosol samples, 40% of the coarse-aerosol samples, and 97% of the NP swabs of enrolled volunteers. For the positive samples, the GM viral RNA content of fine-aerosol samples was 3.8 × 10 4 , for coarse aerosols was 1.2 × 10 4 , and for NP swabs was 8.2 × 10 8 [fig_ref] Figure 2: Viral shedding [/fig_ref]. The adjusted GMs were 1.2 × 10 4 (95% CI 7.0 × 10 3 to 1.9 × 10 4 ) for fine aerosols and 6.0 × 10 2 (95% CI 3.0 × 10 2 to 1.2 × 10 3 ) for coarse aerosols. Quantitative culture was correlated with RNA copies in both NP swabs ( (B) Lower respiratory symptoms (chest tightness, shortness of breath, and cough, score range 0-9). (C) Systemic symptoms (malaise, headache, muscle/joint ache, fever/sweats/chills, and swollen lymph nodes, score range 0-15).
1.5 × 10 3 , 95% CI 4.2 × 10 2 to 5.3 × 10 3 ) and infectious virus to 1.4 × 10 2 FFU per 30-min sample. The few sneezes observed were not associated with greater RNA copy numbers in either coarse or fine aerosols .
Results of regression analyses to identify predictors of viral RNA shedding are shown in [fig_ref] Table 3: Predictors of viral RNA shedding [/fig_ref] , controlled for random effects of subject and repeated observations on individuals.
The day after symptom onset (comparing day 1 postonset with days 2 and 3) was associated with a significant decline in viral RNA shed into fine aerosols (P < 0.05 for day 2 and P < 0.01 for day 3 in adjusted models), a borderline significant decline in coarse-aerosol shedding (P < 0.10), and was not associated with a significant change in shedding detected in NP swabs (P > 0.10).
In regression analyses, cough frequency was significantly associated with increased fine-(P < 0.001 to <0.0001) and coarse-(P < 0.01) aerosol shedding, but was not associated with NP shedding. Fine-aerosol shedding was significantly greater for males. Analysis of an interaction of cough with sex indicated that males produced, on average, 3.2 times more virus than did females per cough. However, females also coughed significantly (P = 0.005) more frequently than males: 33 (SD 39) per 30-min observation and 21 (SD 21), respectively .
BMI was positively associated with shedding in fine and coarse aerosols in unadjusted models (P < 0.10). BMI was retained in the best-fitting adjusted models for both fine and coarse aerosols, where it was significantly associated with fine-aerosol shedding (P < 0.05). However, BMI was not associated with shedding detected in NP swabs (P > 0.10). Standard categories of BMI were not as good a fit as the continuous BMI and were not significantly associated with shedding [fig_ref] Table 3: Predictors of viral RNA shedding [/fig_ref] , although a positive trend is evident for overweight and obese individuals in the adjusted model.
Self-reported vaccination for the current season was associated with a trend (P < 0.10) toward higher viral shedding in fineaerosol samples; vaccination with both the current and previous year's seasonal vaccines, however, was significantly associated with greater fine-aerosol shedding in unadjusted and adjusted models (P < 0.01). In adjusted models, we observed 6.3 (95% CI 1.9-21.5) times more aerosol shedding among cases with vaccination in the current and previous season compared with having no vaccination in those two seasons. Vaccination was not associated with coarse-aerosol or NP shedding (P > 0.10). The association of GSD, geometric SD (only positive samples were included in computation of GM and GSD); NA, not assayed; ND, not detected.
vaccination and shedding was significant for influenza A (P = 0.03) but not for influenza B (P = 0.83) infections .
Viral load in NP swabs was not a significant predictor of aerosol shedding (P = 0.16 for fine and P = 0.48 for coarse aerosols). Temperature measured at the time of sampling, asthma history, smoking, and influenza type were not significantly associated with the extent of measured shedding. While self-reported symptoms were not associated with aerosol shedding, they were significantly associated with shedding measured by the NP swab; only upper respiratory symptoms remained significant when adjusted for other symptoms and age. Increasing age was associated with a significant decrease in shedding in the NP swab; however, age was not associated with aerosol shedding.
# Discussion
We recovered infectious influenza virus from 52 samples of fine aerosols collected from exhaled breath and spontaneous coughs produced by 142 cases of symptomatic influenza infection during 218 clinic visits. Finding infectious virus in 39% of fine-aerosol samples collected during 30 min of normal tidal breathing in a large community-based study of confirmed influenza infection clearly establishes that a significant fraction of influenza cases routinely shed infectious virus, not merely detectable RNA, into aerosol particles small enough to remain suspended in air and present a risk for airborne transmission. Because these data were collected without volunteers having to breathe through a mouthpiece or perform forced coughs, they allow us to provide estimates of average shedding rates, variability, and time course of and risk factors for shedding that can be used to provide well-grounded parameter estimates in future models of the risk of airborne influenza transmission from people with symptomatic illness.
The first published estimates of the numbers of influenza virus variants transmitted from donor to recipient host indicated that the bottleneck for transmission between humans is fairly wide and highly variable (mean 192 with 95% confidence 66-392) [bib_ref] Transmission bottleneck size estimation from pathogen deep-sequencing data, with an application to..., Sobel Leonard [/bib_ref] [bib_ref] Quantifying influenza virus diversity and transmission in humans, Poon [/bib_ref]. Our observation that cases shed considerable quantities of virus into aerosols, GM >10 4 RNA copies per 30 min, and up to 10 3 infectious virus particles per 30 min, suggests that large numbers of variants could be transmitted via aerosols, especially via the short-range mode [bib_ref] Short-range airborne transmission of expiratory droplets between two people, Liu [/bib_ref]. However, longer-range aerosol transmission, as might be observed in less-crowded environments than in the initial report from Hong Kong, would be expected to usually result in lower exposures and transmission of fewer variants, consistent with the narrower bottleneck described in ferret models [bib_ref] Contact transmission of influenza virus between ferrets imposes a looser bottleneck than..., Frise [/bib_ref] [bib_ref] Influenza A virus transmission bottlenecks are defined by infection route and recipient..., Varble [/bib_ref].
Sobel Leonard et al. [bib_ref] Transmission bottleneck size estimation from pathogen deep-sequencing data, with an application to..., Sobel Leonard [/bib_ref] suggested that the width of the bottleneck increased with severity of illness, as indicated by a borderline significant positive association between temperature and number of variants transmitted. We did not see a significant association between measured temperature and shedding by any route. In contrast, symptoms were not a significant predictor of bottleneck size, and in our data, symptoms were not significant predictors for shedding into aerosols. Symptoms were, however, significant predictors for nasal shedding as measured in NP swabs. Thus, if aerosols were the more important route of transmission, our observations would be consistent with the currently available bottleneck analysis.
We observed that influenza cases rarely sneezed, despite having just undergone two NP swab collections (a procedure that Effect estimates are shown as the ratio of male to female, day 2 or day 3 to day 1, type A to B, yes to no vaccination, or fold-increase for an IQR change in age, the number of coughs, symptom reports, or BMI, or ratio of male to female coughs over the IQR (95% confidence interval for the effect estimate). All analyses are controlled for random effects of subject and sample within subject and for censoring by limit of detection using Tobit regression. *P < 0.10, **P < 0.05, ***P < 0.01, ****P < 0.001, *****P < 0.0001 from Tobit regression models with random effect of subject and sample within subject. Adjusted models were selected using the Akaike information criterion from initial models, with all unadjusted parameters having P < 0.10, shown in bold. † Vaccination = self-reported influenza vaccination. ‡ Fever = T ≥ 37.8 measured at visit. generally makes one feel an urge to sneeze). Sneezing was not observed in the absence of cough and was not associated with greater aerosol shedding than we observed with cough alone . Thus, sneezing does not appear to make an important contribution to influenza virus shedding in aerosols. Sneezing might make a contribution to surface contamination. Because sneezes generate considerable amounts of large-droplet spray composed of many ballistic droplets not collected by our sampler, we cannot assess that possibility with our data.
Cough was prevalent and was a strong predictor of virus shedding into both coarse and fine aerosols. However, cough was not necessary for infectious aerosol generation in the ≤5-μm (fine) aerosol fraction; we detected culturable virus in fine aerosols during 48% of sampling sessions when no coughs were observed. This suggests that exhaled droplets, generated by mechanisms other than cough, are responsible for a portion of the viral load observed in the fine-aerosol fraction. Several researchers have recently shown that exhaled aerosol particles are frequently generated from normal healthy lungs by small airway closure and reopening [bib_ref] Effect of airway opening on production of exhaled particles, Almstrand [/bib_ref] [bib_ref] The mechanism of breath aerosol formation, Johnson [/bib_ref] [bib_ref] Origin of exhaled breath particles from healthy and human rhinovirus-infected subjects, Fabian [/bib_ref]. It has been hypothesized that during respiratory infections, airway closure and reopening frequency would be increased due to inflammation with a commensurate increase in aerosol generation and contagiousness [bib_ref] Inhaling to mitigate exhaled bioaerosols, Edwards [/bib_ref].
Cough is thought to produce aerosols from large airways by shear forces that produce relatively coarse-aerosol droplets [bib_ref] Characterization of expiration air jets and droplet size distributions immediately at the..., Chao [/bib_ref]. Our finding that only 13% of cases not observed to have coughed during sample collection produced detectable viral RNA in their coarse aerosols is consistent with that hypothesis. The remaining aerosols may have resulted from speaking; each subject was required to recite the alphabet three times. One might expect that viral replication in the large airways combined with coughgenerated coarse-aerosol droplets would produce the majority of viral aerosols. However, we observed a weak correlation of coarse-aerosol RNA copy number with cough frequency and a much stronger association of fine-aerosol copy number with cough frequency, even though cough would be expected to be the primary source of coarse aerosols. These observations suggest that cough is, at least in part, an epiphenomenon, more of a response to irritation associated with high viral loads in distal airways than a direct source of infectious aerosols.
A striking finding was the association of gender with shedding into fine aerosols. This relationship appears to have resulted from a threefold greater impact of coughing on shedding in males. We observed these gender and gender-by-cough interaction effects only for the fine-aerosol fraction. Absence of a gender effect in the coarse-aerosol fraction suggests that this is not an effect of cough on aerosol generation by shear forces in the upper airway. We did not measure lung volumes and therefore cannot control for a lung size effect. An equally plausible explanation may be that women tend to have more sensitive cough reflexes [bib_ref] Sex-related differences in cough reflex sensitivity in patients with chronic cough, Kastelik [/bib_ref]. Thus, women may have tended to cough in response to lower viral loads and coughed more frequently at a given viral load, which could have produced the observed steeper slope of viral load regressed on cough frequency in males compared with females. Consistent with this suggestion, we did observe a significantly greater cough frequency in females (P = 0.005) and a steeper slope of fine-aerosol viral RNA with cough in males .
BMI was a borderline significant predictor of aerosol shedding in most models, was retained as an important predictor of both coarse and fine aerosols in adjusted models, and reached statistical significance for fine aerosols when adjusted for other factors; it was not a significant predictor of nasal shedding. This observation might be consistent with reports of increased inflammation in models of obesity and influenza and severity of influenza-like illness in obese persons [bib_ref] Obesity as a risk factor for severe influenza-like illness, Cocoros [/bib_ref] [bib_ref] Adiposity and influenza-associated respiratory mortality: A cohort study, Zhou [/bib_ref] [bib_ref] Obesity and respiratory hospitalizations during influenza seasons in Ontario, Canada: A cohort..., Kwong [/bib_ref] [bib_ref] Obesity not associated with severity among hospitalized adults with seasonal influenza virus..., Braun [/bib_ref] [bib_ref] California Pandemic (H1N1) Working Group (2009) Factors associated with death or hospitalization..., Louie [/bib_ref] [bib_ref] Obesity-induced chronic inflammation is associated with the reduced efficacy of influenza vaccine, Park [/bib_ref]. Alternatively, increasing BMI is associated with increased frequency of small airways closure, and the resulting increased aerosol generation during airway reopening as described above may explain the stronger association of BMI with fine than coarse aerosols and lack of association with NP swabs [bib_ref] Physiology of obesity and effects on lung function, Salome [/bib_ref].
Our analysis found a clear separation of factors associated with shedding from the nose and those with shedding into aerosols, especially fine-particle aerosols. Upper airway symptoms, as would be expected, were strongly associated with shedding detected in NP swabs, and greatly reduced the size and significance of lower respiratory and systemic symptoms in the fully adjusted model. Age was negatively associated with nasal shedding but not a predictor of aerosol shedding. More surprisingly, no symptoms, including lower respiratory and systemic systems, were strongly associated with shedding into aerosols, in this population with relatively mild lower respiratory symptoms [fig_ref] Figure 1: Histograms of symptom scores [/fig_ref]. Furthermore, nasal shedding was not a significant predictor of aerosol shedding and none of the strong predictors of aerosol shedding were associated with nasal shedding. Thus, we can conclude that the head airways made a negligible contribution to viral aerosol generation and that viral aerosols represent infection in the lung. Moreover, upper and lower airway infection appear to behave as though infection is compartmentalized and independent. In this context, it is notable that Varble et al. [bib_ref] Influenza A virus transmission bottlenecks are defined by infection route and recipient..., Varble [/bib_ref] observed that intrahost viral variants differ in the nasopharynx and lung of ferrets.
We did not observe a significant decline over time of viral load detected in NP swabs. If day 1 after onset of symptoms (used as baseline for these analyses) in our cases was equivalent to a mixture of day 1 and day 2 after experimental influenza virus inoculation in the report by , then our lack of finding a clear drop in nasal shedding over the next 2 d is reasonably consistent with the pattern reported for experimental infection. There is no available data for comparison of aerosol shedding from published experimental infections. That we saw a much clearer pattern of rapid decline over time in aerosol shedding again suggests a separation of infection into upper and lower airway compartments in humans.
The association of current and prior year vaccination with increased shedding of influenza A might lead one to speculate that certain types of prior immunity promote lung inflammation, airway closure, and aerosol generation. This first observation of the phenomenon needs confirmation. If confirmed, this observation, together with recent literature suggesting reduced protection with annual vaccination, would have implications for influenza vaccination recommendations and policies.
# Materials and methods
Study Population and Sample Collection Procedures. We recruited volunteers with acute respiratory illness on the University of Maryland-College Park campus and surrounding community from December 2012 through March 2013. The University of Maryland Institutional Review Board approved the study, and we obtained a signed consent (or assent and parental verbal assent) from volunteers who reported fever with a cough or sore throat .
During the initial visit, we administered a brief screening questionnaire, measured oral temperature, height, weight, and collected two NP swabs (Copan) for each volunteer screened. One swab was used to perform QuickVue A/B rapid tests for influenza (except when results of a rapid test performed by medical provider were available). The second NP swab was used for viral culture and PCR for those meeting enrollment criteria and for PCR in a random sample of 24 of those not enrolled.
Participants were asked about sex, age, antipyretic use, vaccination status, use of steroid medications, medical and smoking history, to rate current symptoms on a four-level scale (none = 0, mild = 1, moderate = 2, severe = 3), and to rate the worst symptoms during the illness thus far. We defined symptoms as upper respiratory (runny nose, stuffy nose, sneezing, sore throat, and earache), lower respiratory (chest tightness, shortness of breath, and cough), and systemic (malaise, headache, muscle/joint ache, fever/sweats/ chills, and swollen lymph nodes).
Volunteers were enrolled in exhaled breath collection if they met the following criteria: (i) positive QuickVue rapid test, or oral temperature >37.8°C plus cough or sore throat, and (ii) presented within the first 3 d of symptom onset. Exhaled breath samples were collected using the Gesundheit-II human source bioaerosol sampler, as previously described [bib_ref] Influenza virus aerosols in human exhaled breath: Particle size, culturability, and effect..., Milton [/bib_ref] [bib_ref] Development and performance evaluation of an exhaledbreath bioaerosol collector for influenza virus, Mcdevitt [/bib_ref]. We collected exhaled breath for 30 min while the participant was seated with their face inside of the large open end of a cone-shaped inlet for the G-II. The inlet cone draws in 130 L of air per minute and allowed participants to breathe, talk, cough, and sneeze naturally throughout sample collection while maintaining >90% collection efficiency for exhaled and coughed droplets ≤100 μm. Subjects were asked to breathe normally and to recite the alphabet once at 5, 15, and 25 min. We collected "coarse" (>5 μm) aerosol droplets by impaction on a Teflon surface and "fine" droplets (≤5 μm and >0.05 μm) by condensation growth and impaction on a steel surface constantly rinsed into a buffer containing (PBS with 0.1% BSA) liquid reservoir. Audible spontaneous coughs and sneezes during breath collection were counted by direct observation in realtime (n = 59) or by playback of digital recordings (n = 159).
Participants enrolled before the third day after symptom onset were asked to come in for up to two consecutive daily follow-up visits with repeat questionnaire, NP swab, and exhaled breath collections. Final analyses included only visits for enrolled cases occurring on days 1-3 after symptom onset with complete data on cough and sneeze, symptoms, PCR results for swab and aerosol samples.
Laboratory Methods. Detailed methods are described in the SI Materials and Methods. Briefly, NP swabs were eluted in 1 mL of PBS with 0.1% BSA (PBS/ 0.1% BSA) or universal transport medium (Copan), and Teflon impactors were scrubbed with a nylon swab saturated with PBS/0.1% BSA. The swab was eluted in 1 mL PBS/0.1% BSA. Fine-aerosol samples were concentrated to 1 mL using centrifugal ultrafiltration.
RNA was extracted from NP swab, fine-and course-aerosol samples, and whole-virion standards using an automated Qiagen system and viral RNA was quantified by one-step real-time RT-PCR using Taqman primer probe sets designed by the US Centers for Disease Control and Prevention and made available through our cooperative agreement. Standard curves were calibrated for virus copy number using plasmids containing a cDNA copy of the qRT-PCR target amplicon. Experimentally determined limits of detection and quantification for each of the qRT-PCR reactions are shown in .
Virus culture on Madin-Darby canine kidney (MDCK) cells was used to detect infectious virus in NP swab and fine-aerosol samples. Coarse-aerosol samples were not cultured for infectious virus because impaction on a dry Teflon surface was expected to reduce infectivity of those samples. Infectious influenza virus was quantified using an immunofluorescence assay for influenza nucleoprotein, and positive cells were counted as FFU by fluorescence microscopy. Details of laboratory methods can be found in SI Materials and Methods.
Statistical Analysis. We entered and cleaned data using locally hosted REDCap data-capture tools [bib_ref] Research electronic data capture (REDCap)-A metadatadriven methodology and workflow process for providing..., Harris [/bib_ref] and performed data management and analyses in R (v3.2.3 R Development Core Team, Vienna, Austria) and SAS (v9.4, Cary, NC), and produced graphics with Prism Software (PRISM software v7.0; Graph-Pad). We used the delta method to estimate confidence limits for sensitivity and specificity. We used Spearman correlation, generalized linear models (SAS Proc GENMOD), and Tobit regression [bib_ref] Longitudinal tobit regression: A new approach to analyze outcome variables with floor..., Twisk [/bib_ref] with nested random effects of sample within subject in (SAS Proc NLMIXED) to analyze infectious virus counts, RNA copy numbers, and compute GM virus concentrations. Tobit regression accounted for uncertainty and censoring of the observations by the limit of quantification. We included all independent variables with unadjusted P < 0.10 in initial adjusted models and selected final models using the Akaike information criterion while retaining adjustment for age and sex. Regression model results are presented as the ratio of shedding at the 75th percentile to shedding at the 25th percentile of the distribution of the independent variable, so that clinical and epidemiological meaning of the relationship can be more easily interpreted.
[fig] Figure 1: Histograms of symptom scores. (A) Upper respiratory symptoms (runny nose, stuffy nose, sneezing, sore throat, and earache, score range 0-15). [/fig]
[fig] Figure 2: Viral shedding: (A) infectious influenza virus (fluorescent focus counts) in NP swabs and fine aerosols and (B) RNA copies in NP swabs, coarse, and fine aerosols. (C and D) Scatter plots and Spearman correlation coefficients of infectious virus plotted against RNA copies for (C) NP swabs and for (D) fineaerosol samples. (E) The effect of day after symptom onset on RNA copies observed in NP swabs, coarse, and fine aerosols plotted as GM adjusted for missing data using Tobit analysis with error bars denoting 95% CIs. (F-H) The effect of cough frequency on RNA copies observed in (F) NP swabs, (G) coarse aerosols, and (H) in fine aerosols. Coarse: aerosol droplets >5 μm; Fine: aerosol droplets ≤5 μm in aerodynamic diameter. [/fig]
[table] Table 1: Characteristics of study population [/table]
[table] Table 2: Viral shedding [/table]
[table] Table 3: Predictors of viral RNA shedding [/table]
|
The Patent Ductus Arteriosus in Extremely Preterm Neonates Is More than a Hemodynamic Challenge: New Molecular Insights
# Introduction
Preterm birth is the leading cause of neonatal death and there is a considerable risk of lifelong impairment [bib_ref] Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions,..., Murray [/bib_ref]. A patent ductus arteriosus (PDA) is a frequent complication to preterm birth. Much emphasis has been placed on the hemodynamic effects of the PDA. However, preterm birth is associated with complex, mutually overlapping abnormalities resulting from systemic immaturity. Hence, the risk attributable to the PDA is most likely not only dependent on gestational age (GA) and the hemodynamics effects of the PDA, but also on various concomitant perinatal factors, including fetal growth restriction, hypoxia, infection, and possibly inflammation [bib_ref] Chronic inflammation and impaired development of the preterm brain, Bennet [/bib_ref] [bib_ref] Blood Cytokine Profiles Associated with Distinct Patterns of Bronchopulmonary Dysplasia among Extremely..., D'angio [/bib_ref] [bib_ref] Priming Immunity at the beginning of life (PRIMAL) Consortium. Preterm birth and..., Humberg [/bib_ref] [bib_ref] A perfect storm: Fetal inflammation and the developing immune system, Sabic [/bib_ref] [bib_ref] Early-onset sepsis in very low birth weight neonates: A report from the..., Stoll [/bib_ref].
Mounting evidence points to a negative impact of early, sustained or intermittent inflammatory exposure on the outcome of preterm birth [bib_ref] Priming Immunity at the beginning of life (PRIMAL) Consortium. Preterm birth and..., Humberg [/bib_ref]. A major concern is the possible association between inflammation and immune dysfunction and abnormal stress responses, not only in the immediate postnatal period, but also possibly throughout life. Inflammation is thought to have a direct effect on the function and further development of organs with long-term compromised function of the central nervous system (CNS), as well as respiratory and cardiovascular systems. Despite abundant evidence of these associations, there are significant gaps in our biological understanding of these associations [bib_ref] Priming Immunity at the beginning of life (PRIMAL) Consortium. Preterm birth and..., Humberg [/bib_ref] [bib_ref] Perinatal infection, inflammation, preterm birth, and brain injury: A review with proposals..., Reiss [/bib_ref].
Complications to prematurity are complex diseases and their pathogenesis depends on the interaction of a susceptible host with a multitude of environmental and possibly genetic risk factors mutually percolating in a systemic manner, which can be reflected in the plasma protein pattern. Proteins are effectors of biological functions and their levels are not only dependent on corresponding mRNA levels, but also on post translational control and regulation [bib_ref] Proteomics: Technologies and Their Applications, Aslam [/bib_ref]. Therefore, the proteome seems highly relevant to characterize a biological system. Proteomics hold great potential for new insights into pathogenesis related to the continuing presence of a ductus arteriosus and possible novel biomarkers for diagnostic, prognostic, and monitoring purposes in preterm neonates [bib_ref] Recent Developments in Clinical Plasma Proteomics-Applied to Cardiovascular Research, Palstrom [/bib_ref].
We aimed to investigate proteomics analysis and various cytokines and chemokines in plasma samples from 53 neonates born at a gestational age below 28 weeks to achieve a new biological understanding of the PDA.
# Methods
## Study cohort
From 1 June 2010 to 28 February 2012, all neonates born with a GA below 32 completed weeks that were admitted to our level three neonatal intensive care unit (NICU), were eligible for inclusion into the PDA study (n = 184). Neonates with chromosomal abnormalities or congenital heart malformations other than atrial septum defects were excluded. In the present study, we included neonates born with a GA below 28 completed weeks, collected as part of the original and previously presented cohort [bib_ref] Morbidity and mortality in preterm neonates with patent ductus arteriosus on day..., Sellmer [/bib_ref]. Parents provided their informed consent for inclusion of their child in the study. The study was conducted in accordance with the Declaration of Helsinki, and approved by the Central Denmark Region Committees on Health Research Ethics (journal number in January 2010, the Danish Data Protection Agency, and the National Board of Health.
## Echocardiography and clinical information
According to the local guidelines, all neonates born with a GA below 32 weeks had an echocardiography on day 3 after birth to evaluate the presence of a PDA and the structure of the heart.
The echocardiography was performed using two-dimensional, B-mode, and color Doppler in standard neonatal windows [bib_ref] Guidelines and standards for performance of a pediatric echocardiogram: A report from..., Lai [/bib_ref]. A PDA was defined as present if flow could be visualized by color Doppler. The PDA diameter was measured in B-mode at the most narrow point. In addition, the PDA was defined as large if the diameter was above 1.5 mm and small if the diameter was 1.5 mm or below [bib_ref] Morbidity and mortality in preterm neonates with patent ductus arteriosus on day..., Sellmer [/bib_ref]. The ratio of the left atrium to the aorta (LA:Ao ratio) was determined in the parasternal long axis view by M-mode using the leading edge method. A large LA:Ao ratio was defined as a ratio above 1.5. Descending aorta diastolic flow (DADF) was evaluated by Doppler flow patterns in the descending aorta obtained from the suprasternal view. None of the neonates in the present study received Ibuprofen or other medicines to close the PDA and none had surgical closure before sample collection.
The following clinical information was obtained from the patients' medical records: Maternal preeclampsia, antenatal steroid administration, mode of delivery, multiple birth, birth weight, and GA at birth (based on early ultrasound scanning), Apgar score at 1 and 5 min, surfactant administration, intraventricular hemorrhage (IVH) (graded according to Papile et al. [bib_ref] Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants..., Papile [/bib_ref] , packed red blood cell transfusion within the first 3 days of life, inotropes within the first 3 days of life, early onset sepsis (EOS) defined as 7 days of antibiotics initiated before day 3, and the use of mechanical ventilation on day 3.
## Plasma samples
Blood samples for this study were obtained along with routine samples on day 3 after birth. The blood was collected by EDTA tubes and centrifuged. The plasma fraction was collected and stored at −80 - C until analysis [bib_ref] Cardiovascular biomarkers in the evaluation of patent ductus arteriosus in very preterm..., Sellmer [/bib_ref]. N-terminal pro-natriuretic B-type peptide (NT-proBNP) was measured as a routine analysis at the Department of Clinical Biochemistry, Aarhus University Hospital, Denmark [bib_ref] Cardiovascular biomarkers in the evaluation of patent ductus arteriosus in very preterm..., Sellmer [/bib_ref].
## Preparation of samples for proteomics
To avoid masking of potential biomarkers by the high abundant proteins, the samples were depleted for albumin and IgG. Five µL of plasma was mixed with 10 µL of Complete Mini EDTA-free protease inhibitor. The 53 samples were depleted using ProteoPrep ® Immunoaffinity Albumin and IgG Depletion columns. The protein concentration was determined using the Bradford assay (Bio-Rad, Hercules, CA, USA) using technical duplicates, according to the manufacturer's instructions. Amounts corresponding to 15 µg protein were transferred to a new Eppendorf tube and proteins were precipitated overnight with six times sample volume of acetone at −20 - C, then centrifuged at 2600× g, 4 - C for 10 min, and finally air dried. Before mass spectrometry (MS) analysis, the samples were solubilized in 30 µL ammonium bicarbonate with 3% sodium deoxycholate, then reduced with 1.4-dithiothreitol at a final concentration of 5 mM, alkylated in the dark for 30 min with iodoacetamide at a final concentration of 10 mM, and digested with 0.3 µg trypsin. The peptides were subsequently purified with PepClean™ C18 Spin Columns (Thermo Scientific, Waltham, WA, USA), according to the manufacturer's instructions and then evaporated on a miVac Duo Concentrator (Genevac, Ipswich, United Kingdom). Samples were stored at −20 - C.
# Lc-ms/ms analysis
The peptides were re-suspended in 2% ACN, 0.1% formic acid, and were analyzed by nanoLC-MS/MS, essentially as described previously [bib_ref] Proteomics of the Rat Myocardium during Development of Type 2 Diabetes Mellitus..., Edhager [/bib_ref]. The samples were analyzed by nano-liquid chromatography (Easy-nLC 1200, Thermo Scientific)-tandem MS (Q-Exactive HF-X Hybrid Quadrupole Orbitrap, Thermo Scientific). Peptides were trapped by a precolumn (Acclaim PepMap 100 C18, pore size: 100 Å, particle diameter: 3 µm, inner diameter: 75 µm, length: 2 cm, Thermo Scientific) and separated further with a reverse phase analytical column (PepMap RSLC C18, pore size: 100 Å, particle diameter: 2 µm, inner diameter: 75 µm, length: 25 cm, Thermo Scientific) using a 100 min gradient from 5-90% ACN and 0.1 formic acid at a 270 nL/min flowrate. The mass spectrometer was operated in positive mode and higher collision dissociation (HCD) at normalized collision energy of 29 was used for peptide fragmentation. The full scan/MS1 resolution was 60,000, AGC target was 3 × 10 6 , maximum injection time was 80 ms, and scan range was from 340 to 1700 m/z. The fragmentation scan/MS2 resolution was 15,000 and the automatic gain control (AGC) target was 1 × 10 5 . The MS was operated in data dependent mode and up to 10 of the most intense peaks were fragmented. Dynamic exclusion was set to 30 s and both unassigned and single charge ions were excluded.
# Proteomic data analysis
Proteins were identified and quantified using MaxQuant (version 1.5.3.30, https: //www.maxquant.org/, accessed on 5 July 2022) with the building Andromeda algorithm against the human sequence database (Homo Sapiens proteome with 20,129 reviewed sequences from Uniprot.org, accessed on 5 July 2016). Settings included the enzyme trypsin with a maximum of two missed cleavage sites; precursor mass tolerance: 10 ppm; fragment mass tolerance: 0.02 Da; dynamic modification: Oxidation; static modification: carbamidomethyl; and FDR was 0.01 at protein and peptide level. To maximize identifications in MaxQuant, files containing MS spectra from 10 fractions of a pooled study of plasma samples, were uploaded along with MS data from the samples and the "match between runs" were applied.
## Multiplexed luminex analyses
A total of 19 proteins with relation to inflammation were simultaneously quantitated by the bead-, antibody-, and fluorescence-based method Luminex. The 19-plex arrays, including VEGF-A, GM-CSF, TNF-α, IL-1ra, IL-1 alpha, IL-1 beta, IL-4, IL-6, IL-8 (CXCL8), IL-10, IL-12p70, Eotaxin (CCL11), SDF-1 alfa, CD62E (E-selectin), MCP-1 (CCL2), MIP-1, lfa (CCL3), MIP-1 beta (CCL4), RANTES (CCL5), and ICAM-1 were analyzed on the Luminex XMAP platform, Magpix (Millipore Corp, MA, USA), according to the manufacturer's instructions.
## Statistical and bioinformatics analyses
The data analysis was performed in R v4.1.2 (R Core Team), using the LFQ-columns from the MaxQuant generated proteingroups.txt file, which contains protein identifications filtered to the false discovery rate (FDR) < 1%.
Additional filtering was performed to ensure high accuracy qualitative data, by removing (i) proteins tagged as reverse, contaminants, or only identified by modified peptides; (ii) proteins where the quantitation was based on less than one unique peptide to a given protein group; and (iii) proteins that were quantifiable in less than 70% of samples in both of the compared groups, i.e., large PDA compared to no PDA or small PDA compared to no PDA, thereby ensuring that at least 70% of samples in both groups were compared. The remaining missing values were not included in the differential analysis. To include condition-unique proteins, which would have been filtered using the applied filtering strategy, we performed a Fisher's Exact analysis for all proteins, but none passed the statistical filtering criteria (p-value < 0.05). Solely for the purpose of conducting principal component analysis (PCA), missing data were imputed by random draws from a Gaussian distribution centered below the minimal value observed in that sample (q = 0.01, tune.sigma = 0.3) to simulate signals from low-abundant proteins [bib_ref] Super-SILAC allows classification of diffuse large B-cell lymphoma subtypes by their protein..., Deeb [/bib_ref].
In the days after term birth, the plasma concentration of several proteins is known to change, since especially the innate immune system, it rapidly develops in response to new environmental exposures. Therefore, we used linear mixed-effects regression models to identify PDA-associate proteins. A full linear mixed-effects regression model of LFQ values was fitted with fixed effects of PDA status and a random GA week effect, using the lmer function from the lme4 R package [bib_ref] Fitting Linear Mixed-Effects Models using lme4, Bates [/bib_ref]. The p-values were calculated using the ANOVA function against a null model, omitting the PDA status, but including GA to take the ontogeny into account. The p-values were corrected for multiple hypothesis testing using the Benjamini-Hochberg approach [bib_ref] Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple..., Yoav [/bib_ref]. Proteins were considered statistically significantly different at adjusted p-values < 0.05 and +/− 0.3 log2 fold-change. To expand the analysis, we included an analysis of all proteins at unadjusted p-value < 0.05. The analysis was performed in R v4.1.2using Rstudio (2021.09.0). The packages dplyrand mixOmics [bib_ref] mixOmics: An R package for 'omics feature selection and multiple data integration, Rohart [/bib_ref] were used for data formatting, and plots were performed using ggplot2, ggpubr, and cowplot. Significantly differentiating proteins were submitted to String-DB to infer known protein-protein interactions [bib_ref] Protein-protein interaction networks, integrated over the tree of life, Szklarczyk [/bib_ref]. Receiver operating characteristic (ROC) curves were generated using the pROC package, and p-values were calculated using the test.roc function [bib_ref] pROC: An open-source package for R and S+ to analyze and compare..., Robin [/bib_ref].
## Periostin enzyme-linked immunosorbent assay
To validate the MS analysis, an enzyme-linked immunosorbent assay (ELISA) targeting periostin was performed using the Human Periostin ELISA Kit (Thermo Fisher). Periostin was chosen as a protein target for the following reasons: (i) It is currently suggested as a biomarker in preterm neonates with BPD [bib_ref] Can serum periostin predict bronchopulmonary dysplasia in premature infants?, Go [/bib_ref] [bib_ref] Early Elevation of Plasma Periostin Is Associated with Chronic Ventilator-Dependent Bronchopulmonary Dysplasia, Ahlfeld [/bib_ref] and (ii) it was one of the proteins found at higher levels in neonates with PDA by LC-MS proteomics. The test was carried out following the manufacturer's instructions with a minor modification. The instructions suggest that a 1:2 dilution of plasma samples deviations was performed as a preliminary test of the kit and another study of plasma periostin in connection with BPD [bib_ref] Can serum periostin predict bronchopulmonary dysplasia in premature infants?, Go [/bib_ref] indicated that a dilution of 1:33 would be more suitable for the absorbance not to exceed the measurable absorbance of the plate reader. Therefore, samples were diluted 1:33 with the sample diluent provided in the kit. The absorbance was measured at 450 nm using a synergy H1 microplate reader (BioTek, Winooski, VT, USA). A standard curve was based on the periostin standard solutions and concentrations in the samples were determined.
# Results
## Study cohort
The study cohort comprised 53 newborns, 33 (62%) were determined to have a PDA [fig_ref] Table 1: Study cohort [/fig_ref] on day 3 of life. We have previously demonstrated that a PDA with a diameter above 1.5 mm at this day in extremely preterm neonates is associated with adverse outcomes [bib_ref] Morbidity and mortality in preterm neonates with patent ductus arteriosus on day..., Sellmer [/bib_ref]. Accordingly, we divided our cohort into the following three groups: No PDA (n = 20), small (n = 13), and large PDA (n = 20). Neonates with a large PDA had more unfavorable baseline characteristics compared to neonates with no PDA; however, they were not statistically significant.
## Plasma proteomics
LC-MS-based proteomics enabled us to monitor the relative abundance of 219 proteins which passed our stringent filtering, ensuring high-confidence identifications and quantitation. Furthermore, 18 cytokines and chemokines monitored by immunoassays passed the valid value-filtering criteria. To investigate the global proteomics data, we performed an unsupervised principal component analysis (PCA). No clear outlying samples could be identified on the PCA scores plot of the LC-MS proteomics data [fig_ref] Figure 1: Unsupervised principal component analysis [/fig_ref] nor immunoassay data [fig_ref] Figure 1: Unsupervised principal component analysis [/fig_ref]. Moreover, no clear grouping of the samples could be identified based on the available clinical parameters, including GA or PDA. However, the overlap of the no PDA and small PDA samples was generally larger than the large PDA samples, indicating a higher degree of similarity.
## Identifying proteins differentiating pda from no pda samples
To further expand our understanding of the biological mechanisms of PDA, we performed a differential analysis to identify proteins with a PDA-specific concentration profile.
Comparing the small PDA to no PDA, no proteins passed the statistical filtering criteria described in the Methods section, including the adjusted p-value < 0.05 (Supplementary [fig_ref] Figure 1: Unsupervised principal component analysis [/fig_ref] (full list of proteins and cytokines in Supplementary Tables S1 and S2, respectively).
Comparing the large PDA samples to no PDA samples, two proteins and three cytokines passed our statistical correction for multiple hypothesis testing, including the adjusted p-value < 0.05 [fig_ref] Figure 2: Large PDA and the plasma proteome [/fig_ref] ,B for LC-MS proteomics and immunoassay, respectively) (full list of proteins and cytokines in [fig_ref] Table 3: Large PDA significant proteins from immunoassay [/fig_ref].
To investigate if the lack of significant proteins comparing small PDA to no PDA reflected the smaller sample size compared to large PDA vs no PDA, we compared the protein fold-changes between small PDA vs no PDA to large PDA vs no PDA. A statistically significant correlation was found for both LC-MS (R spearman = 0.57, p-value = 2.2 × 10 −16 ) and immunoassay proteins (R spearman = 0.66, p-value = 3.7 × 10 −3 ). However, the protein fold-changes were generally smaller for the small PDA as demonstrated by the < 1x slope of a linear regression for LC-MS (y = 0.47x − 0.02, Supplementary [fig_ref] Figure 2: Large PDA and the plasma proteome [/fig_ref] and immunoassay proteins (y = 0.41x + 0.14, Supplementary [fig_ref] Figure 2: Large PDA and the plasma proteome [/fig_ref] suggesting a dose-response-like relationship between PDA size and protein levels.
Therefore, we focused our study on the comparison between the large PDA and no PDA. For hypotheses generation, we included the additional 19 proteins and 5 cytokines with different abundances between large PDA and no PDA, which passed our statistical filtering criteria, but with the unadjusted p-value < 0.05 in the analysis [fig_ref] Table 2: Large PDA significant proteins from LC-MS [/fig_ref] , respectively). Of the LC-MS proteins, 15 were increased and 6 decreased in large PDA compared to no PDA, whereas all investigated cytokines were increased. The proteins included, but were not limited to, angiotensinogen (AGT), osteopontin (SPP1), periostin (POSTN), and the cytokines, included IL-6, IL-8, IL-10, and IL-1RA. The decreasing proteins were mainly related to the complement system, e.g., C8 (alpha, beta, and gamma chain) and carboxypeptidase N catalytic chain CPN1 and CPN2. To investigate whether the significant proteins were associated with PDA rather than ontogeny, we repeated all of the calculations using only samples from the GA 27 group (Supplementary [fig_ref] Figure 3: LC-MS proteomics and immunoassay data validation [/fig_ref]. A positive correlation was found between the large PDA compared to no PDA protein fold-changes (R spearman = 0.75, p-value = 1.42 × 10 −4 ). Additionally, we plotted the protein profiles for all significant large PDA compared to no PDA proteins [fig_ref] Figure 1: Unsupervised principal component analysis [/fig_ref].
A joint protein-protein interaction analysis of all proteins and cytokines, revealed that known interactions between the proteins and cytokines/chemokines increased and decreased during PDA, respectively, with little overlap [fig_ref] Figure 2: Large PDA and the plasma proteome [/fig_ref]. The finding indicates the presence of PDA interactions between the proteins and cytokines, distinct functional changes in PDA, and the increased involvement of the immune system in neonates with PDA.
## Immunoassay validation of periostin (postn) levels
To ensure the consistency in protein measurements across methods, we compared the relative levels of plasma periostin (POSTN) as determined by the fundamentally different protein measurement techniques LC-MS/MS and immunoassay. A strong and significant samplesample correlation was determined [fig_ref] Figure 3: LC-MS proteomics and immunoassay data validation [/fig_ref] (R spearman = 0.73, p-value = 9.4 × 10 −10 ).
## Pda and increased levels of nt-probnp levels
We re-analyzed the plasma levels of the known PDA-marker NT-proBNP in part of the original cohort [bib_ref] Cardiovascular biomarkers in the evaluation of patent ductus arteriosus in very preterm..., Sellmer [/bib_ref] , which was included in the present study. The marker was found to be significantly higher in neonates with PDA [fig_ref] Figure 3: LC-MS proteomics and immunoassay data validation [/fig_ref] (mean PDA level = 389%, p-value = 1.67 × 10 −8 , q-value = 6.68 × 10 −7 ), as expected for a PDA cohort. Moreover, we generated receiver operating characteristic (ROC) curves [fig_ref] Figure 3: LC-MS proteomics and immunoassay data validation [/fig_ref] , which demonstrated that NT-proBNP can separate large PDA from no PDA with an accuracy of 93.4%. In comparison, the AUC was 60.1% for the ROC compared with the small PDA compared to no PDA.
# Discussion
## Short presentation of main results
In a cohort of 53 extremely preterm neonates, we found that proteomics analysis and multiplex ELISA revealed evidence of protein abundance differences in plasma related to multiple biological immune system associated processes, including coagulation, complement activation, inflammation, and immunomodulation. The two proteins that mainly differed between extremely preterm neonates with large PDA compared to no PDA were angiotensinogen (AGT) with a 1.5-fold increase and periostin with a 1.7-fold increase in the LC-MS analysis. Moreover, IL-1RA, IL-6, IL-8, and IL-10 were found at higher levels in neonates with large PDA in the ELISA analysis. Contrary lower levels of complement factors C8 and carboxypeptidases were found in neonates with large PDA compared to no PDA.
The PDA is a dynamic structure and the small PDA may even close from time to time. We have previously demonstrated that a large PDA in extremely preterm neonates is associated with adverse outcomes [bib_ref] Morbidity and mortality in preterm neonates with patent ductus arteriosus on day..., Sellmer [/bib_ref]. With the present study, we demonstrate that although the general direction of the proteome fold-changes was identical, we found overall more pronounced protein fold differences in neonates with large PDA vs. no PDA, compared to small PDA vs. no PDA.
## Activation of the raas system in neonates with pda
Angiotensinogen (AGT), the precursor of all angiotensin peptides, was found in higher levels (1.5 fold-change) in neonates with large PDA compared to neonates with no PDA. AGT is a member of the serpin superfamily, as well as alpha1 antitrypsin, which is also found at higher levels in neonates with large PDA.
In circulation, AGT is cleaved to angiotensin-I by renin, which is secreted from the juxtaglomerular apparatus in the kidneys in response to a decreased renal perfusion. However, also immune mediators, such as IL-6, are involved [bib_ref] Structure and functions of angiotensinogen, Lu [/bib_ref] [bib_ref] STAT3 NH2-terminal acetylation is activated by the hepatic acute-phase response and required..., Ray [/bib_ref]. We found IL-6 at higher levels in neonates with large PDA compared to neonates with no PDA. Angiotensin I is subsequently converted to angiotensin II (Ang II) by the angiotensin-converting enzyme (ACE).
Ang II is recognized not only as a physiological mediator restoring circulatory integrity, but also as a growth factor that regulates cell growth and fibrosis, organ differentiation, and a key element in the inflammatory process [bib_ref] Inflammation and angiotensin II, Suzuki [/bib_ref] [bib_ref] Angiotensin converting enzyme inhibition decreases cell turnover in the neonatal rat heart, Choi [/bib_ref]. Ang II increases the vascular permeability that initiates the inflammatory process [bib_ref] Role of the renin-angiotensin system in vascular inflammation, Marchesi [/bib_ref] and contributes to the recruitment of inflammatory cells [bib_ref] Inflammation and angiotensin II, Suzuki [/bib_ref]. Moreover, there is an increasing evidence that pro-inflammatory factors enhance the expression of RAAS components [bib_ref] Inflammation as a Regulator of the Renin-Angiotensin System and Blood Pressure, Satou [/bib_ref].
## Ang ii and il-10 stimulated increase in spp1 in neonates with pda
SPP1 was the protein with the second largest fold-change in neonates with large PDA compared to neonates with no PDA in the LC-MS data. Previously, an association between cord blood SPP1 and PDA has been demonstrated [bib_ref] Cord blood levels of osteopontin as a phenotype marker of gestational age..., Joung [/bib_ref] , but this is the first study to demonstrate higher levels in plasma samples from neonates with large PDA.
Secreted phosphoprotein 1 (SPP1), also known as osteopontin (OPN), is a matricellular protein that mediates diverse biological functions. SPP1 functions as a pro-inflammatory cytokine and promotes cell-mediated immune responses [bib_ref] A multifunctional protein at the crossroads of inflammation, atherosclerosis, and vascular calcification, Cho [/bib_ref]. In addition, it has protective functions, such as biomineralization and wound healing. During pathologic processes, SPP1 is produced by various cells [bib_ref] Macrophages express osteopontin during repair of myocardial necrosis, Murry [/bib_ref] [bib_ref] Expression and distribution of osteopontin in human tissues: Widespread association with luminal..., Brown [/bib_ref].
Rat studies have demonstrated that Ang II can induce SPP1 expression. And further that the inhibition of the Ang II type 1 receptor (AT1) attenuate the expression of SPP1 [bib_ref] Angiotensinogen and AT(1) antisense inhibition of osteopontin translation in rat proximal tubular..., Ricardo [/bib_ref] [bib_ref] Expression, roles, receptors, and regulation of osteopontin in the kidney, Xie [/bib_ref] [bib_ref] Osteopontin expression in progressive renal injury in remnant kidney: Role of angiotensin..., Yu [/bib_ref]. Moreover, SPP1 expression is stimulated by inflammatory cytokines, including IL-10 [bib_ref] IL (Interleukin)-10-STAT3-Galectin-3 Axis Is Essential for Osteopontin-Producing Reparative Macrophage Polarization After Myocardial..., Shirakawa [/bib_ref]. We found higher levels of IL-10 in neonates with large PDA compared to neonates with no PDA. It is thought that the acute increase in SPP1 has a protective role in cardiovascular disease in adults, whereas a more chronic increase predicts poor prognosis [bib_ref] Osteopontin in Cardiovascular Diseases, Shirakawa [/bib_ref]. The function and regulation of SPP1 in preterm neonates remain elusive [bib_ref] Osteopontin in Cardiovascular Diseases, Shirakawa [/bib_ref] [bib_ref] Urine biomarkers predict acute kidney injury in newborns, Askenazi [/bib_ref].
## Which cytokines are associated with large pda?
Cytokines are pleiotropic endogenous inflammatory and immunomodulating mediators that exhibit regulatory effects on various target cells [bib_ref] Cytokines in Inflammatory Disease, Kany [/bib_ref]. These cell-derived polypeptides are involved in both acute and chronic inflammatory processes by acting locally or systemically. The IL-1RA was the interleukin with the highest difference in concentration between neonates with large PDA and no PDA amongst the measured cytokines. IL-1RA is a naturally occurring cytokine that inhibits the effects of IL-1α and IL-1β through com-petitively ligand-specific binding to IL-1 receptors without exhibiting detectable agonist activity [bib_ref] The balance between IL-1 and IL-1Ra in disease, Arend [/bib_ref]. We found higher IL-1β levels in neonates with a large PDA compared to neonates with no PDA. IL-1α and IL-1β are potent early response inflammatory cytokines that modulate their own production and induce other pro-inflammatory cytokines, including IL-6, IL-8, and tumor necrosis factor (TNF)-alpha. The biological effects of IL-1 range from inducing specific cell responses to targeting entire systems and may be both important for host responses to injury and infection and pathological in other conditions [bib_ref] Overview of the IL-1 family in innate inflammation and acquired immunity, Dinarello [/bib_ref].
The underlying cause of the association between pro-inflammatory markers and large PDA in very preterm neonates is unknown. Speculatively, in small-scale clinical studies, an alteration in inflammatory response has been found at baseline in children with congenital heart disease (CHD) [bib_ref] Immunity and inflammation: The neglected key players in congenital heart disease?, Wienecke [/bib_ref]. Moreover, an immune contribution to the CHD development has been suggested [bib_ref] Increased inflammatory markers in adult patients born with an atrial septal defect, Schram [/bib_ref] , as well as compensatory remodeling mechanism caused by changes in the cytokine profile. Children born with an atrial septal defect (ASD) undergo cardiac remodeling to compensate for the flow of blood from the left to the right atrium. In these children, the presence of markers for mechanical stress, inflammation, and remodeling are noted. Similar differences in acute-phase reactants are found in relation to ventricular septal defects (VSDs) [bib_ref] Acute phase proteins altered in the plasma of patients with congenital ventricular..., Zhang [/bib_ref].
## Balancing pro-and anti-inflammation in the preterm neonates
We found that neonates with large PDA had higher plasma levels of IL-6, IL-8, IL1RA, and IL-10 compared to neonates with no PDA. This is supported by previous findings, in a study including 47 neonates born at a GA below 28 weeks, which revealed that IL-6, IL-8, IL-10, IL-12 and growth differentiation factor 15 and monocyte chemotactic protein 1, were associated with hemodynamically significant PDA. This was defined as a PDA with ductal diameter of >1.5 mm, or a left atrium to aorta ratio of >1.5, or absent or reversed flow during diastole in the descending aorta [bib_ref] Exploration of potential biochemical markers for persistence of patent ductus arteriosus in..., Olsson [/bib_ref].
A rapid systemic inflammatory response can be an effective defense against microbial invasion. However, failure of inflammation-resolution processes leads to dysregulated and prolonged inflammation, even possibly systemic inflammation [bib_ref] Intermittent or sustained systemic inflammation and the preterm brain, Dammann [/bib_ref]. Several studies indicate that cytokine dysregulation during the first week of life, most notably high levels of IL-8, is associated with long-term morbidity, including BPD [bib_ref] Undetectable interleukin (IL)-10 and persistent IL-8 expression early in hyaline membrane disease:..., Jones [/bib_ref] , NEC [bib_ref] Interleukin 8 may predict surgical necrotizing enterocolitis in infants born less than..., Seo [/bib_ref] , atypical white matter brain development, and executive function limitations in adolescence [bib_ref] Interleukin-8 dysregulation is implicated in brain dysmaturation following preterm birth, Sullivan [/bib_ref] [bib_ref] Executive Dysfunction Early Postnatal Biomarkers among Children Born Extremely Preterm, Leviton [/bib_ref].
We found higher anti-inflammatory IL-10 and IL-1RA levels in neonates with large PDA. This may not be sufficient to balance the pro-inflammatory response present in neonates with large PDA driven by IL-1beta, IL-6, IL-8, and other inflammatory plasma proteins. The premature infant may be at a particularly high risk for unopposed proinflammatory effects, since developmental immaturity may limit the ability to increase endogenous expression of anti-inflammatory mediators sufficiently [bib_ref] Intermittent or sustained systemic inflammation and the preterm brain, Dammann [/bib_ref]. However, other sustained inflammatory stimuli from, e.g., mechanical ventilation, organ damage, and epigenetics processes, that are unable to shift from a pro-inflammatory to an anti-inflammatory state, may also contribute to intermittent or sustained systemic inflammation.
## Complement system
The complement system is an essential part of the innate immune response. In addition, it functions as the first line of defense against pathogens and elicits a pro-inflammatory response, leading to recruitment and activation of immune cells from both the innate and adaptive branches of the immune system [bib_ref] Complement activation, regulation, and molecular basis for complement-related diseases, Bajic [/bib_ref]. In neonates with large PDA, we found complement factor 8 (C8) alpha, beta, and gamma chain at lower plasma levels. A study including 60 infants (average, 4 years) with PDA and no other congenital heart disease and 60 controls with no PDA found that plasma levels of C3, C7, C8, and C9 were lower in infants with PDA compared to infants with no PDA [bib_ref] Altered plasma proteins released from platelets and endothelial cells are associated with..., Hou [/bib_ref].
Furthermore, C3 was decreased in the present study in neonates with large PDA, but with a significantly small fold-change to pass our filtering. Moreover, infants with congenital heart defects have decreased serum levels of C3 and C4 compared to infants without structural heart disease [bib_ref] Immunological profile in congenital heart disease, Khalil [/bib_ref]. The biological properties of C3a are regulated by the protein carboxypeptidase N (CPN). In neonates with large PDAs, we found lower plasma levels of carboxypeptidase N catalytic chain (CPN) 1 and carboxypeptidase N subunit 2 (CPN2). Complement components are synthesized early in fetal life, but with a relative deficiency in comparison with adult levels [bib_ref] Complement profile in neonates of different gestational ages, Grumach [/bib_ref]. It is unclear if the low complement levels in neonates with PDA are related to issues with the liver, the main synthesis location for most complement components, or due to the activation and consumption of complement factors.
## Is periostin a marker of pressure overload or of pulmonary remodeling?
The protein with the highest fold-change concentration in neonates with large PDA compared to neonates with no PDA was periostin. To the best of our knowledge, this is the first study to demonstrate that extremely preterm neonates with large PDA have higher plasma levels of periostin compared to neonates with no PDA. Periostin is a matricellular protein with functions in osteology, tissue repair, oncology, cardiovascular and respiratory systems, and in various inflammatory settings and diseases [bib_ref] The role of periostin in tissue remodeling across health and disease, Conway [/bib_ref].
The hemodynamic effects of a large left-to-right shunt associated with a PDA includes hyperperfusion of the pulmonary vasculature and volume overload of the atrium and ventricles. In the adult heart, periostin is almost undetectable [bib_ref] Periostin paves the way for neonatal heart regeneration, Hudson [/bib_ref] , but is induced in the ventricles following myocardial infarction, pressure overload, or generalized cardiomyopathy [bib_ref] Periostin as a heterofunctional regulator of cardiac development and disease, Conway [/bib_ref]. Speculatively, the hemodynamic effects of the large PDA may, at least in part, explain the higher levels of periostin. Supporting this, NT-proBNP is also known to be increased in relation to circulatory volume overload, and along with other authors, we have demonstrated that NT-proBNP is increased in neonates with PDA compared to neonates with no PDA [bib_ref] Cardiovascular biomarkers in the evaluation of patent ductus arteriosus in very preterm..., Sellmer [/bib_ref]. Data on NT-proBNP were re-analyzed in this cohort, found to be increased, and as a single marker, were able to separate large PDA from no PDA with an outstanding discrimination. We speculate that a large PDA induces the expression and release of periostin and NT-proBNP from the heart due to volume overload.
Periostin levels may represent pulmonary pathology as periostin expression is increased in lungs of neonates that died with severe BPD and is currently evaluated as a marker of BPD [bib_ref] Early Elevation of Plasma Periostin Is Associated with Chronic Ventilator-Dependent Bronchopulmonary Dysplasia, Ahlfeld [/bib_ref]. Periostin has been reported to be associated with neonatal murine lung remodeling [bib_ref] Periostin downregulation is an early marker of inhibited neonatal murine lung alveolar..., Ahlfeld [/bib_ref] and hyperoxic lung injury [bib_ref] Neonatal periostin knockout mice are protected from hyperoxia-induced alveolar simplication, Bozyk [/bib_ref].
In adults, periostin expression in the lungs decreases following acute injury, but then increases substantially during the initiation of repair-mechanisms and even beyond the initial insult, this increase may persist [bib_ref] The role of periostin in tissue remodeling across health and disease, Conway [/bib_ref]. Therefore, periostin may in the context of a PDA be an important structural mediator, conveying tissue adaption in response to insult or injury [bib_ref] The role of periostin in tissue remodeling across health and disease, Conway [/bib_ref] [bib_ref] Ablation of periostin inhibits post-infarction myocardial regeneration in neonatal mice mediated by..., Chen [/bib_ref].
## Do the levels of scd163 indicate inflammation involving macrophages system?
On proteomic analysis, we found higher levels of sCD163 in neonates with large PDA compared to neonates with no PDA. The scavenger receptor CD163 is expressed in macrophages and monocytes and is a receptor for multiple ligands, which are quantitatively important for the haptoglobin-hemoglobin complex [bib_ref] Soluble CD163. Scand, Moller [/bib_ref]. Increased plasma concentration of soluble CD163 (sCD163) [bib_ref] Identification of the hemoglobin scavenger receptor/CD163 as a natural soluble protein in..., Møller [/bib_ref] is observed in diseases related to macrophage activity, including acute and chronic inflammations in adults [bib_ref] Soluble CD163. Scand, Moller [/bib_ref].
In vitro, CD163 is upregulated by glucocorticoids, hemoglobin, and both IL-10 and IL-6 [bib_ref] CD163 and inflammation: Biological, diagnostic, and therapeutic aspects, Etzerodt [/bib_ref]. We speculate that higher sCD163 levels, also supported by higher levels of IL-6 and IL10 in neonates with large PDA, reflect sustained inflammation, or are also related to PRBC transfusion in neonates with large PDA.
# Strengths and limitations
In this study, the cohort investigated was comprised of 53 extremely preterm neonates. Very few studies have reported both clinical information and plasma proteomics data from LC-MS and ELISA analysis from this unique patient group. Our advanced laboratory and statistical methods allowed us to assess both ontogeny and the effect of GA when we identified proteins and inflammatory markers with a statistically significantly PDA associated abundance. Blood was not drawn at the exact same postnatal age (by hours). However, the postnatal ontogenetic effects by age on the plasma proteome is assumed to be minimal compared to prenatal, i.e., taking gestational age into account was crucial. Some heterogenicity may have caused a low statistical power and we included proteins that do not pass multiple hypothesis test corrections in our interpretations. However, many of the included proteins were functionally highly similar in supporting this decision.
Preterm neonates are a heterogenic patient population not only by GA at birth, but also by reason for their preterm birth and their postnatal challenges. We aimed to describe characteristics of the infants present prior to blood sampling and echocardiography postnatal day 3. Preeclampsia, chorioamnionitis, and use of antenatal steroids could impact the risk of both PDA and also influence the proteome. Moreover, both RDS (hence use of surfactant) and sepsis could contribute to an inflammatory response. Unfortunately, we do not have data on chorioamnionitis. Neither preeclampsia, RDS, nor sepsis was found more often in neonates with large PDA compared to neonates with no PDA in this cohort. However, mechanical ventilation was more frequently used in neonates with large PDA and might be an external stimuli for inflammation. What changes PRBC transfusion pose on the proteome has not been fully described.We found coherence between the results obtained with LC-MS and ELISA methods of periostin indicative of a technically robust analysis. Nonetheless, the findings should be validated in an independent study, ideally with a narrower GA distribution and larger sample size. Future studies characterizing heart tissue from newborns with PDA by transcriptome in combination with proteomics will likely enable a mapping of the tissue origins of many of the detected proteins [bib_ref] Degradation of the extracellular matrix is part of the pathology of ulcerative..., Kirov [/bib_ref] , and provide further insights into the PDA etiology.
# Conclusions
The present study investigated the proteome, which fluctuates over time in response to internal and external stimuli, making it particularly valuable for understanding various complications in preterm neonates. We found that a large PDA in extremely preterm neonates was associated with differences in angiotensinogen, periostin, and measures of both immune-and complement systems. A small PDA was associated with more subtle biological differences in the plasma proteome. Moreover, our findings indicate that the PDA may interfere with or be driven by an imbalance in core bio-physiological systems. Therefore, the bio-physiological understanding of mechanisms behind inflammatory stimuli and the neonate response to inflammation should be the focus of future PDA research. In particular, it may increase our understanding of pathogenesis, reveal potential tools for early diagnosis, drug development and prognostication, and present ways to monitor disease severity or improvement. Informed Consent Statement: All parents provided informed, written consent for their child to participate in the study.
# Data availability statement:
The data can be made available by contacting the authors.
[fig] Figure 1: Unsupervised principal component analysis (PCA). Unsupervised PCA scores plot of the (A) LC-MS proteomics data and (B) immuno-based assay data. Each dot represents a sample and the color denotes the PDA size. Explained variance is provided in percentages. No clear outliers were identified. [/fig]
[fig] Figure 2: Large PDA and the plasma proteome. Data from 53 neonates born at a gestational age below 28 weeks. (A) Plasma proteome changes large PDA compared to no PDA as determined by LC-MS proteomics, and (B) immunoassay. Red (blue) proteins are increased (decreased) in PDA. Grey proteins are not significantly different. Gene names are provided. Horizontal solid line: p-value = 0.05, horizontal dotted line: q-value = 0.05, vertical dotted lines: Fold-change cutoff. (C) Combined analysis of the large PDA compared to no PDA changing plasma proteins. Color and protein size indicates PDA fold-change and statistical significance, respectively. Known protein-protein interactions from the STRING database are indicated by lines, where the width signifies the number of known interactions. [/fig]
[fig] Figure 3: LC-MS proteomics and immunoassay data validation. (A) Correlation of periostin (POSTN) plasma levels in all samples as determined by LC-MS and immunoassay. Pearson's correlation coefficient (R), p-value, and 95% confidence interval (grey) is provided. (B) NT-proBNP plasma abundance variation over the gestational ages. (C) Sensitivity and specificity of NT-proBNP to identify large PDA (red) and small (green) PDA compared to no PDA, respectively. Area under the curve (AUC) and p-value for differences between the ROCs are provided. [/fig]
[fig] Author: Contributions: Conceptualization, A.S., T.B.H., B.H.B. and V.E.H.; methodology, A.S., T.B.H., J.P., B.H.B., T.B.B. and V.E.H.; validation, A.S., J.A., J.P. and T.B.B.; formal analysis, A.S., J.P., J.A. and T.B.B.; investigation, A.S., T.B.H., J.A., J.P. and T.B.B.; resources, T.B.H., J.P. and V.E.H.; data curation, A.S. and T.B.B.; visualization, A.S. and T.B.B.; supervision, T.B.H., B.H.B., J.P. and V.E.H.; project administration, A.S.; funding acquisition, A.S., T.B.H., V.E.H. and J.P. All authors participated in writing and preparing the original draft preparation. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: The work was supported financially by Arvid Nilssons Foundation (A.S.); Grosserer L.F. Foghts Foundation (A.S.); Aarhus University Hospital (A.S.); Novo Nordisk Foundation, grant no. NFF17SA0030576 (V.E.H.); and The John and Birthe Meyer Foundation (J.P.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Institutional Review Board Statement: The study was approved by the Central Denmark Region Committee on Health Research Ethics (journal number M-20090243), the Danish Data Protection Agency, and the National Board of Health. [/fig]
[table] Table 1: Study cohort.Characteristics of 53 neonates born before 28 + 0 gestational weeks by presence of PDA postnatal day 3. Small: PDA diameter ≤ 1.5 mm, large: PDA diameter ≥ 1.5 mm. * p-values listed for large PDA to no PDA comparison. Gestational age (GA) weeks, packed red blood cell transfusion (PRBC) within the first 3 days of life, inotropes within the first 3 days of life, early onset sepsis (EOS) defined as 7 days of antibiotics initiated before day 3. Mechanical ventilation used day 3. Intraventricular hemorrhage (IVH). Large LA:Ao ratio (left atrium to aorta ratio) > 1.5. DADF descending aorta diastolic flow. [/table]
[table] Table 2: Large PDA significant proteins from LC-MS. Statistically significant proteins (p-value < 0.05) differentiating large PDA from no PDA, as determined by proteomics. [/table]
[table] Table 3: Large PDA significant proteins from immunoassay. Statistically significant cytokines (p-value < 0.05) differentiating large PDA from no PDA, as determined by immunoassay. [/table]
|
Dasiglucagon, a next‐generation ready‐to‐use glucagon analog, for treatment of severe hypoglycemia in children and adolescents with type 1 diabetes: Results of a phase 3, randomized controlled trial
Background: Dasiglucagon, a next-generation, ready-to-use aqueous glucagon analog formulation, has been developed to treat severe hypoglycemia in individuals with diabetes.Objective: The aim of this trial was to evaluate the safety and efficacy of dasiglucagon in pediatric individuals with type 1 diabetes (T1DM). Participants were children and adolescents (6-17 years) with T1DM.Methods: In this randomized double-blind trial, 42 participants were randomly allocated (2:1:1) to a single subcutaneous (SC) injection of dasiglucagon (0.6 mg), placebo, or reconstituted glucagon (GlucaGen; dosed per label) during insulin-induced hypoglycemia. The primary endpoint was time to plasma glucose (PG) recovery (first PG increase ≥20 mg/dL after treatment initiation without rescue intravenous glucose). The primary comparison was dasiglucagon vs. placebo; glucagon acted as a reference.Results: The median time (95% confidence interval) to PG recovery following SC injection was 10 min (8-12) for dasiglucagon vs. 30 min (20 to -) for placebo (P < .001); the median time for glucagon was 10 min (8-12), which did not include the time taken to reconstitute the lyophilized powder. PG recovery was achieved in all participants in the dasiglucagon and glucagon groups within 20 min of dosing compared to 2 out of 11 patients (18%) with placebo. The most frequent adverse events were nausea and vomiting, as expected with glucagon treatment.Conclusions: Consistent with adult phase 3 trials, dasiglucagon rapidly and effectively restored PG levels following insulin-induced hypoglycemia in children and adolescents with T1DM, with an overall safety profile similar to glucagon.K E Y W O R D Sdouble-blind trial, glucagon, hypoglycemia, hypoglycemic agent, type 1 diabetes
# | introduction
Insulin therapy is central to the treatment of type 1 diabetes (T1DM); as pancreatic β-cell function declines, insulin treatment is also required to achieve glycemic control in many people with type 2 diabetes. Hypoglycemia is a common side effect of insulin treatment, with severe hypoglycemia often requiring prompt medical intervention to prevent potentially life-threatening complications (seizure, loss of consciousness, and/or coma). [bib_ref] Assessment and management of hypoglycemia in children and adolescents with diabetes, Abraham [/bib_ref] The incidence of severe hypoglycemia in pediatric patients has fallen in recent years with improved hypoglycemia education and the increased use of insulin analogs, insulin pump therapy, and continuous glucose monitoring, 1 but clinically significant hypoglycemia remains a challenge. It has been suggested that multiple hypoglycemic episodes may have negative cognitive effects in pediatric patients, particularly during early development. [bib_ref] Frequency and timing of severe hypoglycemia affects spatial memory in children with..., Hershey [/bib_ref] Therefore, it is of paramount importance, in terms of both acute and potentially more long-term complications, that treatment options are available that can rapidly reverse episodes of severe hypoglycemia.
Glucagon is a well-established and long-standing first-line treatment for severe hypoglycemia in people with diabetes. American Diabetes Association (ADA) treatment guidelines recommend that glucagon is prescribed for all individuals at increased risk for clinically significant hypoglycemia so that it is available in emergencies.Caregivers, school personnel, and family members of patients are furthermore advised that they should know where glucagon treatment is stored and be trained in when and how to administer it.The majority of glucagon for prescription use is provided in glucagon emergency kits,where the need to reconstitute the lyophilized drug product via multiple steps immediately prior to injection represents a significant barrier to timely and accurate administration and has led to underutilization of glucagon for the treatment of hypoglycemia. [bib_ref] Glucagon administrationunderevaluated and undertaught, Harris [/bib_ref] [bib_ref] Treatment of severe diabetic hypoglycemia with glucagon: an underutilized therapeutic approach, Kedia [/bib_ref] To overcome these limitations, glucagon products have recently been developed that do not require reconstitution for the treatment of severe hypoglycemia, with human glucagon delivered nasally as a powder 8 or via the subcutaneous (SC) injection of a liquid formulation using dimethyl sulfoxide as diluent.An alternative treatment option is dasiglucagon, a next-generation glucagon analog. [bib_ref] Pharmacokinetic and Pharmacodynamic Characteristics of Dasiglucagon, a Novel Soluble and Stable Glucagon..., Hovelmann [/bib_ref] Like human glucagon, dasiglucagon comprises 29 amino acids, with seven amino acid substitutions introduced to improve its physical and chemical stability in aqueous media. [bib_ref] Pharmacokinetic and Pharmacodynamic Characteristics of Dasiglucagon, a Novel Soluble and Stable Glucagon..., Hovelmann [/bib_ref] In addition to enabling continuous infusion via SC pump delivery for glycemic control in diabetes and for other unique indications currently being pursued, the improved stability in aqueous media has enabled dasiglucagon development and approval (US trade name Zegalogue) as a ready-to-use, aqueous product for treatment of severe hypoglycemia via SC injection in pediatric and adult patients with diabetes aged 6 years and older. Zegalogue is stored in the refrigerator and can be kept at room temperature between 20 C and 25 C for a single period of up to 12 months. Two recent phase 3 trials in adults with T1DM have shown dasiglucagon to be a rapid and effective treatment for severe hypoglycemia when administered as a single SC dose of 0.6 mg, with a median time to plasma glucose (PG) recovery (defined as the time to first observed increase in PG of ≥20 mg/dL [1.1 mmol/L] after SC injection) of 10 min in both trials. [bib_ref] Dasiglucagon:A Next-generation Glucagon Analog for Rapid and Effective Treatment of Severe Hypoglycemia...., Pieber [/bib_ref] [bib_ref] Dasiglucagon Hypopal autoinjector as a fast and effective treatment for severe hypoglycemia:..., Bailey [/bib_ref] We report the results of a phase 3 trial that evaluated the efficacy and safety of dasiglucagon relative to placebo and glucagon for treatment of severe hypoglycemia in children and adolescents (age range: 6-17 years) with T1DM.
# | methods
## | study design
This multicenter, randomized, placebo-controlled, double-blind, parallel-group phase 3 trial in children and adolescents with T1DM was conducted at five sites in three countries (Germany, Slovenia, and the US) between September 28, 2018, and June 28, 2019. Participants were randomly allocated (2:1:1) to receive a single SC injection of dasiglucagon 0.6 mg, placebo, or reconstituted glucagon (GlucaGen [glucagon for injection], Novo Nordisk A/S). A dasiglucagon dose of 0.6 mg (same dose used in adults) was selected based on pharmacokinetics (PK)/pharmacodynamics (PD) modeling and simulation, which showed that the higher total exposure and PD response at lower weight is partially compensated by a higher clearance, lower bioavailability, and practically saturated PD response, all contributing to a safe and effective drug response. Glucagon was dosed according to label (1.0 mg for body weight > 25 kg; 0.5 mg for body weight < 25 kg).
The trial protocol, consent form, and other information provided to participants and parents/legal representatives were approved by independent ethics committees or institutional review boards at participating sites and by competent authorities. The trial was conducted according to the Declaration of Helsinki and Good Clinical Practice with written informed consent obtained from parents/caregivers and assent from participants (as required) before trial enrollment. The trial is registered at www.clinicaltrials.gov (NCT03667053).
## | participants
Eligible participants were aged between 6 and 17 years (both inclusive) and had diagnosed T1DM per ADA criteria for at least 1 year prior to trial participation, were receiving daily insulin, and were ≥ 20 kg in body weight.
Children and adolescents with a presence or history of pheochromocytoma or insulinoma were excluded from participation, as were those who had hypoglycemic events associated with seizures or hypoglycemia unawareness (as assessed at the investigators' discretion) in the prior year or severe hypoglycemia in the prior month. Trial exclusion criteria also included regular use of beta blockers, indomethacin, warfarin, or anticholinergic drugs in the 28 days prior to screening or use of prescription or nonprescription medications known to cause QT prolongation (see for a list of all inclusion and exclusion criteria).
## | randomization and blinding
Randomization was stratified by age group and injection site (abdomen; thigh) using a central, dynamic variance minimization randomization method via an interactive web response system. Both investigators and participants were blinded to the investigational therapy. However, due to differences in the appearance of dasiglucagon (aqueous formulation) and glucagon (lyophilized powder for reconstitution), the preparation and administration of trial medication were performed by unblinded trial personnel who were not involved in any other trial procedures or assessments. Treatment assignment was also blinded for trial personnel involved in medical and safety monitoring, data cleaning, and defining analysis sets until the database was released for statistical analysis.
## | procedures
Participants attended an on-site dosing visit where they were randomized to receive a single SC dose of dasiglucagon 0.6 mg, placebo, or glucagon in an insulin-induced hypoglycemic state. The on-site dosing visit included an overnight stay prior to dosing (participants fasted overnight from 10 PM).
The participants' regular insulin therapy was stopped in advance of dosing according to predefined timelines (see. Hypoglycemia was induced by intravenous (IV) infusion of insulin glulisine (Apidra, 100 U/mL), using a dose and infusion rate chosen by the investigator to facilitate a controlled decline in PG concentration to a target of 80 mg/dL (4.4 mmol/L). This PG threshold for stopping insulin infusion was deliberately set conservatively to ensure safety. PG was monitored using glucose analyzers (YSI 2300, Yellow Springs Instruments, or the Super GL analyzer, Dr. Müller Gerätebau GmbH). After the start of the insulin infusion, PG was monitored approximately every 10 min while PG was >110 mg/dL and approximately every 5 min once PG was ≤110 mg/dL. The insulin infusion was stopped once the PG concentration was <80 mg/dL (4.4 mmol/L). After 5 min, if PG was ≥54 mg/dL and < 80 mg/dL (3.0-4.4 mmol/L), the study drug was administered by SC injection. If the PG concentration was outside this range, sufficient IV glucose or insulin was administered to bring the PG concentration within the target range. The needle length of the prefilled syringe used for administering the study drug was 0.5 inch (12.7 mm). The trial protocol instructed that dosing be done subcutaneously but did not otherwise actively discriminate between intramuscular and SC injections, as the risk of an inadvertent intramuscular injection was considered low. Blood samples for PG measurement at a central laboratory (efficacy assessments) were taken within 2 min prior to dosing and at predefined intervals at 4, 6, 8, 10, 12, 15, 17, 20, 30, and 45 min (as well as 60 min if the patient's body weight was ≥21 kg) postdose.
Blood samples for PK measurements (dasiglucagon and glucagon)
were taken prior to dosing and at predefined intervals up to 300 min postdose, after which the participant was discharged from the trial facility. Safety assessments including recording of adverse events (AEs) were done at screening, at the dosing visit, and at a safety follow-up visit 28 days after dosing.
## | statistical analyses
The primary endpoint was time from dosing to PG recovery, defined as time to first observed increase in PG of ≥20 mg/dL (1.1 mmol/L) from time of administration (baseline) without IV glucose rescue treatment.
PG was considered "not recovered" if IV glucose was administered prior to recovery or if recovery was not achieved by 45 min. Sample size was determined based on the primary objective of confirming superiority of dasiglucagon to placebo with respect to the primary endpoint. Phase 2 results in adults were used to determine sample size. [bib_ref] Pharmacokinetic and Pharmacodynamic Characteristics of Dasiglucagon, a Novel Soluble and Stable Glucagon..., Hovelmann [/bib_ref] These results showed that the median time to an increase in PG of 20 mg/dL was approximately 10 min for dasiglucagon 0.6 mg compared to an assumed median time of 50 min for placebo. Assuming an exponential distribution with median times of 10 and 50 min, this difference can be detected with 90% power at a 5% significance level using a two-sided log-rank test in 20 participants randomized to dasiglucagon and 10 participants randomized to placebo with a follow-up time of 45 min. Glucagon was included as a reference to compare the responses and AE profile of dasiglucagon with those of a marketed product; no formal statistical comparison between dasiglucagon and glucagon was conducted, and a glucagon group of 10 participants was considered sufficient for this comparison.
The primary endpoint was summarized by treatment group using survival analysis methods (median time to event). Participants who received rescue IV glucose before 45 min (as well as those without PG recovery within 45 min of dosing) were censored at 45 min. The treatment difference for dasiglucagon versus placebo was evaluated inferentially using a two-sided log-rank test stratified by age group and injection site. Plots show the estimated probability of having recovered ("inverted" Kaplan-Meier curves) for each treatment group, with results for the primary endpoint expressed as Kaplan-Meier estimates of the median time to PG recovery. Kaplan-Meier plots are also presented by age group and injection site. Two sensitivity analyses were performed, whereby the primary analysis was repeated without censoring for participants receiving rescue IV glucose within 45 min and with censoring at the time of rescue for participants receiving rescue IV glucose within 45 min.
In addition to determining the observed time from dosing to PG recovery, a prespecified supportive analysis was performed to calculate the true time from dosing to PG recovery for each participant using linear interpolation between the two time points before and after PG recovery had occurred (i.e., to determine the predicted time of an exact 20-mg/dL increase in PG). As for the primary endpoint, results were summarized by treatment group using survival analysis methods (median time to event).
Confirmatory secondary endpoints were achievement of PG recovery within 30, 20, 15, and 10 min of dosing and change in PG from baseline at these time points. As for the primary endpoint, formal statistical comparisons were made between dasiglucagon and placebo. The 30-, 20-, 15-, and 10-min recovery rates were compared at each time point using a Cochran-Mantel-Haenszel test stratified by age group and injection site. Change from baseline in PG was compared at each time point using an analysis of covariance model with treatment group, age group, and injection site as fixed effects and baseline PG as a covariate. For participants who required rescue IV glucose treatment, the PG value at the time of rescue was carried forward.
For the primary and secondary confirmatory endpoints, the overall type 1 error was controlled via a prespecified hierarchical (fixed sequence) inferential testing procedure (Supplementary .
# | results
## | participant disposition and characteristics
Of the 59 children and adolescents screened for the trial, 42 eligible participants were randomly assigned in a 2:1:1 ratio to dasiglucagon (n = 21), placebo (n = 11), and glucagon (n = 10). One participant (dasiglucagon group) withdrew for logistical reasons prior to receiving the investigational product. In total, 41 participants received investigational product (dasiglucagon: n = 20; placebo: n = 11; glucagon: n = 10), all of whom completed the trial (see for further details).
Overall, baseline characteristics at randomization were well matched among treatment groups , with the exception of a higher proportion of male participants in the glucagon group. Most of the participants were White (95.1%), and 80.5% of participants were non-Hispanic/Latino. PG at baseline following the hypoglycemic clamp procedure was similar across treatment groups, with median values in the range of 72-73 mg/dL across groups . . In the glucagon group, the median time to PG recovery was 10 min (95% CI, [bib_ref] Pharmacokinetic and Pharmacodynamic Characteristics of Dasiglucagon, a Novel Soluble and Stable Glucagon..., Hovelmann [/bib_ref] [bib_ref] Dasiglucagon:A Next-generation Glucagon Analog for Rapid and Effective Treatment of Severe Hypoglycemia...., Pieber [/bib_ref] [bib_ref] Dasiglucagon Hypopal autoinjector as a fast and effective treatment for severe hypoglycemia:..., Bailey [/bib_ref] , that is, similar to the results obtained for dasiglucagon. Using linear interpolation to estimate the true time from dosing to PG recovery, the median time to recovery was 8.7 min (95% CI, 6.9-10.6) for dasiglucagon, 29.3 min (95% CI, 18.5; upper limit not estimable) for placebo, and 9.8 min (95% CI, 7.4-10.6) for glucagon .
## T a b l e 1 participant baseline characteristics
## | proportion of participants in whom pg recovery was achieved
## | pg change from baseline
Mean PG increase from baseline over time is shown in . After
## | safety
No serious or severe AEs were reported. Nausea and vomiting were the most frequently reported AEs for dasiglucagon and the active comparator, with a higher proportion of participants experiencing these transient events with dasiglucagon than with glucagon ( imbalance was apparent in the 6-to-11-year age group . There was no apparent correlation between dasiglucagon exposure (area under the curve or C max ) and nausea and/or vomiting by either age (Supplementary or body weight (Supplementary .
For both dasiglucagon and glucagon, most AEs of nausea had an onset within 1.5-3 h of dosing, with the majority of events lasting less than 2 h. Vomiting tended to occur later than nausea. The majority of events occurred between 2 and 3 h of dosing and had a duration of less than 5 min. There was no apparent difference between age groups (6-11 years; 12-17 years) in the time to onset or duration of nausea or vomiting.
Three injection site erythema events were reported; all were mild and transient events following glucagon treatment. Hypoglycemia AEs did not appear temporally related to the clamp procedure or investigational product dosing. One hypoglycemia AE (in the placebo group) was recorded as having an onset at 5 min after dosing;
the remainder had an onset between approximately 5 and 33 days of investigational product administration.
There were no clinically significant changes in laboratory evalua- Consistent with the results for the two pivotal adult trials, [bib_ref] Dasiglucagon:A Next-generation Glucagon Analog for Rapid and Effective Treatment of Severe Hypoglycemia...., Pieber [/bib_ref] [bib_ref] Dasiglucagon Hypopal autoinjector as a fast and effective treatment for severe hypoglycemia:..., Bailey [/bib_ref] superiority was confirmed for dasiglucagon relative to placebo with regard to the primary endpoint and all confirmatory secondary endpoints. The median time from injection to PG recovery was 10 min for dasiglucagon versus 30 min for placebo and 10 min for glucagon, with no clinically relevant differences seen for dasiglucagon with respect to age group (6-11 years; 12-17 years) or injection site (abdomen; thigh). It is noteworthy that the median time to PG recovery was similar for dasiglucagon in pediatric and adult participants (approximately 10 min). Furthermore, dasiglucagon increased PG levels by a similar magnitude from baseline in pediatric and adult participants. [bib_ref] Dasiglucagon:A Next-generation Glucagon Analog for Rapid and Effective Treatment of Severe Hypoglycemia...., Pieber [/bib_ref] [bib_ref] Dasiglucagon Hypopal autoinjector as a fast and effective treatment for severe hypoglycemia:..., Bailey [/bib_ref] Using data interpolation to obtain a more precise estimate of recovery times, the median true time from dosing to PG recovery was year-old patients with these events in the dasiglucagon group may be a coincidental finding in this relatively small trial with 42 participants. This is further supported by the results from a larger phase 3 trial in adult patients testing dasiglucagon in a similar trial setting, which showed no difference between dasiglucagon and glucagon with respect to incidences of nausea and vomiting. [bib_ref] Dasiglucagon:A Next-generation Glucagon Analog for Rapid and Effective Treatment of Severe Hypoglycemia...., Pieber [/bib_ref] Pediatric guidelines recommend that glucagon should be readily accessible to all parents and caregivers. [bib_ref] Assessment and management of hypoglycemia in children and adolescents with diabetes, Abraham [/bib_ref] Currently available treatments for severe hypoglycemia, when a child or adolescent is unable to safely swallow oral carbohydrates, are limited to IV dextrose and glucagon. Patients reported various administration issues (unsuccessful injection, delays in injection) in using injectable glucagon emergency kits that need complex, time-consuming procedures for reconstitution. [bib_ref] Faster Use and Fewer Failures with Needle-Free Nasal Glucagon Versus Injectable Glucagon..., Yale [/bib_ref] [bib_ref] Human Factors Usability and Validation Studies of a Glucagon Autoinjector in a..., Valentine [/bib_ref] [bib_ref] Nasal Glucagon Delivery Is More Successful Than Injectable Delivery: A Simulated Severe..., Settles [/bib_ref] Nasal glucagon was recently approved for children in a fixed-dose drug-device combination for children older than 4 years but may cause side effects such as headache, upper airway discomfort, lacrimation, or nasal congestion in addition to the known side effects of glucagon. [bib_ref] Glucagon Nasal Powder: A Promising Alternative to Intramuscular Glucagon in Youth With..., Sherr [/bib_ref] Thus, an aqueous, ready-to-use glucagon analog formulation may be a welcome addition to the options for the treatment of severe hypoglycemia in childhood.
The strengths of this trial include its randomized, placebo-controlled, multicenter design as well as the blinding of the investigator and participants to study treatment to reduce potential bias. However, it is acknowledged that the trial was conducted in a highly controlled investigational inpatient setting that may not fully reflect conditions in the real world where, for example, the PG level at which dasiglucagon is used in clinical practice may be lower than that used in the trial (where a PG target of 80 mg/dL [4.4 mmol/L] was set for the hypoglycemic clamp).
In conclusion, dasiglucagon is rapid, effective, and reliable in restoring PG levels following insulin-induced hypoglycemia in children
# Author contributions
Zealand Pharma A/S sponsored the study and was involved in the design (Ramin Tehranchi and Thomas Danne designed the trial) and
conduct of the study and analysis and interpretation of the data, including collection, management, and statistical analysis of the data.
## Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of this article.
[fig] 30: min, mean PG had increased from baseline by 98.2 mg/mL with dasiglucagon compared to 17.3 mg/mL with placebo; for glucagon the mean increase was 84.4 mg/mL. The increase in mean PG from baseline was significantly greater for dasiglucagon versus placebo at 10, 15, 20, and 30 min postdose (P < .001 at each time point). [/fig]
|
Gallbladder Neuroendocrine Carcinoma: A Rare Endocrine Tumor
Gallbladder neuroendocrine neoplasms (GB-NEN) are very rare neuroendocrine tumors (NETs). GB-NEN can present as carcinoid or typical/atypical carcinoid or small cell carcinoma. Most of the GB-NENs present as gall bladder polyps or stones with right upper quadrant pain, nausea and non-specific symptoms which leads to clinical misdiagnosis. Considering the rare occurrence of GB-NENs, and lack of multi-center research data there is no unified standard for identification and treatment. We here present an 84-year-old male presenting with right upper quadrant and epigastric pain, and eventually diagnosed with mixed cell (more of small cells mixed with intermediate to large cells) neuroendocrine cancer of gall bladder.
# Introduction
Neuroendocrine tumors (NETs) are a group of neoplasms that originate from neuroendocrine cells present in the various organs but more commonly found in gastrointestinal (GI) tract, lungs, and thyroid. In the GI tract, they are more common in ileum, jejunum, and pancreas, but very rarely reported in gallbladder as well. As it is very rare aggressive neoplasm, we need more research to identify it earlier and develop a standardized treatment option.
## Case presentation
An 84-year-old Caucasian male presented to the emergency room with one day history of abdominal pain, more localized to epigastric and right upper quadrant area. Pain is sharp in intensity, radiating to substernal area. Pain was associated with few episodes of nausea and vomiting. He denied any hematemesis, diarrhea, shortness of breath, wheezing, fever and chills. He never had any similar complaints in the past. He denied any weight loss or loss of appetite. His pain slightly got better in the emergency room with intravenous opiates. His past medical history was significant for essential hypertension, Alzheimer's dementia, coronary artery disease, atrial fibrillation and stroke. Past surgical history was significant for appendectomy, coronary artery by-pass grafting and joint surgeries.
His initial vitals were within normal limits. On physical exam he had irregularly irregular rhythm with good breath sounds bilaterally. Abdomen examination was significant for right upper quadrant tenderness with positive murphy sign (tenderness and guarding in the right upper quadrant of abdomen on palpation and exacerbated by inspiration).
Significant initial laboratory findings on initial presentation include elevated alkaline phosphate 273 IU/ml (reference range: 40-150 IU/ml), elevated aspartate aminotransferase (AST) -63 IU/ml (reference range: 5-34 IU/ml), elevated total bilirubin of 1.4 mg/dl (reference range: 0.2-1.2 mg/dl), mildly elevated lipase of 101 IU/ml (reference range: 8-78 IU/ml), normal alanine transaminase (ALT) 44 IU/ml (reference range: 0-55 IU/ml), and normal white blood cell count 5.9 k/cmm (reference range: 5-10 k/cmm).
Because of clinical suspicion of acute cholecystitis, the patient had an ultrasound of gallbladder which showed cholelithiasis, and sludge/inspissated gall bladderwith no radiographic evidence of acute cholecystitis, non-specific, soft, non-dependent tissue along the gall bladder wall . The patient was started on intravenous Levofloxacin and admitted to the hospital with a surgical consultation.
The patient had a laparoscopic cholecystectomy with normal intra-operative cholangiogram during that hospitalization. He was found to have a gangrenous gallbladder with a large Calot node (sentinel lymph node of gall bladder) and nodule on the wall of gallbladder and it was sent for biopsy after removal. He had an uneventful postoperative state and was discharged home the next day.
Biopsy results showed high-grade neuroendocrine carcinoma of gallbladder with predominantly small cell cancer. It was poorly differentiated, invading lymphovascular structures, perimuscular fat and serosa. Unfortunately, lymph nodes were not sampled during the surgery as the suspicion for malignancy was low.
Pathological staging showed pT3, pNx. Sections showed abundant neuroendocrine tumor with associated necrosis. Most of the cells were small in size with few intermediate to large cells.
## Cells) with hyperchromatic nuclei and scant cytoplasm (magnified slide) with lymphocytic invasion (pink colored lymph vessel with blue dot cells)
Immunohistochemistry reveals that malignant cells are positive for pan keratin, low molecular weight keratin, synaptophysin, TTF1, CK7 and CA19-9. Cells are negative for CD20, CD3, CK20 and p63.
His staging studies including computerized tomography (CT) scanning of chest, abdomen, pelvis, head, and nuclear medicine whole body bone scan were negative for any metastatic lesions. Oncologist diagnosed him with Stage IIIA (pT3, Nx Mx) small cell carcinoma of the gallbladder and arranged for adjuvant chemoradiation with Capecitabine as radio sensitizing agent. The patient was supposed to be started on a small cell cancer chemotherapy regimen, however he decompensated in the next few days and died after family has opted for comfort care measures only.
# Discussion
Neuroendocrine neoplasms (NEN) are a group of neoplasms that can arise in various organs such as gastrointestinal tract, lungs, thyroid that usually have predominate neuroendocrine differentiation. Incidence of neuroendocrine tumors is about 5.25/100,000. Gallbladder NETs only comprise 5% of those. There are no clear classification systems as some studies classify based on organ where they arise from, and some classify based on histological grading and staging. International Agency for Research on cancer (IARC) and World Health Organization (WHO) experts proposed a classification system for NENs in 2018. Most of the physicians use a classification system endorsed for gastroenteropancreatic (GEP) system NETs by WHO. WHO classification is mainly based on the tumor proliferative rate as assessed using mitotic counts or Ki-67 labelling index. Basically, there are two main categories of GEP NENs: Well-differentiated neuroendocrine tumors (NETs) and poorly differentiated neuroendocrine carcinomas (NECs).
## Morphology
1) Well-differentiated NETs are well circumscribed with relatively uniform cells and nuclei with granular cytoplasm and stippled chromatin. Most of them secrete neuro-secretory granules and express markers such as synaptophysin and chromogranin A. There are some tumors that specifically secrete hormones such as insulin, glucagon, gastrin, somatostatin, etc. Most of the time, we cannot correlate the histological features with morphological location except for few specific NETs like insulin secreting pancreatic NETs.
2) Poorly-differentiated NECs on the other hand have non-uniform cells with irregular nuclei and less cytoplasmic granularity. They can still have some immune reactivity but not as much as NETs. There is no definitive theory on origin for gallbladder NECs. Some researchers think that gallbladder epithelium becomes metaplastic and changes to neuroendocrine cells because of chronic inflammation in patients with cholelithiasis, while some others think that NEC is transformed from gallbladder adenocarcinoma.
## Clinical features and diagnosis
There are no specific symptoms and signs that are characteristic for gallbladder NECs. Patients most commonly present with right upper quadrant pain, nausea and vomiting, and positive Murphy's sign. Some patients present with weight loss and loss of appetite. Very rarely, patients present with symptoms of carcinoid syndrome including flushing, edema, and wheezing. Liver function tests, complete blood counts, basic metabolic panel are usually done in these patients on presentation. Imaging studies like ultrasonography, CT or magnetic resonance imaging (MRI) help in identifying the lesions. Most of the times we cannot distinguish them from other types of gallbladder cancers just based on imaging studies. A definitive diagnosis of gallbladder NECs is made only based on pathology reports and immunohistochemistry staining. It also helps with tumor grading and staging based on WHO classification. Patients also need other tests like bone scans, positron emission tomography (PET) scans, serum tumor markers, etc. to look for metastatic lesions that will help in staging and eventually with treatment.
## Treatment
The treatment options for gallbladder NEC vary based on the stage of tumors. Early in situ tumors usually respond well to cholecystectomy alone. But late stage tumors need radical surgeries along with dissection of nodes and even removal of some close metastatic lesions. As most of these patients present in advanced stages, they need concomitant chemoradiation. Despite all above treatment options, gallbladder NEC still has very high mortality with poor prognosis.
# Conclusions
Neuroendocrine carcinoma of gallbladder is an exceedingly rare malignancy with poor prognosis. The poor prognosis, in part, is based on the vague symptoms on presentation and advanced stage of presentation. Because of the rarity of the disease, we need more research to diagnose these patients early and come up with a standardized treatment plan.
# Additional information disclosures
Human subjects: Consent was obtained by all participants in this study.
## Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.
|
Neural correlates in the development of and recovery from dysphagia after supratentorial stroke: A prospective tractography study
A B S T R A C TBackground: Swallowing impairment after stroke may be related to the state of the corticobulbar tract (CBT), which is the motor projection fiber responsible for deglutition, but evidence is still lacking regarding which parameter could relate to poststroke swallowing recovery as measured by videofluroscope findings. This prospective study evaluated diffusion tensor imaging (DTI) parameters among dysphagic stroke patients compared with those of nondysphagia stroke patients and age-matched healthy subjects and followed swallowing recovery in dysphagic patients as assessed with the Modified Barium Swallow Impairment Profile (MBSImP©). Methods: Diffusion tractography was performed in 69 subjects, consisting of 27 S patients with dysphagia, 18 healthy subjects and 24 S patients with no evidence of dysphagia. DTI was performed within 14 days of stroke onset. Follow-up DTI was performed in the dysphagic group at three months. The tract volume (TV) of the CBT and frontal operculum as determined by fractional anisotropy (FA) was compared among the 3 groups. Correlations of these parameters with initial dysphagia severity and swallowing parameters at baseline and 3 months postonset were assessed. Results: All stroke patients showed lower CBT TV on the affected and unaffected sides than those in the control group, even in those who showed no evidence of clinical dysphagia. The dysphagia group showed a greater reduction in CBT TV on the affected side (P < 0.001). Receiver operating characteristic analysis showed that cutoff values of 4.1 cm 3 for TV and 0.24 for FA from the affected side could classify dysphagia with good accuracy (AUC = 0.77, 0.75, respectively) and specificity levels. FA values in the unaffected frontal operculum showed a significant correlation (rho = − 0.40, P = 0.02) with swallowing outcome as observed by the total scores of MBSImP©. In addition, these values proved to be significant variables to predict swallowing outcome in multiple regression analysis (R 2 = 0.6317, adjR 2 = 0.5815, F = 12.58, p < 0.001, AIC = 203.65). Conclusions: Even when clinical dysphagia is not apparent, individuals with a supratentorial stroke may show reduced CBT parameters compared to healthy controls. Supratentorial stroke may manifest with dysphagia if a certain extent of CBT volume and white matter tract integrity is involved, with a greater degree of CBT injury in the affected sides determining poststroke dysphagia severity. In contrast, recovery was independent of the affected parameters, and an initial lower FA value in the unaffected frontal operculum was indicative of a poorer 3-month dysphagia outcome. DTI parameters obtained within two weeks of stroke onset may help classify those with dysphagia, predict recovery and help plain therapeutic strategies to maintain the adaptive role of the white matter tract, which is crucial in swallowing recovery.
# Introduction
Approximately 50% of all patients with stroke experience dysphagia [bib_ref] Aspiration in patients with acute stroke, Daniels [/bib_ref]. Dysphagia is associated with increased stroke mortality [bib_ref] Dysphagia in acute stroke: incidence, burden and impact on clinical outcome, Arnold [/bib_ref] and medical complications such as malnutrition, dehydration [bib_ref] Dysphagia, nutrition, and hydration in ischemic stroke patients at admission and discharge..., Burn [/bib_ref] and aspiration pneumonia [bib_ref] Dysphagia after stroke: incidence, diagnosis, and pulmonary complications, Martino [/bib_ref]. Long-term recovery of functional swallowing is independently associated with poor outcomes after stroke [bib_ref] Development and validation of a prognostic model of swallowing recovery and enteral..., Galovic [/bib_ref]. Therefore, in addition to quickly and accurately detecting dysphagia, it is imperative to identify patients manifesting poor swallowing recovery after stroke.
Stroke lesions involving the frontal operculum, the posterior limb of the internal capsule, and bilateral hemispheres have been associated with the occurrence and prognosis of poststroke dysphagia. [bib_ref] Lesion location predicts transient and extended risk of aspiration after supratentorial ischemic..., Galovic [/bib_ref] [bib_ref] Development and validation of a prognostic model of swallowing recovery and enteral..., Galovic [/bib_ref] [bib_ref] Development of a Novel Prognostic Model to Predict 6-Month Swallowing Recovery After..., Lee [/bib_ref] Past research has focused on the lesion location causing dysphagia. However, strokes show highly variable lesion locations, lesion extents, and degrees of neural tract injury. Recently, the importance of analyzing neural pathways and their integrity related to swallowing has been highlighted [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref] [bib_ref] Prognostic prediction of dysphagia by analyzing the corticobulbar tract in the early..., Jang [/bib_ref] [bib_ref] Dysphagia prognosis prediction via corticobulbar tract assessment in lateral medullary infarction: a..., Jang [/bib_ref].
The corticobulbar tract has been shown to be a potent neural pathway involved in swallowing, and any lesions disrupting this pathway can lead to dysphagia. [bib_ref] Diverging lesion and connectivity patterns influence early and late swallowing recovery after..., Galovic [/bib_ref] Fibers from the caudal part of the premotor and primary motor cortex pass through the corona radiata, internal capsule, and pons and finally project bilaterally to the motor neurons of cranial nerve nuclei V, VII, IX, X, XI, and XII. [bib_ref] McGraw Hill, Health Professional Division, Afifi [/bib_ref] Swallowing musculatures receive innervation from corticobulbar projections from both hemispheres, while at the same time, one hemisphere can control the bilateral swallowing nuclei of the brainstem.
Diffusion tensor imaging (DTI) can provide an assessment of the whole structural integrity of the corticobulbar tract [bib_ref] The anatomical location of the corticobulbar tract at the corona radiata in..., Jang [/bib_ref]. A recent study has shown that the tract integrity of the unaffected corticobulbar tract may indicate the severity of poststroke dysphagia [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref]. However, the retrospective study primarily compared tractography findings between patients with dysphagic stroke and agematched healthy subjects, without direct comparison with nondysphagic stroke patients. Most DTI studies on the corticobulbar tract (CBT) made comparisons between healthy controls and lacked a detailed comparison to stroke patients who do not manifest any evidence of dysphagia. Not all stroke patients manifest swallowing disturbances, and little is known about how CBT involvement in patients with dysphagia differs from its involvement in those who show no dysphagia after a supratentorial brain lesion. Likewise, little is known about how these DTI parameters relate to the degree of swallowing recovery.
In addition, past studies have reported the primary outcome of recovery based on the level of tube feeding. Instrumental swallowing tests, such as the videofluoroscopic swallowing study (VFSS), are gold standards for the comprehensive assessment of dysphagia and allow global rating to a scoring system. However, neuroanatomical correlations from DTI to recovery as assessed by standardized instrumental tests are still lacking. Research elucidating the precise relationship between neuroimaging analysis and instrumental swallowing tests is needed to apply these parameters to predicting dysphagia development and swallowing recovery after a supratentorial stroke.
In this prospective study, we compared DTI-derived parameters among 1) age-matched healthy subjects 2) stroke patients with dysphagia and 3) stroke patients without dysphagia to elucidate the role of the bilateral corticobulbar tracts on poststroke swallowing function and outcome. In those with dysphagia, swallowing parameters, including the modified barium swallow impairment profile (MBSImP©) based on the VFSS, were obtained within two weeks of stroke onset and at the 3-month follow-up. We aimed to determine which DTI parameters best reflect dysphagia development after supratentorial stroke and explore which parameters relate to swallowing recovery at 3 months poststroke.
# Methods
## Subjects
The participants were first-ever stroke patients who were admitted to the stroke center between March 2017 and February 2018. The inclusion criteria were as follows: (i) de novo ischemic stroke; (ii) infarction restricted to the supratentorial territory; and (iii) brain DTI performed within the first 2 weeks of stroke onset. The exclusion criteria were as follows: (i) history of prior ischemic or hemorrhagic stroke, (ii) bihemispheric stroke, and (iii) additional disorders that may cause swallowing impairment other than stroke, such as Parkinson's disease, Alzheimer's disease, Guillain-Barre syndrome, and myasthenia gravis.
This study was performed at a university affiliated hospital, where stroke patients are assessed for dysphagia presence upon admission. Stroke patients are referred for a VFSS or FEES if oral feeding is deemed unsafe and requires tube feeding or modification of diet or liquid consistency after testing the Gugging Swallowing Screen (GUSS). Patients with stroke manifesting dysphagia were investigated after validation with a VFSS performed within the first 2 weeks, which required dietary modification or tube feeding and availability for follow-up assessment at 3 months after stroke onset. Stroke patients who showed no evidence of dysphagia requiring dietary modification or treatment at the initial stroke onset were defined as the nondysphagia stroke group. They were required to have an initial GUSS of 20 assessed upon admission to the stroke unit. Stroke patients were enrolled by voluntary participation. They were included in the study if they fully consented to the study protocols and were available for all required assessments at baseline and at follow-up.
Healthy subjects were recruited prospectively for DTI after informed consent was obtained. They were excluded if they had (i) a history of psychopathology or neurological disorders ascertained via a health questionnaire and medical examination, (ii) structural abnormalities on their scan, or (iii) a prior history of swallowing defects. The institutional review board approved the study protocols, which were in accordance with the Declaration of Helsinki (HC17OESI0033).
## Assessment of swallowing and functional outcomes
All baseline swallowing assessments, including the VFSS, from the dysphagia stroke patients were evaluated within two days, ideally on the same day, of performing DTI. Swallowing function was evaluated using screening tools, including the GUSS [bib_ref] Aspiration and dysphagia screening in acute stroke -the Gugging Swallowing Screen revisited, Warnecke [/bib_ref] and the Mann Assessment of Swallowing Ability (MASA). The Penetration-Aspiration Scale (PAS) [bib_ref] A penetrationaspiration scale, Rosenbek [/bib_ref] and Modified Barium Swallow Impairment Profile (MBSImP©) [bib_ref] MBS measurement tool for swallow impairment-MBSImp: establishing a standard, Martin-Harris [/bib_ref] were assessed via VFSS performed by a certified specialist with more than 10 years of experience. VFSS was performed via digitalized fluoroscopy (AXIOM LUMINOS DRF; SIEMENS, Germany) on continuous setting with image capture settings of 30 frames per second. Thin/ thick fluid consisting of 35% v/v low/70% v/v high concentration liquid containing 300 mL of normal saline mixed with 140 g/100 mL of barium sulfate (Baritop HD; Kaigen Pharm, Osaka, Japan) [bib_ref] Dysphagia in Tongue Cancer Patients, Son [/bib_ref] was used along with semisolid and solid boluses according to standard protocols. We used a modified version of the MBSImP© protocols [bib_ref] MBS measurement tool for swallow impairment-MBSImp: establishing a standard, Martin-Harris [/bib_ref]. Patients were presented with one 5-ml teaspoon of nectar-thick liquid barium and one 5-ml teaspoon of pudding-thick barium. If patients were tolerable, larger volumes of thin liquid barium (two trials of 5-ml cup sip, followed by sequential swallows), nectar-thick liquid barium, honey-thick and pudding thick barium were presented. If allowed, rice or cookies coated with thick barium followed were presented.
While the Penetration-Aspiration Scale indicated the presence and severity of penetration or aspiration, the MBSImP© is a standardized and validated tool used to rate 17 different swallowing components that are scored in terms of 11 different swallowing tasks (such as teaspoon thin, sequential nectar and solid) during the VFSS [bib_ref] MBS measurement tool for swallow impairment-MBSImp: establishing a standard, Martin-Harris [/bib_ref] [bib_ref] Identification of swallowing tasks from a modified barium swallow study that optimize..., Hazelwood [/bib_ref]. The MBSImP© quantifies physiological swallow impairments, while the Functional Oral Intake Scale (FOIS) [bib_ref] Initial psychometric assessment of a functional oral intake scale for dysphagia in..., Crary [/bib_ref] describes the level of oral intake of liquid and food. If multiple swallows occurred, scoring was based on the initial swallow. The sum of the oral components was used as the oral total sum score (min 0, max 22), and the sum of the pharyngeal components was used as the pharyngeal total sum score (min 0, max 29) [bib_ref] MBS measurement tool for swallow impairment-MBSImp: establishing a standard, Martin-Harris [/bib_ref].
Baseline stroke severity was evaluated using the National Institutes of Health Stroke Severity Scale [bib_ref] Reliability of the national institutes of health stroke scale. extension to non-neurologists..., Goldstein [/bib_ref]. The Minimental State Exam [bib_ref] Modification of the mini-mental state examination for use in the elderly in..., Park [/bib_ref] and Berg balance scale [bib_ref] Measuring balance in the elderly: preliminary development of an instrument, Berg [/bib_ref] scores at baseline were assessed to evaluate baseline cognitive and balance impairment. The modified Rankin scale [bib_ref] Dysphagia, nutrition, and hydration in ischemic stroke patients at admission and discharge..., Burn [/bib_ref] and modified Barthel index (MBI) [bib_ref] Improving the sensitivity of the Barthel Index for stroke rehabilitation, Shah [/bib_ref] were assessed at baseline to evaluate the participants' mobility and ability to perform the activities of daily living.
## Diffusion tensor imaging and tractography
The patient underwent MRI by using a 3 T Philips MRI scanner (Phillips Health care, Best, The Netherlands) with a 32-channel head coil. Data were acquired in the form of single-shot spin-echo echo-planar images, with axial slices covering the whole brain across 75 interleaved slices of 2.0 mm thickness [no gap; repetition time/echo time = 1000/ 75 ms; field of view = 230.4 × 230.4 mm 2 ; matrix = 144 × 144; voxel size = 2 × 2 × 2 mm 3 (isotropic), number of excitations = 1]. Diffusionsensitizing gradients were applied in 32 noncollinear directions with a bvalue of 1,000 ms/mm 2 . The b = 0 images were scanned before acquisition of the diffusion-weighted images, with 33 volumes acquired in total.
## Image processing
The images were processed using the FMRIB Software Library (FSL; ver. 5.0.9; https://www.fmrib.ox.ac.uk/fsl). Source data were corrected for eddy currents and head motion by registration to the first b = 0 image using an affine transformation. Probability distributions in two fiber directions were modeled at each voxel for probabilistic tractography using the BEDPOSTX program, which is based on a multifiber diffusion model. The fractional anisotropy (FA) maps were prepared using the DTIFIT program.
Probabilistic tractography was performed using the FSL ProbtrackX program to reconstruct the corticobulbar tract. The fiber-tracking parameters were as follows: number of samples 5000; curvature threshold 0.1 (cosine 84.3 - ); and step length 0.5 mm [bib_ref] Oropharyngeal dysphagia in secondary normal pressure hydrocephalus due to corticobulbar tract compression:..., Jo [/bib_ref]. A tworegion of interest (ROI) approach was used to reconstruct the corticobulbar tract. The seed ROI was set on the frontal operculum and the caudal part of the premotor and primary motor cortex [bib_ref] Somatotopic organization of corticospinal/corticobulbar motor tracts in controls and patients with tumours:..., Hazzaa [/bib_ref] [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref] [bib_ref] Dysphagia prognosis prediction via corticobulbar tract assessment in lateral medullary infarction: a..., Jang [/bib_ref]. The way ROI was placed at the mid-pontine level, where the corticobulbar/corticospinal tract is represented by the blue fibers of the anterior pons on the color fractional anisotropy (FA) map. [fig_ref] Figure 1: Flow chart of participants with supratentorial stroke recruited for diffusion tractrography image... [/fig_ref]. The FA and tract volume (TV) of the corticobulbar tract in each hemisphere were determined [bib_ref] The anatomical location of the corticobulbar tract at the corona radiata in..., Jang [/bib_ref] [bib_ref] Identification of the corticobulbar tracts of the tongue and face using deterministic..., Jenabi [/bib_ref]. FA values were measured on the seed ROIs from the frontal operculum [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref]. The TV values of the corticobulbar tract were calculated by multiplying the voxel volume by the number of traced voxels during fiber tracking performed based on a robust minimum intensity threshold of 1 [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref]. The DTI of all participants was analyzed by an expert with 5 years of experience in tractography analysis, who was blinded to swallowing assessments. We overlaid the participants' lesion map with a probabilistic tract derived from diffusion tensor images of the patients.
# Statistical analysis
All statistical analyses were performed using R Statistical Software (version 2.15.3; R Foundation for Statistical Computing, Vienna, Austria). The Shapiro-Wilk test for normality was used to evaluate the distribution of the continuous variables. Between-group analyses were conducted using the Kruskal-Wallis test, followed by a post hoc Dunn test. P values were adjusted with Bonferroni correction. Spearman correlation analysis was used to assess the association between tractography parameters and swallowing. Within the dysphagia stroke group, a paired t test or Wilcoxon signed-rank sum test was conducted to determine significant differences in swallowing parameters obtained at baseline and at 3 months of follow-up. Finally multiple linear regression analysis was used to predict initial and follow-up swallowing function including the covariates of lesion size, affected and unaffected FA values, MMSE and initial NIHSS. Only variables with a p value <0.2 on simple analysis were included. The significance level was determined as P < 0.05. Continuous variables are expressed as the mean and standard deviation or median with interquartile range, and categorical variables are expressed as percentages. Optimal cutoff values were computed from the receiver operating characteristic (ROC) curve for CBT TV and FA values obtained with data regarding the presence of dysphagia.
# Results
## Participants
A total of 79 participants were enrolled in the study. [fig_ref] Figure 1: Flow chart of participants with supratentorial stroke recruited for diffusion tractrography image... [/fig_ref] shows a flowchart of the participants included [fig_ref] Figure 1: Flow chart of participants with supratentorial stroke recruited for diffusion tractrography image... [/fig_ref]. The baseline characteristics of the 3 groups along with the clinical characteristics of the two stroke groups showed that the dysphagia group showed greater neurological severity, as reflected by their initial National Institutes of Health Stroke Severity Scale and modified Rankin Scale scores [fig_ref] Table 1: Baseline demographics of enrolled participants [/fig_ref]. The dysphagia group showed a larger mean (±SD) total lesion volume of 21.7 (3.3) cm 3 than the nondysphagia stroke group (7.4 (1.2) cm 3 ), although statistical analysis tests failed to identify statistical significance (P = 0.067). Both groups showed similar degrees of deep white matter lesions as assessed by the Fazekas scale. The enrolled participants in the dysphagia group showed severe dysphagia, with the majority on tube feeding.
## Comparison of the dti-derived parameters among the three groups with threshold values
DTIs were acquired at a mean of 10.5 (SD = 4.25) days after stroke. All three groups showed significant differences (P < 0.001) in the TV of the affected corticobulbar tract, with the dysphagia stroke group showing the greatest reduction [fig_ref] Figure 2: Graphical illustration of the corticobulbar tract involvement in a [/fig_ref]. In contrast, on the unaffected side, the two stroke groups showed a smaller TV than the healthy control TV (<0.001), but no intergroup differences were observed between the two stroke groups (P = 0.224) .
ROC curve analysis showed that the CBT TV cutoff value of 4.10 cm 3 (area under the receiver operating characteristic curve (AUC) = 0.77; sensitivity 61.5%; specificity 86.0%) was associated with dysphagia.
When FAs were obtained from the affected frontal operculum, the dysphagia stroke group showed the greatest reduction among the three groups (P < 0.001). Additionally, the dysphagic stroke groups showed greater FA reductions from the unaffected frontal operculum than those of the healthy control group (P = 0.005). However, no significant differences in the FA values were observed between the two stroke groups .
ROC curve analysis revealed that the optimal cutoff value of FA of the affected side to classify dysphagia was 0.241 with modest accuracy levels but high specificity levels. (Sensitivity: 61.5%, specificity: 86.0%, AUC = 0.75 [0.63-0.87]) [fig_ref] Figure 5: ROC analysis show that cut-off values of the [/fig_ref]. Analysis of laterality showed no differences between the right and left lesions, although a tendency of intragroup differences between the affected and nonaffected CBT volume was shown, although it did not reach statistical significance (Supplementary [fig_ref] Table 1: Baseline demographics of enrolled participants [/fig_ref].
## Correlation with swallowing parameters with outcome prediction at 3 months
At baseline, a large proportion of patients in the dysphagia stroke group (65.4%, n = 17) were at near tube feeding status (FOIS level 1,2), with only 34.6% (n = 9) showing partial oral feeding ability. Significant improvement was found at follow-up, with 84.6% (n = 22) of patients on some form of oral feeding showing improvement in all swallowing parameters, including aspiration severity.
Total stroke volume was correlated with the initial total MASA scores (rho = − 0.57, P = 0.002) and aspiration severity (rho = 0.49, P = 0.012).
The TV of the corticobulbar tract failed to show any significant correlation with any swallowing parameters assessed at the 3-month follow-up. The FA values of the affected frontal operculum showed a significant correlation with the initial swallowing level as reflected by the FOIS scores (rho = 0.49, p = 0.011) but not the follow-up parameters. In contrast, the initial FA values of the unaffected frontal region correlated with the total MBSImP© values (rho = − 0.40, P = 0.02) at 3 months and with the oral and pharyngeal MBSImP© subscores (Supplementary [fig_ref] Table 2: Multiple regression model to predict FOIS and MASA score changes [/fig_ref].
# Multiple regression analysis
Variables of the affected FA from the frontal operculum, stroke lesion, NIHSS and MMSE accounted for 62.38% of the total variation of the initial FOIS level. Affected FA values (p = 0.045) and MMSE (p = 0.002) were significant predictors in this model after controlling lesion size and NIHSS (R 2 = 0.6238, adjR 2 = 0.5521, F = 8.7, p < 0.001).
Variables of the unaffected FA from the frontal operculum and initial NIHSS accounted for 58.09% of the total variation in the FOIS level at 3 months. A final model showed that the nonaffected FA values (p = 0.043) and NIHSS (p < 0 0.001) were significant predictors, after controlling for lesion size and MMSE (R 2 = 0.5809, ad jR 2 = 0.5444, F = 15.94, p < 0.001, AIC = 91.51) [fig_ref] Table 2: Multiple regression model to predict FOIS and MASA score changes [/fig_ref].
Additional multiple regression analysis was performed for the MASA, to be significant predictors in the MASA, PAS and MBSImP© at 3 months, a final model that included the unaffected FA(p = 0.006) accounted for 63.1% of the total variation for predicting the degree of improvement as reflected in the degree of MASA score changes between baseline and follow-up (R 2 = 0.6317, adj R 2 = 0.5815, F = 12.58, p < 0.001, AIC = 203.65).
## Follow-up dti parameters and correlation with outcome
For the three-month follow-up analysis, only data obtained from 74% of the dysphagic stroke patients (n = 20), who completed the second DTI on the same week the VFSS was performed, were used for analysis. The DTI parameters from baseline to follow-up were compared, but no significant differences were observed, although there was a tendency for the affected CBT volume to increase and the unaffected FA values to decrease (Supplementary . Unlike the initial FA values, the absolute FA values at the 3-month follow-up did not show any meaningful association with the follow-up swallowing outcomes (results not shown). However, a greater degree of FA value changes across follow-up showed a significant correlation with PAS (rho = − 0.46, P = 0.041) and oral component score of the MBSImP© (rho = 0.47, P = 0.035), indicating that greater reduction of the unaffected FA across the three months could lead to poor outcome. . Bar plot comparison of the corticobulbar tract volume among the three groups of the affected and non-affected sides showing that supratentorial stroke influences corticobulbar tract in group B and C but greatest involvement observed in group C. Abbreviations: A. healthy control, B. supratentorial stroke without dysphagia and C supratentorial stroke with dysphagia. ** P value < 0.01 ***P value < 0.001.
## Stroke lesion overlap with cbt
An overlapping density map of the stroke lesion and reconstructed corticobulbar tract is shown [fig_ref] Figure 6: Overlapping density map of the stroke lesion and reconstructed corticobulbar tract [/fig_ref]. The two stroke groups showed similar overlap of stroke lesions and CBT.
# Discussion
We analyzed the DTI-derived parameters of the bilateral corticobulbar tracts assessed within two weeks of stroke onset and explored whether these parameters were related to swallowing function in the . Bar plot comparison of the corticobulbar fractional anisotropy among the three groups of the affected and non-affected sides showing that supratentorial stroke influences corticobulbar tract in group B and C but greatest involvement observed in group C. Abbreviations: A. healthy control, B. supratentorial stroke without dysphagia and C supratentorial stroke with dysphagia. * P value < 0.05 ***P value < 0.0001. acute phase of stroke and to recovery at three months. Supratentorial infarctions affected CBT of both the affected and unaffected sides in all stroke patients, but the optimal cutoff value of CBT TV values associated with dysphagia was 4.10 cm 3 . Similarly, the FA reduction of the affected frontal operculum was associated with dysphagia presence and severity, with cutoff values of 0.24 associated with dysphagia. These two cutoff values showed good specificity levels. In contrast, the lower FA value of the unaffected frontal operculum was indicative of a poorer 3-month dysphagia outcome. Our results suggest that initial dysphagia presence depends on the involvement of the affected CBT TV. However, long-term swallowing recovery may depend on the integrity of the unaffected FA which showed early changes within two weeks of stroke onset. The greater involvement of the unaffected white matter integrity at three months predicted the degree of dysphagia recovery even after controlling for other confounding factors. The fact that the degree of CBT involvement and total stroke lesion volume determined the initial severity of dysphagia as manifested by the MASA scores and the aspiration severity as manifested by the PAS scores corresponds well with the results of past studies [bib_ref] Aspiration after stroke: Lesion analysis by brain MRI, Alberts [/bib_ref] [bib_ref] Lesion location predicts transient and extended risk of aspiration after supratentorial ischemic..., Galovic [/bib_ref] that indicated larger stroke volume to be associated with dysphagia and aspiration severity. These data are in line with recently published studies investigating those lesions that disrupt the CBT structural integrity as the primary mechanisms for impaired swallowing after stroke [bib_ref] Aspiration in patients with acute stroke, Daniels [/bib_ref] [bib_ref] Parallel cortical networks for volitional control of swallowing in humans, Mosier [/bib_ref]. However, because supratentorial stroke by itself may cause CBT TV reduction, without any evidence of dysphagia, precaution is warranted in interpreting any reduction in CBT TV compared to healthy controls as indicative of dysphagia. According to our results, a CBT TV higher than the cutoff value of 4.10 cm 3 could classify those without dysphagia with high specificity.
Interestingly, the FA of the unaffected side of the frontal operculum showed a positive correlation with a 3-month dysphagia outcome, suggesting the compensatory role of the contralesional corticobulbar tract in swallowing recovery after unilateral hemispheric stroke. Recent studies have also suggested the role of the unaffected hemisphere in poststroke swallowing recovery in patients with intracerebral hemorrhage and medullary infarctions. shown that injuries to the corticobulbar tract in both hemispheres prevented removal of the nasogastric tube until six months after hemorrhagic stroke and lateral medullary infarction [bib_ref] Prognostic prediction of dysphagia by analyzing the corticobulbar tract in the early..., Jang [/bib_ref] [bib_ref] Dysphagia prognosis prediction via corticobulbar tract assessment in lateral medullary infarction: a..., Jang [/bib_ref]. Similar results have been reported after middle cerebral artery infarction by Im et al., who reported that a lower corticobulbar tract volume in the unaffected hemisphere was indicative of worse swallowing function [bib_ref] Role of bilateral corticobulbar tracts in dysphagia after middle cerebral artery stroke, Im [/bib_ref]. Our findings are in line with the idea that swallowing function was dependent on a network connecting both cortical and subcortical lesions from both hemispheres with asymmetrical connectivity [bib_ref] Brain imaging correlates of recovered swallowing after dysphagic stroke: A fMRI and..., Mihai [/bib_ref] and that the cortical reorganization of the nondominant hemisphere plays a crucial role in the recovery of swallowing function [bib_ref] Recovery of swallowing after dysphagic stroke relates to functional reorganization in the..., Hamdy [/bib_ref].
Of note, the unaffected FA values were also reduced in the nondysphagic stroke group compared to the healthy control group. Some have suggested abnormal FA values of the contralateral stroke lesion to reflect generalized aberrant connectivity, diffuse dysregulation of neural dynamics or possible compensatory changes in response to the primary deficit (Alba-Ferrara and Erausquin, 2013). Such bilateral hemispheric reorganization can be observed as early as 12 days poststroke onset [bib_ref] Dynamics of language reorganization after stroke, Saur [/bib_ref]. Therefore, the reduced FA values of the contralesional side in all the stroke participants were expected. Based on our results, these early adaptive changes in the unaffected white matter integrity within two weeks of stroke seem crucial in dysphagia outcome. Additionally, more significant FA reductions of the unaffected frontal operculum, indicating poorer white matter integrity, were associated with poorer swallowing recovery, even after controlling for multiple confounding factors that may affect recovery. Therefore, the unaffected parameters are helpful as early biomarkers in predicting dysphagia outcome. More critically, they may help to guide strategies to maintain the adaptative role of the contralesional white matter, which would prove essential in enhancing swallowing recovery.
Unlike the corticospinal tract, a strong biomarker of motor recovery, which is characterized by competitive interhemispheric connections, bilateral corticobulbar tracts may act noncompetitively. Swallowing is unique from other motor activities in that it is a midline function with bilateral cortical representation [bib_ref] The cortical topography of human swallowing musculature in health and disease, Hamdy [/bib_ref] that is mediated by a complex bilateral neural network [bib_ref] Supratentorial regions of acute ischemia associated with clinically important swallowing disorders: a..., Gonzalez-Fernandez [/bib_ref].reported that cortical activation in the unaffected hemisphere during swallowing was a predictor of dysphagia in the subacute phase of stroke. Similarly, a functional MRI study showed that following unilateral hemispheric stroke, the region of the motor cortex involved in swallowing showed reorganization with a compensatory shift in the area activated during swallowing in the unaffected hemisphere [bib_ref] Functional magnetic resonance imaging study on dysphagia after unilateral hemispheric stroke: a..., Li [/bib_ref]. Based on these findings, studies on the target location for repetitive transcranial magnetic stimulation revealed that facilitatory high-frequency stimulation on the bilateral or unaffected hemisphere showed positive therapeutic effects . Our results are a natural extension of recent studies indicating that the unaffected hemisphere plays a role in swallowing modulation and neural plasticity and that swallowing recovery depends on an extended swallowing network [bib_ref] Parallel cortical networks for volitional control of swallowing in humans, Mosier [/bib_ref] involving bilateral cortices and subcortical connections.
Our results also support the role of the affected FA from the fontal operculum in predicting poststroke dysphagia, which showed a significant correlation with the initial level of functional swallowing. The importance of the microstructural organization of the frontal operculum is strongly supported by recent emerging evidence that indicates that this stroke location influences swallowing function [bib_ref] Mapping acute lesion locations to physiological swallow impairments after stroke, Wilmskoetter [/bib_ref]. The frontal and insular lobes were most commonly affected in patients with dysphagia persisting for 2 weeks [bib_ref] Predicting prolonged dysphagia in acute stroke: the Royal Adelaide Prognostic Index for..., Broadley [/bib_ref] , with the frontal operculum being the only ROI significantly associated with an extended risk of aspiration in the univariate analysis [bib_ref] Lesion location predicts transient and extended risk of aspiration after supratentorial ischemic..., Galovic [/bib_ref]. The insula plays a singular role in swallowing by contributing to the processing of food taste and texture [bib_ref] Functions of the anterior insula in taste, autonomic, and related functions, Rolls [/bib_ref] and controlling the timing and synchronization of swallowing motor events by integrating sensorimotor information [bib_ref] Parallel cortical networks for volitional control of swallowing in humans, Mosier [/bib_ref]. The frontal operculum is known to play a role in peri-infarct tissue recruitment after insular stroke, and parts of it in the zone of transition to the insular cortex may exhibit swallowing features similar to those of lesions in the insula [bib_ref] The role of the insular cortex in dysphagia, Daniels [/bib_ref].
An important characteristic of this study is that stroke patients with/ without dysphagia were enrolled consecutively. For the former, full serial VFSS based on MBSImP© scores with DTI metrics were collected. VFSS allows bolus flow visualization in relation to structural movement throughout the upper aerodigestive tract in real time [bib_ref] MBS measurement tool for swallow impairment-MBSImp: establishing a standard, Martin-Harris [/bib_ref]. Past CBT DTI studies have used the presence of tube feeding as the primary outcome parameter for swallowing recovery [bib_ref] Prognostic prediction of dysphagia by analyzing the corticobulbar tract in the early..., Jang [/bib_ref] [bib_ref] Dysphagia prognosis prediction via corticobulbar tract assessment in lateral medullary infarction: a..., Jang [/bib_ref]. However, successful tube removal after a stroke may depend on other clinical factors, such as age, nutritional status, or frailty, and may not fully reflect the degree of swallowing recovery. Thus, the outcome parameter used to determine swallowing recovery should be based on a parameter that is less ambiguous, more modality specific, standardized and validated to assess swallowing. The changes reported in our study are clinically relevant since the MBSImP© is a standardized semiobjective measure for evaluating oral and pharyngeal swallowing physiology. Although some studies have used this scoring system with lesion localization in poststroke dysphagia [bib_ref] Mapping acute lesion locations to physiological swallow impairments after stroke, Wilmskoetter [/bib_ref] , our study is the one of the few studies to monitor score changes and demonstrate positive associations with neuroanatomical correlates. In addition, our study provides longitudinal DTI follow-up data that were obtained in a prospective manner and indicates the positive association of swallowing recovery with the degree of white matter changes of the unaffected sides at three months postonset.
The study has a few limitations. First, the DTIs were obtained within two weeks of acute ischemic stroke, during which FA may be subject to confounding effects of tissue edema and acute injury and require additional time to reflect Wallerian degeneration [bib_ref] Corticospinal tract lesion load: An imaging biomarker for stroke motor outcomes, Feng [/bib_ref]. According to the literature, Wallerian degeneration can be detected histopathologically within the first week, [bib_ref] Quantitative Biochemical Studies of Wallerian Degeneration in the Peripheral and Central Nervous..., Mccaman [/bib_ref] and DTI can detect these early changes within the first two weeks even before they can be visualized in T2-weighted MRI [bib_ref] Diffusion tensor imaging detects early Wallerian degeneration of the pyramidal tract after..., Thomalla [/bib_ref] [bib_ref] Natural history and prognostic value of corticospinal tract Wallerian degeneration in intracerebral..., Venkatasubramanian [/bib_ref]. Therefore, we feel that the two-week time frame was adequate to reflect the early changes. Additionally, the reliability of using the DTI parameters after a stroke has been supported by previous studies [bib_ref] Detection and Predictive Value of Fractional Anisotropy Changes of the Corticospinal Tract..., Doughty [/bib_ref] [bib_ref] An overview of fractional anisotropy as a reliable quantitative measurement for the..., Zolkefley [/bib_ref]. Second, although a similar proportion of right or left lesions was observed in the two stroke groups, lateralization was not considered in this study. Lateralization involving left-side hemispheres has been shown to be associated with oral stage dysfunction, lateralization involving the right side has been shown to be associated with pharyngeal stages and aspiration [bib_ref] Swallowing after unilateral stroke of the cerebral cortex: preliminary experience, Robbins [/bib_ref] , and lateralization involving the right hemisphere stroke lesions has been shown to be associated with a higher rate of dysphagia [bib_ref] The impact of lesion location on dysphagia incidence, pattern and complications in..., Suntrup [/bib_ref]. However, the role of lateralization is still disputed and inconclusive [bib_ref] Lesion location predicts transient and extended risk of aspiration after supratentorial ischemic..., Galovic [/bib_ref] , and although our study showed no significant differences in all the parameters between the right and left side lesions, a tendency for right side dominancy was suggested. Third, although a previous study conducted voxel-based lesion mapping with each MBSImP© element separately [bib_ref] Mapping acute lesion locations to physiological swallow impairments after stroke, Wilmskoetter [/bib_ref] , our results presented the oral and pharyngeal subscores. Since the main objective of this study was to demonstrate overall improvement of swallowing rather than track changes in specific types of impairment, these subscores were selected for analysis. Fourth, the CBT TV of the affected side failed to show any association with the swallowing parameters at baseline. This finding may be related to the fact that all patients were already classified with severe dysphagia. Fifth, factors other than swallowing recovery may be influenced by clinical features such as age, degree of brain atrophy [bib_ref] MRI-based neuroanatomical predictors of dysphagia, dysarthria, and aphasia in patients with first..., Flowers [/bib_ref] , genetic polymorphism [bib_ref] Role of rs6265 BDNF polymorphisms and post-stroke dysphagia recovery-A prospective cohort study, Oh [/bib_ref] , sex, or type of therapy. Our findings are limited to brain lesions and structural integrity of the CBT and cannot be generalized to all clinical scenarios involving supratentorial infarction. Nevertheless, the results from the multiple regression analysis showed that even after controlling for NIHSS, stroke lesion size and MMSE; factors known to affect the degree of swallowing recovery; the nonaffected FA helped predict swallowing recovery. Additionally, the reconstructed TV may be influenced by the angle threshold or b-values, and anatomical differences may lead to high intersubject variability [bib_ref] Diffusion-based tractography atlas of the human acoustic radiation, Maffei [/bib_ref] ; therefore, the cutoff values provided in this study should be interpreted with caution. Further studies investigating the optimal TV under different threshold b values that best reflect clinical swallowing are warranted. The interaction of other clinical factors with tractography findings in swallowing recovery is a topic that requires further exploration in future studies. Finally, the sample size of the dysphagia stroke group may have been limited to reflect the influence of all the independent variables on the follow-up swallowing outcome parameters. Further validation of this imaging biomarker in another cohort with a large sample size is needed.
Despite the initial severity of dysphagia, most patients recover well within the first three months, but 15% of patients may have persistent dysphagia and increased mortality [bib_ref] Dysphagia, nutrition, and hydration in ischemic stroke patients at admission and discharge..., Burn [/bib_ref]. Conventional CT or MRI results have a limited role in predicting poor swallowing recovery. Our findings highlight that the microstructural organization of the unaffected CBT within the two weeks of stroke onset may help predict dysphagia outcomes. After supratentorial stroke, the degree of the TV of the affected CBT could classify those with dysphagia, but the FA values of the unaffected sides were associated with swallowing recovery. These parameters could be potentially relevant to developing neuroimaging biomarkers that predict those that may benefit from neuromodulation techniques that enhance neuroplasticity.
In summary, this prospective study demonstrates that, a supratentorial stroke may lead to changes in DTI parameters regardless of the presence of dysphagia. Our study indicates that patients may exhibit dysphagia when CBT parameters are involved past a certain cutoff value. However, despite severe CBT injury after a supratentorial stroke, the bihemispheric organization of the tract is crucial in swallowing recovery.
# Conclusion
Prolonged dysphagia may lead to increased poststroke mortality. Delineating potential biomarkers that allow early identification of those with severe dysphagia and protracted swallowing recovery is clinically relevant to prevent respiratory complications and prolonged morbidity. Analysis of CBT via tractography within the first two weeks may be considered a viable method for predicting poststroke swallowing recovery to tailor treatment to the individual patient.
## 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.
[fig] Figure 1: Flow chart of participants with supratentorial stroke recruited for diffusion tractrography image analysis. Abbreviation: DTI, diffusion tensor imaging. [/fig]
[fig] Figure 2: Graphical illustration of the corticobulbar tract involvement in a.healthy control, supratentorial stroke b. without dysphagia and c.with dysphagia. Abbreviations: FA, fractional anisotropy; TV, tract volume. [/fig]
[fig] Figure: 3. Bar plot comparison of the corticobulbar tract volume among the three groups of the affected and non-affected sides showing that supratentorial stroke influences corticobulbar tract in group B and C but greatest involvement observed in group C. Abbreviations: A. healthy control, B. supratentorial stroke without dysphagia and C supratentorial stroke with dysphagia. ** P value < 0.01 ***P value < 0.001. [/fig]
[table] Table 1: Baseline demographics of enrolled participants. [/table]
[table] Table 2: Multiple regression model to predict FOIS and MASA score changes. = 0.6317, adj R 2 = 0.5815, F = 12.58, p < 0.001, AIC = 203.65 Abbreviations: F, Frontal; FA, fractional anisotropy; FOIS, Functional Oral Intake Scale; MASA, Mann assessment of swallowing ability; NIHSS, National Institutes of Health Stroke Severity Scale; MMSE, Mini-mental state exam; AIC, Akaike Information Criterion. [/table]
|
Alternative splicing controls teneurin-latrophilin interaction and synapse specificity by a shape-shifting mechanism
The trans-synaptic interaction of the cell-adhesion molecules teneurins (TENs) with latrophilins (LPHNs/ADGRLs) promotes excitatory synapse formation when LPHNs simultaneously interact with FLRTs. Insertion of a short alternatively-spliced region within TENs abolishes the TEN-LPHN interaction and switches TEN function to specify inhibitory synapses. How alternative-splicing regulates TEN-LPHN interaction remains unclear. Here, we report the 2.9 Å resolution cryo-EM structure of the TEN2-LPHN3 complex, and describe the trimeric TEN2-LPHN3-FLRT3 complex. The structure reveals that the N-terminal lectin domain of LPHN3 binds to the TEN2 barrel at a site far away from the alternatively spliced region. Alternative-splicing regulates the TEN2-LPHN3 interaction by hindering access to the LPHN-binding surface rather than altering it. Strikingly, mutagenesis of the LPHN-binding surface of TEN2 abolishes the LPHN3 interaction and impairs excitatory but not inhibitory synapse formation. These results suggest that a multi-level coincident binding mechanism mediated by a cryptic adhesion complex between TENs and LPHNs regulates synapse specificity.
N eural circuit assembly and function in the central nervous system requires precise formation and specification of diverse excitatory and inhibitory synapse subtypes. Imbalances in the ratio of excitatory to inhibitory synapse function are thought to be a major component of brain disorders such as autism, mental retardation, and attention deficit hyperactivity disorder (ADHD) [bib_ref] Autism's cause may reside in abnormalities at the synapse, Garber [/bib_ref]. Recent work suggested that combinatorial sets of trans-synaptic interactions between cell-adhesion molecules, including teneurins (TENs or ODZs) and latrophilins (LPHNs or ADGRLs), mediate synapse formation and regulate the exquisite specification of synapses, but the underlying molecular mechanisms remain largely unexplored [bib_ref] Structural basis for teneurin function in circuit-wiring: a toxin motif at the..., Li [/bib_ref] [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref].
TENs and LPHNs are evolutionarily conserved cell-surface proteins. While the roles of TENs and LPHNs in early organisms remain unclear, they have critical roles in embryonic development and brain wiring in higher eukaryotes. TENs (TEN1-4 in mammals) are large type-II transmembrane proteins that are composed of an N-terminal cytoplasmic sequence, a single transmembrane region, and a large extracellular region (ECR) composed of >2000 amino acids with partial homology to bacterial Tc toxins [bib_ref] Drosophila gene related to tenascin, shows selective transcript localization, Baumgartner [/bib_ref] [bib_ref] Chiquet-Ehrismann R. Tenm, a Drosophila gene related to tenascin, is a new..., Baumgartner [/bib_ref] [bib_ref] Odd Oz: a novel Drosophila pair rule gene, Levine [/bib_ref] [bib_ref] Mouse ten-m/Odz is a new family of dimeric type II transmembrane proteins..., Oohashi [/bib_ref]. They form constitutive cis-dimers via highly conserved disulfide bonds formed in proximity to their transmembrane helix [bib_ref] Mouse ten-m/Odz is a new family of dimeric type II transmembrane proteins..., Oohashi [/bib_ref] [bib_ref] All four members of the Ten-m/Odz family of transmembrane proteins form dimers, Feng [/bib_ref] [bib_ref] The mechanism of regulated release of Lasso/Teneurin-2, Vysokov [/bib_ref]. TENs have central roles in tissue polarity, embryogenesis, heart development, axon guidance, and synapse formation [bib_ref] Odd Oz: a novel Drosophila pair rule gene, Levine [/bib_ref] [bib_ref] Latrophilin-2 is a novel component of the epithelialmesenchymal transition within the atrioventricular..., Doyle [/bib_ref] [bib_ref] Teneurins instruct synaptic partner matching in an olfactory map, Hong [/bib_ref] [bib_ref] Trans-synaptic teneurin signalling in neuromuscular synapse organization and target choice, Mosca [/bib_ref] [bib_ref] Teneurin-3 controls topographic circuit assembly in the hippocampus, Berns [/bib_ref] [bib_ref] Latrophilin 1 and its endogenous ligand Lasso/Teneurin-2 form a high-affinity transsynaptic receptor..., Silva [/bib_ref] [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref] [bib_ref] Teneurins, TCAP, and latrophilins: roles in the etiology of mood disorders, Woelfle [/bib_ref] ; and are linked to various diseases including neurological disorders, developmental problems, various cancers, and congenital general anosmia [bib_ref] Homozygous null mutation in ODZ3 causes microphthalmia in humans, Aldahmesh [/bib_ref] [bib_ref] A role for TENM1 mutations in congenital general anosmia, Alkelai [/bib_ref] [bib_ref] Missense mutations in TENM4, a regulator of axon guidance and central myelination,..., Hor [/bib_ref] [bib_ref] Teneurin protein family: an emerging role in human tumorigenesis and drug resistance, Ziegler [/bib_ref] [bib_ref] Confirmation of TENM3 involvement in autosomal recessive colobomatous microphthalmia, Chassaing [/bib_ref] [bib_ref] ODZ1 allows glioblastoma to sustain invasiveness through a Myc-dependent transcriptional upregulation of..., Talamillo [/bib_ref] [bib_ref] Expression of teneurins is associated with tumor differentiation and patient survival in..., Graumann [/bib_ref]. LPHNs (LPHN1-3 in mammals) belong to the adhesion-type G-protein coupled receptor (GPCR) family [bib_ref] The calcium-independent receptor of alphalatrotoxin is not a neurexin, Krasnoperov [/bib_ref] [bib_ref] Alpha-latrotoxin receptor, latrophilin, is a novel member of the secretin family of..., Lelianova [/bib_ref] [bib_ref] alpha-Latrotoxin receptor CIRL/latrophilin 1 (CL1) defines an unusual family of ubiquitous Gprotein-linked..., Sugita [/bib_ref] and have a large N-terminal ECR (>800 amino acids) in addition to their signaling seven-pass transmembrane domain and cytoplasmic tail [bib_ref] Alpha-latrotoxin receptor, latrophilin, is a novel member of the secretin family of..., Lelianova [/bib_ref] [bib_ref] alpha-Latrotoxin stimulates exocytosis by the interaction with a neuronal G-protein-coupled receptor, Krasnoperov [/bib_ref] [bib_ref] Neurexins are functional alphalatrotoxin receptors, Sugita [/bib_ref] [bib_ref] alpha-Latrotoxin and its receptors: neurexins and CIRL/ latrophilins, Sudhof [/bib_ref] [bib_ref] Alpha-latrotoxin stimulates a novel pathway of Ca2 +-dependent synaptic exocytosis independent of..., Deak [/bib_ref]. LPHNs have roles in embryogenesis, tissue polarity, synaptic development, and neural circuit connectivity, interestingly almost identical to the functions of TENs [bib_ref] Latrophilin 1 and its endogenous ligand Lasso/Teneurin-2 form a high-affinity transsynaptic receptor..., Silva [/bib_ref] [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref] [bib_ref] Alpha-latrotoxin receptor, latrophilin, is a novel member of the secretin family of..., Lelianova [/bib_ref] [bib_ref] alpha-Latrotoxin stimulates exocytosis by the interaction with a neuronal G-protein-coupled receptor, Krasnoperov [/bib_ref] [bib_ref] Neurexins are functional alphalatrotoxin receptors, Sugita [/bib_ref] [bib_ref] alpha-Latrotoxin and its receptors: neurexins and CIRL/ latrophilins, Sudhof [/bib_ref] [bib_ref] Alpha-latrotoxin stimulates a novel pathway of Ca2 +-dependent synaptic exocytosis independent of..., Deak [/bib_ref] [bib_ref] Postsynaptic adhesion GPCR latrophilin-2 mediates target recognition in entorhinal-hippocampal synapse assembly, Anderson [/bib_ref] [bib_ref] International union of basic and clinical pharmacology. XCIV. Adhesion G protein-coupled receptors, Hamann [/bib_ref]. LPHN3 mutations are linked to ADHD, as well as numerous cancers in humans [bib_ref] Diverse somatic mutation patterns and pathway alterations in human cancers, Kan [/bib_ref] [bib_ref] The emerging mutational landscape of G proteins and Gprotein-coupled receptors in cancer, O'hayre [/bib_ref] [bib_ref] A common variant of the latrophilin 3 gene, LPHN3, confers susceptibility to..., Arcos-Burgos [/bib_ref].
The large ECRs of TENs and LPHNs form a tight trans-cellular adhesion complex [bib_ref] Odd Oz: a novel Drosophila pair rule gene, Levine [/bib_ref] [bib_ref] Latrophilin 1 and its endogenous ligand Lasso/Teneurin-2 form a high-affinity transsynaptic receptor..., Silva [/bib_ref] [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref]. The most N-terminal Lectin (Lec) and Olfactomedin (Olf) domains of LPHN interact with TEN2, with the Lec domain contributing most of the binding affinity . A four or five amino acid splice insert (MEQK or KVEQK) between these domains decreases the affinity of the TEN/LPHN interaction [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref]. In addition to TENs, LPHNs form trans-cellular interactions also with homodimeric cell-adhesion molecules called fibronectin leucine rich repeat transmembrane proteins (FLRTs), which further interact with Uncoordinated5 (UNC5s) [bib_ref] alpha-Latrotoxin receptor CIRL/latrophilin 1 (CL1) defines an unusual family of ubiquitous Gprotein-linked..., Sugita [/bib_ref] [bib_ref] FLRT proteins are endogenous latrophilin ligands and regulate excitatory synapse development, O'sullivan [/bib_ref]. The LPHN3/TEN2 interaction as well as the LPHN3/FLRT3 interaction were individually reported to be important for synapse formation and organization [bib_ref] Latrophilin 1 and its endogenous ligand Lasso/Teneurin-2 form a high-affinity transsynaptic receptor..., Silva [/bib_ref] [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref] [bib_ref] FLRT proteins are endogenous latrophilin ligands and regulate excitatory synapse development, O'sullivan [/bib_ref]. Recent work showed that in transgenic mice in vivo, postsynaptic LPHN3 promotes excitatory synapse formation by simultaneously binding to TEN and FLRT, two unrelated presynaptic ligands, which is required for formation of synaptic inputs at specific dendritic localizations [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref]. Conversely, LPHN3 deletion had no effect on inhibitory synapse formation [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref]. The precise molecular mechanisms of neither excitatory nor inhibitory synapse formation are known.
In addition to the critical role of coincident TEN2 and FLRT3 binding to LPHN3 for specification of excitatory synapses, alternative splicing of TEN2 also plays a crucial role in specifying excitatory vs. inhibitory synapses 2 . An alternatively spliced sevenresidue region (NKEFKHS) within TEN2 acts as a switch to regulate trans-cellular adhesion of TEN2 with LPHNs and to induce different types of synapses in vitro 2 . The TEN2 −SS splice variant that lacks the splice insert can bind to LPHN3 in trans , left side) 2 . However, TEN2 +SS, the splice variant that includes the seven amino acids is unable to interact with LPHN3 in trans in identical experiments [bib_ref] Structural basis for teneurin function in circuit-wiring: a toxin motif at the..., Li [/bib_ref]. Similarly, the same alternatively spliced site also may regulate TEN2 transhomodimerization [bib_ref] Teneurin-3 controls topographic circuit assembly in the hippocampus, Berns [/bib_ref] , although no such trans-homodimerization could be detected in some assays [bib_ref] Latrophilins function as heterophilic cell-adhesion molecules by binding to teneurins: regulation by..., Boucard [/bib_ref]. However, the molecular mechanism of how alternative splicing regulates ligand interactions is unclear.
In agreement with in vivo transgenic mice experiments, only the splice variant of TEN2 that can interact with LPHN3 (TEN2 −SS) was able to promote excitatory synapses when co-expressed with FLRT3 in cultured neurons in vitro [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref]. The other TEN2 splice variant that cannot interact with LPHN3 (TEN2 +SS) did not promote excitatory synapse formation when expressed alone or co-expressed with FLRT3. Instead, this TEN2 isoform induced inhibitory postsynaptic specifications in a LPHN-independent manner likely by interacting with other unknown ligands, suggesting that alternative splicing of TEN2 regulates excitatory vs. inhibitory synapse specification 2 . These results indicate a multilevel coincidence signaling mechanism for the specification of synaptic connections that requires the presence of the proper combination of molecules and their appropriate alternatively spliced isoforms to colocalize in order to induce the formation of a specific type of synapse. However, the molecular details of the TEN/LPHN/FLRT interaction are not known. Furthermore, the structural basis for the lack of the TEN2/LPHN3 trans interaction in the presence of a short splice insert in TEN2 is unclear.
Here, we have determined the 2.9 Å resolution cryo-EM structure of the TEN2/LPHN3 complex and described the direct and simultaneous interaction of LPHN3 with both TEN2 and FLRT3. The TEN2/LPHN3 complex structure revealed that the N-terminal Lec domain of LPHN3 interacts with the β-barrel domain of TEN2. Both the Lec and the preceding Olf domains of LPHN3 face away from the alternatively spliced site within the βpropeller domain of TEN2, indeed providing no explanation for how the short splice insert may regulate TEN2/LPHN3 interaction. Using a series of experimental setups that mimic either trans-cellular interactions between opposing cell-membranes, or cis-like-interactions in solution, we showed that alternative splicing of TEN2 indirectly regulates the TEN2/LPHN3 interaction by altering the accessibility of the LPHN-binding site on TEN2 with the help of membranes, rather than directly interfering with the LPHN-binding site. Mutagenesis of the LPHN-binding site on TEN2 abolished the TEN2/LPHN3 interaction and had a severe and specific effect on excitatory synapse formation, but had no effect on inhibitory synapse formation. These results provide a molecular and mechanistic understanding of the multi-level coincidence binding mechanism that mediates specificity in synapse formation and circuit-wiring.
# Results
Structure of the TEN2/LPHN3 complex. To determine the structure of the TEN2/LPHN3 complex, the ECR of human TEN2 lacking the EGF repeats that are responsible for cisdimerization (TEN2 −SS ECRΔ1, encoding residues 727-2648, and the full ECR of human LPHN3 (LPHN3 +SS ECR, encoding residues S21-V866, were co-expressed. The complex structure was determined by single-particle cryo-EM. After multiple rounds of 3D classification, two cryo-EM maps were obtained: one corresponding to the monomeric TEN2 −SS ECRΔ1 in complex with LPHN3 ECR at a nominal resolution of 2.9 Å (from 9.7% of particles, [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref] , and the other corresponding to the monomeric TEN2 ECRΔ1 with a better resolved β-propeller domain at a nominal resolution of 3.0 Å (from 3.8% of particles). A near-atomic resolution model of the protein complex was built using the available TEN2 structure (PDB: 6CMX) and the Lec Olf structure (PDB:5AFB) , Supplementary Figs. 1-4, [fig_ref] Table 1: Cryo-EM data collection, refinement and validation statistics [/fig_ref].
Our TEN2/LPHN3 complex structure comprises a heterodimer of~100 × 50 × 115 Å in which the Lec domain of LPHN3 interacts with the side of the barrel domain of TEN2 .
The TEN2 ECR is assembled as a large cylindrical barrel sealed by the β-propeller and Ig-like domains at the bottom; and the toxin-like domain protrudes from and attaches to the side of the barrel as previously reported [bib_ref] Structural basis for teneurin function in circuit-wiring: a toxin motif at the..., Li [/bib_ref] [bib_ref] Structures of teneurin adhesion receptors reveal an ancient fold for cell-cell interaction, Jackson [/bib_ref]. The Lec domain of LPHN3 and the toxin-like domain of TEN2 bind to opposite faces of the β- Tox-like (5) Ig-like (2) β-propeller ( ARTICLE barrel in a seemingly parallel orientation to each other .
No major conformational changes are observed when the complex structure is compared with the individual structures of TEN2 or LPHN3. Analysis of the cryo-EM maps at a lower threshold also revealed continuous density for the Olf domain that extended from the Lec domain towards the opposite side from the βpropeller domain of TEN2 [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. In spite of the lower resolution, it was possible to fit the available LPHN3 Olf domain structure in this density [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. The Olf domain is positioned in close proximity to the top of the TEN2 β-barrel, although it is not in contact with TEN2 [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. The presence of the splice insert (KVEQK) between the Lec and OLF domains in our LPHN3 construct likely causes the lack of interaction of Olf domain with TEN. The remaining C-terminal domains of LPHN3 for5 which there is no EM density likely extend from the opposite side, away from the TEN2 TM domain located at the TEN2 N-terminus. This orientation positions the membrane-anchored TM domain of LPHN3 on the opposite side from the membrane-anchored TM domain of TEN, and thus is compatible with a trans-cellular interaction of TEN with LPHN [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref].
LPHN3, TEN2, and FLRT3 form a trimeric complex. LPHN proteins are involved in heterodimeric interactions with TENs and FLRTs and coincidence binding of both FLRTs and TENs is required for excitatory synapse formation . Additionally, FLRTs interact with UNC5s and form homodimers that are incompatible with their UNC5 binding. However, whether these interactions are compatible is unclear. Thus, we investigated whether the TEN2, LPHN3, and FLRT3 interactions are compatible with each other, or in other words, whether TEN2, LPHN3, and FLRT3 can form a trimeric complex. The availability of the LPHN3/FLRT3 complex structures and of the LPHN3 Lec-Olf structure enabled us to compare structures, and to predict and test the compatibility of the possible interactions of TEN, LPHN3, and FLRT3 39 . Intriguingly, superimposition of the Lec domain from the LPHN3/FLRT3 structure with the Lec domain from the LPHN3/TEN2 complex structure showed that FLRT3 and TEN2 bind to distinct domains on LPHN3 and that there are no clashes between TEN2 and FLRT3, suggesting that LPHN3 can simultaneously bind to TEN2 and FLRT3 [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref].
In order to test whether this model is correct, we co-expressed FLRT3 LRR, LPHN3 full ECR, and TEN2 −SS full ECR, purified the complex by gel filtration chromatography and analyzed the fractions by SDS-PAGE [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. Both LPHN3 ECR and FLRT3 LRR elution volumes shifted to the left as compared to the elution volumes of the individual proteins. All three proteins eluted in the same fractions, indicating the formation of a trimeric TEN2/ LPHN3/FLRT3 complex [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. These results suggest that TEN2 and FLRT3, both ligands of LPHN3, can simultaneously bind to LPHN3 and form a trimeric complex in vitro, supporting the in vivo observations that coincident binding of both TEN2 and FLRT3 to LPHN3 is required for excitatory synapse formation [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref].
The binding interface of the TEN2/LPHN3 complex is conserved. To visualize conserved and variable regions of the TEN2 βbarrel and the LPHN3 Lec/Olf domains, we mapped the conservation of residues on the TEN2/LPHN3 complex structure, and colored residues from most conserved (magenta) to least conserved (cyan) [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. The interaction surfaces of TENs and LPHNs correspond to one of the most conserved regions (yellow ovals in [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. As the TEN2 β-barrel is homologous to bacterial Tc toxins, we also analyzed the conservation between bacterial toxins by mapping the conservation of residues between bacterial toxins on the homologous bacterial TcC toxin structure (PDB ID: 4O9X) [bib_ref] Mechanism of Tc toxin action revealed in molecular detail, Meusch [/bib_ref] and observed that the identical surface of bacterial Tc toxins is not conserved [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. Our cryo-EM map thus revealed that LPHN3 binds to a highly conserved surface on the barrel of TEN2 that likely evolved to bind to LPHNs after diverging from bacterial toxins.
The Lec domain of LPHN3 belongs to the sea urchin egg lectin (SUEL) related Lec family. It adopts a kidney shape with dimensions of 20 Å × 20 Å × 50 Å, and is composed of five βstrands and a single alpha helix, interconnected by four conserved disulfide bonds [bib_ref] Solution structure and sugar-binding mechanism of mouse latrophilin-1 RBL: a 7TM receptor-attached..., Vakonakis [/bib_ref]. In our map, the Lec domain was not as well resolved as TEN2 (Supplementary Figs. 1d, 4c). Therefore, the crystal structure of the Lec domain was docked as a rigid body without fitting the side chains. The docking of the complementary surfaces of the β-sandwich of the LPHN3 Lec domain to the concave surface of the TEN2 β-barrel creates an average interface area of 690 Å 2 . The high affinity of the TEN2/LPHN3 complex is achieved by a combination of tentative interactions, comprised of salt bridges, hydrogen bonds, and long-range electrostatic interactions [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. Notably, salt bridges located at the top, middle and the bottom of the interface stabilize the interaction [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. The extensive network of salt bridges likely helps achieve the high affinity of the TEN2/LPHN3 complex. In addition, two residues on TEN2 (D1737 and H1738) play important roles by interacting with a disulfide bond (C36, C66) and D67 of the Lec domain [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. Interestingly, the cryo-EM map showed clear density of the Lec domain interacting with a glycan originating from glycosylation at N1681 [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. The glycan inserts into a well conserved sugar-binding pocket of the SUEL-related Lec domain of LPHN3 [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref] , suggesting that in contrast to previous notions [bib_ref] Solution structure and sugar-binding mechanism of mouse latrophilin-1 RBL: a 7TM receptor-attached..., Vakonakis [/bib_ref] , the Lec domain of LPHN3 may still be able to bind carbohydrates.
In order to specifically abolish the interaction of TEN2 with LPHN3, and to confirm the validity of the binding interface that we observed in the TEN2/LPHN3 complex structure, we designed mutations on full-length TEN2 that change only a few atoms on the protein surface. Several TEN2 mutations were designed, including the DHR (D1737N, H1738T, R1739T) mutation that alters residues at the LPHN3 Lec domain binding interface [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref]. To ensure that the mutant proteins are properly folded, we first examined the expression levels and surface transport of all TEN2 mutants, and exclusively used mutants that had no localization problems [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. Nonpermeabilized HEK293T cells transfected with TEN2 constructs were stained with an antibody suitable to react with an extracellular tag on the proteins, and the amount of surface-exposed TEN2 was assessed TEN2 and LPHN3 are colored cyan and orange, respectively. FLRT3 molecules in the FLRT3 dimer are colored magenta and gray. The second LPHN3 molecule that might be bound to the FLRT3 dimer is not shown. An UNC5 molecule that might be bound to the magenta FLRT3 is not shown. e SD200 size-exclusion chromatography profile of the TEN2/LPHN3/FLRT3 complex (Blue line) as compared to the profiles of individual proteins (dashed gray lines) showing that LPHN3 binds to TEN2 and FLRT3, simultaneously. Size-exclusion fractions are run on an SDS-PAGE gel. Triple complex is indicated by blue arrow. Source data are provided as a Source Data file.
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-16029-7 ARTICLE NATURE COMMUNICATIONS | (2020) 11:2140 | https://doi.org/10.1038/s41467-020-16029-7 | www.nature.com/naturecommunications by indirect immunofluorescence. Importantly, the DHR mutant was properly folded and trafficked to the cell-surface. Binding experiments showed that the DHR mutant does not have the ability to interact with LPHN3. The binding experiment results are discussed in detail below within the context of geometrical restraints that act on the TEN2/LPHN3 interaction (see below).
LPHN binding site is away from the splice site on TEN. A seven amino acid alternatively spliced site on the β-propeller domain of TEN2 regulates the TEN2/LPHN3 interaction, and, consequently, excitatory vs. inhibitory synapse formation 2 . A striking observation from the TEN2/LPHN3 complex structure was that the LPHN3 binding site on TEN2 is located distal to the alternatively spliced sequence in the TEN2 β-propeller (Figs. 1d-f, 2c, d). This observation is very surprising because in other protein-ligand interactions regulated by alternative splicing, the alternatively spliced sequence is located at the ligand-binding interface [bib_ref] Structures of neuroligin-1 and the neuroligin-1/neurexin-1 beta complex reveal specific protein-protein and..., Arac [/bib_ref] [bib_ref] Structures of neurexophilin-neurexin complexes reveal a regulatory mechanism of alternative splicing, Wilson [/bib_ref]. Thus, in the case of the TEN2/LPHN3 interaction, alternative splicing regulates this interaction remotely in a manner that was previously not described.
We hypothesized that the membranes of two opposing cells may impose a docking geometry on TEN2 and LPHN3 that is critical for their trans-cellular adhesion in the extracellular space between the membranes, and that alternative splicing might control the docking geometry of TEN2. Consequently, with an altered geometry, LPHN3 may not be able to access its binding site on TEN2. This hypothesis suggests that insertion or deletion of the alternatively spliced sequence does not affect the LPHN3 binding surface on TEN2. It also suggests that when one or more membranes are removed from the experimental system, LPHN3 and TEN2 would interact with each other independent of alternative splicing because they will not be restricted by the membranes to which they are anchored. In order to test this hypothesis, we designed various experimental setups in which the restraints that act on the docking geometries of TEN2 and LPHN3 vary from high to low [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref].
First, we conducted cell-aggregation assays with HEK293 cells in which a population of HEK cells expressing full-length TEN2 are mixed with a different population of HEK cells expressing full-length LPHN3, and cell aggregation is monitored as a function of TEN2/LPHN3 interaction [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref]. These experiments mimic trans-cellular interaction as they detect the binding of two full-length proteins that are anchored on opposing cell membranes (referred to as trans hereon). Cell-aggregation experiments apply high restraints on the docking geometries of the proteins because both proteins can diffuse only laterally in two dimensions within the plane of the membrane bilayer. The cell-aggregation experiments are the best imitation for the in vivo interaction of TEN2 and LPHN3, where full-length proteins are on the cell surfaces of neighboring cells during development or synapse formation. Second, we used flow cytometry experiments and cell-surface staining experiments in which the binding of a soluble protein to a membrane-anchored protein is tested (referred to as cis-like hereon, although it should not be confused with commonly used meanings of cis) [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref]. Soluble fragments of the LPHN3 ECR were tested for their ability to bind to HEK293T cells expressing full-length TEN2 on the cell surface. These experiments apply intermediate restraints on the docking geometries of the proteins because the membrane-anchored TEN2 can diffuse only laterally in two dimensions within the plane of membrane bilayer, while the soluble LPHN3 ECR fragment can freely diffuse in three dimensions in solution [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref]. Third, we used size-exclusion chromatography in which the binding of two soluble proteins experience no restraints. Here, binding of the ECRs of TEN2 and LPHN3 is tested in the absence of any membranes (referred to as cis-like as well) [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref]. Importantly, any intrinsic restraints that may be originating from the intrinsic conformation of the proteins may still act on any of the above experiments. We used a combination of these experimental setups to investigate the effect of various restraints on the ability of TEN2 to interact with LPHN3. We examined two sets of TEN2 constructs in these experimental setups: (i) TEN2 −SS carrying LPHN3 binding site mutations (TEN2 −SS DHR) was compared to WT TEN2 −SS to observe the effect of LPHN-binding site mutations on TEN2/ LPHN3 interaction [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. We expect that this mutant should abolish TEN2/LPHN3 interaction in all experimental setups because it directly disrupts the LPHN-binding site on TEN. (ii) WT TEN2 +SS was compared to WT TEN2 −SS to observe the effect of inclusion of the splice insert on TEN2/LPHN3 interaction . We expect that, if the inclusion of the alternative splice insert is disrupting the binding interface on TEN2 for LPHN3, then TEN2 +SS variant should not bind LPHN3 in any of the experimental setups. However, if the insertion of the alternative splice insert is acting by a different mechanism, such as changing the docking geometry of TEN2 onto LPHN3, then, TEN2 +SS isoform may bind LPHN3 in cislike setups where binding restraints on TEN2 and LPHN3 are relaxed [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref].
The first set of experiments testing the effect of point mutations on the LPHN3 binding surface showed that TEN2 −SS DHR mutant was unable to bind to LPHN3 in all experimental setups, including cell-aggregation experiments [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref] , flow cytometry experiments [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref] , cell-surface staining experiments [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref] , right) and gel filtration experiments [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. These results show that this mutation destroys the binding interface on TEN2 for LPHN3 and abolishes complex formation in trans and cis-like [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. The second set of experiments testing the effect of the alternatively spliced sequence of TEN2 on LPHN3 binding, however, displayed differential effects in trans and cis-iike experimental setups . In cell-aggregation experiments, full-length TEN2 lacking the β-propeller splice insert robustly induced trans-cellular aggregation with full-length LPHN3 . Intriguingly, as we previously showed, inclusion of the seven amino-acid splice insert in the β-propeller eliminated trans-cellular adhesion with LPHN3 . In flow cytometry experiments, however, the affinity of TEN2 for the soluble Lec domain (that is unattached to the membranes) was not affected by inserting the seven-residue segment . Cell-surface staining experiments also showed that the soluble Lec domain of LPHN3 binds to both TEN2 splice isoforms, confirming the flow cytometry experiments . Finally, gel filtration chromatography experiments performed with the full-length TEN2 and LPHN3 ECRs as soluble proteins unattached to membranes showed that when no restraints are applied on TEN2 and LPHN3, both TEN2 splice isoforms robustly interacted with LPHN3 . Thus, we suggest that trans-cellular TEN2-LPHN3 interactions are regulated by β-propeller alternative splicing likely due to conformational restrains in the context of full-length proteins.
Binding mutants abolish excitatory synapse formation. Previous work showed that the splice isoforms of TEN2 (TEN2 +SS and TEN2 −SS) induce different synaptic specifications in artificial synapse formation assays. In these assays, HEK293 cells expressing TEN2 variants were co-cultured with primary neurons; and inhibitory and excitatory synapse formation was monitored for pre-and postsynaptic differentiation for both types of synapses 2 . The results showed that TEN2 +SS induced GABAergic (inhibitory) postsynaptic specializations but failed to induce glutamatergic (excitatory) postsynaptic specifications 2 . On the other hand, initially, TEN2 −SS failed to recruit both excitatory and inhibitory synaptic markers 2 . However, when FLRT3,
[formula] Normalized UV absorbance (%) [/formula]
Elution volume (mL) [bib_ref] Latrophilin 1 and its endogenous ligand Lasso/Teneurin-2 form a high-affinity transsynaptic receptor..., Silva [/bib_ref] another LPHN3 ligand that on its own is also unable to induce pre-or postsynaptic specializations, was co-expressed in HEK293 cells with the TEN2 −SS, these molecules together potently induced excitatory but not inhibitory postsynaptic specializations 3 .
As only the TEN2 −SS isoform is capable of interacting with LPHN3 in a trans-configuration, we speculated that the DHR mutation that abolishes the interaction of TEN2 with LPHN3 by demolishing the binding site should affect excitatory synapse formation, but should not impair inhibitory synapse formation that is mediated by the TEN2 +SS isoform because inhibitory synapse formation is independent of the TEN2/LPHN3 interaction. To test this hypothesis, we engineered the DHR mutation on both the full-length TEN2 −SS and the TEN2 +SS isoforms, and tested its effect on the induction of either excitatory (−SS variant) or inhibitory (+SS variant) postsynaptic specializations in the artificial synapse formation assay [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref]. As previously shown, WT TEN2 −SS induced excitatory postsynaptic specializations when co-expressed with FLRT3 [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref] ; and WT TEN2 +SS induced inhibitory postsynaptic specializations [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref] [bib_ref] Structural basis for teneurin function in circuit-wiring: a toxin motif at the..., Li [/bib_ref] [bib_ref] Latrophilin GPCRs direct synapse specificity by coincident binding of FLRTs and teneurins, Sando [/bib_ref]. We observed that the DHR mutant attenuated the formation of excitatory synapses when compared with wild-type TEN2 −SS [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref]. However, the same mutant triggered inhibitory postsynaptic specializations similar to that of the wild-type TEN2 +SS [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref] , as predicted. These results indicate that LPHN3 binding mediates the excitatory synapse formation of TEN2 −SS, whereas binding of LPHN3 to TEN2 is not involved in inhibitory synapse formation.
# Discussion
Teneurins and latrophilins are multifunctional transmembrane proteins that perform important biological roles via trans-cellular interactions. The function of LPHN3 in excitatory synapse formation requires simultaneous binding of LPHN3 to both TENs and FLRTs, suggesting that a coincidence signaling mechanism mediates specificity of synaptic connections. Synaptic specificity is further regulated by alternative splicing of TEN2 because only the LPHN3-binding splice variant of TEN2 can induce excitatory synapses, but not the other variant that induces inhibitory synapses likely in a LPHN3-independent manner. A molecular understanding of the TEN2-LPHN3 complex and its critical regulation by alternative splicing to specify excitatory vs. inhibitory synapse specification is essential for progress in understanding synapse formation.
Here, we determined the cryo-EM structure of the LPHN3-TEN2 complex which revealed that the N-terminal Lec domain of LPHN3 binds to the side of the TEN2 barrel opposite to the toxin-like domain . Previously, we reported that the toxin-like domain of TEN2 is needed for LPHN3 binding because a toxin domain deletion construct (TEN2 ΔTox) abolished Lec binding in cell-aggregation and cell-surface staining experiments, and lacked any defects in cell-surface localization 2 .
Our further experiments suggested that this mutant is unable to bind LPHN likely because it is misfolded and escaped the protein quality control system and was still trafficked to the cell surface [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref]. During the revision of this manuscript, the structure of the chicken Ten2 in complex with mouse Latrophilin2 was published revealing a similar structure to our complex structure [bib_ref] Structural basis of teneurin-latrophilin interaction in repulsive guidance of migrating neurons, Toro [/bib_ref]. Both structures agree that LPHN binds to the side of the TEN2 barrel and not the toxin domain. The nearby Olf domain of LPHN3 faces away from the N-terminus of TEN2, positioning the membrane-anchored domains of LPHN3 and TEN2 opposite from each other, consistent with a transcellular interaction, rather than cis [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. A FLRT3 molecule can simultaneously bind to the Olf domain of LPHN3 and form a trimeric TEN2-LPHN3-FLRT3 complex [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. Whether the trimer may accommodate binding of a second FLRT3 molecule to enable FLRT3 dimerization or binding of an UNC5 molecule on the FLRT monomer may depend on the alternatively spliced sequence of LPHN3 between the Lec and Olf domains. FLRT3 dimerization or the FLRT3/UNC5 interaction may lead to further rearrangement of the protein-protein interaction network at the synapse.
Importantly, the LPHN3-TEN2 complex structure revealed that the LPHN3-binding site on TEN2 is away from the alternatively spliced site that is on the TEN2 propeller [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref] , raising the question of how a seven amino acid splice insert within the >2000 amino acid ECR of TEN2 could dictate LPHN3 binding and synapse specificity without being close to the binding interface. The crystal structure of the TEN2 +SS isoform showed that the splice insert lies at the crystal contact site and likely mediate TEN homodimerization [bib_ref] Structures of teneurin adhesion receptors reveal an ancient fold for cell-cell interaction, Jackson [/bib_ref]. Alternative splicing in the coding region of proteins expands the functional and regulatory capacity of metazoan genomes [bib_ref] Dynamic integration of splicing within gene regulatory pathways, Braunschweig [/bib_ref] [bib_ref] Mechanisms of alternative splicing regulation: insights from molecular and genomics approaches, Chen [/bib_ref] [bib_ref] Functional consequences of developmentally regulated alternative splicing, Kalsotra [/bib_ref] [bib_ref] Structural basis for adhesion G protein-coupled receptor Gpr126 function, Leon [/bib_ref]. In addition to TEN2, numerous proteins such as DSCAMs, protocadherins, neurexins and neuroligins use alternative splicing for diversifying their functions, such as their ability to bind ligands [bib_ref] Presynaptic neurexin-3 alternative splicing trans-synaptically controls postsynaptic AMPA receptor trafficking, Aoto [/bib_ref] [bib_ref] A highly conserved program of neuronal microexons is misregulated in autistic brains, Irimia [/bib_ref] [bib_ref] Alternative splicing of P/Q-Type Ca(2+) channels shapes presynaptic plasticity, Thalhammer [/bib_ref] [bib_ref] Single-cell mRNA profiling reveals cell-type-specific expression of neurexin isoforms, Fuccillo [/bib_ref]. In most proteins, the alternatively spliced sites localize to the ligandbinding site in order to directly enable or disturb ligand binding [bib_ref] Structures of neuroligin-1 and the neuroligin-1/neurexin-1 beta complex reveal specific protein-protein and..., Arac [/bib_ref] [bib_ref] Structures of neurexophilin-neurexin complexes reveal a regulatory mechanism of alternative splicing, Wilson [/bib_ref] [bib_ref] Regulation of neurexin 1beta tertiary structure and ligand binding through alternative splicing, Shen [/bib_ref]. Thus, it is unusual that the LPHN3 binding site is localized away from the alternatively spliced sequence. In the case of TEN2, alternative splicing allows the protein to act as a switch in regulating ligand binding despite the ligand-binding site being away from the seven residue alternatively spliced site 2 , and this switch disables LPHN3 binding that is required for excitatory [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref] is on the TEN2 β-barrel located at the LPHN3-binding interface (black dots). Results for TEN2 and LPHN3 binding in different experimental setups [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref] are summarized in the table. The DHR mutation breaks the interaction of TEN2 with LPHN3 in all experimental setups. b Representative images for cell-aggregation assays with TEN2 constructs and full-length LPHN3. WT TEN2 −SS induces cell aggregation with LPHN3, while TEN2 DHR −SS abolishes cell aggregation. HEK293 cells were co-transfected with TEN2 or LPHN3 and either tdTomato or EGFP as indicated. Scale bar indicates 100 µm. Quantification of aggregation index (%) is shown on the right (***p < 0.001 by one-way ANOVA). c TEN2 constructs expressed in mammalian cells were tested for their ability to bind soluble biotinylated LPHN3 or LPHN1 Lec domain using flow cytometry experiments (left) or cell-surface staining assays (right). The DHR mutation abolishes the cis-like interaction between TEN2 and LPHN. TEN2 construct expression was determined by HA tag fluorescence (Y axis) and purified Lec binding to TEN2-expressing cells was measured by fluorescence of DyLight attached to neutravidin (X axis). Dot plots represent the correlation between TEN2 expression and LPHN3 binding. Black cross indicates "high TEN2 expression and high LPHN3 binding" gate. Scale bar indicates 20 µm. Quantification of cell-surface-binding assays are shown next to the image (***p < 0.001 by one-way ANOVA). Although it is intuitively difficult to understand the relationship between alternative splicing and LPHN3 binding, our synapse formation experiments demonstrated a clear requirement of TEN2-LPHN3 interaction for the excitatory synapse specification function of TEN2 −SS, since the DHR mutation on TEN2 −SS isoform that is unable to bind to LPHN3 was unable to induce excitatory postsynaptic specializations [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref]. The same mutation on the TEN2 +SS isoform, however, behaved like wild-type TEN2 +SS and successfully induced inhibitory synapse formation [fig_ref] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R [/fig_ref]. These results suggest that interaction of TEN2 with LPHN3 is required for excitatory but not for inhibitory synapse formation. Moreover, the observation that the DHR LPHN-binding mutant had no effect on the ability of TEN2 +SS to induce inhibitory postsynaptic specializations suggests a LPHN-independent mechanism that requires unidentified TEN2 interaction partners at inhibitory synapses.
Our results show that the interaction of LPHN3 with TEN2 can be disrupted in at least two ways: (1) by point mutations on the LPHN-binding interface on TEN2 (but not by mutations that are not at the interface, , and (2) by insertion of the seven alternatively spliced residues in the propeller domain of TEN2. The mutagenesis of the LPHN-binding interface abolished the TEN2/LPHN3 interaction in all experimental setups as expected from a binding site mutant [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. However, the effect of alternatively spliced site on the TEN2/LPHN3 interaction depended on whether one or both proteins experienced restraints due to their attachment to cell membranes; or they could freely rotate and tumble in solution . Specifically, alternative splicing abolished the TEN2/LPHN3 interaction in cellaggregation experiments where proteins approach each other from opposing membranes; but not in cell-surface staining or insolution experiments where one or more proteins are in solution. Altogether these results suggest that alternative splicing regulates the TEN2/LPHN3 interaction via a mechanism that differs from disrupting the binding interface. These results enable us to suggest a model for how alternative splicing regulates TEN2 interactions and functions:
TEN2 forms a cis-dimer on the presynaptic membrane that is mediated by two disulfide bonds formed between the 2nd and 5th EGF repeats (black lines, that extend the globular cytoplasmic C-terminal heads of TEN2 (TEN2 ECRΔ1) towards the opposite membrane . Previous cryo-EM images of the dimeric TEN2 −SS showed that the globular heads have the rotational flexibility around the EGF/head linker (arrow) that enables TEN2 −SS to sample the 3D space 2 and to successfully bind the Lec domain of LPHN3 in cis-like and trans . In this conformation, FLRT3 is also able to interact with LPHN3 and form a trimeric complex, consequently leading to excitatory synapse formation . However, the crystal structure of the TEN2 +SS isoform showed that, in the presence of the splice insert, the two globular heads form a dimer that is facilitated by the interactions between the splice inserts . The presence of the splice inserts enables the formation of two salt bridges (E1306-H1315 and E1301-R1337 in chicken Ten2, red dashed lines in and five hydrogen bonds between the β-propellers. These newly generated interactions of the propeller domains would zipper-up the molecule introducing rigidity to the TEN2 +SS cis-dimer and restrict rotational flexibility around the EGF/head linker preventing TEN2 +SS from sampling the 3D space . On the other hand, the ECR of LPHN3 consists of two globular regions separated by a Ser-Thr-Pro rich glycosylated linker region that is reported to be semirigid [bib_ref] LPHN3, a presynaptic adhesion-GPCR implicated in ADHD, regulates the strength of neocortical..., O'sullivan [/bib_ref]. As a result, the Lec domain of LPHN3 on the opposite membrane would have limited or no access to the LPHN-binding site on TEN2 +SS and fail to bind, although the binding site is intact and functional. In order to test the validity of this model, we introduced mutations to break the two salt bridges that are newly generated in the TEN2 +SS isform (E1154A, H1161A, R1183A, E to A in human TEN2 splice site NKEFKHS) and to decrease the rigidity introduced by the seven amino acid splice insert . Cell-aggregation experiments showed that the TEN2 +SS salt bridge mutant restored the LPHN3 binding ability of TEN2 +SS partially suggesting that the rigidity introduced by the salt bridges that are formed upon insertion of the splice site limits accessibility of TEN2 to LPHN3.
As alternative splicing prevents the formation of the TEN2/ LPHN3 interaction by hiding the binding site rather than destroying it, it is plausible that the +SS isoform adopts a shape that enables other TEN2 interactions that the −SS isoform cannot mediate. Indeed, the trans-dimerization of TEN2 was reported to occur only by the +SS isoform [bib_ref] Teneurin-3 controls topographic circuit assembly in the hippocampus, Berns [/bib_ref] ; and previous studies suggested that TEN2 +SS isoform should interact with unknown ligands in order to induce inhibitory synapses 2 . This structure of a teneurin-latrophilin complex in combination with our biochemical results demonstrate the clear mechanistic difference of excitatory vs. inhibitory synapse specification and lead to previously unimagined new directions in both the synapse formation and alternative splicing fields.
# Methods
Cell culture. High-Five insect cells (Trichoplusia ni, female, ovarian. Thermo Fisher, B85502) cultured in Insect-Xpress medium (Lonza, 04351Q) supplemented with 10 μg/mL gentamicin at 27°C were used for production of recombinant proteins. HEK293T mammalian cells (ATCC, CRL-3216) were used for cell-surface expression assays and flow cytometry binding assays and were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, 11965092) supplemented with 10% FBS (Sigma-Aldrich, F0926) at 37°C in 5% CO Cloning and expression in insect cells. TEN2 splice variant Lasso (UniProt: Q9NT68-2) and LPHN (LPHN1, UniProt: O88917; LPHN3, UniProt: Q9HAR2) constructs were cloned into a pAcGP67a vector and expressed in High-Five insect cells using the baculovirus expression system. Sf9 cells (Thermo Fisher, 12659017) were co-transfected with the linearized baculovirus DNA (Expression Systems, 91-002) and the constructed plasmid using the Cellfectin II (Thermo Fisher, 10362100) transfection reagent. Baculovirus was amplified in Sf9 cells in SF900-III medium containing 10% (v/v) FBS [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. Large-scale protein Membrane anchoring restricts alternative splice-dependent interaction of TEN2 to LPHN3. Same three experimental setups as in [fig_ref] Figure 4: Experimental setups to investigate the effect of various restraints [/fig_ref] -c were used to test the effect of alternative splicing on TEN2/LPHN3 interaction. Figure outline is identical in principle to that of in [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref]. a Diagram for WT TEN2 −SS and WT TEN2 +SS constructs that were used in the below experiments. The seven amino acid splice site on the TEN2 β-propeller is indicated by empty or filled red stars. Results for the interaction of TEN2 and LPHN3 in different experimental setups (as in Figs. 4a-c and 5b-d) are summarized in the table. The insertion of the splice site breaks the interaction of TEN2 with LPHN3 only in the cell-aggregation assays, but not in the other experimental setups. b Representative images for cell-aggregation assays with TEN2 −SS or TEN2 +SS and full-length LPHN3. TEN2 −SS induces cell aggregation with LPHN3, while TEN2 +SS abolishes cell aggregation. Scale bar indicates 100 µm. Figure modified from ref. [bib_ref] Structural basis for teneurin function in circuit-wiring: a toxin motif at the..., Li [/bib_ref]. c TEN2 −SS and TEN2 +SS expressed in mammalian cells were tested for their ability to bind soluble biotinylated LPHN3 or LPHN1 Lec domain using flow cytometry experiments (left) and using cell-surface staining assays (right). Both −SS and +SS mediate the interaction between TEN2 and LPHN in cis-like. Quantification of cell-surface-binding assays are shown next to the image. The cell-surface staining assays in c was performed in the same experiment as in [fig_ref] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and... [/fig_ref] , and thus the control images are identical. Scale bar indicates 20 µm. (***p < 0.001 by one-way ANOVA.) d Size-exclusion chromatograms showing the formation of binary complexes between soluble full TEN2 ECR and full LPHN3 ECR (left, blue and green lines). Elution profile for individual TEN2 −SS ECR, TEN2 +SS ECR and LPHN3 ECR are shown for reference (gray lines). Both TEN2 −SS ECR and TEN2 +SS ECR bind to LPHN3 ECR (green and blue lines, respectively), as also observed by co-elution in the fractions ran on SDS-PAGE gel. Colors of the chromatograms match the colors of box around the SDS-PAGE gel. Data in b and c are presented as mean ± SEM, n = 3, and are representative of at least three independent experiments. Source data are provided as a Source Data file. For the structural studies, TEN2 ECRΔ1 (residues T727-R2648) and LPNH3 ECR (residues S21-V866) were cloned with carboxyl-terminal 6XHis-tags separately and co-expressed in High-Five insect cells. The following primers were used for amplification of High Five cells expressed human TEN2 ECRΔ1: F: 5′- R1337 R1337 Alternatively spliced insert within the β-propeller mediates the TEN2 +SS dimer interface. a Structure of TEN2 +SS dimer shows the splice inserts from each protomer (yellow and magenta residues) creates a binding interface and leads to TEN2 dimerization via the β-propeller. Close-up views of the dimer interface show two salt bridges and five hydrogen bonds are at the interface. One of the salt bridges is directly mediated by the glutamate (E1306) within the splice insert NKEFKHS and four of the hydrogen bonds also require splice site residues. Two salt bridges align almost parallel to each other and to the disulfide bonds between the EGF repeats and restricts the conformational flexibility of the TEN2 head significantly (see . The Ntermini of both protomers face the same direction towards the EGF repeats, and thus, the dimer is positioned as a cis-dimer that will extend the zippering of the already existing EGF-mediated cis-dimer, although it was reported to form as a trans-homodimer, previously 38 . TEN protomers (PDB: 6FB3) are colored as cyan and palegreen, respectively, and splice sites are colored as yellow and magenta, respectively. b Cell-aggregation experiments show the LPHN3 binding ability of TEN2 +SS is partially restored when the two salt bridges are broken in the TEN2 +SS mutant (*P ≤ 0.05; ***P ≤ 0.001; by one-way ANOVA). Data are presented as mean ± SEM, n = 3, and are representative of at least three independent experiments. Source data are provided as a Source Data file.
purified in a final buffer comprised of 10 mM Tris pH 8.5, 150 mM NaCl. For the flow cytometry binding assays, LPHN1 Lec (residues S26-Y131) and LPHN3 Lec (residues S21-Y126) were cloned with carboxyl-terminal 6XHis-AVI-tags and captured on nickel-nitrilotriacetic resin as described above. Following a wash with HBS buffer containing 20 mM imidazole, final concentrations of 50 mM Bicine pH 8.3, 100 mM NaCl, 10 mM Mg-acetate, 10 mM ATP, 0.5 mM biotin and 5 mM BirA were added to the resin, which was then rotated for 2 h at 27°C. After removing residual BirA and ATP by washing with HBS buffer containing 20 mM imidazole, the biotinylated lectin was eluted with HBS buffer containing 200 mM imidazole. Purified protein was applied to size-exclusion chromatography. The efficiency of biotinylation was assessed using a streptavidin bead pulldown assay. Model for the splice variant-dependent interaction of TEN2 with LPHN3. The model depicts how alternative splicing acts as a molecular switch to determine which adhesion partner TEN2 binds to and, accordingly, which type of synapse TEN2 specifies. Both TEN2 isoforms form a cis-dimer on the presynaptic membrane through two disulfide bonds formed between the 2nd and 5th EGF repeats (black sticks). a TEN2 −SS isoform has rotational flexibility (arrows) that enables TEN2 to find the correct docking geometry in order to bind to the Lec domain of LPHN3 expressed on the neighbor cell 2 . Such rotational flexibility also allows FLRT3 to bind to the Olf domain of LPHN3 and, altogether, to induce excitatory synapse formation. The DHR mutation breaks the interaction of TEN2 −SS with LPHN3 and abolishes excitatory synapse formation. b The TEN2 +SS isoform does not have rotational flexibility around the linker between the EGF repeats and the rest of the extracellular head as observed in the crystal structure of the TEN2 +SS isoform, which shows that the splice insert mediates a dimeric interaction between the two TEN2 +SS protomers 38 , PDB ID: 6FB3). Instead, the TEN2 +SS protomers are zipped-up due to the additional two salt bridges (black balls) between the propellers of the TEN2 cis-dimer. Thus, the LPHN-binding site on TEN2 +SS is not at the right docking geometry to interact with the Lec domain of LPHN3 expressed on the neighboring cell (though it can still bind soluble Lec domain). The geometry of TEN2 +SS likely enables other hetero-or homophilic protein interactions that were not possible in the −SS isoform, such as TEN2 transhomodimerization, and mediates inhibitory synapse formation. The DHR mutation on TEN2 +SS has no effect on these unknown interactions and thus, does not affect the ability of TEN2 +SS to induce inhibitory synapses. Model partially drawn to scale. The LPHN ECR structure (orange), is based on Lec and Olf domain structure (PDB: 5AFB), connected by a STP-rich stalk to the GAIN and HormR domain structure (PDB: 4DLQ). TEN2 protomers are colored as cyan and palegreen (PDB: 6CMX and 6FB3), FLRT protomers are colored as magenta and gray (PDB: 5CMN).
carboxyl-terminal FLAG-tag were cloned into a pcDNA3.1 vector for cell-surface expression assays and flow cytometry binding assays in HEK293T cells. The following primers were used for amplification of HEK cells expressed human TEN2 ECR: F: 5′-GGATGACGACGATAAAGGCGGTAAGCTTAGCCCACCTCTC-3′ and R: 5′-TTACTTATCGTCGTCATCCTTGTAATCCCTCTTTCCCATCTCATT CTGTCTT-3′. The following primers were used for amplification of HEK cells expressed human TEN2 ΔTox: F: 5′-GGATGACGACGATAAAGGCGGTAAGC TTAGCCCACCTCTC-3′ and R: 5′-TTACTTATCGTCGTCATCCTTGTAATCTT CATAGGGAGGAGGCACGAAATACAT-3′. The following primers were used for amplification of HEK cells expressed human TEN2 ΔToxΔBarrel: F: 5′-GGATGA CGACGATAAAGGCGGTAAGCTTAGCCCACCTCTC-3′ and R: 5′-TTACTTA TCGTCGTCATCCTTGTAATCGAAGGCATTAAGAACAGGCTTGTTC-3′. TEN2 DHR mutants were generated using a standard two-step PCR-based strategy with primers: F: 5′-ATTCGGACTGAAAAGATCTATGATAACACCACGAAGTT CACCCTGAGGATCATTTATG-3′ and R: 5′-CATAAATGATCCTCAGGGTGA ACTTCGTGGTGTTATCATAGATCTTTTCAGTCCGAAT-3′. TEN2 LR mutants were generated using a standard two-step PCR-based strategy with primers: F: 5′-AGTGAGACTCCCCTCCCCGTTGACAACTACACCTATGATGAGATTTCT GGCAAGGTG-3′ and R: 5′-CACCTTGCCAGAAATCTCATCATAGGTGTAG TTGTCAACGGGGAGGGGAGTCTCACT-3′. After wash with PBS + 2% BSA, cells were stained with a secondary antibody mixture: donkey anti-mouse Alexa Fluor 488 (nvitrogen, A21202) 1:3000 and goat anti-rabbit Alexa Fluor 647 (Invitrogen, A32733) 1:3000 for 30 min. After washing, cell pellets were resuspended in PBS + 2% BSA immediately before flow cytometry data acquisition (Accuri C6 flow cytometer, 10000 events measured) after washing. Acquired data were analyzed using the FlowJo analysis software (FlowJo LLC). For the binding assays, His-Avi-tagged Lec was captured on nickelnitrilotriacetic resin and purified as described above. Biotinylated Lec was tetramerized and fluorescently labeled through incubation with NeutrAvidin DyLight 488 (Thermo, 22832) on ice for 20 min. Cultured cells expressing HAtagged TEN2 were detached and then washed as described above. Next, the cells were stained with rabbit anti-HA 1:1000 antibody and, following two wash cycles, stained with goat anti-rabbit Alexa Fluor 647 antibody in the presence of the 100 nM NAV488 labeled Lec mixture. The following primers were used for amplification of His-Avi-tagged human LPHN1 Lec: F: 5′-CGGCGGCGCATTC TGCCTTTGCGGCGAGCCGGGCTGGACTCCCATTTGG-3′ and R: 5′-CTTCTG AGCCTCGAAAATATCATTAAGACCGCGGTAAGGGACACAGTCGTACT GC-3′. The following primers were used for amplification of His-Avi-tagged human LPHN3 Lec: F: 5′-GGCGGCGCATTCTGCCTTTGCGGCGTCCCG CGCACCCATTCCTATGGCCG-3′ and R: 5′-TTCTGAGCCTCGAAAATATCA TTAAGACCGCGATATGGCACGCACTCGTACTGCACT-3′.
Cell-aggregation assays. HEK293T cells (ATCC) were grown to 90% confluence in a T-75 flask. Cells were trypsinized with 3 mL 0.05% trypsin-EDTA (Gibco, 25300-054) and resuspended to 10 mL with DMEM/10% FBS/1% Penicillin-Streptomycin media (Complete DMEM). Three-hundred µL of the cell suspension was added to each well of a 6-well plate containing 3 mL of Complete DMEM media and incubated overnight at 37°C. Cells in each well were then cotransfected with 2 µg of either pCMV (empty vector) + pEmerald, pCMV LPHN3 + pEmerald, pCMV (empty vector) + pCMV dsRed, or dsRed and the indicated TEN2 construct using the Calcium Phosphate method. All cDNAs were encoded in the pCMV5 or pcDNA3 vector and driven by the CMV promoter. Three days after transfection, the media was aspirated and cells were gently washed with 2 mL of PBS. Cells were resuspended by adding 1 mL of Resuspension Solution (PBS containing 1 mM EGTA) and then incubated for 5 min at 37°C. Fifteen µL of 1 mg/20 µL DNAse (Sigma, D5025) was then added to each well and cells were triturated by pipetting up-and-down (16 times) in each well to resuspend cells off the plate bottom and create single-cell suspensions. Cells were then transferred to a new Eppendorf tube and another 15 µL of DNAse solution was added to each sample. Cells were mixed in 1:1 ratio by adding 70 µL of pCMV (empty vector) + pEmerald or LPHN3 + pEmerald with 70 µL of pCMV (empty vector) + pCMV-dsRed or TEN2 Construct + dsRed in a new Eppendorf that contained 360 µL of Incubation Solution (DMEM containing 50 mM HEPES-NaOH pH 7.4, 10% FBS, 10 mM CaCl 2 and 10 mM MgCl 2 ) for a final volume of 500 µL. The mixture was triturated and the entire volume was transferred to one well in a non-coated 12well plate (Costar, 3737). Images were taken immediately (time = 0) using a Leica Fluorescent DMIL LED Microscope with a 10x objective. Cells were then placed on a shaking incubator at 125 rpm at 37°C for 20 min and imaged again (time = 20).
Aggregation index at time = 20 was calculated using ImageJ, measuring the percentage of signal/frame occupied by cells forming complexes of two or more cells relative to the total signal of the frame.
Cell-surface-binding assays. HEK293T cells (ATCC) were grown to 90% confluence in a T-75 flask. Cells were trypsinized with 3 mL 0.05% trypsin-EDTA (Gibco, 25300-054) and resuspended to 10 mL of DMEM + 10% FBS + 1% Penicillin-Streptomycin (complete DMEM) media. Fifty µL of cell suspension was added to each well of a 24-well plate that contained a Matrigel-coated coverslip and 1 mL complete DMEM and incubated overnight. Cells were then co-transfected with 1 µg of either empty pCMV, wild-type Teneurin 2 or the indicated mutant Teneurin construct and 1 µg of pEmerald using the Calcium Phosphate method and incubated for 2 days at 37°C. Transfection media was gently removed and 500 µL of chilled DMEM containing 250 µM of purified, Avi-fusion, biotinylated, rat LPHN1 Lec or human LPHN3 Lec was added to each well. Plates were wrapped in foil and incubated overnight at 4°C to reduce endocytosis, with gentle shaking. This was performed essentially as described in [bib_ref] High affinity neurexin binding to cell adhesion G-protein-coupled receptor CIRL1/latrophilin-1 produces an..., Boucard [/bib_ref]. Cells were gently washed 2x using 1 mL of PBS and fixed with 300 µL of ice-cold 4% PFA/4%sucrose/PBS. Plates were wrapped in foil and incubated for 20 min at 4°C during the fixation. Cells were gently washed 3× using 1 mL of room temperature PBS and blocked with 300 µL of 5% BSA (Sigma, 10735086001)/PBS (blocking buffer) for 1 h at room temperature. Bound biotinylated Lec was detected by immunofluorescence using 300 µL per well of Streptavidin (AlexaFluor-555 conjugated, Invitrogen, S21381, at 1:10,000 dilution) diluted into blocking buffer for 1 h at room temperature. Cells were gently washed 3× with 1 mL of PBS. Cells were re-blocked, and HA-tagged, surface Teneurins were detected by adding 300 µL of rabbit anti-HA antibody (Cell Signaling Technologies, 3724) at 1:1,000 dilution in blocking buffer. Cells were gently washed 3× with 1 mL PBS. Goat anti-rabbit secondary antibodies (Alexa-Fluor 633 conjugated, Invitrogen) and DAPI (Sigma, 10236276001) staining was done for 30 min at 1:10,000 and 1:5,000, respectively, in blocking buffer, followed by 3× gentle washes with 1 mL of PBS. Coverslips were mounted onto slides (UltraClear microscope slides Denville Scientific, M1021) in mounting media (Fluoromount-G, Southern Biotech, 010020). Images were acquired using a Nikon A1 Eclipse Ti2 confocal microscope with a ×60 oil-immersion objective, operated by NIS-Elements AR acquisition softw×are. The same confocal acquisition settings were applied to all samples of the experiment. Collected z-stacks at a 0.4 µm z-step size were analyzed blindly using Nikon Elements Analysis software. Co-localization was calculated using the Pearson's correlation coefficient of Lec-Streptavidin-555 to Teneurin-HA-633 emission.
Artificial synapse formation assay. HEK293T cells were transfected with the expression vectors of the cell-adhesion molecules. 24 h later, HEK293T cells were co-cultured with cultured cortical neurons (DIV16) from P0 mice. After 24 h, cells were fixed with 4% PFA and immunostained with rabbit anti-Flag (Sigma; 1:1000 both) together with mouse anti-PSD95 (Sysy, 124011, 1:500) or mouse anti-GABAA α2 (Sysy; 224211, 1:500) respectively. Images were collected with a Nikon A1 confocal microscope using a ×60 objective. A human NPR mutant (1-118 aa of the full-length protein) which comprises an A domain containing low-complexity sequences is used as the negative control in the artificial synapse formation assays. The signals of the synaptic markers that were recruited to the surface of the HEK293T cells were quantified using Image J. Normalized values equal the fluorescent intensity of the synaptic marker that was examined (GABAα2/PSD95) / the fluorescent intensity of the Flag-tagged protein expressed in the HEK cells (TENs /Nrn1β).
Cryo-EM data acquisition. 2.5 μl purified human TEN2 ECRΔ1 and human LPHN3 ECR complex (0.22 mg/ml) was applied on glow-discharged holey carbon grids (Quantifoil R1.2/1.3, 300 mesh), and vitrified using a Vitrobot Mark IV (FEI Company). The specimen was visualized using a Titan Krios electron microscope (FEI) operating at 300 kV and equipped with a K3 direct electron detector (Gatan, Inc.). Images were recorded with a nominal magnification of ×81,000 in superresolution counting mode, corresponding to a pixel size of 0.54 Å on the specimen level. To maximize data collection speed while keeping image aberrations minimal, image shift was used as imaging strategy using one data position per hole with four holes targeted in one template with one focus positio. In total, 4967 images with defocus values in the range of −1.0 to −2.5 μm were recorded using a dose rate of 14.6 electrons/Å 2 /s. The total exposure time was set to 4.2 s with frames recorded every 0.105 s, resulting in an accumulated dose of about 60.1 electrons per Å 2 and a total of 40 frames per movie stack.
Image processing and 3D reconstructions. Stack images were subjected to beaminduced motion correction using MotionCor2 [bib_ref] Distortion correction of EPI data using multimodal nonrigid registration with an anisotropic..., Glodeck [/bib_ref]. CTF parameters for each micrograph were determined by CTFFIND4 [bib_ref] Accurate determination of local defocus and specimen tilt in electron microscopy, Mindell [/bib_ref]. Particle selection, two-and threedimensional classifications were initially performed on a binned dataset with a pixel size of 4.32 Å using RELION-3. In total, 4,475,958 particle projections were selected using automated particle picking and subjected to reference-free twodimensional classification to discard false-positive particles or particles categorized in poorly defined classes, resulting in 3,307,148 particle projections for further processing. The initial 3D maximum-likelihood-based classification was performed on a binned dataset with a pixel size of 4.32 Å using the previously reported TEN2 structure 2 as the reference model. The detailed data processing flow is shown in [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. Briefly, for Tenurin-Letrophilin complex, 1,309,684 particles that showed well-defined density of Lec domain were selected after initial rounds of 3D classification. Then, two rounds of focused 3D classification with mask around Lec domain were performed without alignment. 3D refinement and post-processing was performed on the best class with clear features for the Lec domain [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. The final map for TEN2_Lec was resolved at 2.97 Å [fig_ref] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously [/fig_ref]. To resolve domain 3, the same dataset was reprocessed with a total of 1,137,765 particles showing well-resolved domain 3. Two rounds of focused 3D classification with mask around domain 3 were performed, followed by 3D refinement and post-processing. The final map for TEN2_domain 3 was resolved at 3.07 Å [fig_ref] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues [/fig_ref].
Reported resolutions are based on the gold-standard Fourier shell correlation (FSC) using the 0.143 criterion . All density maps were corrected for the modulation transfer function (MTF) of the K3 direct detector and then sharpened by applying a temperature factor that was estimated using postprocessing in RELION-3. Local resolution was determined using ResMap 59 with half-reconstructions as input maps .
Model building and refinement. Model building was based on the structure of human TEN2 ECR (PDB: 6CMX) and the Lec domain from human LPHN3 (PDB: 5AFB and 5FTT). The models were first docked into the EM density maps using Chimera 60 and then manually checked and adjusted residue-by-residue to fit the density using COOT [bib_ref] Features and development of Coot, Emsley [/bib_ref]. The ECR of TEN2 was built based on that of chicken TEN2 (PDB: 6FB3) and manually adjusted to human sequence and splice form. Note that the Lec domain was not as well resolved as TEN2, so it was docked as a rigid body without fitting and manipulating the side chains. Both maps (TEN2/LPHN3 and TEN2 focusing on domain 3) were used for model building. There is a slight shift between the two maps from reconstruction, so they were aligned based on TEN-Lec before model building. The final model containing both ECR of TEN2 and Lec domain of LPHN3 was subjected to global refinement and minimization in real space using the phenix_real_space_refine module in Phenix 62 first against TEN2 domain 3 map while keeping the Lec domain as a rigid body, and then against TEN2-Lec map while keeping the domain 3 as a rigid body. FSC curves were calculated between the resulting model and either maps using Phenix M-triage . The final model statistics are provided in [fig_ref] Table 1: Cryo-EM data collection, refinement and validation statistics [/fig_ref].
Nine N-linked glycosylation sites (on residues N1490, N1586, N1647, N1681, N1766, N1867, N2071, N2211, N2522.) and five disulfide bonds (C1394-C1402), (C1396-C1404), (C1106-C1109), (C1210-C1218), (C1277-C1330) were observed in TEN2.
Quantification and statistical analysis. Error bars in Figs. 5, 6, 7, 8 and Supplementary represent means ± SEM. Each measurement was repeated at least three times independently. Data were analyzed using software GraphPad Prism and ImageJ.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.
## Data availability
The cryo-EM density map has been deposited in the Electron Microscopy Data Bank (https://www.ebi.ac.uk/pdbe/emdb/) under accession code EMD-21205 and the model coordinates have been deposited in the Protein Data Bank (http://www.rcsb.org) under accession number PDB 6VHH. Data supporting the findings of this manuscript are available from the corresponding authors upon reasonable request. A reporting summary for this Article is available as a Supplementary Information file. The source data underlying Figs. 2e, 5b-d, 6c, d, 7c, e, 8b and , c, and 7 are provided as a Source Data file.
Received: 3 December 2019; Accepted: 6 April 2020;
[fig] Figure 2: LPHN3 interacts with TEN2 and FLRT3 simultaneously. a Continuous density C-terminal to the Lec domain of LPHN3 revealed by analysis of the cryo-EM maps at a lower threshold. b Manual fitting of the Olf domain from the LPHN3 Lec-Olf structure (PDB: 5AFB) into the extra EM density. Ribbon diagram of TEN2/LPHN3-Lec complex is colored cyan and the LPHN3 Lec-Olf domains are colored orange. Lec domains are superimposed. c Schematic diagram of the trans-cellular positioning of the TEN2/LPHN3 complex; and the formation of a ternary complex between TEN2, LPHN3, and FLRT3. d Superimposition of the TEN2/LPHN3-Lec-Olf complex with the LPHN3/FLRT3 complex structure (PDB: 5CMN). The Olf domains are superimposed. [/fig]
[fig] Figure 3: TEN2 and LPHN3 interaction is mediated by conserved residues. a The TEN2/LPHN3 binding interface is conserved. The structure of the TEN2/ LPHN3 complex is shown in surface representation on which the conservation of residues is mapped from most conserved (magenta) to least conserved (cyan) (using the ConSurf server63 ). The LPHN-binding site on TEN2 and the TEN2-binding site on LPHN3 are indicated by yellow circles. b Ribbon diagram of the TEN2/LPHN3 heterodimer showing the interface between TEN2 and LPHN3. Close-up view of the binding interface shows tentative residues involved hydrogen bonds and salt bridge shown by yellow dashes and red dashes, respectively. The TEN2 mutation DHR mutation (D1737N, H1738T, R1739T) which disrupt the TEN2/LPHN3 are shown as sticks. Two salt bridges between R44 and E2001; two between D49 and R2043; one between D49 and R2043; and one between D67 and K1712 are observed. c Close-up view for the EM density showing the close packing of the conserved N-linked glycosylation on TEN2-N1681 with LPHN3-S38 on the Lec domain. Due to the uncertainty of the glycan topology, we only modeled the first three sugars of N-glycosylation (NAG-NAG-BMA) into the density. [/fig]
[fig] Figure 4: Experimental setups to investigate the effect of various restraints. Three experimental setups with decreasing restraints on the docking geometry of LPHN3 and TEN2 during their interaction. a Setup for trans-cellular interaction of full-length TEN2 with full-length LPHN3 in cell-aggregation experiments. Both proteins are anchored on the cell-membranes and their mobility is restricted by their lateral diffusion within the membrane. b Setup for cis-like-interaction of full-length TEN2 and soluble biotinylated Lec domain of LPHN3 in either flow cytometry or cell-surface staining experiments. TEN2 is anchored on the cell-membrane, but the Lec domain of LPHN3 freely rotates in solution. c Setup for in-solution experiments for cis-like-interaction of soluble TEN2 and soluble LPHN3 in the absence of any membranes. | (2020) 11:2140 | https://doi.org/10.1038/s41467-020-16029-7 | www.nature.com/naturecommunications [/fig]
[fig] Figure 5: Binding site mutations on TEN2 abolish LPHN3 binding in both trans and cis-like. a Diagram for WT TEN2 −SS and TEN2 DHR −SS constructs. DHR mutation [/fig]
[fig] Figure 7: CATTCTGCCTTTGCGGCGGATCCCACTTCCTGTGCTGATAACAAGGAT AATGAG-3′ and R: 5′-GGATCAGATCTGCAGCTTAGTGATGGTGATGGTGA TGCCTCTTTCCCATCTCATTCTGTC-3′. The following primers were used for amplification of High Five cells expressed human LPHN3 ECR: F: 5′-CATTCTG CCTTTGCGGCGGATCCCTCCCGCGCACCCATTCC-3′ and R: 5′-GGATCAGA TCTGCAGCTTAGTGATGGTGATGGTGATGCACGTCCAGCAGCAGATCG TG-3′. Seventy-two hours after viral infection, the medium containing secreted Binding site mutations on TEN2 selectively abolish excitatory but not inhibitory synapse formation. a Diagram for TEN2 DHR −SS and TEN2 DHR +SS constructs that were used in the below experiments. The seven amino acid splice site on the TEN2 β-propeller is indicated by empty or filled red stars; DHR mutation is indicated by black dots. b, c Artificial synapse formation assay showing that LPHN-binding mutant (DHR) of TEN2 −SS attenuated excitatory synapse formation. HEK293T cells are co-transfected with indicated cell-adhesion molecules and GFP, and co-cultured with cortical neurons. Cultures were subsequently immunostained for the excitatory postsynaptic synapse marker PSD95. Representative images (b) and quantifications of PSD95 signals (c) are shown. d, e LPHN-binding mutant (DHR) of TEN2 +SS did not affect inhibitory synapse formation. Similar in b and c, except that immunostaining for the inhibitory postsynaptic synapse marker GABA(A)α2 was performed. Scale bar in b and d indicates 10 µm. Data in c and e are presented as mean ± SEM, n = 3, and are representative of at least three independent experiments. ns, P > 0.05; *P < 0.05 (one-way ANOVA). Source data are provided as a Source Data file.glycosylated proteins was collected and centrifuged at 900 g for 15 min at room temperature. The supernatant was transferred into a beaker and mixed with (final concentrations): 50 mM Tris pH 8.0, 5 mM CaCl 2 and 1 mM NiCl 2 and stirred for 30 min. After centrifugation at 8000 g for 30 min, the clarified supernatant was incubated with nickel-nitrilotriacetic agarose resin (QIAGEN) for 3 h at room temperature. The resin was collected with a glass Buchner funnel and rinsed with HBS buffer containing 20 mM imidazole, then transferred to a poly-prep chromatography column (Bio-rad). The protein was eluted with HBS buffer containing 200 mM imidazole and run on size-exclusion chromatography (Superdex 200 10/300 GL; Superose 6 Increase 10/300 columns; GE Healthcare), [/fig]
[fig] Flow cytometry: HEK293T cells were cultured in 6-well plates and were transfected 2 μg cDNA using LipoD293 transfection reagent. Cells at 50-60% confluence were transiently transfected as follows: 2 µg cDNA was diluted in 50 µl serum-free DMEM, and 3 µl LipoD293 transfection reagent (SignaGen, SL100668) was diluted with 47 µl serum-free DMEM. The diluted LipoD293 was added to the diluted cDNA and incubated for 10 min. Then, the transfection mixture was added dropwise to each well. The cells were detached using citric saline solution (50 mM sodium citrate, 135 mM KCl) after 48 h incubation and washed with PBS + 2% BSA. To test TEN2 WT and mutant cell-surface expression, cells were stained with a primary antibody mixture: mouse anti-FLAG M2 (Sigma, F3165) 1:1000 and rabbit anti-HA (Life Technologies, 715500) 1:1000 for 30 min at room temperature. [/fig]
[table] Table 1: Cryo-EM data collection, refinement and validation statistics. [/table]
|
A systematic review of provider-and system-level factors influencing the delivery of cardiac rehabilitation for heart failure
Background: There is a longstanding research-to-practice gap in the delivery of cardiac rehabilitation for patients with heart failure. Despite adequate evidence confirming that comprehensive cardiac rehabilitation can improve quality of life and decrease morbidity and mortality in heart failure patients, only a fraction of eligible patients receives it. Many studies and reviews have identified patient-level barriers that might contribute to this disparity, yet little is known about provider-and system-level influences. Methods: A systematic review using narrative synthesis. The aims of the systematic review were to a) determine provider-and system-level barriers and enablers that affect the delivery of cardiac rehabilitation for heart failure and b) juxtapose identified barriers with possible solutions reported in the literature. A comprehensive search strategy was applied to the MEDLINE, Embase, PsycINFO, CINAHL Plus, EThoS and ProQuest databases. Articles were included if they were empirical, peer-reviewed, conducted in any setting, using any study design and describing factors influencing the delivery of cardiac rehabilitation for heart failure patients. Data were synthesised using inductive thematic analysis and a triangulation protocol to identify convergence/contradiction between different data sources.Results: Seven eligible studies were identified. Thematic analysis identified nine overarching categories of barriers and enablers which were classified into 24 and 26 themes respectively. The most prevalent categories were 'the organisation of healthcare system', 'the organisation of cardiac rehabilitation programmes', 'healthcare professional' factors and 'guidelines'. The most frequent themes included 'lack of resources: time, staff, facilities and equipment' and 'professional's knowledge, awareness and attitude'.Conclusions: Our systematic review identified a wide range of provider-and system-level barriers impacting the delivery of cardiac rehabilitation for heart failure, along with a range of potential solutions. This information may be useful for healthcare professionals to deliver, plan or commission cardiac rehabilitation services, as well as future research.
# Background
Heart failure is a debilitating progressive clinical syndrome, that due to increasing life expectancy and more widespread adoption of a western lifestyle has seen a steady increase in prevalence across the globe. The cost of treating patients with heart failure by the National Health Service is estimated at two billion pounds per year, with most of the cost associated with hospital admissions. There is also a substantial human cost of heart failure, as many patients experience a diminished quality of life related to their illness. Improving health-related quality of life is a fundamental aim of heart failure management.
Key strategies for improving health-related quality of life include self-management of symptoms and psychological consequences of heart failure and exercise-based rehabilitation of physical functioning, all of which are part of comprehensive cardiac rehabilitation programmes. Several trials and systematic reviews have confirmed the safety and effectiveness (reduction in hospital admissions and improvement in health-related quality of life) of cardiac rehabilitation for heart failure. Thus, cardiac rehabilitation programmes are an effective and cost-effective strategy for improving health-related quality of life in people with heart failure.
Despite the strong evidence for effectiveness, according to a recent global survey, cardiac rehabilitation is available in only half the countries of the world. Furthermore, even in countries that do offer cardiac rehabilitation services, coverage is low. Globally only 30% of eligible patients access cardiac rehabilitationand there are large regional variations in the content of cardiac rehabilitation programmes. The European Cardiac Rehabilitation Inventory Survey 2010also highlighted that less than 20% of patients with heart failure receive cardiac rehabilitation.
The low proportion of eligible patients receiving cardiac rehabilitation may reflect a lack of service availability or it may reflect low uptake by patients of services. For example in the UK, uptake of cardiac rehabilitation is estimated to be around 50% on average, with lower uptake in women, ethnic minorities and people living in rural areas and areas of high deprivation. There is a large body of evidence exploring patient-level factors impacting cardiac rehabilitation enrolment/attendance, compliance/adherence, completion and drop-out rates amongst general cardiac population. These factors include distance required to travel, financial constraints and work obligations. However, to the best of our knowledge, there have been no systematic reviews of non-patient level factorsi.e., provider-and systemlevel barriers affecting the delivery of cardiac rehabilitation for patients with heart failure.
The current systematic review, therefore, aimed to answer the following research question: 'What are the factors influencing the offer, referral, delivery, implementation, and provision of cardiac rehabilitation for heart failure?'. The purpose of the study was to identify and qualitatively describe barriers and enablers affecting the delivery of cardiac rehabilitation for patients with heart failure.
# Methods
The systematic review has been registered with PROS-PERO (CRD42019153247), conducted according to Guidance on the Conduct of Narrative Synthesis in Systematic Reviewsand reported in concordance with PRISMA guidance.
## Inclusion and exclusion criteria
The scope for the systematic review is illustrated in.
## Search strategies
The full search strategy is provided in Additional file 1. The following databases were searched using a combination of free-text search terms and controlled vocabulary (Medical Subject Headings): MEDLINE (OVID interface), Embase (OVID interface), PsycINFO (OVID interface), CINAHL Plus, and EThoS and ProQuest libraries. The only exclusion criterion applied to the search strategies was for studies in languages other than the English language.
## Study selection
PD conducted all searches and the initial screening of all titles and abstracts; CG and TW screened 20% each of Intervention 'A coordinated and structured programme designed to remove or reduce the underlying causes of cardiovascular disease' to 'include a range of interventions with health education, lifestyle advice, stress management and physical exercise components'.
Comparison None
## Outcome barriers and enablers
Study type Any empirical the total titles and abstracts. Following the initial screening, PD read the full text of all potentially eligible articles. CG and TW reviewed 50% each of the total of fulllength articles against the eligibility criteria. To ensure saturation in sources, extensive backward and forward citation tracking was applied to reference lists of relevant articles and key texts. Any discrepancies in selection were discussed between the reviewers and a fourth reviewer (JVvZ) was available for arbitration if needed. No additional information had to be sought from study authors to inform eligibility decisions. The review authors were not blind to the journal titles, study authors or institutions of the full-text articles.
## Study appraisal
We used four different quality assessment tools for different study designs in line with the National Institute for Health and Care Excellence manual on developing guidelines. The chosen study appraisal tools are listed in. Using the most suitable quality assessment tool, a total numerical score obtained for each study was re-calculated into percentages and assigned into the following categories of quality: low (below 20%), low-to-medium (20-44%), medium (45-69%), mediumto-high (70-89%) and high (above 90%). The quality assessment was conducted by PD and TW who independently scored all of the included studies.
## Data extraction
PD extracted study characteristics and any relevant data on factors influencing delivery of cardiac rehabilitation from the included studies using a data extraction template. Data extraction for all included studies was verified by CG and TW. The extracted study characteristics included: author, year, study design/methods of data collection, country/setting, sample size, study/report aim and healthcare professional population. Extraction of the data pertaining to provider-and system-level barriers and enablers associated with the delivery of cardiac rehabilitation for heart failure included first-order constructs (data from the original study participants) and second-order constructs (assumptions and observations made by the studies' researchers). The review team only included reported data (i.e., a lack of barrier was not entered as an enabler unless the article clearly stated that).
Passages of text describing barriers and enablers were inputted and organised in the nVivo softwareand summarised into a table available in the Additional file 2.
## Data synthesis
In developing our analytic approach, we followed guidance on the selection of qualitative evidence synthesis methods for health technology assessments of complex interventionsand the seven-domain RE-TREAT framework. The following components of the framework were considered: the type of the review question, the review's purpose and the targeted audience, the timeframe, availability of resources and expertise, and the type of available data. Consequently, we conducted a narrative review of the qualitative data, using the following tools and techniques as described by Popay et al. For building preliminary synthesestextual descriptions, tabulation, groupings and clusterings and thematic analysis, for exploring relationships within and between studiesconcept triangulation and consideration for variability in outcomes, study design and study population, and for assessing robustness of findingscritical reflection. Additionally, categories identified during the thematic analysis were further considered according to the level of influence from the social ecological model.
# Thematic analysis
All data relevant to the research question was entered into the nVivio software. The verbatim text of first and second-order constructs representing barriers and enablers was organised thematically using thematic coding procedures described by Braun and Clarke. First and second-order constructs were given the same weight in the final analysis. The coding scheme emerged inductively following reading and rereading of the original data sources and discussions between the core review team (PD and CG). The final coding scheme consisted of a small number of overarching categories and a larger set of more granular themes within each category. The identified themes were further analysed in terms of their frequency and prominence (identifying the most common themes across the data set and their spread).
## Triangulation protocol
A triangulation protocol was used to summarize similarities and differences between different data sources. Each theme was considered in each data source and categorised as being in agreement, partial agreement or dissonance. An additional category (isolation) was created for themes that were neither confirmatory nor contradictory, as they simply added a concept that was not identified in other studies. In case of disagreement between data sources, further data within the articles (e.g. year of publication, differences in populations or methods used) was considered as potential explanations of such discrepancies.
# Results
All searches were conducted in October 2019 by PD and updated in March 2021. The searches identified 9654 articles, of which 3444 were duplicates. Following the screening of titles and abstracts of 6210 articles, 46 fulltext articles were obtained, and seven articles were included for analysis. The full search results are presented in.
## Study characteristics
The characteristics of the included studies and the quality assessment tools/scores are presented in. There was a little demarcation between studies in terms of setting (centre-based cardiac rehabilitation programmes taking place in hospitals or community settings), healthcare professionals involved in the study (members of multidisciplinary teams that ordinarily care for patients with heart failure), methods of data collection (mostly qualitative methods utilising document analysis, survey questionnaires, interviews, focus groups and observations) or evidence quality (medium to medium-to-high). The included studies were published between 2010 and 2020 and represented mostly European healthcare systems (i.e., Denmark, Netherlands, the UK and the European Society of Cardiology affiliated countries) or western healthcare systems (i.e., Australia, Canada and the USA). Five studies were rated as being of medium quality and two were rated as medium-to-high quality.
# Thematic analysis
During the process of thematic analysis, the identified barriers were organised into nine categories and 24 themes. The same categories, except one ('the origins of cardiac rehabilitation and previous practices') emerged in the thematic analysis of reported enablers; the enablers were further divided into 26 themes.contains a summary of the thematic analysis, the main analysis used to analyse available data. This table lists the identified categories and themes, highlights each theme frequency and coverage and, where possible, matches a theme related to a barrier with a counteracting enabler. 'The organisation of healthcare system' was the most frequent category for both barriers (15 instances) and enablers (15 instances) and this category was mentioned at least once in all of the included articles. The other most frequent categories related to barriers were 'the organisation of cardiac rehabilitation programmes', 'healthcare professional' and 'guidelines'. The same categories were the most frequent categories describing enablers. Themes pertaining to barriers that were quoted most frequently in the included studies were 'lack of resources: time, staff, facilities and equipment' and 'professional's knowledge, awareness and attitude'. The latter was also the most CR Cardiac rehabilitation, HF Heart failure, N/R Not reported frequently identified enabler.apportions the identified categories relating to barriers and enablers.
## Triangulation of themes across the data sources
Convergence analysisrevealed that 50% (12) of themes related to barriers and 53% (14) of themes related to enablers appeared as isolated concepts. There was agreement or partial agreement for 50% (12) of the identified barriers and dissonance was identified for 8%: 'poor professional's knowledge, awareness and attitude' and 'safety concerns'themes that showed the most complex convergence relationship (agreement, partial agreement and dissonance). Piepoli et al. concluded that 'perceived lack of importance, safety concerns, physicians not being confident or not having sufficient skill or knowledge and uncertainties about the usefulness all played a marginal role'. Similarly, Dalal et al. found that 'more than half (54%) of the centres expressed confidence in the skill mix and knowledge of their staff to provide cardiac rehabilitation in heart failure', as well as that 'a lack of evidence on safety or clinical benefit was not a factor that influenced most centres' ability to offer cardiac rehabilitation'.
Thus, Piepoli et at. and were in agreement about a marginal influence of 'poor professional's knowledge, awareness and attitude' and 'safety concerns', that was at odds with the remaining data sources, which recognised those as substantial barriers. Additionally, Piepoli et al. concluded that 'lack of resources: time, staff, facilities and equipment' was a barrier affecting non-Western regions of the European Society of Cardiology affiliated countries only. This partial agreement with two other studies might be linked with Piepoli et al. considering in their analysis several distinct geographical areas and therefore capturing a more nuanced picture in the results.
Fifty-seven percent (4) of sources were aligned regarding the top potential factor positively impacting the delivery of cardiac rehabilitation for heart failure (i.e., 'professional's knowledge, awareness and attitude'). Twelve (46%) enabler themes were classified as being in agreement with at least one additional data source. Only implementation of the existing ones'. This was in conflict with two other studies, which reported guideline endorsement as a potentially enabling factor.
# Discussion
The systematic review identified a wide range of provider-and system-level barriers and enablers affecting The origins of CR and previous practices
The outdated practise of bed restEvidence-base Poor evidence-base supporting CR for HFSufficient evidence-base supporting CR for HFGuidelines
Guidelines not tailored to the end-userBetter tailoring of guidelinesVolume and complexity of guidelinesTranslating guidelines into clinical algorithmsLack of inclusion of CR in local guidelinesGuideline endorsementCross-institutional guidelinesGuideline implementationEducation Lack of formal education on exercise trainingEducation programmes on the importance of exercise trainingKnowledge sharing opportunitiesAwareness-raisingMedical insurance Lack of medical insurance coverMedical insurance eligibility criteria and sufficient coverResources Lack of resources: time, staff, facilities and equipmentAdequate resources: time, staff, facilities and equipmentThe organisation of healthcare system Lack of commissioningSufficient commissioningBlurred professional rolesClear professional roles and responsibilitiesLack of integration between organisationsBetter integration between organisationsLack of patient pathwaysReferral systemInadequate IT systemsAdequate IT systemsLack of integration between departmentsBetter integration between departmentsLack of care standardisationCare standardisationLack of implementation strategiesLack of referralsHealthcare legislationPerformance and target measuresUse of clinical algorithmsThe organisation of CR programmes
Lack of different modes of deliveryAvailability of different modes of deliveryLack of programmesAvailability of programmes (specialised and communitybased)Limiting eligibility criteriaBroadened eligibilityDifficult to choose a suitable programmeConfusing referral proceduresHealthcare professional Poor professional's knowledge, awareness and attitudeSufficient professional's knowledge, awareness and attitudeSafety concernsImproving the doctor-patient relationshipCR Cardiac rehabilitation, HF Heart failure the delivery of cardiac rehabilitation for heart failure and linked the identified barriers with possible solutions. The broad array of factors identified may reflect the complexity of the phenomenon or it may reflect the range of healthcare systems and implementation contexts studied. Encouragingly, most of the identified barriers were matched with potential 'enablers' or solutions. The most prevalent barriers were 'poor professional's knowledge, awareness and attitude', 'lack of resources: time, staff, facilities and equipment' and 'safety concerns'. Interestingly, the most prevalent themes also showed some dissonance, with one of the most recent studiespresenting a more nuanced and updated picture relating to those factors. Namely, that lack of resources might not be as much of a barrier in Western regions of the European Society of Cardiology affiliated countries as opposed to more poorly-resourced areas and that professionals' knowledge and safety concerns may no longer be as prevalent as they have been previously reported. The latter dissonance might be linked with changing attitudes of healthcare professionals as a result of a gradually improving evidence-base for offering cardiac rehabilitation to heart failure patients.
The majority of identified barriers were consistent with literature outlining more generic barriers to implementation of healthcare services. Examples of this are the system, staff and intervention-level barriers affecting implementation of novel interventions identified by Geerligs et al.or barriers to change identified by the National Institute for Health and Care Excellence (e.g. staff awareness, knowledge, workforce skills, resources and political environments). A barrier identified in the review that might be particularly pertinent to the delivery of cardiac rehabilitation for heart failure patients that has not been considered extensively in other literature is 'the origins of cardiac rehabilitation'. The awareness of healthcare staff of the benefits of cardiac rehabilitation (as opposed to the outdated practice of bed rest) is a strong predictor of cardiac rehabilitation referral.
The identified categories of barriers and potential solutions fit well with the social ecological model which has previously been used to identify influences impactingTriangulation of reported barriers and enablers across the data sources (Continued)
M-H Medium-to-high, M Medium, ✓ Agreement, ✓ Partial agreement, ✗ Dissonance An empty field Silence, □ Isolated idea, CR Cardiac rehabilitation, HF Heart failure healthcare delivery at several different levels. These include the macrosystem encompassing widely shared cultural/social values, beliefs, customs and laws (e.g. public policies, enabling environments), the exosystem capturing the indirect environment (e.g. economic system, political system, educational system, governmental system, community-level influences), the microsystem describing the interpersonal environment (e.g. a small group of professionals who work together on a regular basis) and the mesosystem capturing the interactions between microsystem and exosystem (e.g. organisation-level influences). The most granular level of influence is the individual level, in this case, understood as an intrapersonal environment (e.g. a healthcare professional providing care to individual patients).
Barriers to the delivery of cardiac rehabilitation for heart failure patients are varied and multi-levelled and overcoming them will involve changes at different levels. This reflects the suggested 're-engineering of health care system' and 'progressive policy' in the recently published Journal of the American College of Cardiology expert panel report. Individual and microsystem-level initiatives include creating inter-professional knowledge-sharing opportunities or in-house monitoring and evaluation of the management of heart failure patients. These solutions can be implemented by individual cardiac rehabilitation teams.
An example of a practical solution from the mesosystem of influence was introducing an automated referral system to mitigate barriers linked with poor clinical knowledge. Such organisational level solutions may also facilitate the development of local patient pathways (which in turn may lead to the provision of more integrated healthcare).
Exosystem and macrosystem-level solutions related to the availability of resources and the creation of further evidence require collaborations between many different stakeholders and rely on policy-level changes and improvements (e.g., development of cross-institutional guidelines or increasing insurance cover).
In recent years, healthcare systems have been described as complex and adaptive. A change in one part of the system can lead to changes to other components, for example offering education to healthcare professionals on the benefits of cardiac rehabilitation in heart failure patients may lead to development of inter- - Better implementation of existing guidelines CR Cardiac rehabilitation, HF heart failure, HFpEF Heart failure with preserved ejection fraction professional collaborations or inspire service providers to use novel delivery systems.
# Strengths and limitations
To the best of our knowledge, this is the first systematic review investigating provider-and system-level factors affecting the delivery of cardiac rehabilitation for heart failure. The review applied robust methods, i.e., systematic search strategy, second coding of study selection and study quality procedures, use of comprehensive narrative synthesis techniques that included thematic analysis and triangulation of identified themes to maximise depth and robustness of the findings. Additionally, the included studies used different methodologies leading to triangulation of available data and increasing rigour of the systematic review findings. Despite applying a very inclusive search strategy the review identified only seven studies meeting the inclusion criteria. The paucity of empirical studies and/or relatively poor quality of empirical data limits the findings and increases the possibility of a publication bias being present in the final synthesis. Additionally, although including second-order constructs increased the overall amount of data, the origins and robustness of the second-order constructs were difficult to establish.
Due to limitations of the data reported in the reviewed literature, we were unable to consider how representative the sample was of professionals involved in the delivery of cardiac rehabilitation for heart failure. However, we were able to identify that the sample was restricted mainly to European and Western healthcare systems. Therefore the generalisability of the identified barriers and enablers is limited to this context. Furthermore, the literature that we reviewed did not report characteristics of the patient populations served or consider how barriers might vary depending on patient characteristics (e.g. some healthcare professionals may be less willing to invite more frail patients for cardiac rehabilitation).
# Future research
Further research is needed to identify barriers in other healthcare systems and in a wider, more clearly defined range of healthcare professionals. Future implementation studies could also seek to identify any barriers and enablers that apply differently to different patient groups. Further research is also needed to qualitatively investigate barriers that are unique to the heart failure population (e.g. the origins of cardiac rehabilitation) and barriers that showed divergent relationships between sources included in our review (e.g. the impact of professional's knowledge, guidelines, safety concerns and lack of resources).
The gaps in the literature, uncovered by the systematic review, confirmed a continuing dearth of implementation studies on the topic of cardiac rehabilitation for heart failure and an ongoing need for further high quality research that goes beyond patient-level factors affecting the delivery of cardiac rehabilitation for heart failure. Such research is acutely needed in the light of initiatives to improve access to and uptake of cardiac rehabilitation for heart failure, such as the National Health Service Long Term Plan that aims to increase the proportion of eligible heart failure patients accessing cardiac rehabilitation from less than 10 to 33% by 2028.
# Conclusions
This systematic review identified a broad range of provider-and service-level factors affecting the delivery of cardiac rehabilitation for heart failure. The identified barriers and enablers operate on multiple levels of influence from the knowledge and views of individual healthcare professionals to the organisation of cardiac rehabilitation teams and the wider healthcare system. Consequently, efforts to increase the delivery of cardiac rehabilitation for patients with heart failure will likely require intervention at all these levels. Strategies for improving delivery of cardiac rehabilitation for heart failure may include increasing inter-professional collaboration, providing choice between hospital and home-based rehabilitation programmes, inclusion of cardiac rehabilitation for heart failure in local commissioning contracts and staff-education initiatives to raise awareness of the importance of cardiac rehabilitation and of the evidencebase on the benefits and safety of cardiac rehabilitation in patients with heart failure. Funding PD's time is funded by a PhD studentship from the University of Birmingham.
## Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
|
Competition between bridged dinucleotides and activated mononucleotides determines the error frequency of nonenzymatic RNA primer extension
Nonenzymatic copying of RNA templates with activated nucleotides is a useful model for studying the emergence of heredity at the origin of life. Previous experiments with defined-sequence templates have pointed to the poor fidelity of primer extension as a major problem. Here we examine the origin of mismatches during primer extension on random templates in the simultaneous presence of all four 2-aminoimidazole-activated nucleotides. Using a deep sequencing approach that reports on millions of individual template-product pairs, we are able to examine correct and incorrect polymerization as a function of sequence context. We have previously shown that the predominant pathway for primer extension involves reaction with imidazoliumbridged dinucleotides, which form spontaneously by the reaction of two mononucleotides with each other. We now show that the sequences of correctly paired products reveal patterns that are expected from the bridged dinucleotide mechanism, whereas those associated with mismatches are consistent with direct reaction of the primer with activated mononucleotides. Increasing the ratio of bridged dinucleotides to activated mononucleotides, either by using purified components or by using isocyanidebased activation chemistry, reduces the error frequency. Our results point to testable strategies for the accurate nonenzymatic copying of arbitrary RNA sequences.GRAPHICAL ABSTRACT
# Introduction
The emergence of RNA-based life would have required a mechanism for RNA replication without enzymes [bib_ref] Was RNA the first genetic polymer?, Orgel [/bib_ref] [bib_ref] The antiquity of RNA-based evolution, Joyce [/bib_ref] [bib_ref] The origins of the RNA world, Robertson [/bib_ref] [bib_ref] On the emergence of RNA, Krishnamurthy [/bib_ref] [bib_ref] The narrow road to the deep past: in search of the chemistry..., Szostak [/bib_ref] [bib_ref] Protocells and RNA self-replication, Joyce [/bib_ref]. Replication, in turn, depends on template-directed polymerization of nucleotide building blocks--a copying step during which information stored in one piece of RNA is transferred to another. Nonenzymatic primer extension has been used extensively to study template-directed polymerization [bib_ref] Enhanced nonenzymatic RNA copying with 2-aminoimidazole activated nucleotides, Li [/bib_ref] [bib_ref] Enzyme-free ligation of dimers and trimers to RNA primers, Sosson [/bib_ref] [bib_ref] Template-Directed Copying of RNA by Non-enzymatic Ligation, Zhou [/bib_ref] [bib_ref] The virtual circular genome model for primordial RNA replication, Zhou [/bib_ref]. Nonenzymatic primer extension does not work with nucleoside 5 triphosphates because the associated activation energy is too high [bib_ref] Kinetic and mechanistic analysis of nonenzymatic, template-directed oligoribonucleotide ligation, Rohatgi [/bib_ref] , so more reactive chemistries have been developed [bib_ref] Template-Directed Copying of RNA by Non-enzymatic Ligation, Zhou [/bib_ref] [bib_ref] Non-enzymatic template-directed synthesis on RNA random copolymer: Poly(C,U) templates, Joyce [/bib_ref] [bib_ref] Non-enzymatic template-directed synthesis on RNA random copolymers: poly(C, G) templates, Joyce [/bib_ref] [bib_ref] Non-enzymatic template-directed synthesis on random copolymers: Poly(C,A) templates, Joyce [/bib_ref] [bib_ref] Copying of RNA sequences without pre-activation, Jauker [/bib_ref] [bib_ref] The effect of leaving groups on binding and reactivity in enzyme-free copying..., Kervio [/bib_ref] [bib_ref] Enzyme-free replication with two or four bases, Hanle [/bib_ref] [bib_ref] Template-directed nonenzymatic primer extension using 2-methylimidazole-activated morpholino derivatives of guanosine and cytidine, Zhang [/bib_ref] [bib_ref] Nonenzymatic template-directed synthesis of mixed-sequence 3 '-NP-DNA up to 25 nucleotides long..., O'flaherty [/bib_ref]. A prebiotically plausible activating group for mononucleotide-based primer extension is 2-aminoimidazole (2AI) because it can be synthesized together with 2-aminooxazole, a nucleotide precursor, and can mediate the copying of short mixed sequence templates [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref] [bib_ref] Common and potentially prebiotic origin for precursors of nucleotide synthesis and activation, Fahrenbach [/bib_ref]. Efficient primer extension with imidazole-based nucleotide activation requires the formation of a 5 -5 phospho-imidazolium-phospho bridged dinucleotide intermediate [bib_ref] The mechanism of nonenzymatic template copying with imidazole-activated nucleotides, Walton [/bib_ref] [bib_ref] Potentially prebiotic activation chemistry compatible with nonenzymatic RNA copying, Zhang [/bib_ref]. The bridged dinucleotide binds the template adjacent to a primer, the 3 hydroxyl of which attacks the proximal bridging phosphate and displaces an activated nucleotide as the leaving group. Consequently, nonenzymatic RNA primer extension proceeds one base at a time even though the bridged dinucleotide pathway involves two covalently bonded nucleotides.
Mapping a prebiotic route to the emergence of heredity requires an understanding of the origin and frequency of mismatches during nonenzymatic copying. A high error rate during copying steps would quickly corrupt any beneficial sequences that may have emerged from a pool of replicating RNA molecules [bib_ref] Selforganization of matter and the evolution of biological macromolecules, Eigen [/bib_ref] [bib_ref] Effect of stalling after mismatches on the error catastrophe in nonenzymatic nucleic..., Rajamani [/bib_ref]. Nonenzymatic copying is error-prone because it relies entirely on base-pairing, without the benefit of substrate-selective enzymes and mismatch-repair machinery. Consequently, copying G/C templates with activated G and C nucleotides occurs with higher fidelity than combinations that include A/U/T [bib_ref] Enzyme-free replication with two or four bases, Hanle [/bib_ref] [bib_ref] Chemical primer extension: efficiently determining single nucleotides in DNA, Hagenbuch [/bib_ref] [bib_ref] Templating efficiency of naked DNA, Kervio [/bib_ref] [bib_ref] The prebiotic evolutionary advantage of transferring genetic information from RNA to DNA, Leu [/bib_ref] [bib_ref] Replacing Uridine with 2-Thiouridine Enhances the Rate and Fidelity of Nonenzymatic RNA..., Heuberger [/bib_ref]. Furthermore, mismatches have been found to slow down the incorporation of subsequent nucleotides, which are also more likely to be mismatched [bib_ref] Non-enzymatic primer extension reactions: stalling factors for mismatch extensions and misincorporations, Rajamani [/bib_ref]. This stalling effect may enable correct copying to kinetically outcompete incorrect copying [bib_ref] Effect of stalling after mismatches on the error catastrophe in nonenzymatic nucleic..., Rajamani [/bib_ref]. These and many other studies have relied on templates with defined sequences and usually one nucleotide species as a reactant so that the identities of primer extension products could be readily interpreted. In experiments with multiple activated nucleotides, or even all four, templates still had to be defined so that mismatch identities could be inferred from the known template sequences [bib_ref] Enzyme-free replication with two or four bases, Hanle [/bib_ref] [bib_ref] Replacing Uridine with 2-Thiouridine Enhances the Rate and Fidelity of Nonenzymatic RNA..., Heuberger [/bib_ref] [bib_ref] Cascade of reduced speed and accuracy after errors in enzyme-free copying of..., Leu [/bib_ref]. A more complete understanding of how sequences are copied and how errors arise demands an assay that can characterize products on random templates with all four nucleotide reactants.
We recently developed NonEnzymatic RNA Primer Extension Sequencing (NERPE-Seq) to investigate heterogeneous reaction systems containing arbitrary template sequences and reactant nucleotides [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. NERPE-Seq uses a self-priming RNA hairpin with a user-defined template. For the work presented here, the template is six random bases, and the hairpin construct is incubated with a mixture of all four 2-aminoimidazole-activated RNA nucleotides (2AIrN), isolated, processed, and deep-sequenced. The sequencing data consists of each individual templateproduct pair because the template and primer are physically connected in the parent hairpin. A custom open-source analysis software package is used to filter, sort and process the raw sequencing reads to generate information about complementary (correct) and mismatched (incorrect) products and their sequence contexts [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. These experiments tell us the yields of heterogeneous reactions, encompassing all combinations of base-dependent reactivities, competitive inhibitions, and mismatches. Here we use NERPE-Seq to identify complementary products that are preferentially synthesized, their sequence contexts, and how well primer extension samples sequence space. We also identify all mismatches generated by primer extension, and their sequence contexts. We find that complementary products harbor sequence patterns consistent with a bridged dinucleotide pathway for nucleotide incorporation, whereas se- quence patterns associated with mismatches are consistent with an activated mononucleotide pathway. Finally, we test how a prebiotically plausible isocyanide-based activation chemistry [bib_ref] Potentially prebiotic activation chemistry compatible with nonenzymatic RNA copying, Zhang [/bib_ref] [bib_ref] A light-releasable potentially prebiotic nucleotide activating agent, Mariani [/bib_ref] , which has been shown to increase the yield of bridged dinucleotides, affects the reaction.
# Materials and methods
## General
Reagents, reaction conditions, syntheses, the NERPE-Seq protocol, and the data analysis code package were as previously described [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. Minor changes to the final step of sample preparation prior to sequencing submission, and the addition of a new module to the analysis code, are described below. We summarize the general procedures here in condensed form. Bicine buffer was prepared from the Na + salt and HEPES from the free acid (Sigma-Aldrich), adjusted to pH 8 with NaOH, and filtered. Enzymes and DNA/RNA ladders were purchased from New England BioLabs (NEB). Incubations at a specified temperature were performed in a Bio-Rad T100 thermal cycler. Reactions with methyl isocyanide were carried out at room temperature (∼23 - C).
## Synthesis of 2-aminoimidazole-activated monoribonucleotides
0.63 mmol (1 equivalent) of the nucleoside-5monophosphate free acid (Santa Cruz Biotechnology or ACROS Organics) and 5 equiv. of 2-aminoimidazole hydrochloride (Combi-Blocks) were dissolved in water, the pH adjusted to 5.5 with NaOH, and the mixture lyophilized. 30 ml dry dimethyl sulfoxide (DMSO, Sigma-Aldrich) and 1.2 ml dry triethylamine (TEA, Sigma-Aldrich) were mixed under argon. To this was added the lyophilized nucleoside-5 -monophosphate and 2-aminoimidazole hydrochloride, 9 equiv. of triphenylphosphine (TPP, Sigma-Aldrich), and 10 equiv. of 2,2 -dipyridyldisulfide (DPDS, Combi-Blocks). After 30 min the mixture was poured into an ice-chilled solution of 400 ml acetone (Fisher Scientific), 250 ml diethyl ether (Fisher Scientific), 30 ml TEA, and 1.6 ml acetone saturated with NaClO 4 (Sigma-Aldrich). After 30 min the supernatant was removed. The remaining mixture with the flocculant precipitate was centrifuged, the supernatant discarded, and the pellet washed in a solution of 1:8.3:13.3 of TEA:diethyl ether:acetone and centrifuged again. The wash was repeated twice with just acetone, and the pellets dried overnight under vacuum. Each sample was dissolved in 5 ml water, and the combined volume purified by reverse-phase flash chromatography (CombiFlash, Teledyne ISCO) over a 50 g RediSep Rf Gold C18Aq column (Teledyne ISCO) using gradient elution between (A) water and (B) acetonitrile. Target fractions were placed on ice, pooled, adjusted to pH 10 with NaOH, lyophilized, and stored at -80 - C. To prepare experimentally convenient quantities, purified material was resuspended in water and the nucleotide concentration measured by a NanoDrop 2000c spectrophotometer (ThermoFisher Scientific). For experiments with all four activated nucleotides, aliquots were prepared with equimolar concentrations of each. Aliquots were lyophilized, stored at -80 - C, and resuspended in water immediately prior to use.
Oxyazabenzotriazole-activated monoribonucleotides were prepared as in [bib_ref] Template-directed catalysis of a multistep reaction pathway for nonenzymatic RNA primer extension, Walton [/bib_ref]. Briefly, a lyophilized mixture of 0.2 g 1-hydroxy-7-aza-benzotriazole (Combi-Blocks), 0.1 g nucleoside-5 -monophosphate free acid, and TEA was resuspended in 20 ml DMSO and reacted with 1.5 g of DPDS and 1.5 g TPP. After 1-2 h, the nucleotide was precipitated with a solution of 120 ml cold acetone, 60 ml diethyl ether, and 4.5 g sodium perchlorate. For OAtrG, the reaction mixture was left overnight, followed by a second addition of 1 g DPDS and 1 g TPP, incubated for 1-2 h, and then precipitated. The precipitate was purified by reverse-phase flash chromatography and lyophilized.
## Synthesis and isolation of 5 -5 phospho-imidazoliumphospho bridged dinucleotide intermediates
Each bridged dinucleotide, N1*N2, was prepared by first synthesizing 2AIrN1 and OAtrN2. 2AIrN1 was then mixed with at least a 1.2× excess of OAtrN2 in water, and adjusted to pH 8.0 with NaOH/HCl. The mixture was stirred at room temperature for 1 h, then purified by reverse-phase flash chromatography using gradient elution between (A) 2 mM TEAB, pH 8 and (B) acetonitrile and lyophilized.
## Gels and gel analysis
Polyacrylamide gels (20%) were prepared with the SequaGel-urea gel system (National Diagnostics) and cast at 0.75 mm thick (1.5 mm for preparative) using 20 × 20 cm plates. Gels used to assay the sequencing constructs, which have no attached dyes, were removed from the glass plates and incubated in ∼140 ml of TBE + 14 l SYBR Gold Nucleic Acid Gel Stain (Invitrogen) for several minutes then destained in TBE for ∼10 min. Gels were imaged with a Typhoon 9410 scanner (GE Healthcare). For analysis and visualization, TIFF-formatted images were imported into Fiji [bib_ref] Fiji: an open-source platform for biological-image analysis, Schindelin [/bib_ref]. Bands were quantified using the Gel Analysis function, with relative band intensities reported as the ratio of the band intensity to the total lane intensity. Band edges were excluded. Contrast and color changes were applied uniformly to the entire gel image in all cases.
## Nmr measurement of bridged dinucleotide concentrations
Deuterated solvents were purchased from Cambridge Isotope Laboratories. 1 H and 31 P spectra were acquired on a Varian Oxford AS-400 NMR spectrometer (400 MHz for 1 H, 161 MHz for 31 P). Chemical shifts are reported in parts per million (ppm) on the ␦ scale. 1 H NMR was referenced using the solvent resonance as the internal standard (HDO, 4.79 ppm at 25 - C). All NMR spectra were recorded at 25 - C. Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quadruplet, m = multiplet), integration and coupling constants. A 600 l sample was prepared with 10 mM of each of the four 2AI-activated monomers and 200 mM deuterated Tris-Cl, pH 8. 1 H and 31 P spectra were acquired immediately after the sample was prepared, then at 1, 3, 6 and 24 h. Trace TEA was used as an internal standard for integration (␦ 3.20 (q, J = 7.3 Hz, 2H), 1.28 (t, J = 7.3 Hz, 3H)). 1 H spectra were baseline corrected in MNova (Mestrelab Research), and calibrated based on TEA present in the sample from flash chromatography. The TEA concentration was determined by its integration ratio to the known concentration of mononucleotides at the initial timepoint. Activated mononucleotide and bridged dinucleotide concentrations at subsequent time points were measured by integrating corresponding 1 H peaks. Some of the bridged dinucleotides exhibit overlapping peaks, in which case their concentrations were determined from the 31 P spectra instead, using the peaks of other bridged dinucleotides as internal standards. Each individual bridged dinucleotide accumulates to a relatively low concentration in the 1 mM range, such that background noise contributes to the error associated with measuring their concentrations. Baseline correction was used to eliminate spectral artifacts that would otherwise incorrectly contribute to peak integrations [bib_ref] Recommended strategies for spectral processing and post-processing of 1D 1H-NMR data of..., Emwas [/bib_ref] [bib_ref] Baseline correction for NMR spectroscopic metabolomics data analysis, Xi [/bib_ref]. The error associated with each peak integration was calculated using the signal-to-noise (SNR) script in Mnova.
## Methyl isocyanide activation
The absolute concentrations of methyl isocyanide (MeNC, prepared as in (25)) and 2-methylbutyraldehyde (Tokyo Chemical Industry Co., highest available purity) were determined by comparing the integrals of their 1 H-NMR peaks to those of adenosine-5 -monophosphate, a calibrant of known concentration. Primer extension reactions in the presence of methyl isocyanide-mediated bridge-forming activation were carried out in the same manner as for standard primer extension but with 200 mM HEPES, pH 8, as the buffer, 200 mM MeNC, 200 mM 2-methylbutyraldehyde and 30 mM MgCl 2 .
## Rna sample preparation for deep-sequencing
Supplemental lists oligonucleotides used in this study, and Supplemental lists specific conditions for each deep sequencing experiment. Generally, 1 M hairpin construct and 1.2 M 5 Handle Block were annealed in 200 mM buffer by incubation at 95 - C for 3 min, and cooling to 23 - C at 0.2 - C/s. The activated monoribonucleotide reactants, typically 20 mM 2AI-activated nucleotides (2AIrN), and 50 mM MgCl 2 (unless otherwise indicated) were added to initiate the reaction. The total final reaction volume was 30 l. The mixture was briefly vortexed and incubated at 23 - C with a 42 - C heated lid, quenched, desalted, and the NPOM (caged) bases uncaged. Samples were PAGE-purified (ZR small-RNA PAGE Recovery Kit, Zymo Research). The RT Handle (template for the reverse transcription primer) was ligated with T4 RNA Ligase 2, truncated KQ. The mixture was treated with Proteinase K, phenol-chloroform extracted, and concentrated with an Oligo Clean & Concentrator spin column (Zymo Research).
The RT Primer was annealed to the purified sample and the RNA reverse transcribed with ProtoScript II. The mixture was purified with an Oligo Clean & Concentrator spin column, and the eluted cDNA stock concentration measured by spectrophotometry. 0.1 g cDNA was added to a 50 l Q5 Hot Start High-Fidelity DNA Polymerase PCR reaction with 0.2 M each of NEBNext SR Primer for Illumina and NEBNext Index Primer for Illumina and run for 6 cycles with a 15 s 62 - C extension step. PCR product was purified by preparative agarose gel electrophoresis (Quantum Prep Freeze 'N Squeeze spin column, Bio-Rad) and magnetic beads (Agencourt AMPure XP, 1.7:1 volume ratio of magnetic bead suspension to sample volume). Samples were validated and concentrations measured by TapeStation (Agilent) instead of qPCR [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. Paired-end sequencing by MiSeq (Illumina, MiSeq Reagent Kit v3, 150 cycle) produced ∼20 million reads with ∼95% passing the instrument's quality filter. Previous work has evaluated the effects of 2 -5 linkages on the sequencing protocol and found that they do not measurably alter the reported sequencing data [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref] [bib_ref] A mechanistic explanation for the regioselectivity of nonenzymatic RNA primer extension, Giurgiu [/bib_ref] [bib_ref] Reverse-transcriptase Reads Through a 2 -5 -linkage and a 2 -thiophosphate in..., Lorsch [/bib_ref]. The experimentally established error associated with each reported base identity is 0.062 ± 0.13% [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref].
Sequencing data analysis was performed with the NERPE-Seq custom code package written in MATLAB (MathWorks) and described in [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. Briefly, the code filters raw data by using read quality scores, and by checking that reads agree with defined-sequence regions of the hairpin construct and that forward and reverse reads (which overlap) agree with each other. Template and product sequence pairs are extracted and characterized. The template sequence of the construct does not contain precisely equal fractions of rA, rC, rG and rU [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref] , so the template from a control experiment in which no activated nucleotides were added was used to generate normalization factors. All presented data are normalized--showing results as they would appear if the template were perfectly random with equal base ratios--unless otherwise indicated. A new code module reports the local sequence context of mismatches (a key for interpreting the printed results of this module has been added to the digital resource; see link at Data Availability). The numerical data used to generate heat maps is included in the Supplementary data.
# Results
## Products of primer extension copying of random sequence templates containing all four bases
Complementary product sequence patterns reflect reaction with the bridged dinucleotide intermediate. To begin exploring the efficiency and accuracy with which different sequences can be copied by nonenzymatic primer extension, we incubated 1 M of our random-template hairpin RNA with 20 mM 2AI-activated nucleotides (2AIrN) and 50 mM MgCl 2 for 24 h. The hairpin was then subjected to NERPE-Seq (Supplementary . [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref] shows the frequencies of complementary and mismatched nucleotide incorporations at each position. The reaction yield is low compared with experiments that use defined templates and reactants (9,34), but there is nonetheless a clear accumulation of products up to +3 and even +4. The distribution of product lengths is 13.5% +1, 1.5% +2, 0.5% +3 and <0.13% +4 (Supplementary . Increasing the concentration of 2AIrN from 20 to 100 mM yields more products (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref] , with a product length distribution of 27% +1, 10.4% +2, 6.6% + 3 and < 3.1% +4. Similarly, there are significantly fewer products in the presence of 100 mM free 2AI, which suppresses the accumulation of bridged dinucleotides (22) (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref].
To facilitate more detailed analyses, the data were sorted into products that are fully complementary and products with at least one mismatch. Based on prior results with defined templates, we anticipated a uniform bias in favor of primer extension with rG and rC [bib_ref] Enhanced nonenzymatic RNA copying with 2-aminoimidazole activated nucleotides, Li [/bib_ref]. However, among fully complementary products, the incorporation of the different nucleotides was highly non-uniform. There is a marked increase in the proportion of rG and especially rC incorporations as a function of extension length [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. The G/C-enrichment is less pronounced at position 1, which has a higher frequency of rA and rU products than downstream positions. To understand the reason for the increasing G/C bias with length, we examined the sequence context of each nucleotide incorporation event.
Because the bridged dinucleotide binds two tandem bases in the template, we examined the effect of both template positions on incorporation frequency [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. (Recall that in the bridged dinucleotide pathway, the first nucleotide--adjacent to the extending primer--becomes incorporated as the +1 product, but the second downstream nucleotide functions as a leaving group and diffuses away.) We identified two major patterns among inferred bridged dinucleotides [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. First, the frequencies of inferred bridged dinucleotides with a rG or rC in the first position are higher than the frequencies of those with a rA or rU in the first position, meaning that rG and rC incorporations are more common. Second, the highest frequency inferred bridged dinucleotides for each possible first position base have a rG or rC in the second position. These features are not merely due to some bridged dinucleotides forming more readily in solution than others (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. To verify that the observed G/C bias in downstream templating positions comes from bridged dinucleotides, we performed a NERPE-Seq experiment using oxyazabenzotriazole (OAt)-based activation. OAtrNs cannot form bridged dinucleotide intermediates, and poly-merize by a single-nucleotide-only pathway [bib_ref] Enzyme-free genetic copying of DNA and RNA sequences, Sosson [/bib_ref]. We did not find any second-position biases in the OAt-based experiment (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. We conclude that bridged dinucleotides with rG or rC in the second position are more likely to bind the template and react with the primer. Consequently, products extended to some length have a high probability of harboring a downstream templating rC or rG. This generates a position-dependent enrichment of rC and rG template bases among extended products, explaining the patterns of terminal product bases and bases downstream of terminal products [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref] , Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. The simplest explanation for these observations is that the relative affinities of bridged dinucleotides determine the templates on which primer extension proceeds.
Inferred bridged dinucleotide frequencies and template composition change with time and determine the composition of product sequences. We next considered how bridged dinucleotide frequencies, template composition and complementary product sequences are related [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. We began by asking whether the inferred bridged dinucleotide frequencies could be explained by known base pairing and stacking preferences, so we turned to the empiricallyderived Nearest Neighbor DataBase (NNDB) which catalogs the energetic stabilities of adjacent pairs of nucleotides [bib_ref] NNDB: the nearest neighbor parameter database for predicting stability of nucleic acid..., Turner [/bib_ref]. Our sequencing analysis reports cumulative yields and depends on out-of-equilibrium reactions, whereas the NNDB reports free energy changes associated with annealing. Nonetheless, if the various bridged dinucleotides are in comparable concentrations (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref] and competing for sites on a randomized template based on their intrinsic binding affinities, then we expect a correlation between their inferred frequencies derived from the sequencing analysis and their predicted equilibrium binding constants. We therefore plotted the frequencies of inferred bridged dinucleotides at position 1 (20 mM 2AIrN, 24 h incubation) against their corresponding predicted equilibrium constants [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. The inferred bridged dinucleotides cluster into four groups: combinations of rA and rU; combinations of rA/rU and rG/rC; A*G, C*A, G*G and G*C; and finally C*G and C*C, with C*C having both the highest measured frequency and the highest predicted equilibrium constant. Overall, we find an R 2 correlation coefficient of 0.76 between inferred bridged dinucleotide frequency and estimated K d . We conclude that the binding affinities of bridged dinucleotides are primarily responsible for driving complementary product formation.
We next sought to assess how the time-dependent changes in the base composition of the remaining available templates--i.e. template positions not occupied by a product nucleotide--affect inferred bridged dinucleotide frequencies. We performed a time series of the 20 mM 2AIrN experiment, taking periodic samples up to 3 days, and then measured the frequencies of all remaining available template bases in positions 1 and 2 at each timepoint [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. The frequencies of available rG and rC drop as those of rA and rU rise. (This analysis is for all available remaining templates, not just templates on which some extension has already occurred and where rG and rC are enriched as described above.) Similarly, more available sites have been . (C) rG and rC templates are used up by the more reactive bridged dinucleotides, leaving behind higher proportions of rA and rU templates. (D) More reactive inferred bridged dinucleotides participating in primer extension become less frequent with time as less reactive inferred bridged dinucleotides--dominated by combinations of rA and rU--become slightly more frequent.
consumed with 100 mM 2AIrN after 24 h than at the same timepoint with 20 mM 2AIrN because the reaction has been pushed forward. These changes are necessarily entangled with inferred bridged dinucleotide frequencies [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref] : as more of the tighter binding bridged dinucleotides lead to products on rG and rC templates, those bases become less frequent among available templates, prompting reductions in the frequencies at which rC-and rG-containing bridged dinucleotides can participate in extension. There is also a slight increase in rA-and rU-containing inferred bridged dinucleotide frequencies, with C*A increasing most dramatically, probably due to its relatively high stability [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref] and the steady rise in templating rU at position 2 [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. Finally, we sought to visualize how these phenomena cooperate to determine which complementary product sequences are accessed by primer extension. We measured the frequencies of products at least three bases long and plotted them by base identity at positions 1, 2 and 3 [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref] and Supplementary [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. The volume of each sphere at each coordinate is proportional to the frequency of that triplet. Most triplets are represented, excepting a few rA/rU-rich combinations. The most prominent feature for 20 mM 2AI is the dominance of rC in the third position, which is expected given the progressive selection of rGtemplates among extended products by the bridged dinucleotides [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. The distribution becomes more uniform with 100 mM 2AIrN and several of the triplets not accessed at 20 mM 2AI appear at low frequencies (Supplementary [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. Despite the challenges of copying random templates, we see examples of 59 and 64 out of 64 possible triplet sequences having been copied at 20 mM and 100 mM 2AIrN, respectively; the remaining triplet sequences may presumably be present but at levels so low that they were not represented in our sequencing data.
## Mismatches
Mismatch frequencies depend on position-dependent changes in template composition. Having characterized the properties of complementary products, we turned to examining mismatches. The frequency of mismatches relative to all incorporations is 11% with 20 mM 2AIrN and 13% with 100 mM 2AIrN [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. The proportion of complementary incorporations is much higher than that of mismatched incorporations at each position, demonstrating that in this system errors are not responsible for the majority of primer extension terminations. Although a mismatch is less likely to extend further than a correct incorporation [bib_ref] Cascade of reduced speed and accuracy after errors in enzyme-free copying of..., Leu [/bib_ref] , the majority of terminal base-pairs are correct (86% for 20 mM 2AIrN). This suggests that competition among reactants for binding sites [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref] [bib_ref] Potentially prebiotic activation chemistry compatible with nonenzymatic RNA copying, Zhang [/bib_ref] [bib_ref] Efficient enzyme-free copying of all four nucleobases templated by immobilized RNA, Deck [/bib_ref] [bib_ref] Adenosine residues in the template do not block spontaneous replication steps of..., Vogel [/bib_ref] and the lower reactivities of some reactants may inhibit extension in the presence of all four activated nucleotides [bib_ref] Replacing Uridine with 2-Thiouridine Enhances the Rate and Fidelity of Nonenzymatic RNA..., Heuberger [/bib_ref]. Furthermore, the position-dependent increase in template G/C content correlates with an increased fidelity at these positions [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. In position 1, rG is copied correctly at 95.1%, rC at 94.2%, rU at 88.4%, and rA at 54% (see below). In contrast, previous work has shown that with a defined G/C template and only 2AIrC and 2AIrG as reactants, rG is copied correctly at 99.9% and rC at 99.7% [bib_ref] Deep sequencing of non-enzymatic RNA primer extension, Duzdevich [/bib_ref]. This underscores the significant effect on fidelity of using all four nucleotides. It is likely that the copying of rC and rG in the template is more accurate than the copying of rA and rU because the G-C base-pair stability is greater than that of the A-U base-pair, G-U wobble-pairs and of other mismatched pairings.
To understand mismatch patterns in greater detail, we measured the position-dependent frequency of each possible mismatch [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. G-U wobble pairs are known to be the most stable mismatches [bib_ref] Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure, Mathews [/bib_ref] , but rG templates at position 1 are largely copied correctly [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. Consequently the A:G (template:product) pair is the most frequent mismatch at position 1 under these conditions. The G:U mismatch dominates downstream positions where there are fewer products overall [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. The bridged dinucleotide mechanism also enriches extended products for downstream templating rG, explaining the prominent streak of G:U mismatches after position 1 [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. The low frequency of G:A mismatches relative to A:G mismatches can also be explained by rG template depletion by correct rC incorporations at position 1. At downstream positions A:G and G:A take on comparable frequencies. U:G is also less common than G:U in position 1, but that seems to result from the identity of the priming base, which affects mismatch stabilities [bib_ref] Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure, Mathews [/bib_ref] and therefore incorporation preferences (Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. NERPE-Seq also enables us to measure the effect of the incorporation of a mismatched nucleotide on subsequent incorporations. For 20 mM 2AIrN at 24 h, 95% of mismatches are terminal, compared with 73% of correct incorporations. Of the mismatches that go on to prime further extension, 49% are followed by mismatches, which is much higher than the overall 11% mismatch frequency. This confirms previous observations in orthogonal systems that errors potentiate errors [bib_ref] Effect of stalling after mismatches on the error catastrophe in nonenzymatic nucleic..., Rajamani [/bib_ref] [bib_ref] Cascade of reduced speed and accuracy after errors in enzyme-free copying of..., Leu [/bib_ref]. Thus, the phenomenon of a cascade of errors also applies in an all-RNA, 2AI-activation based primer extension reaction with randomized components. Furthermore, mismatches that are followed by correct incorporations tend to have rC as the product nucleotide, presumably because rC stacks favorably with incoming bases. In addition the two G-U wobble base-pairs, which are energetically stable and are geometrically most similar to a correct base-pair, are often followed by incorporation of a correct nucleotide (Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. Tandem mismatches show a different pattern: they are depleted in product rU and to a lesser extent rA, combinations of which do not stack as favorably as combinations of rG and rC (43,47) (Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref].
Mismatch frequency depends on the ratio of activated mononucleotides to bridged dinucleotides. Having considered the distribution of mismatches, we next sought to discern how they form. We hypothesized that if they arise through the bridged dinucleotide pathway, then templating bases downstream of mismatches should carry a discernable signature, just as for complementary products [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. In particular, we expected that because mismatches are less stable than correct base-pairs, the second base of any associated inferred bridged dinucleotides would be strongly enriched for rG or rC; in effect, to drive the mismatch. This would appear as a bias for templating rC and rG downstream of mismatches. However, if mismatches originate from the reaction of activated mononucleotides, there should be no such enrichment among templating bases downstream of mismatches. We therefore measured the frequencies of templating bases downstream of all mismatches [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. Instead of a G/C bias we found roughly equal proportions of rA, rC and rU, with rG slightly less well represented. This result suggests that mismatches arise from activated mononucleotides, and not the bridged dinucleotide.
To further test the hypothesis that mismatches arise preferentially from the reaction of a primer with an activated mononucleotide, we repeated the measurement of template bases downstream of mismatches on sequence data from the experiment with 100 mM free 2AI, which reduces the concentration of bridged dinucleotides [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref]. These conditions drastically reduce the amount of extended product (Supplementary [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref] , but sufficient data was obtained to allow for an analysis of mismatch sequence context. The results show a similar unbiased distribution of template bases in the downstream position to that found above [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. Finally, a similar result was also obtained from the experiment with 20 mM OAt-activated mononucleotides, which cannot form bridged dinucleotides. The relative frequencies of mismatches, especially at position 1, show a similar pattern to that obtained with 20 mM 2AIrN [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref] and Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref] , respectively). This similarity is striking given that the error frequency in the 20 mM OA-trN experiment was 44%, compared with 11% in the 20 mM 2AIrN experiment (Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref] and [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. These results support our hypothesis that most mismatches result from the reaction of activated mononucleotides with the primer.
The mononucleotide model for the origin of mismatches is consistent with the observation that the mismatch frequency is correlated with the ratio of activated mononucleotide to bridged dinucleotide [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. A salient prediction of this model is that a higher ratio of bridged dinucleotides to mononucleotides should increase fidelity. To test this, we performed experiments using purified bridged dinucleotides instead of relying on their formation in situ. Because bridged dinucleotides hydrolyze much more rapidly than the activated mononucleotides [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref] , comparisons at the 24 h timepoint would not be useful. We therefore considered two earlier timepoints instead. After one hour the frequency of mismatches in the experiment initiated with purified bridged dinucleotides was 5.8%. Although some activated mononucleotides are expected at even early time points due to bridged dinucleotide hydrolysis, we can conclude that bridged dinucleotides contribute to mismatches at some frequency <5.8%. This value is still lower than the 8.5% mismatch frequency at the same timepoint under standard conditions. As expected due to further bridged dinucleotide hydrolysis, after three hours the difference shrank to 6.4% versus 7.3%. These results agree with the model that higher proportions of bridged dinucleotides lead to more accurate template copying. Finally, our timecourse dataset under standard conditions, initiated with activated mononucleotides, reveals that the ratio of correct to incorrect incorporations begins low, rises for the first several hours, and then drops again (Supplementary [fig_ref] Figure 4: Activated mononucleotides are responsible for mismatches [/fig_ref]. A similar curve has been observed for the concentration of the bridged dinucleotide under comparable conditions [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref] [bib_ref] Potentially prebiotic activation chemistry compatible with nonenzymatic RNA copying, Zhang [/bib_ref]. Collectively, these results show that the reaction of activated mononucleotides contributes to mismatches and that the ratio of mononucleotides to bridged dinucleotides is predictive of overall primer extension fidelity.
## Methyl isocyanide-mediated activation chemistry improves fidelity
Our model of mismatch formation prompted us to consider whether a prebiotically plausible activation chemistry could improve fidelity. Methyl isocyanide (MeNC) combined with a simple aldehyde has been found to pro-Nucleic Acids mote nucleotide activation [bib_ref] A light-releasable potentially prebiotic nucleotide activating agent, Mariani [/bib_ref] [bib_ref] Common origins of RNA, protein and lipid precursors in a cyanosulfidic protometabolism, Patel [/bib_ref] [bib_ref] The origin of life-out of the blue, Sutherland [/bib_ref]. Subsequent work with the same chemistry identified an additional pathway that promotes the formation of bridged dinucleotides from 2AI-activated mononucleotides (bridge-forming activation) [bib_ref] Potentially prebiotic activation chemistry compatible with nonenzymatic RNA copying, Zhang [/bib_ref]. Bridge-forming activation exhibits several features that our model for mismatch formation predicts should reduce the frequency of errors. With bridge-forming activation, a larger fraction of the input activated mononucleotides are converted to bridged dinucleotides. Furthermore, bridge-forming activation requires less Mg 2+ , thus prolonging the lifetime of the bridged dinucleotide. We therefore incubated our random-template hairpin RNA with 10 mM 2AIrN, 30 mM MgCl 2 , and the bridge-forming activation chemistry for 24 hours. Despite the lower concentration of reactants, the product distribution is comparable to that obtained with 20 mM 2AIrN under standard conditions [fig_ref] Figure 5: Advantages of a prebiotically plausible bridge-forming activation chemistry [/fig_ref] ; compare with [fig_ref] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences [/fig_ref]. Satisfyingly, the frequency of incorrect incorporations is 7.3%, compared with 11% for the standard case. The sequence features associated with primer extension, including inferred bridged dinucleotide frequencies [fig_ref] Figure 5: Advantages of a prebiotically plausible bridge-forming activation chemistry [/fig_ref] , mismatch frequencies [fig_ref] Figure 5: Advantages of a prebiotically plausible bridge-forming activation chemistry [/fig_ref] , and complementary product sequences (Supplementary [fig_ref] Figure 5: Advantages of a prebiotically plausible bridge-forming activation chemistry [/fig_ref] are in agreement with those measured under standard conditions. These results further reinforce the conclusion that the ratio of activated mononucleotides to bridged dinucleotides dictates the prevalence of mismatches.
# Discussion
Deep-sequencing nonenzymatic RNA primer extension enables us to measure the averaged consequences of many processes that contribute to product generation: differences in the reactivities of components, competition among them, and changes to template makeup that feed back to the behaviors of reactants. Such experiments are more challenging to analyze than those with defined templates and few reactants, but bring us a step closer to realistic scenarios, which are expected to be heterogeneous [bib_ref] An optimal degree of physical and chemical heterogeneity for the origin of..., Szostak [/bib_ref].
We have found that inferred bridged dinucleotide frequencies explain complementary product features, especially in selecting templates with rC or rG in the position downstream of an incorporation. This in turn affects the position-and time-dependent accumulation of products. We also discovered that the sequence pattern of downstream templating rG and rC associated with the bridged dinucleotide pathway is absent for mismatches, suggesting that mismatches arise from activated mononucleotides. An attendant insight is that some of the advantageous properties of bridged dinucleotides stem from their capacity to out-compete mononucleotides for binding sites. The two nucleotides of a bridged dinucleotide both base-pair with the template, leading to tighter binding. This contributes to the relatively low error rates observed with 2AI-based activation, even though the activated mononucleotides are at much higher concentrations than the bridged dinucleotides over the reaction time course. 2AI-activated mononucleotides are also less reactive than bridged dinucleotides [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref] , effectively selecting against error-prone incorporations. Our assay cannot account for A B C untemplated and non-primer-linked reactant oligomerization. We anticipate that short oligomers may form over the course of the reaction, but previous characterization of this chemical system by 31 P NMR did not detect any such oligomers [bib_ref] A kinetic model of nonenzymatic RNA polymerization by cytidine-5 '-phosphoro-2-aminoimidazolide, Walton [/bib_ref]. Taken together, our results indicate that the ratio of bridged dinucleotides to activated mononucleotides determines the fidelity of nonenzymatic RNA primer extension, suggesting that bridged dinucleotides are more effective at directing correct incorporations. This conclusion points to several strategies for mitigating errors during nonenzymatic RNA copying of random templates. Bridged dinucleotides are more effective than activated mononucleotides at highfidelity template copying presumably because two covalently attached nucleotides bind template sequences more selectively than individual mononucleotides. Crystal struc-ture analysis has shown that the bridged dinucleotide binds the primer-template complex by Watson-Crick base pairs, analogously to a 3 -5 linked dimer [bib_ref] The mechanism of nonenzymatic template copying with imidazole-activated nucleotides, Walton [/bib_ref] [bib_ref] Insight into the mechanism of nonenzymatic RNA primer extension from the structure..., Zhang [/bib_ref] [bib_ref] Deciphering nonenzymatic RNA polymerization through crystallography, Zhang [/bib_ref] We therefore anticipate that oligos made up of two or more nucleotides should exhibit increasing selectivity for correct templates as they get longer, so that the ligation of short oligos may lead to even higher fidelity copying [bib_ref] Enzyme-free ligation of dimers and trimers to RNA primers, Sosson [/bib_ref] [bib_ref] Nonenzymatic, template-directed ligation of oligoribonucleotides is highly regioselective for the formation of..., Rohatgi [/bib_ref] [bib_ref] Surprising fidelity of template-directed chemical ligation of oligonucleotides, James [/bib_ref] [bib_ref] Heteropolynucleotides as templates for nonenzymatic polymerizations, Ninio [/bib_ref] [bib_ref] Self-condensation of activated dinucleotides on polynucleotide templates with alternating sequences, Lohrmann [/bib_ref]. This approach has not been attempted in a heterogeneous system, which would also contain mononucleotides to be outcompeted. Activated oligos could also serve as helpers to the mononucleotides, chaperoning them to correct templates. (The helper pathway is now understood to involve the formation of a bridged intermediate between a mononucleotide and a downstream oligo instead of a second downstream mononucleotide [bib_ref] Template-directed catalysis of a multistep reaction pathway for nonenzymatic RNA primer extension, Walton [/bib_ref].
Our results also show that fidelity with all four canonical nucleotides as templates and reactants is worse than with just rG and rC, indicating that the addition of rA and rU to the mix is responsible for the increase in overall error frequency. Previous work has identified 2-thio-uridine (s2U) as a potential alternative to rU. The s2U-rA base-pair is more stable than the rU-rA base-pair, and the s2U-rG wobble base-pair is less stable than the rU-rG wobble base-pair. These properties lead to higher fidelity copying in defined template and reactant systems where rU is replaced by s2U [bib_ref] Replacing Uridine with 2-Thiouridine Enhances the Rate and Fidelity of Nonenzymatic RNA..., Heuberger [/bib_ref] [bib_ref] Crystal structure studies of RNA duplexes containing s(2)U:A and s(2)U:U base pairs, Sheng [/bib_ref] [bib_ref] Thermodynamic insights into 2-thiouridine-enhanced RNA hybridization, Larsen [/bib_ref] [bib_ref] Thiolated uridine substrates and templates improve the rate and fidelity of ribozyme-catalyzed..., Prywes [/bib_ref]. As with ligations and helper oligos, s2U has also not been tested with random templates and a full complement of reactants. NERPE-Seq is an ideal method to examine these heterogeneous conditions.
Our sequencing assay was also found to be compatible with isocyanide-based activation chemistry, and this too promises new avenues of research. Our results show that with 2AI-based activation, some sequences are more frequent than others among complementary products, but otherwise less frequent sequences begin to appear with time or at a higher reactant concentration [fig_ref] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space [/fig_ref]. This indicates that even in a randomized sequence context, most short sequences have the potential of appearing among products. There is no inherent reason why they cannot form. Isocyanide-based activation opens the possibility of facilitating access to these sequences by iteratively re-activating hydrolyzed bridged intermediates. Future experiments will explore the extent to which this prebiotically plausible activation chemistry combined with realistic heterogeneity will enable nonenzymatic RNA primer extension to operate with the efficiency and fidelity required for the emergence and evolution of functional sequences.
## Data availability
The NERPE-Seq analysis code is available in the GitHub repository: https://github.com/CarrCE/NERPE-Seq Raw sequencing data and NERPE-Seq analysis output data are available at OSF.io: https://osf.io/racsp/?view only= 90aa541d46e14e1abf970517ccbf51fe
[fig] Figure 2: The bridged dinucleotide intermediate determines complementary product sequences. (A) Frequencies of complementary and mismatched nucleotide incorporations (20 mM 2AIrN, 24 h; n = unextended hairpins + total nucleotide incorporation events). (B) Position-dependent base frequencies of complementary products. (C) Inferring bridged dinucleotide identities. The first nucleotide adjacent to the primer becomes incorporated, whereas the downstream second nucleotide functions as a leaving group and diffuses away. (D) Position-dependent frequencies of inferred bridged dinucleotides that participated in generating complementary products. [/fig]
[fig] Figure 3: Inferred bridged dinucleotide frequencies and complementary product sequence space. (A) A log-linear correlation between predicted equilibrium binding constants and inferred bridged dinucleotide frequencies (20 mM 2AIrN, 24 h incubation, position 1; least squares unconstrained linear fit, R 2 = 0.76, dashed lines indicate 95% confidence interval on the fit). The Nearest Neighbor Database (NNDB,(43)) was used to calculate predicted equilibrium binding constants for each stretch of priming base, correctly incorporated base, and downstream base of a bridged dinucleotide. These values are shown plotted against the measured frequencies of inferred bridged dinucleotides at position 1 (Figure 2D). (B) The sequence space of primer extension with 20 mM 2AIrN, 24 h incubation. The volume of each sphere is proportional to the frequency of each complementary product at least three bases long, beginning at position 1. Note that the frequencies of some triplets are too low to be visualized (Supplemental [/fig]
[fig] Figure 4: Activated mononucleotides are responsible for mismatches. (A) The position-dependent frequency of each mismatch, relative to all mismatches at the indicated position (20 mM 2AIrN, 24 hours; T:P = Template:Product). (B1-3) Mismatches could originate from bridged dinucleotides (B1), or the incorporation of activated mononucleotides (B2). (B3) The templating base distribution downstream of mismatches can be ascertained from the sequencing data. (C) The templating base distribution downstream of position 1 mismatches. (D) The overall mismatch frequency depends on the ratio of activated mononucleotides to bridged dinucleotides (increasing across experiments left to right). At the time of peak bridged dinucleotide concentration, the ratio ∼ = 14 for 20 mM 2AIrN (Supplementary Figure S2C), ∼ = 50 for 20 mM 2AIrN + 100 mM AI (22), and bridged dinucleotides do not form with 20 mM OAtrN. [/fig]
[fig] Figure 5: Advantages of a prebiotically plausible bridge-forming activation chemistry. (A) Frequencies of complementary and mismatched nucleotide incorporations (10 mM 2AIrN + MeNC-mediated bridge-forming activation, 24 h; n = unextended hairpins + total nucleotide incorporation events). Product yields with bridge-forming activation but only 10 mM 2AIrN are comparable to that with 20 mM 2AIrN but no activation chemistry(Figure 2A), and the products are less error-prone. Other features, including bridged dinucleotide (B) and mismatch patterns (C) remain the same. [/fig]
|
Improving Therapeutic Ratio in Head and Neck Cancer with Adjuvant and Cisplatin-Based Treatments
Advanced head and neck cancers are difficult to manage despite the large treatment arsenal currently available. The multidisciplinary effort to increase disease-free survival and diminish normal tissue toxicity was rewarded with better locoregional control and sometimes fewer side effects. Nevertheless, locoregional recurrence is still one of the main reasons for treatment failure. Today, the standard of care in head and neck cancer management is represented by altered fractionation radiotherapy combined with platinum-based chemotherapy. Targeted therapies as well as chronotherapy were trialled with more or less success. The aim of the current work is to review the available techniques, which could contribute towards a higher therapeutic ratio in the treatment of advanced head and neck cancer patients.
# Introduction
The major goal of cancer treatment is to improve the clinical outcome by increasing the therapeutic ratio (TR). Most commonly, the therapeutic ratio is quantitatively defined as the ratio between tumour control probability (TCP) and normal tissue complication probability (NTCP). In order to maximise the therapeutic ratio, tumour control needs to increase while normal tissue complications need to decrease. As with any other malignancy, the objective in the treatment of advanced head and neck cancer is to improve TR through both components: TCP and NTCP.
After decades of treatment optimisation via novel irradiation techniques, new cytotoxins and several adjuvant agents to improve tumour response to therapy, advanced unresectable head, and neck cancers are still a clinical challenge. Although the locoregional control showed improvement along the implementation of new treatment techniques, the death rate does not seem to decline for this malignancy [bib_ref] Cancer statistics, Siegel [/bib_ref].
Several randomised clinical trials showed a significant improvement in locoregional tumour control and diseasefree survival when radiation was combined with cisplatin, as compared to radiation as a single agent [bib_ref] Hyperfractionated radiation therapy with or without concurrent low-dose daily cisplatin in locally..., Jeremic [/bib_ref] [bib_ref] A metaanalysis of hyperfractionated and accelerated radiotherapy and combined chemotherapy and radiotherapy..., Budach [/bib_ref] , reason why concurrent cisplatin-based chemoradiotherapy is nowadays the standard of care for advanced head and neck cancer patients.
Cisplatin is a platinum compound with complex properties when it comes to radiation-drug interaction. Through inhibition of DNA repair and cell cycle arrest cisplatin demonstrates radiosensitizing properties [bib_ref] The mechanism of action of radiosensitization of conventional chemotherapeutic agents, Lawrence [/bib_ref]. Furthermore, cisplatin exhibits radiosensitization of hypoxic cells due to scavenging of hydrated electrons by the platinum complex and formation of local concentrations of OH radicals, which eventually damage the DNA. Cisplatin was shown to have cytostatic properties by blocking the cells in the G2 phase of the mitotic cycle. It was demonstrated that cell cycle arrest at G2 is relevant to the in vivo action of cisplatin as subsequent lethal mitosis may be the most significant mechanism of cell death induced by this drug. One of cisplatin's properties which is yet to be elucidated is the suppression of tumour neovascularization [bib_ref] A mechanism of cisplatin action: antineoplastic effect through inhibition of neovascularization, Yoshikawa [/bib_ref]. Yoshikawa et al. examined the effect of cisplatin on endothelial cell proliferation observing significant inhibition of endothelial cell growth for clinical drug concentrations. To date, the most important property of cisplatin as confirmed by preclinical and clinical studies is the ability to form DNA adducts. Cisplatin can form both intrastrand and interstrand adducts with the DNA. Despite the low number of interstrand crosslinks (less than 2 BioMed Research International 1% of the total adducts) it was considered that these adducts are responsible for cisplatin's cytotoxic effect. At the same time, there are studies relating cisplatin's cytotoxicity to the DNA-intrastrand crosslinks [bib_ref] Analysis of events associated with cell cycle arrest at G2 phase and..., Sorenson [/bib_ref].
Irrespective of the exact mechanism that leads to radiosensitization cisplatin is a powerful drug, and since its clinical implementation it remains a fundamental cytotoxic agent for the management of head and neck cancer.
## Challenges imposed by radiotherapy
Tumour hypoxia and accelerated proliferation of tumour cells during therapy (both radiotherapy and chemotherapy) remain some of the biggest challenges concerning the treatment of advanced head and neck cancer. The unpredictability of acute hypoxia in tumours often leads to treatment failure, and so does the rapid proliferation of tumour cells after the initiation of therapy. Cellular recruitment from the quiescent phase, accelerated accelerated stem cell division, abortive division, and loss of asynchronous stem cell division are thought to be the main mechanisms responsible for accelerated regrowth in squamous cell carcinomas of the head and neck [bib_ref] Three A's of repopulation during fractionated irradiation of squamous epithelia: asymmetry loss,..., Dörr [/bib_ref] [bib_ref] Treatment-induced accelerated human tumor growth, Withers [/bib_ref] [bib_ref] Modelling of postirradiation accelerated repopulation in squamous cell carcinomas, Marcu [/bib_ref] [bib_ref] Radiobiological modeling of interplay between accelerated repopulation and altered fractionation schedules in..., Marcu [/bib_ref]. Perhaps the most efficient method to overcome this burden is the alteration of standard fractionation in radiotherapy into hyperfractionated and/or accelerated radiotherapy. Several clinical trials confirmed the superiority of altered fractionation schedules in regard to tumour control as compared to standard (conventional) fractionation [bib_ref] A radiation therapy oncology group (RTOG) phase III randomized study to compare..., Fu [/bib_ref] [bib_ref] Five compared with six fractions per week of conventional radiotherapy of squamous-cell..., Overgaard [/bib_ref].
Although altered fractionation increased locoregional control, there were trials that showed no treatment gain because of normal tissue complications. For instance, in the EORTC 22851 randomised trial, where hyperfractionated-accelerated radiotherapy was compared with the conventionally fractionated regimen, late toxicity nullified the gain in tumour control. The most common acute toxicities after radiotherapy reported in head and neck cancer patients are weight loss due to difficulties in swallowing, mucositis, xerostomia, and stomatitis, while dysphagia and late xerostomia are listed among late toxicities. There is, therefore, a price to pay for a better tumour control, though several times, normal tissue toxicity is a dose-limiting factor in both radiotherapy and combined chemoradiotherapy. The result is often treatment interruption, which further compromises tumour control.
## Challenges imposed by the administration of cisplatin
3.1. Normal Tissue Toxicity. While being one of the most potent chemotherapeutic agents, cisplatin is highly toxic to various organs. Nephrotoxicity and ototoxicity are some of the most commonly reported side effects during cisplatinbased chemotherapy. Other side effects, such as hematologic and central nervous system-related toxicities, are often doselimiting factors or reason for treatment interruption. The rate of treatment completion is frequently reported to be below 100% due to adverse events. New radiotherapy delivery techniques might be a possible solution in reducing side effects due to better tumour conformity, as shown by studies comparing helical tomotherapy to conventional IMRT [bib_ref] Significant improvement in normal tissue sparing and target coverage for head and..., Fiorino [/bib_ref]. However, there is need for further studies to evaluate the real benefit of these techniques as the latest results show no significant difference among tomotherapy, IMRT, and 3D-CRT regarding normal tissue effects [bib_ref] Acute hematologic and mucosal toxicities in head and neck cancer patients undergoing..., Kruser [/bib_ref].
## Risk of second primary cancers.
The risk of developing a second primary cancer after head and neck cancer treatment was shown to be strongly linked to the original risk factors, which initiated the first primary cancer, that is, smoking and alcohol consumption as well as the oncogenic human papillomavirus [bib_ref] Synchronous cancers in patients with head and neck cancer: risks in the..., Jain [/bib_ref]. Thus the most common site for developing a second primary cancer in these patients is also the head and neck region. However, there is evidence that chemotherapy, particularly cisplatin, can induce carcinogenesis in patients treated with this agent for their primary malignancy. While proven to be toxic to normal tissue since the beginning of its clinical use, cisplatin was not shown to be carcinogenic until later. A large clinical study undertaken by Travis et al. showed an increased risk of leukaemia in patients previously treated with cisplatin [bib_ref] Risk of leukemia after platinum-based chemotherapy for ovarian cancer, Travis [/bib_ref] [bib_ref] The epidemiology of second primary cancers, Travis [/bib_ref]. [fig_ref] Figure 1: Risk of leukaemia as a function of the cumulative dose of platinum... [/fig_ref] shows the increase in leukaemia risk together with the increase in cumulative dose of cisplatin. Although the data was based on a study undertaken on ovarian cancer patients treated with cisplatin, the findings may be applicable to head and neck cancer patients also [bib_ref] Risk of leukemia after platinum-based chemotherapy for ovarian cancer, Travis [/bib_ref].
## Techniques to maximise the therapeutic ratio in head and neck cancer
The most commonly used methods to optimise the therapeutic ratio are reviewed below. While some of these techniques are widely accepted among the medical community (altered fractionation, cisplatin-based radiochemotherapy, and image-guided radiotherapy) others either are less successful in their clinical implementation or await more conclusive results (bioreductive drugs, normal tissue radioprotectors, and chronotherapy).
## Combined chemoradiotherapy.
There is an extensive number of trials on head and neck cancer comparing the clinical effect of combined chemoradiotherapy versus radiotherapy alone. In order to collate and examine the results, a metaanalysis of the role of chemotherapy in head and neck cancer (MACH-NC) was published based on 93 randomised trials (conducted between 1965 and 2000) [bib_ref] Metaanalysis of chemotherapy in head and neck cancer (MACH-NC): an update on..., Pignon [/bib_ref] , which confirmed the superior outcome with combined chemoradiotherapy as compared to radiotherapy as a sole agent. While with concomitant chemotherapy a notable benefit was achieved, the results showed no clear justification for induction chemotherapy [bib_ref] Metaanalysis of chemotherapy in head and neck cancer (MACH-NC): an update on..., Pignon [/bib_ref]. Still, the absolute benefit for chemotherapy after a 5-year follow-up period was only 4.5% (for concomitant chemotherapy the absolute benefit was 6.5%).
The advances in knowledge concerning head and neck radiobiology over the last few decades lead to clinical implementation of various altered fractionation schedules with and without chemotherapy. Altered fractionation radiotherapy combined with cisplatin-based chemotherapy became a common practice within the management of advanced head and neck cancer patients [bib_ref] Evolution of clinical trials in head and neck cancer, Yang [/bib_ref]. Intensity modulated techniques (IMRT) are employed for better tumour conformity either via hyperfractionated or accelerated radiotherapy concurrently with cisplatin-based chemotherapy. [fig_ref] Table 1: Cisplatin-based chemoradiotherapy regimens employing IMRT techniques [/fig_ref] presents the most recent phases II and III clinical trial results employing intensity-modulated radiotherapy and cisplatinbased chemotherapy for advanced head and neck cancers. Both tumour control and normal tissue toxicity are presented, with variable results.
The studies presented in [fig_ref] Table 1: Cisplatin-based chemoradiotherapy regimens employing IMRT techniques [/fig_ref] have proven, once again, that combined chemoradiotherapy leads to increased locoregional tumour control when compared to radiation as a sole agent [bib_ref] Concomitant cisplatin and hyperfractionated radiotherapy in locally advanced head and neck cancer:..., Ghadjar [/bib_ref]. As also proven before, the outcome of nasopharyngeal cancer treatment tends to be better than that of other head and neck cancers [bib_ref] Feasibility and efficacy study of weekly cisplatin with concurrent intensity-modulated radiation therapy..., Lu [/bib_ref]. Significantly improved treatment outcome was reported in the SAKK 10/94 trial in the combined arm (locoregional failure-free survival at 10 years: 40% versus 32%) [bib_ref] Concomitant cisplatin and hyperfractionated radiotherapy in locally advanced head and neck cancer:..., Ghadjar [/bib_ref].
The addition of bevacizumab [bib_ref] A phase 2 study of bevacizumab with cisplatin plus intensity-modulated radiation therapy..., Fury [/bib_ref] or cetuximab [bib_ref] A phase II study of concurrent cetuximab-cisplatin and intensity-modulated radiotherapy in locoregionally..., Ma [/bib_ref] [bib_ref] IMRT with simultaneous integrated boost and concurrent chemotherapy for locoregionally advanced squamous..., Montejo [/bib_ref] for an improved tumour control resulted in tolerable acute and late toxicity. Cetuximab-specific rash was reported by both trials.
Although locoregional control is significantly improved with the conversion of conventional therapy into IMRT, longterm survival remains at low rates, mainly due to tumour recurrence.
The mixed results achieved with altered fractionation combined with chemotherapy suggest that there is need for a careful selection of patients who would benefit from such therapy. The RTOG 0129 phase III randomised trial showed that combining cisplatin with altered fractionation (accelerated concomitant boost) does not give superior clinical results to standard fractionation combined with cisplatin [bib_ref] A phase III trial to test accelerated versus standard fractionation in combination..., Ang [/bib_ref]. The followup of 721 patients included in the RTOG 0129 trial presented no differences in the 5-year overall survival (59% versus 56%, = 0.18), disease-free survival (45% versus 44%; = 0.42), or locoregional failure (31% versus 28%; = 0.76). Also, there was no significant difference in the overall grade 3-4 acute mucositis (33% versus 40%) or grade 3-4 late toxicity (26% versus 21%) [bib_ref] A phase III trial to test accelerated versus standard fractionation in combination..., Ang [/bib_ref]. Differences in radiobiological parameters, such as hypoxia and proliferation status might give indication on the patient group that would benefit from a more aggressive treatment.
One of the main challenges with injectable cisplatin is normal tissue toxicity. Therefore, the idea of oral administration of cisplatin was worth testing in order to investigate its level of tolerability. Tao et al. have designed a dose-escalation trial where oral cisplatin (CP Ethypharm) was administered in combination with radiotherapy to 18 head and neck cancer patients [bib_ref] A phase i trial combining oral cisplatin (CP Ethypharm) with radiotherapy in..., Tao [/bib_ref]. Four cisplatin dose levels were tested from 10 mg/m 2 /day to 25 mg/m 2 /day. Dose limiting toxicities in the form of gastrointestinal disorders were experienced for the highest dose hence the dose recommended for phase II trial was 20 mg/m 2 /day. Daily small doses of cisplatin have demonstrated good tolerability in the past even in injectable form [bib_ref] Hyperfractionated radiation therapy with or without concurrent low-dose daily cisplatin in locally..., Jeremic [/bib_ref] ; therefore further studies are warranted to investigate the advantages of daily oral cisplatin in head and neck cancer patients.
A new platinum compound is on the horizon, namely, mitaplatin, which is a fusion between cisplatin and the orphan drug dichloroacetate previously developed to treat lactic acidosis [bib_ref] Mitaplatin, a potent fusion of cisplatin and the orphan drug dichloroacetate, Dhar [/bib_ref] [bib_ref] Mitaplatin increases sensitivity of tumor cells to cisplatin by inducing mitochondrial dysfunction, Xue [/bib_ref]. It was shown that mitaplatin elicits cytotoxic effects upon cisplatin-resistant head and neck tumour cells [bib_ref] Mitaplatin, a potent fusion of cisplatin and the orphan drug dichloroacetate, Dhar [/bib_ref] by a dual killing mode: cisplatin interacts with the DNA while the action of the dichloroacetate is focused on the mitochondria by reversing the mitochondrial changes that confers cancer cells resistance to apoptosis. The ability of mitaplatin to selectively target cancer cells in vitro also with less normal tissue toxicity emerges as a strong foundation for further in vivo experiments.
## Image-guided
Radiotherapy. Image-guidance represents today an important tool for increasing the therapeutic gain by better targeting the tumour, especially during fractionated radiotherapy when tumour shrinkage is expected over time and by better sparing of the surrounding normal tissue. Image-guided radiotherapy can be optimally achieved by employing cone beam CT during the course of radiotherapy then adapting the treatment plan according to the new tumour parameters [bib_ref] Repeat CT imaging and replanning during the course of IMRT for head-and-neck..., Hansen [/bib_ref].
Another imaging method, which assists in improving tumour control in head and neck cancer patients, is PET/CT. The functional properties supplied by PET together with the anatomical tumour delineation offered by CT provide a powerful tool in tumour classification and prediction of treatment outcome as well as selective targeting of hypoxic regions within the tumour. Beside 18F-FDG, which is still considered the standard radiotracer for PET imaging, there are hypoxia-specific radiotracers (F-MISO, F-FAZA) as well as proliferation-specific radiotracers (F-FLT) successfully used in clinical settings [bib_ref] The role of PET imaging in overcoming radiobiological challenges in the treatment..., Marcu [/bib_ref].
A clinical study conducted by Rothschild et al. [bib_ref] PET/CT Staging followed by Intensity-Modulated Radiotherapy (IMRT) improves treatment outcome of locally..., Rothschild [/bib_ref] employ-ing image-guided IMRT was undertaken on 131 patients with locally advanced pharyngeal carcinoma. The aim was to investigate the role of FDG-PET/CT guidance in predicting treatment outcome. One treatment arm comprised of 45 patients treated with PET/CT-based IMRT was controlmatched with the second arm that included 86 patients treated with CT-based 3D conformal radiotherapy without image guidance. The 2-year overall survival for the PET/CT-IMRT arm was 91%, while in the control group only 54% of the patients were still alive 2 years after treatment. The significant increase in survival among the image-guided group shows the high potential of PET/CT in personalizing radiotherapy for head and neck cancer patients.
An innovative radiation therapy trial is currently accruing patients for analysing the predictive value of biological markers and 89Zr-cetuximab uptake in head and neck cancer treated with cisplatin versus cetuximab and standard radiotherapy versus redistributed radiotherapy. The focus of the ARTFORCE trial is on individualised treatment, using functional imaging for the assessment of dose escalation to the FDG-PET positive region and adaptive replanning accounting for anatomical changes during treatment [bib_ref] Adaptive and innovative Radiation Treatment FOR improving Cancer treatment outcomE (ARTFORCE), a..., Heukelom [/bib_ref].
## Epidermal growth factor receptor (egfr) inhibitors.
The epidermal growth factor receptor (EGFR) plays a vital role in head and neck cancer development, growth, and metastatic spread and angiogenesis, owing to promotion of epidermal cell growth and regulation of cell proliferation. It was clinically proven that overexpression of EGFR leads to increased tumour proliferation and other growth-promoting behaviour. Of all head and neck squamous cell carcinomas a value as high as 90% exhibits overexpression of epidermal growth factor receptor [bib_ref] Epidermal growth factor receptor biology in head and neck cancer, Kalyankrishna [/bib_ref].
A clinical trial conducted by Bonner et al. [bib_ref] Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck, Bonner [/bib_ref] aimed to investigate the efficacy of cetuximab when given concurrently with radiation as compared to radiotherapy alone. In the trial, 213 patients were randomised to radiotherapy alone and 211 patients to radiotherapy with cetuximab. The addition of cetuximab to radiotherapy significantly increased both 3year locoregional control rates (47% versus 34%) and overall survival rates (55% versus 45%). The main cetuximab-related toxicities were acneiform rash and hypomagnesemia [bib_ref] Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck, Bonner [/bib_ref]. The group has also reported on the 5-year survival data while attempting to find a relationship between survival and the cetuximab-related rash [bib_ref] Radiotherapy plus cetuximab for locoregionally advanced head and neck cancer: 5-year survival..., Bonner [/bib_ref]. Based on their clinical observations, patients enrolled in the cetuximab arm who developed grade 2+ cetuximab-induced rash had an improved overall survival compared to those patients who only presented with a mild rash.
The combined effect of cetuximab and cisplatin-based chemoradiotherapy was studied in a phase II trial for patients with advanced head and neck cancer [bib_ref] Concurrent cetuximab, cisplatin, and concomitant boost radiotherapy for locoregionally advanced, squamous cell..., Pfister [/bib_ref]. Cetuximab was administered on a weekly basis concurrently with cisplatin (on weeks 1 and 4) and radiotherapy (with a boost dose for the last 2 weeks). Although the 3-year overall survival rate and locoregional control were 76% and 71%, respectively, the study was closed due to unexpected and significant adverse events.
Another anti-EGFR monoclonal antibody investigated as an agent for targeted therapies in head and neck cancer is panitumumab. As shown by preclinical studies, panitumumab has higher affinity for EGFR than other monoclonal antibodies developed to target EGFR [bib_ref] Preclinical and clinical evaluations of ABX-EGF, a fully human anti-epidermal growth factor..., Foon [/bib_ref]. This might explain the results of a phase I dose-escalation trial of panitumumab and paclitaxel combined with intensity modulated radiotherapy and carboplatin which indicated a high activity of the EGFR blocking antibody with well-tolerated side effects and an overall complete response rate of 95% [bib_ref] Phase I dosefinding study of paclitaxel with panitumumab, carboplatin and intensity-modulated radiotherapy..., Wirth [/bib_ref]. Further studies to elucidate the long-term clinical effect of this agent are warranted.
## Antiangiogenic drugs.
Angiogenesis inhibitors target those signalling molecules which stimulate the endothelial cells to migrate, divide, and form new blood capillaries. One such signalling molecule is the vascular endothelial growth factor (VEGF). Antiangiogenic drugs bind to VEGF before they could connect with the receptors of the endothelial cell to initiate the angiogenic process. An angiogenic inhibitor that recognises and binds to VEGF is bevacizumab, a monoclonal antibody which was the first agent of its kind to slow tumor growth in glioblastoma, nonsmall cell lung cancer, and metastatic colorectal cancer patients [bib_ref] Bevacizumab: an angiogenesis inhibitor for the treatment of solid malignancies, Shih [/bib_ref]. Fury et al. [bib_ref] A phase 2 study of bevacizumab with cisplatin plus intensity-modulated radiation therapy..., Fury [/bib_ref] reported the results of a phase II clinical trial for advanced head and neck cancer that involved bevacizumab and cisplatin-based chemotherapy combined with intensity-modulated radiotherapy (see also [fig_ref] Table 1: Cisplatin-based chemoradiotherapy regimens employing IMRT techniques [/fig_ref]. Progression-free survival rate at 2 years was 75.9% with an overall survival rate at 2 years of 88%. No increased toxicity with the addition of bevacizumab was reported.
An interesting study undertaken by Wang et al. [bib_ref] Investigation of the efficacy of a bevacizumab-cetuximabcisplatin regimen in treating head and..., Wang [/bib_ref] investigated the effect of bevacizumab, cetuximab, and cisplatin in various combinations on head and neck carcinoma in mice. Based on the results, the highest tumour control (expressed by the delay in tumour growth) and the maximum survival were achieved in a double-agent combination (bevacizumab-cisplatin) rather than the tripleagent combination of cetuximab-bevacizumab-cisplatin. The high apoptotic index in the triple-agent study as compared to the double-agent one (31.6 ± 12.0% versus 6.9 ± 1.3%) suggests that there might be an antagonistic effect of the monoclonal antibodies when administered together and also in combination with the platinum agent.
The same research group showed that bevacizumab has antitumour effect beyond antiangiogenesis, potentiating the cytotoxicity of cisplatin on squamous cell carcinomas [bib_ref] Beyond antiangiogenesis: intratumorally injected bevacizumab plays a cisplatin-sensitizing role in squamous cell..., Wang [/bib_ref]. The chemosensitizing property of bevacizumab was evidenced by in vivo studies via intratumoral injection. While the mechanism behind this behaviour is unclear, the results warrant further investigations.
## Normal tissue radioprotectors.
The rate and degree of normal tissue toxicity in head and neck cancer patients treated with radiation have justified the clinical need for a normal tissue radioprotector for reduction of side effects.
Amifostine is a commonly used normal tissue radioprotector, which has been trialled in combination with both radiotherapy and chemoradiotherapy. Amifostine was also proven to be an effective cytoprotector against the side effects caused by cisplatin [bib_ref] Amifostine: chemotherapeutic and radiotherapeutic protective effects, Santini [/bib_ref] ; therefore it was widely administered in cisplatin-based chemoradiotherapy for head and neck cancer [bib_ref] The role of amifostine in the treatment of head and neck cancer..., Marcu [/bib_ref]. Nevertheless, the results of clinical trials employing amifostine were inconclusive regarding the effect of this agent on the clinical outcome. To shed some light on the controversies raised by the use of amifostine, Bourhis et al. [bib_ref] Effect of amifostine on survival among patients treated with radiotherapy: a metaanalysis..., Bourhis [/bib_ref] analysed 12 trials encompassing 1119 patients (65% head and neck cancer patients) treated with radiotherapy ± chemotherapy. The conclusion of this metaanalysis was that amifostine did not have a negative impact on overall survival or progression-free survival.
The future use of amifostine remains uncertain owing to conflicting results regarding normal tissue toxicities. While some of amifostine's properties such as selectivity (i.e., lack of tumour protection) and safety were confirmed by several studies, the reduction in normal tissue toxicity was not always demonstrated [bib_ref] Intravenous amifostine during chemoradiotherapy for head-and-neck cancer: a randomized placebocontrolled phase III..., Buentzel [/bib_ref].
## Chronotherapy.
It is a known and accepted fact that biological functions in humans are organised around a circadian (day/night) clock. Circadian rhythms are controlled by the suprachiasmatic nucleus of the hypothalamus, also known as the "master pacemaker". Consequently, several biological processes are dictated by this clock including sleep, hormone secretion, and cell proliferation. Experimental findings suggest the existence of crosstalk between clock gene molecules and those molecules, which are responsible for cellular progression through the cell cycle [bib_ref] Control mechanism of the circadian clock for timing of cell division in..., Matsuo [/bib_ref].
Chronotherapy refers to treatment that is timed around this biological clock, which was shown to differ between normal and cancer cells [bib_ref] Circadian gating of S phase in human ovarian cancer, Klevecz [/bib_ref] [bib_ref] Circadian-system alterations during cancer processes: a review, Mormont [/bib_ref]. Chronotherapy could assist in reducing normal tissue toxicity in cancer patients if drugs pharmacokinetics and their target organs are known and understood.
One of the most common side effects during and after chemoradiotherapy of head and neck cancers are those involving the oral mucosa. The goal of chronotherapy is to take into account the biological clock of various tissues in trying to schedule treatment in the most opportune time for the tumour and the least harmful time for the normal cells. Therefore, investigating the peak times of oral mucosa cells could dictate the timing of treatment to diminish side effects. Bjarnason et al. have investigated the circadian variation of human oral epithelium through the expression of cellcycle proteins [bib_ref] Circadian variation in the expression of cell-cycle proteins in human oral epithelium, Bjarnason [/bib_ref] and verified the findings in a randomised clinical trial of 205 head and neck cancer patients [bib_ref] Comparison of toxicity associated with early morning versus late afternoon radiotherapy in..., Bjarnason [/bib_ref]. The study concerning the circadian rhythm of the oral mucosa cells involved samples from healthy male volunteers and showed that cells in the S phase peaked around 3 pm, whereas cells in the M phase had their peak in the evening [fig_ref] Table 2: The circadian rhythm of normal oral mucosa cells [/fig_ref].
Experimental studies support the evidence whereby the DNA synthesis rhythm in normal cells is in phase opposition with that of tumour cells, observation that is valid for the mitotic phase as well [bib_ref] Circadian gating of S phase in human ovarian cancer, Klevecz [/bib_ref] [bib_ref] The days and nights of cancer cells, Canaple [/bib_ref]. Consequently, the most optimum treatment time for a high tumour cell kill would be in the morning, when most target cells are transitioning G 2 /M. During this time, normal cells would receive better protection, as they peak in G 1 .
There are clinical studies supporting the evidence whereby treatment timing according to cellular circadian rhythm can improve treatment outcome in head and neck cancer patients. Bjarnason et al. [bib_ref] Comparison of toxicity associated with early morning versus late afternoon radiotherapy in..., Bjarnason [/bib_ref] conducted a randomised trial to verify the advantage of morning radiotherapy versus afternoon radiotherapy from a normal tissue perspective. They hypothesised that morning radiotherapy will lead to less normal tissue effects involving the oral mucosa, as oral epithelial cells peak in G 1 . Patients were randomised to morning (8-10 am) versus afternoon (4-6 pm) radiotherapy. The results showed a significant improvement in weight loss for the morning group as compared to patients receiving radiotherapy in the afternoon and reduced incidence of oral mucositis among male patients (49.4% versus 64.1%). It was suggested that a greater treatment time difference between the two arms could have lead to a greater advantage of morning radiotherapy.
It was shown that circadian dosing time could also influence drug-related toxicities [bib_ref] Chronotherapeutics: the relevance of timing in cancer therapy, Lévi [/bib_ref]. Consequently, cisplatin should be less toxic to normal tissue if administered around 4 pm when the target organs (such as kidney and bone marrow) are less susceptible to cisplatin-caused damage. Positive results confirming the preclinical findings of platinum chronotherapy were reported by Focan et al. [bib_ref] Interest of chronotherapy in multidisciplinary management of oesophageal and gastric cancers, Focan [/bib_ref] in a study undertaken on oesophageal cancer patients. Normal tissue toxicity was considered excellent, with grade 3-4 oral mucositis occurring in only 11-23% patients, leucopenia in 6-19%, and thrombopenia in 18-50%.
The results achieved with chronotherapeutics in leukaemia patients, ovarian cancer patients, and metastatic colorectal cancer patients are even more pronounced. For instance, a twofold increase in survival and disease free rate after 5 years was reported in a chronotherapy trial involving children with acute lymphoblastic leukaemia when chemotherapy (antimetabolites) was administered in the evening (80% survival) versus morning treatment (40% survival) [bib_ref] Maintenance chemotherapy for childhood acute lymphoblastic leukaemia: better in the evening, Rivard [/bib_ref] [bib_ref] Impact of morning versus evening schedule for oral methotrexate and 6-mercaptopurine on..., Schmiegelow [/bib_ref]. The results indicate that malignant lymphoblasts are more susceptible to antimetabolites in the evening than in the morning hours.
To obtain conclusive results with head and neck cancer chronotherapy, there is need for further studies involving a multidisciplinary approach and an open-minded attitude towards less orthodox treatment methods that showed promising results in the past.
# Conclusions
Advanced head and neck cancers are difficult to manage despite the large treatment arsenal currently available. The multidisciplinary effort to increase disease-free survival and diminish normal tissue toxicity is rewarded with better 7 locoregional control and sometimes fewer side effects. Nevertheless, locoregional recurrence is still one of the main reasons for treatment failure.
In order to increase therapeutic ratio, there are methods to improve tumour control as well as normal tissue sparing. Some of the techniques to achieve these goals are listed below.
Techniques to increase TCP:
(i) optimum fractionation schedules;
(ii) optimum timing between radiotherapy and cisplatin administration based on cisplatin's pharmacokinetics and pharmacodynamics as well as the interaction between cisplatin and radiation;
(iii) knowledge of pretreatment radiobiological tumour parameters such as oxygenation status and cellular proliferative capacity;
(iv) image-guidance during radiotherapy;
(v) EGFR inhibitors such as cetuximab;
(vi) angiogenic inhibitors;
(vii) chronotherapy involving knowledge of tumour circadian rhythm.
Techniques to decrease NTCP:
(i) optimum timing between radiotherapy and cisplatin administration;
(ii) more conformal radiotherapy (IMRT);
(iii) image-guidance during treatment;
(iv) normal tissue radioprotectors such as amifostine;
(v) chronotherapy involving knowledge of normal tissue circadian rhythms, especially of the oral mucosa and bone marrow to diminish side effects of both radiotherapy and chemotherapy.
The latest treatment techniques combined with adjuvant and/or targeted therapies succeeded in increasing locoregional control in advanced head and neck cancer patients. The downside, however, is the increased rate of side effects. Furthermore, overall survival in this patient group has not seen any considerable improvement over the last decades. While there are promising results with targeted therapies involving monoclonal antibodies as well as with chronotherapy, the optimal treatment for advanced head and neck cancer patients is yet to be established.
[fig] Figure 1: Risk of leukaemia as a function of the cumulative dose of platinum agents (based on Travis 1999 data[20]). [/fig]
[table] Table 1: Cisplatin-based chemoradiotherapy regimens employing IMRT techniques. [/table]
[table] Table 2: The circadian rhythm of normal oral mucosa cells (data from Bjarnason et al.[37]). [/table]
|
Mechanisms of Transcranial Doppler Ultrasound phenotypes in paediatric cerebral malaria remain elusive
Background: Cerebral malaria (CM) results in significant paediatric death and neurodisability in sub-Saharan Africa. Several different alterations to typical Transcranial Doppler Ultrasound (TCD) flow velocities and waveforms in CM have been described, but mechanistic contributors to these abnormalities are unknown. If identified, targeted, TCDguided adjunctive therapy in CM may improve outcomes.Methods: This was a prospective, observational study of children 6 months to 12 years with CM in Blantyre, Malawi recruited between January 2018 and June 2021. Medical history, physical examination, laboratory analysis, electroencephalogram, and magnetic resonance imaging were undertaken on presentation. Admission TCD results determined phenotypic grouping following a priori definitions. Evaluation of the relationship between haemodynamic, metabolic, or intracranial perturbations that lead to these observed phenotypes in other diseases was undertaken. Neurological outcomes at hospital discharge were evaluated using the Paediatric Cerebral Performance Categorization (PCPC) score.Results:One hundred seventy-four patients were enrolled. Seven (4%) had a normal TCD examination, 57 (33%) met criteria for hyperaemia, 50 (29%) for low flow, 14 (8%) for microvascular obstruction, 11 (6%) for vasospasm, and 35 (20%) for isolated posterior circulation high flow. A lower cardiac index (CI) and higher systemic vascular resistive index (SVRI) were present in those with low flow than other groups (p < 0.003), though these values are normal for age (CI 4.4 [3.7,5] l/min/m2, SVRI 1552 [1197,1961 dscm-5m2). Other parameters were largely not significantly different between phenotypes. Overall, 118 children (68%) had a good neurological outcome. Twenty-three (13%) died, and 33 (19%) had neurological deficits. Outcomes were best for participants with hyperaemia and isolated posterior high flow (PCPC 1-2 in 77 and 89% respectively). Participants with low flow had the least likelihood of a good outcome (PCPC 1-2 in 42%) (p < 0.001). Cerebral autoregulation was significantly better in children with good outcome (transient hyperemic response ratio (THRR) 1.12 [1.04,1.2]) compared to a poor outcome (THRR 1.05 [0.98,1.02], p = 0.05).Conclusions:Common pathophysiological mechanisms leading to TCD phenotypes in non-malarial illness are not causative in children with CM. Alternative mechanistic contributors, including mechanical factors of the cerebrovasculature and biologically active regulators of vascular tone should be explored.
# Background
There were an estimated 241 million malaria cases and 627,000 deaths worldwide in 2020.
The burden is heaviest in sub-Saharan Africa, where 94% of the deaths occur, primarily in children younger than 5 years of age. Cerebral malaria (CM) is a severe manifestation of the disease with case fatality rates of 15-40%, even with effective treatment [bib_ref] Brain swelling and death in children with cerebral malaria, Seydel [/bib_ref] [bib_ref] Pathogenesis, clinical features and neurological outcome of cerebral malaria, Idro [/bib_ref] [bib_ref] Cerebral malaria; mechanisms of brain injury and strategies for improved neuro-cognitive outcome, Idro [/bib_ref]. Deficits in gross motor or sensory function, cognition, behavior, and/or subsequent epilepsy occur in more than 50% of survivors [bib_ref] Neurodevelopmental impairments 1year after cerebral malaria, Langfitt [/bib_ref] [bib_ref] Blantyre Malaria Project Epilepsy Study (BMPES) of neurological outcomes in retinopathy-positive pediatric..., Birbeck [/bib_ref]. As such, CM is a leading cause of death and disability in African children [bib_ref] Pathogenesis, clinical features and neurological outcome of cerebral malaria, Idro [/bib_ref]. While magnetic resonance imaging has provided improved understanding of the anatomic abnormalities that occur in paediatric CM, pathohysiological contributors to these neuroradiologic findings remain less clear [bib_ref] Brain swelling and death in children with cerebral malaria, Seydel [/bib_ref] [bib_ref] Tesla magnetic resonance imaging to investigate potential etiologies of brain swelling in..., Potchen [/bib_ref] [bib_ref] Neuroimaging findings in children with retinopathy confirmed cerebral malaria, Potchen [/bib_ref]. In order to develop efficacious adjunctive therapeutic approaches that improve outcomes in CM, mechanisms of neurological injury must determined.
Transcranial Doppler Ultrasound (TCD) is a portable, non-invasive method to assess the cerebral blood flow velocities (CBFVs) and haemodynamics in the major cerebral vessels [bib_ref] Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries, Aaslid [/bib_ref] [bib_ref] Age dependence of flow velocities in basal cerebral arteries, Bode [/bib_ref] [bib_ref] Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements, Lindegaard [/bib_ref] [bib_ref] Cerebral autoregulation dynamics in humans, Aaslid [/bib_ref] [bib_ref] Transcranial Doppler Ultrasound during critical illness in children: survey of practices in..., Larovere [/bib_ref] [bib_ref] Transcranial Doppler in children, Verlhac [/bib_ref]. TCD derived CBFVs and morphologic waveforms are determined by the mean arterial pressure, the tone and patency of the systemic and cerebral vessels, and the intracranial and central venous pressures [bib_ref] Treasure Island (FL): StatPearls Publishing, Silverman [/bib_ref] [bib_ref] Parasympathetic innervation of vertebrobasilar arteries: is this a potential clinical target?, Roloff [/bib_ref] [bib_ref] Effects of venous pressure elevation on myogenic vasoconstrictive responses to static and..., Iida [/bib_ref] [bib_ref] Acid-base regulation and sensing: accelerators and brakes in metabolic regulation of cerebrovascular..., Boedtkjer [/bib_ref] [bib_ref] Mechanisms of hypoxic cerebral vasodilatation, Pearce [/bib_ref] [bib_ref] Influence of cerebrovascular parasympathetic nerves on resting cerebral blood flow, spontaneous vasomotion,..., Morita [/bib_ref] [bib_ref] Interaction between cerebrovascular sympathetic, parasympathetic and sensory nerves in blood flow regulation, Morita-Tsuzuki [/bib_ref]. Alterations to any of these factors results in distinct changes to measured TCD parameters and waveform morphology [fig_ref] Table 1: Physiological or pathological factors that contribute to Transcranial Doppler Ultrasound [/fig_ref]. Thus, TCD is used as a point of care tool to determine specific mechanisms of focal or global cerebral dysfunction in several clinical scenarios [bib_ref] Transcranial Doppler monitoring and causes of stroke from carotid endarterectomy, Spencer [/bib_ref] [bib_ref] Role of transcranial Doppler and FOUR score in assessment of sepsis-associated encephalopathy, Zidan [/bib_ref] [bib_ref] Toward automated classification of pathological transcranial Doppler waveform morphology via spectral clustering, Thorpe [/bib_ref] [bib_ref] Transcranial doppler ultrasound criteria for recanalization after thrombolysis for middle cerebral artery..., Burgin [/bib_ref] [bib_ref] Thrombolysis in Brain Ischemia (TIBI) transcranial Doppler flow grades predict clinical severity,..., Demchuk [/bib_ref] [bib_ref] Velocity curvature index: a novel diagnostic biomarker for large vessel occlusion, Thorpe [/bib_ref] [bib_ref] Decision decision criteria for large vessel occlusion using Transcranial Doppler Waveform morphology, Thorpe [/bib_ref] [bib_ref] American Society of Neuroimaging Practice Guidelines Committee. Practice standards for transcranial Doppler..., Alexandrov [/bib_ref]. Therefore, TCD may also be of aid in determining mechanisms of neurological injury in CM.
Five different patterns of abnormal TCD flow velocities and waveforms have been observed in children with CM in the Democratic Republic of the Congo [bib_ref] Transcranial Doppler Ultrasonography provides insights into neurovascular changes in children with cerebral..., O'brien [/bib_ref]. Serial assessments showed that the admission phenotype was generally sustained until the patient recovered or died. TCD phenotypes were also associated with distinct risks of neurological morbidity or mortality. Evaluation of the association between mechanisms that commonly lead to these observed phenotypes in non-CM illness was not done.
This prospective, observational study was performed to evaluate the presence and frequency of each TCD phenotype in Malawian children with CM. Additionally, the associations of common pathophysiological mechanisms known to contribute to each phenotype in other disease states were assessed. If mechanisms are determined, TCD may be used as a bedside tool to direct, in real-time, individualized mechanism-specific adjunctive therapy in CM.
# Methods
Malawi is a low-income country in sub-Saharan Africa with a population of approximately 18 million people, including over three million children under 5 years of age. Queen Elizabeth Central Hospital (QECH) is a 1250bed public tertiary hospital in Blantyre, Malawi, with a catchment area of approximately six million people in the southern part of the country. This prospective, observational study was performed from January 2018 to June 2021 at QECH in conjunction with the "Treating Brain Swelling in Paediatric Cerebral Malaria" clinical trial (5U01AI126610-02, ClinicalTrials.gov NCT03300648). The study was approved by the ethics committee at Michigan State University and at the University of Malawi College of Medicine Research Ethics Committee (COMREC). All subjects' guardians provided verbal and written informed consent.
Children 6 months to 12 years of age who met the World Health Organization case definition of cerebral malaria (Plasmodium falciparum parasitaemia, Blantyre Coma Score (BCS) ≤ 2, and no other discernable cause of encephalopathy) were approached for enrollment. Direct and indirect ocular fundoscopy was performed at admission, and patients were subgrouped as retinopathy positive or negative based on the presence or absence of characteristic retinal findings previously reported in CM [bib_ref] Malarial retinopathy: a newly established diagnostic sign in severe malaria, Beare [/bib_ref]. Children with sickle cell disease (known or suspected) were excluded, given the high frequency of abnormal TCD examinations in this population. Likewise, given the unknown impact of severe malnutrition (mid-upper arm circumference < 11 cm) or advanced HIV disease (known HIV positive status with severe wasting) on TCD examinations, these children also were excluded.
Demographic data, vital signs, and physical examination findings were collected. Finger-prick samples were analysed to determine parasite species and density, packed-cell volume, and blood glucose and lactate concentrations (Aviva Accu-Check, Zurich, Switzerland and Arkray Lactate Pro 2, Kyoto, Japan). Finger-prick samples were also obtained for blood gas analysis (Abbot iSTAT, Chicago, Illinois, USA). Venous blood was drawn to obtain a complete blood count and electrolyte analysis (Coulter Counter; Beckman Coulter, Brea, California, USA). Parasite counts, packed-cell volume, glucose, and lactate concentrations were evaluated every six hours until the BCS was 5 or for 72 h, whichever came later. An admission lumbar puncture was performed, opening pressure measured, and the cerebrospinal fluid was analysed [bib_ref] Caring for children with cerebral malaria: insights gleaned from 20 years on..., Taylor [/bib_ref]. All patients underwent an admission electroencephalogram (EEG)(Ceegraph digital machine, BioLogic, Natus Medical Incorporated, Pleasanton, California, USA) with a modified 10-20 system to evaluate for non-convulsive status epilepticus. EEGs were clinically reviewed by a neurologist with fellowship training in EEG to evaluate for seizures/subclinical status epilepticus. When imaging capabilities were available (2018-2019 and March-June 2021), participants underwent a brain magnetic resonance imaging (MRI)(0.35-T Signa Ovation Excite, General Electric, Boston, Massachusetts or 0.064-T Hyperfine Swoop ® Guilford, Connecticut, USA) to evaluate the brain volume [bib_ref] Brain swelling and death in children with cerebral malaria, Seydel [/bib_ref]. MRIs were systematically reviewed by radiologists experienced in radiographic findings of children with CM. All patients underwent daily, non-invasive, evaluation of their systemic haemodynamics including cardiac output, cardiac index (CI), stroke volume, stroke volume index (SVI), and systemic vascular resistance using a handheld portable ultrasound device (Butterfly IQ, Guilford, CT, USA). CI was calculated as = Heart rate x SVI and SVI as = End diastolic volume -End systolic volume. Optic nerve sheath diameter was also measured daily (Butterfly IQ, Guilford, CT, USA).
All patients received intravenous artesunate according to national guidelines. Patients received 20 mL/kg of whole blood if admission packed cell volume was < 15% or > 15% but with signs of intolerance (defined as respiratory distress or haemodynamic compromise with capillary refill time > 2 s, weak pulse, and/or cool extremities). Intravenous dextrose (1 mL/kg of Dextrose 50%) was given when documented hypoglycaemia occurred (< 3 mmol/L). Clinical or sub-clinical seizure activity identified on EEG was treated with 0.2 mg/kg of diazepam followed by phenobarbital 20 mg/kg.
## Tcd examinations
TCD was performed using a commercially available unit (NovaSignal, Los Angeles, California, USA). One limitation of TCD is that it is operator dependent with diagnostic accuracy depending on the skill and experience of the examiner. All study personnel who performed TCD examinations for this study participated in 10 h of online didactic training, completed > 50 proctored TCD examinations, and demonstrated a coefficient of variation < 10% for each study measurement compared to the trainer (author NO) before being considered proficient for independent TCD scanning. The initial TCD examination occurred within 4 h of admission. TCD was performed after initial blood and dextrose infusions, if prescribed, were complete. Middle cerebral arteries (MCAs), extracranial internal carotid arteries (Ex-ICA), and basilar arteries were insonated at 2-mm intervals using previously described methods [bib_ref] Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries, Aaslid [/bib_ref] [bib_ref] Age dependence of flow velocities in basal cerebral arteries, Bode [/bib_ref] [bib_ref] Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements, Lindegaard [/bib_ref] [bib_ref] Cerebral autoregulation dynamics in humans, Aaslid [/bib_ref] [bib_ref] Transcranial Doppler Ultrasound during critical illness in children: survey of practices in..., Larovere [/bib_ref] [bib_ref] Transcranial Doppler in children, Verlhac [/bib_ref]. Systolic (Vs), diastolic (Vd), and mean flow (Vm) velocities were recorded at each interval. Pulsatility index (PI = (Vs-Vd/Vm)), a marker of downstream cerebrovascular resistance (CVR), was automatically calculated by the TCD unit at each depth in each vessel. To differentiate causes of high CBFV values, the Lindegaard ratio (LR = MCA Vm/Ex-ICA Vm) was calculated [bib_ref] Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements, Lindegaard [/bib_ref]. A LR < 3 was considered to represent hyperaemia whereas a LR > 3 was considered to represent vascular narrowing. Autoregulation is the capacity of the cerebrovasculature to maintain constant cerebral blood flow over a wide range of mean arterial blood pressures. Autoregulation can be impaired or lost in several clinical scenarios. Therefore, the transient hyperemic response ratio (THRR) was used to interrogate cerebral autoregulation in study subjects [bib_ref] Cerebral autoregulation dynamics in humans, Aaslid [/bib_ref]. THRR < 1.1 represented impaired autoregulation and ≥ 1.1 represented intact autoregulation. Based on the admission TCD findings, subjects TCD findings are dependent on cerebral perfusion pressure (CPP) and inversely proportional to the cerebrovascular resistance (CVR) [bib_ref] Treasure Island (FL): StatPearls Publishing, Silverman [/bib_ref] [bib_ref] Parasympathetic innervation of vertebrobasilar arteries: is this a potential clinical target?, Roloff [/bib_ref] [bib_ref] Effects of venous pressure elevation on myogenic vasoconstrictive responses to static and..., Iida [/bib_ref] [bib_ref] Acid-base regulation and sensing: accelerators and brakes in metabolic regulation of cerebrovascular..., Boedtkjer [/bib_ref] [bib_ref] Mechanisms of hypoxic cerebral vasodilatation, Pearce [/bib_ref] [bib_ref] Influence of cerebrovascular parasympathetic nerves on resting cerebral blood flow, spontaneous vasomotion,..., Morita [/bib_ref] [bib_ref] Interaction between cerebrovascular sympathetic, parasympathetic and sensory nerves in blood flow regulation, Morita-Tsuzuki [/bib_ref] ǂTo cause low flow, vasospasm would need to result in > 80% vessel lumen diameter reduction * If associated with waveform characteristics such as systolic spikes and absent or reversed diastolic flow ¥It is unknown if PRES represents vasodilation or vasoconstriction in the posterior circulation were classified into the following phenotypes: normal, hyperaemia, low flow, microvascular obstruction, vasospasm, isolated posterior circulation high flow, or terminal intracranial hypertension . Participants underwent daily TCD examinations through discharge, death, or hospital day 8, whichever came later.
## Outcomes
The Paediatric Cerebral Performance Category (PCPC) scoring system is a tool that was developed to measure and quantify morbidity after paediatric critical illness [bib_ref] Assessing the outcome of pediatric intensive care, Fiser [/bib_ref] [bib_ref] Relationship of illness severity and length of stay to functional outcomes in..., Fiser [/bib_ref]. Scores range from 1 to 6, with 1 being a normal functional level and 6 being death. Other values represent progressive impairment: 2 = mild disability (alert and able to interact at an age appropriate level but with mild cognitive, behavioral, or neurological deficits), 3 = moderate disability (alert and able to carry out age appropriate activities of daily life but with obvious cognitive or neurological deficits that limit function), 4 = severe disability (conscious but dependent on others for all daily functions), and 5 = vegetative state (any degree of coma or an inability to interact with the environment). PCPC was scored at the time of hospital discharge. Children with a PCPC of 1 or 2 were considered to have a good outcome while those with a PCPC of 3 to 6 were considered to have a poor outcome.
## Statistical analyses
Variables were summarized using medians with interquartile ranges and frequencies with percentages. Differences by phenotype were explored using Kruskal-Wallis tests, with Dwass, Steel, Critchlow-Fligner corrections for multiple comparisons for continuous and ordinal variables, and chi-square or Fisher's exact tests for categorical variables. All analyses were conducted using R for Statistical Computing and SAS 9.4.
# Results
A total of 245 potential participants were screened and 174 were enrolled [fig_ref] Figure 1: Flow diagram of patient screening and enrollment [/fig_ref]. Demographics, admission physical examination findings, and admission laboratory results are summarized in [fig_ref] Table 3: Demographics, laboratory investigations, imaging, and outcomes for the cohort [/fig_ref].
## Transcranial doppler ultrasound examinations
On admission, seven children (4%) had a normal TCD examination. Fifty-seven children (33%) met criteria for hyperaemia, 50 (29%) for low flow, 14 (8%) for microvascular obstruction, 11 (6%) for vasospasm, and 35 (20%) for isolated high flow in the posterior circulation [fig_ref] 2: Definitions used to categorize participants into Transcranial Doppler Ultrasound phenotypes a Lindegaard... [/fig_ref]. No participant met criteria for terminal intracranial hypertension on the admission TCD. Eleven participants (7%) transitioned from one phenotype to another on subsequent evaluation: 5 with isolated posterior high flow (IPH) changed to hyperaemia, 3 with IPH moved to low flow, 1 with IPH subsequently developed middle cerebral artery vasospasm, and 2 with low flow transitioned to vasospasm. TCD phenotype did not change in the remaining children. Normalization of flow velocities and morphology occurred at significantly different time points depending on the underlying phenotype [fig_ref] 2: Definitions used to categorize participants into Transcranial Doppler Ultrasound phenotypes a Lindegaard... [/fig_ref]. By hospital day 2, 97% of children with MO, 63% with hyperaemia, and 67% with IPH had normalized their TCD findings, whereas only 42% of those with low flow and 22% of those with vasospasm had (p = 0.02). By day 4, most surviving children in each phenotype had normalized (hyperaemia 85%, IPH 93%, Low flow 85%, MO 100%, Vasospasm 95%. p = 0.67). Three children with low flow and one with vasospasm had not normalized TCD by hospital day 8. [fig_ref] Figure 3: Representative images of Transcranial Doppler Ultrasound phenotypes in children with cerebral malaria [/fig_ref] displays representative images of children classified into each of the phenotypes.
Differences in demographics, physical examination, laboratory results, and imaging findings in children with different TCD phenotypes are in [fig_ref] Table 4: Demographics, laboratory investigations, imaging, and outcomes by Transcranial Doppler Ultrasound phenotype [/fig_ref]. Plasma Plasmodium falciparum histidine rich protein (PfHRP2) was lowest in children meeting criteria for the microvascular obstruction phenotype (p = 0.006). The pulse pressure, or the difference between the systolic and diastolic blood pressure, was significantly higher in children with hyperaemia (median pulse pressure 43 [bib_ref] Plateletendothelial cell interactions in cerebral malaria: the end of a cordial understanding, Faille [/bib_ref] [bib_ref] New insights into the causes and therapy of cerebral vasospasm following subarachnoid..., Crowley [/bib_ref] mmHg) than other phenotypes (p = 0.003). The stroke volume index was on the low end of normal or reduced compared to published values for age (40-55 ml/m2) in all groups. While not statistically significant, there was a trend to children with low flow also having the lowest SVI (SVI 32 , p = 0.08) [bib_ref] The normal ranges of cardiovascular parameters in children measured using the Ultrasonic..., Cattermole [/bib_ref]. For most phenotypes, the cardiac index (CI) was generally above the published normal range (CI 3.5-5 l/min/m 2 ) and systemic vascular resistive index was normal (SVRI 1000-1600 d.s.cm-5 m 2 ). Children with low flow had significantly lower CI and higher SVRI than other groups (median CI 4.4 l/min/m 2 (p = < 0.001), median SVRI 1552 [1197, 1961] d.s.cm −5 m 2 (p = 0.003)). No other statistically significant differences of known contributors to TCD changes (outlined in [fig_ref] Table 1: Physiological or pathological factors that contribute to Transcranial Doppler Ultrasound [/fig_ref] were identified between TCD phenotypic groups.
## Outcomes
Overall, 118 children (68%) had a good neurologicalal outcome at the time of hospital discharge. Twenty-three (13%) died, and 33 (19%) had moderate to severe deficits [fig_ref] Table 3: Demographics, laboratory investigations, imaging, and outcomes for the cohort [/fig_ref]. Outcomes were best for participants with TCD-defined hyperaemia and IPH (PCPC 1-2 in 77 and 89% respectively). Participants with TCD-defined low flow had the highest day 1 mortality and the least likelihood of a good outcome (PCPC 1-2 in 42%)
# Discussion
Previous work in the Democratic Republic of the Congo (DRC) described five different Transcranial Doppler Ultrasound phenotypes in a cohort of children with CM [bib_ref] Transcranial Doppler Ultrasonography provides insights into neurovascular changes in children with cerebral..., O'brien [/bib_ref]. The current study, in a unique group of children with CM in Malawi, identified the same five phenotypes in similar proportions to what was previously reported: hyperaemia (28% in DRC, 33% in Malawi), low flow (28% in DRC, 29% in Malawi), microvascular obstruction (23% in DRC, 8% in Malawi), vasospasm (14% in DRC, 6% in Malawi), and isolated posterior high flow (7% in DRC, 20% in Malawi). In both studies, hyperaemia was associated with a higher likelihood of favourable outcome whereas low flow was associated with increased mortality. Impaired autoregulation was also identified in both studies as being significantly associated with worse outcomes. It is unusual to identify multiple distinct changes to TCD flow velocities and morphology in a clinical diagnosis with a single underlying pathologic mechanism. Thus, given the number of phenotypes again identified in this population of children with CM, the hypothesis that multiple different mechanisms contribute to neurological injury and neuroimaging findings in CM must be considered.
Cerebral blood flow (CBF), and thus TCD flow velocities and waveforms, are dependent on cerebral perfusion pressure (CPP) and inversely proportional to the cerebrovascular resistance (CVR) [bib_ref] Treasure Island (FL): StatPearls Publishing, Silverman [/bib_ref]. CPP is determined by the pressure gradient between the brain's supplying arteries (mean arterial pressure, MAP) and the central venous pressure, which is approximately equivalent to the intracranial pressure (ICP) so that: CBF = CPP/ CVR = (MAP − ICP)/CVR. Thus, increases in MAP may result in increased CBF, particularly if the blood pressure is elevated above the autoregulatory threshold OR if less elevated but autoregulation is impaired. Low MAP has the opposite effect on CBF, which again, may be particularly significant when blood pressures fall below the lower limit of autoregulatory capacity or if autoregulation is not intact. When hyperaemia is identified on TCD, significant hypertension or elevated blood pressures with impaired autoregulation should be considered as potential mechanisms of that phenotype. Hypotension or relatively low blood pressures with impaired autoregulation may result in a low flow phenotype on TCD.
Increases in ICP will also alter CBF. With mild to moderate elevations in ICP, normal systolic flow is generally maintained but a preferential reduction in diastolic blood flow occurs as small cerebral vessels are compressed. This results in a high pulsatility index identified on TCD (with the combination of these alterations equating to the "microvascular obstruction" phenotype described in this study). Significant intracranial hypertension can result in low flow of all measured velocities accompanied by characteristic alterations to the TCD waveform (systolic spikes and absent or reversal of diastolic flow).
The CVR is determined by the smooth muscle tone of the cerebral vessels. This tone is controlled by a multiplicity of components that cross talk to maintain brain homeostasis over a range of physiologic conditions and in response to changing cerebral metabolic demand . For example, blood viscosity is inversely related to CBF; reduction of shear force applied to the cerebrovascular endothelium as viscosity falls reduces CVR and CBF increases. Thus, anaemia can result in increased CBF and be identified as hyperaemia on TCD. Hypoxia increases endothelial production of vasodilating substances, reduces CVR, and increases CBF, again resulting in the hyperaemia phenotype on TCD. Hypercapnia and hypocapnia, likely through modulating nitric oxide, decrease and increase CVR respectively. As such, hypercapnia is frequently identified as hyperaemia on TCD and hypocapnia as reduced diastolic flow and increased PI ("microvascular obstruction/alteration" in this study). Additionally, circulating, parenchymal, and endothelially derived vasodilatory and vasoconstricting compounds alter vascular tone/CVR and increase or decrease CBF to meet metabolic demand locally. Thus, fever or seizures that increase demand will increase production of vasodilatory compounds and result in hyperaemia on TCD. These classic physiologic or pathologic factors that contribute to specific TCD flow velocity or waveform alterations in most situations [fig_ref] Table 1: Physiological or pathological factors that contribute to Transcranial Doppler Ultrasound [/fig_ref] were not clearly causative of the identified phenotypes in children with CM. Patients categorized as having hyperaemia and isolated posterior circulation high flow were not more hypertensive, anaemic, hypercapnic, febrile, or more likely to be having seizures than those classified into a different TCD phenotype [fig_ref] Table 4: Demographics, laboratory investigations, imaging, and outcomes by Transcranial Doppler Ultrasound phenotype [/fig_ref]. Those with low flow were not more likely to have indirect evidence of significant increased intracranial pressure (ICP) (opening pressure on lumbar puncture, optic nerve sheath diameter, or brain volume score) than other phenotypes. Hypocapnia/ alkalosis and signs of early increased intracranial pressure were no more likely in children with microvascular obstruction than other phenotypes. Differences in some cardiovascular parameters from normal as well as between TCD phenotypic groups were observed. Across the cohort, cardiac index (CI) was within or above the published normal value for age whereas stroke volume index (SVI) was at the low end of normal or reduced. CI is calculated as CI = Heart rate x SVI and SVI as SVI = End diastolic volume -End systolic volume. Low SVI in CM patients likely represents decreased preload (and hence reduced end diastolic volume) secondary to some component of decreased circulating blood volume and/or dehydration [bib_ref] Hyponatraemia and dehydration in severe malaria, English [/bib_ref]. A compensatory increase in heart rate maintains or increases CI to meet high systemic metabolic demands, thus explaining the normal to elevated CI identified in the cohort. SVI and CI were lowest in children with the low flow phenotype, potentially due to greater reductions in preload than in other groups. Additionally, systemic vascular resistance was highest in children categorized as low flow, which by increasing end systolic volume, could have also contributed to the lower CI in this group.
## Percent of patients with each tcd phenotype
However, CI was still within the normal range for age in children identified as having low flow on TCD, decreasing the likelihood that poor cardiac output completely contributed to the low flow velocities observed in the cerebrovasculature. Therefore, alternative potential mechanisms leading to TCD phenotypes in paediatric CM must be considered. Examination of the brain tissue of children who have died of CM reveals sequestration, a multifocal microvascular obstruction by adherent, parasitized red blood cells [bib_ref] Differentiating the pathologies of cerebral malaria by postmortem parasite counts, Taylor [/bib_ref]. Sequestration results in endothelial cell activation, increased cytokine production, neurovascular inflammation, and blood-brain barrier disruption [bib_ref] Plateletendothelial cell interactions in cerebral malaria: the end of a cordial understanding, Faille [/bib_ref]. Neuroinflammation is known to affect multiple metabolic pathways in the central nervous system [bib_ref] Cerebrospinal fluid metabolites in tryptophan-kynurenine and nitric oxide pathways: biomarkers for acute..., Yan [/bib_ref] [bib_ref] Cerebrospinal fluid metabolomics: detection of neuroinflammation in human central nervous system disease, Yan [/bib_ref] [bib_ref] What is the tryptophan kynurenine pathway and why is it important to..., Davis [/bib_ref] [bib_ref] The kynurenine pathway activities in a sub-Saharan HIV/AIDS population, Bipath [/bib_ref] [bib_ref] Tryptophan pathway catabolites (serotonin, 5-hydroxyindolacetic acid, kynurenine) and enzymes (monoamine oxidase and..., Troché [/bib_ref] [bib_ref] Quinolinic acid: an endogenous neurotoxin with multiple targets, Lugo-Huitrón [/bib_ref]. Overactivation or dysregulation of these metabolic pathways in the central nervous system may result in the accumulation or depletion of local circulating, parenchymal, or endothelially derived vasoactive compounds. These factors may contribute to observed TCD phenotypes through the alteration of neurovascular tone [bib_ref] Glutamine supplementation for critically ill adults, Tao [/bib_ref] [bib_ref] Acquired amino acid deficiencies: a focus on arginine and glutamine, Morris [/bib_ref] [bib_ref] Role of L-glutamine in critical illness: new insights, Kelly [/bib_ref] [bib_ref] Monocyte dysregulation and systemic inflammation during pediatric falciparum malaria, Dobbs [/bib_ref] [bib_ref] Nitric oxide in Tanzanian children with malaria: inverse relationship between malaria severity..., Anstey [/bib_ref] [bib_ref] Pro-and anti-inflammatory cytokines in children with malaria in Franceville, Oyegue-Liabagui [/bib_ref] [bib_ref] A new NOS2 promoter polymorphism associated with increased nitric oxide production and..., Hobbs [/bib_ref] [bib_ref] Low plasma arginine concentrations in children with cerebral malaria and decreased nitric..., Lopansri [/bib_ref] [bib_ref] Metabolites of the kynurenine pathway of tryptophan metabolism in the cerebrospinal fluid..., Medana [/bib_ref] [bib_ref] New insights into the causes and therapy of cerebral vasospasm following subarachnoid..., Crowley [/bib_ref] [bib_ref] Kinetic and cross-sectional studies on the genesis of hypoargininemia in severe paediatric..., Rubach [/bib_ref] [bib_ref] Dimethylarginines: endogenous inhibitors of nitric oxide synthesis in children with falciparum malaria, Weinberg [/bib_ref]. Future work should examine the relationships between potential putative compounds and TCD phenotypes. n number, SD standard deviations, hrs hours, RR respiratory rate, SBP systolic blood pressure, mmHg millimeters mercury, DBP diastolic blood pressure; MBP mean blood pressure, IQR interquartile range, SVRI systemic vascular resistive index, PfHRP2 Plasmodium falciparum histidine rich protein 2, CO 2 carbon dioxide, NIRS nearinfrared spectroscopy, SO 2 cerebral oxygen saturation, EEG electroencephalogram, OP opening pressure, ONSD optic nerve sheath diameter, THRR transient hyperemic response ratio, MRI magnetic resonance imaging, TCD transcranial doppler ultrasound, CSF cerebrospinal fluid MRI data are available for 94 participants
[fig] -: Posterior reversible encephalopathy syndrome (PRES)¥ ǂ To cause low flow, vasospasm would need to result in >80% vessel lumen diameter reducƟon * If associated with waveform characterisƟcs such as systolic spikes and absent or reversed diastolic flow ¥ It is unknown if PRES represents vasodilaƟon or vasoconstricƟon in the posterior circulaƟon Vertebrobasilar CirculaƟon [/fig]
[fig] 2: Definitions used to categorize participants into Transcranial Doppler Ultrasound phenotypes a Lindegaard Ratio (LR) = (Mean flow velocity in the middle cerebral artery/ mean flow velocity in the extra-cranial carotid artery) b Pulsatility Index (PI) = (Systolic flow velocity-Diastolic flow velocity/Mean flow velocity) Normal Flow(1) Systolic, diastolic, and mean flow velocity in the middle cerebral artery ± standard deviations (SD) from the age normal value Hyperaemia (1) Systolic, diastolic, and mean flow velocity in the middle cerebral artery ≥ SD above the age normal value Systolic, diastolic, and mean flow velocity in the middle cerebral artery ≤ SD below the age normal value flow velocity in middle cerebral artery within SD of the age normal value AND () Diastolic flow velocity in middle cerebral artery ≤ SD below the age normal value flow velocity in the middle cerebral artery ≥ SD above the age normal value flow velocity in the basilar artery ≥ SD above the age normal value AND () Mean flow velocity in both middle cerebral arteries within SD of the age normal value Terminal intracranial hypertension (1) Systolic flow velocity in the middle cerebral artery ≤ SD below the age normal value WITH associated systolic spikes on waveform analysis AND () Absence of or reversal of diastolic flow provide informed consent n=57 ·No research ward bed space available n=5 ·Aparasitemic on arrival n=4 ·Died before consent obtained n=10 [/fig]
[fig] Figure 1: Flow diagram of patient screening and enrollment [/fig]
[fig] Figure 3: Representative images of Transcranial Doppler Ultrasound phenotypes in children with cerebral malaria. A Normal middle cerebral artery (MCA) TCD flow velocities and waveform for a 3-year-old child. B TCD with increased systolic flow velocity, increased diastolic flow velocity, Lindegaard ratio (LR) < 3. These findings represent a child categorized as having hyperaemia. C TCD with decreased systolic flow velocity, decreased diastolic flow velocity, decreased mean flow velocity. These findings represent a child categorized as having low flow. D TCD with normal systolic flow velocity, reduced diastolic flow velocity, increased pulsatility index. These findings represent a child categorized as having microvascular obstruction. E TCD with increased systolic flow velocity, increased diastolic flow velocity, LR > 3. These findings represent a child categorized as having cerebral vasospasm. F TCD with increased systolic flow velocity, increased diastolic flow velocity, increased mean flow velocity in the basilar artery. At the same time, all measurements in the MCAs were normal. These findings represent a child categorized as having isolated posterior circulation high flow [/fig]
[table] Table 1: Physiological or pathological factors that contribute to Transcranial Doppler Ultrasound (TCD) flow velocity or waveform alterations [/table]
[table] Table 3: Demographics, laboratory investigations, imaging, and outcomes for the cohort (n = 174) [/table]
[table] Table 4: Demographics, laboratory investigations, imaging, and outcomes by Transcranial Doppler Ultrasound phenotype [/table]
[table] Table 5: Number of deaths by day by Transcranial Doppler Ultrasound phenotype [/table]
|
Sarcopenia and Sarcopenic Obesity
The aging process is associated with progressive loss of muscle mass and strength, as well as decline in physical functioning. Although consensus diagnosis has not been reached, sarcopenia is increasingly defined by both loss of muscle mass and loss of muscle function or strength. The cause of sarcopenia is suggested as multifactorial, including hormonal changes, inflammatory pathway activation, fatty infiltration, poor nutrition, and decreased physical activity. Sarcopenia is often associated with visceral obesity. Sarcopenic obesity in the elderly impacts metabolic complications and represents a major public health challenge in a rapidly aging society. Further research about sarcopenia and sarcopenic obesity may be needed to confront the influence of aging society in Korea.
# Introduction
Aging is associated with a progressive loss of muscle mass, quality and strength, which results in a condition known as sarcopenia. In 1989, Rosenberg and Roubenoff [bib_ref] Stalking sarcopenia, Rosenberg [/bib_ref] proposed the term sarcopenia, originating from the Greek words sarx (flesh) and penia (loss). Sarcopenia has been defined as "the age-associated loss of skeletal muscle mass, which results in decreased strength and aerobic capacity and thus functional capacity" [bib_ref] Sarcopenia: an undiagnosed condition in older adults. Current consensus definition: prevalence, etiology,..., Fielding [/bib_ref]. Lean muscle mass contributes up to 50% of total body weight in young adults but declines with age to 25% at 80 years old [bib_ref] Age and aerobic exercise training effects on whole body and muscle protein..., Short [/bib_ref]. After 50 years of age, approximately 1% to 2% of muscle mass is expected to be lost per year, and muscle strength decreases at an even greater rate [bib_ref] Departures from linearity in the relationship between measures of muscular strength and..., Ferrucci [/bib_ref]. Sarcopenia is characterized by atrophy of type II muscle fiber and reduction in muscle fiber satellite cells with aging [bib_ref] Satellite cell content is specifically reduced in type II skeletal muscle fibers..., Verdijk [/bib_ref]. Interestingly, young men have twice as much muscle mass as fat mass, whereas this ratio is almost reversed in older men [bib_ref] Compartmental body composition based on total-body nitrogen, potassium, and calcium, Cohn [/bib_ref]. Aging is also related to increased visceral fat mass, which is an important factor in the development of metabolic syndrome, type 2 diabetes and cardiovascular disease. Sarcopenia and visceral obesity may have a synergistic impact on both chronic metabolic disorders and physical disability [bib_ref] Sarcopenic obesity: definition, cause and consequences, Stenholm [/bib_ref].
## Definition of sarcopenia
Several different definitions of sarcopenia and sarcopenic obesity have been proposed in previous studies. Baumgartner et al. [bib_ref] Epidemiology of sarcopenia among the elderly in New Mexico, Baumgartner [/bib_ref] defined sarcopenia as a two or greater standard deviation (SD) reduction in appendicular skeletal muscle (ASM) divided by height squared (ASM/height 2 ) below the normal mean for a young reference group measured using dual X-ray absorptiometry [bib_ref] Epidemiology of sarcopenia among the elderly in New Mexico, Baumgartner [/bib_ref]. Janssen et al. [bib_ref] Low relative skeletal muscle mass (sarcopenia) in older persons is associated with..., Janssen [/bib_ref] proposed a definition of sarcopenia as skeletal muscle mass index (skeletal muscle mass [kg]/weight [kg]×100) one or two SD below the mean for a younger reference group. Newman et al. [bib_ref] Sarcopenia: alternative definitions and associations with lower extremity function, Newman [/bib_ref] introduced an alternative definition of sarcopenia, using appendicular lean mass adjusted for height and body fat mass (residuals). Recently, www.e-enm.org 87 the European Working Group on Sarcopenia in Older People developed a practical clinical definition and recommends using the presence of both low muscle mass and low muscle function (strength or performance) for the diagnosis of sarcopenia. [fig_ref] Table 1: Measurements of Muscle Mass, Strength, and Function [/fig_ref] summarizes the measurement of muscle mass, strength, and function. The International Working Group on Sarcopenia proposed that a diagnosis of sarcopenia is consistent with a gait speed of less than 1 m/sec and an objectively measured low muscle mass (e.g., ASM mass relative to height 2 that is ≤7.23 kg/m 2 in men and ≤5.67 kg/m 2 in women) [bib_ref] Sarcopenia: an undiagnosed condition in older adults. Current consensus definition: prevalence, etiology,..., Fielding [/bib_ref].
## Mechanisms of sarcopenia
## Changes in hormone levels
Aging is associated with changes in hormone levels, such as those of growth hormone (GH), insulin-like growth factor (IGF)-I, insulin, androgens, estrogens, and corticosteroids, all of which affect the anabolic and catabolic conditions for muscle protein metabolism [bib_ref] Sarcopenia in the elderly: basic and clinical issues, Wang [/bib_ref]. Elderly people show a decrease in GH and IGF-I levels that is paralleled by changes in body composition [bib_ref] The influence of age on the 24-hour integrated concentration of growth hormone..., Zadik [/bib_ref]. Reduced testosterone and estrogen levels might cause decreased muscle mass as well as diminishing bone strength. Increased insulin resistance with aging is associated with augmentation of intramyocellular fat mass and loss of muscle function [bib_ref] Sarcopenia in the elderly: basic and clinical issues, Wang [/bib_ref].
## Inflammation
Cytokines play a pivotal role not only in muscle homeostasis, but also in the pathogenesis of clinical conditions characterized by alterations in protein metabolism [bib_ref] The role of cytokines in regulating protein metabolism and muscle function, Zoico [/bib_ref]. An increase in proinflammatory cytokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6, results in muscle breakdown, which is caused by decreasing muscle protein synthesis and increasing myofibrillar protein degradation [bib_ref] TNF-alpha impairs heart and skeletal muscle protein synthesis by altering translation initiation, Lang [/bib_ref]. TNF-α induces nuclear factor-κB, which leads to increasing proteolysis by activating the ubiquitin-dependent proteolysis system. IL-6 is considered to be a catabolic cytokine due to its participation in the regulation of muscle protein turnover [bib_ref] The role of cytokines in regulating protein metabolism and muscle function, Zoico [/bib_ref].
## Other risk factors
Several previous studies have demonstrated that serum 25-hydroxyvitamin D (25(OH)D) levels are inversely correlated with various parameters of obesity [bib_ref] The impact of vitamin D deficiency on diabetes and cardiovascular risk, Baz-Hecht [/bib_ref]. 25(OH)D levels decrease longitudinally with aging [bib_ref] Longitudinal changes in serum 25-hydroxyvitamin D in older people, Perry [/bib_ref]. Recently, we found that insulin resistance and 25(OH)D levels were independently associated with sarcopenic obesity in men, while insulin resistance and high sensitivity C-reactive protein were significant factors predicting sarcopenic obesity in women [bib_ref] Relationships between sarcopenic obesity and insulin resistance, inflammation, and vitamin D status:..., Kim [/bib_ref]. On the other hand, oxidative metabolism generates reactive oxygen species, which damage cell components, particularly mitochondria. Alterations to mtDNA increase with age in skeletal muscle and are affected by sarcopenia [bib_ref] The role of mitochondrial DNA mutations in aging and sarcopenia: implications for..., Hiona [/bib_ref]. Another important factor in the regulation of skeletal muscle mass is myostatin, which suppresses differentiation and proliferation of myocytes [bib_ref] Myostatin negatively regulates satellite cell activation and self-renewal, Mccroskery [/bib_ref]. Myostatin suppression may have therapeutic potential for improvement of sarcopenia [bib_ref] Sarcopenia: etiology, clinical consequences, intervention, and assessment, Lang [/bib_ref].
## Influence of sarcopenia and sarcopenic obesity
Previous studies have reported that sarcopenia is associated with risk of adverse outcomes, such as physical disability, poor quality of life, and death. In an early study by Baumgartner et al. [bib_ref] Epidemiology of sarcopenia among the elderly in New Mexico, Baumgartner [/bib_ref] , sarcopenia was significantly associated with a 3-fold to 4-fold increased risk of physical disability in both men and women [bib_ref] Epidemiology of sarcopenia among the elderly in New Mexico, Baumgartner [/bib_ref]. They also reported a relationship between sarcopenia and falls in the previous year. After adjustment for other confounding factors, the odds ratio for falls was statistically significant in men at 2.58 (95% confidence interval [CI], 0.60 to 2.67) but not in women at 1.28 (95% CI, 0.60 to 2.67) [bib_ref] Epidemiology of sarcopenia among the elderly in New Mexico, Baumgartner [/bib_ref].
In the Korean Sarcopenic Obesity Study (KSOS) of 810 subjects (414 patients with type 2 diabetes and 396 control subjects), we found that type 2 diabetes was independently associated with increased risk of sarcopenia [bib_ref] Prevalence and determinant factors of sarcopenia in patients with type 2 diabetes:..., Kim [/bib_ref]. Furthermore, we observed that sarcopenic obesity was associated with the risk of metabolic syndrome [bib_ref] Prevalence of sarcopenia and sarcopenic obesity in Korean adults: the Korean sarcopenic..., Kim [/bib_ref]. Park et al. [bib_ref] Newman AB; Health, Aging, and Body Composition Study. Excessive loss of skeletal..., Park [/bib_ref] showed that type 2 diabetes is associated with excessive loss of skeletal muscle, and that older women with type 2 diabetes in the Health ABC cohort are at especially high risk for loss of skeletal muscle [bib_ref] Newman AB; Health, Aging, and Body Composition Study. Excessive loss of skeletal..., Park [/bib_ref]. Recently, Lim et al. [bib_ref] Sarcopenic obesity: prevalence and association with metabolic syndrome in the Korean Longitudinal..., Lim [/bib_ref] found that sarcopenic obesity, defined by ASM/weight, was more closely associated with metabolic syndrome than either sarcopenia or obesity alone. In a study including 1,396 men and women aged 70 years and older, low arm muscle area was associated with an elevated mortality rate during an 8-year follow-up period (hazard ratio, 1.95; 95% CI, 1.25 to 2.00) [bib_ref] Corrected arm muscle area: an independent predictor of long-term mortality in community-dwelling..., Miller [/bib_ref]. Heitmann et al. [bib_ref] Mortality associated with body fat, fat-free mass and body mass index among..., Heitmann [/bib_ref] reported that lower levels of fat-free mass were associated with an increased risk of mortality among 787 men aged 60 years and older who were followed for 22 years. Recent studies suggest that subjects with less common body-size phenotypes, such as metabolically abnormal but normal weight or metabolically healthy obese, seem to be more prone or more resistant, respectively, to the development of obesity-associated metabolic disorders. We observed that low muscle mass might be a factor associated with different metabolic consequences according to body-size phenotype in the KSOS cohort [bib_ref] Body size phenotypes and low muscle mass: the Korean sarcopenic obesity study..., Kim [/bib_ref].
# Conclusions
Both sarcopenia and obesity are becoming major threats to aging society. The concept of sarcopenic obesity may help to elucidate the interrelationship between physical disability, metabolic disorders and mortality in the elderly population. More research about sarcopenia and sarcopenic obesity is needed to improve morbidity and mortality as consequences of rapidly aging society.
## Conflicts of interest
No potential conflict of interest relevant to this article was reported.
[table] Table 1: Measurements of Muscle Mass, Strength, and Function [/table]
|
Persistent Hypoglycemia in Seven-year-old Saudi Child: A Case Report
## C a s e r e p o rt
A seven-year-old boy with a history of frequent hypoglycemia episodes (over one year) was referred to our tertiary care center for a diagnostic opinion. The patient was maintained on small doses of diazoxide (3 mg/kg/day) from the referring hospital.
The patient was admitted for a diagnostic induced hypoglycemia test. After eight hours of fasting, he developed hypoglycemia (glucose level 1.9 mmol/l), and a critical sample showed his insulin level was 11.7 µu/ml (normal level 6-27 µu/ml) equivalent to 70 pmol/l. His C-peptide level was found to be 533 pmol/l (normal level 364-1655 pmol/l). Other elements of the samples (ketones, cortisol, adrenocorticotropic hormone, growth hormone, lactic acid, ammonia, free fatty acid level, serum acetylcarnitine profile, and urine organic acid screening) 8 were all normal. The blood glucose (BG) response to glucagon was 5 mmol/l. However, his hypoglycemia persisted and required dextrose 10% bolus and infusion in addition to diazoxide and octreotide. We were able to wean him from intravenous (IV) fluids after 24 hours. Magnetic resonance imaging (MRI) of the abdomen showed a normal pancreas and liver, as reported alongside the computed tomography (CT) images taken by the referring hospital reported as normal. DNA sample was extracted to determine any genetic causes of hyperinsulinism.
During follow-up over months in the clinic, the patient started to require increasing doses of diazoxide 75 mg orally three times daily (10 mg/kg/ day) and octreotide 150 mcg subcutaneous twice daily (15 mcg/kg/day).
One day, the patient missed the evening dose of octreotide, and the next morning he developed severe hypoglycemia symptoms and was disoriented.
A home glucose test was 1 mmol/l (20 mg/dl), and he rapidly lost consciousness. He received 1 mg glucagon and was brought to our emergency department. He was unconscious and severely hypoglycemic (1.9 mmol/l).
Patient's hypoglycemia was very resistant to multiple glucose D10% boluses (5 ml/kg) and infusion, three doses of glucagon 1 mg, two doses of octreotide 150 mcg, two doses of hydrocortisone 50 mg/m 2 /dose, two doses of diazoxide 125 mg, and glucose D12.5% infusion (15 mg/kg/min). With all these measures, his maximum glucose level was 4-6 mmol/l, and it was fluctuating rapidly down to 1 mmol/l. His level of consciousness only showed some improvement when his glucose level was > 2 mmol/l.
We observed the patient in the pediatric intensive care unit. In the next 24 hours, he needed maximum doses of diazoxide treatment (15 mg/kg/day six hourly) and octreotide treatment (40 mcg/kg/day four hourly) with tapering of D12.5% infusion. IV fluids were discontinued when glucose level maintained > 4 mmol/l.
Investigations during this severe hypoglycemia revealed BG was 1.9 mmol/l, and a very high insulin level of 20 µu/ml (105 pmol/l).
Abdominal CT scan was repeated with frequent sequencing to assess the pancreas, which revealed a 2.5 × 1 cm mass at the tail of the pancreas, suggesting an insulinoma [].
The patient underwent laparoscopic enucleation to remove the tumor from the pancreatic tail. The patient was observed post-operatively in the pediatric intensive care unit for 24 hours and was then shifted to the ward. During the post-operative time, he was off medication and IV fluids, tolerating oral intake, and BG were within normal levels (6-8 mmol/l). He was discharged home in good condition four days post-surgery.
Specimen histopathology report confirmed the diagnosis of pancreatic tail insulinoma welldifferentiated neuroendocrine tumor.
Outpatient follow-up for the last 12 months showed a well-controlled blood sugar level with no hypoglycemia and an improvement in school performance and lifestyle with the disappearance of hirsutism that occurred as a side effect of diazoxide use.
There was no significant family history of similar illness. All molecular genetic analysis of known familial hyperinsulinism genes (ABCC8, KCNJ11, GCK, GluD1, HADH, HNF4A, SlC16A1, and uCP2) by next-generation sequencing showed no mutation. In addition, MeN1 mutation screening was negative [].
## D i s c u s s i o n
CHI is the most common cause of hypoglycemic hyperinsulinemia in infants usually caused by genetic defects, though it can be at any age.In contrast, insulinoma is the most prevalent form of endogenous hypoglycemic hyperinsulinemia in adults, but only in 20% of cases is the correct diagnosis was established within one year of symptoms onset.To date, < 30 cases of insulinoma at a young age have been published, and the rarity makes a timely and accurate diagnosis even more difficult.A diagnostic fasting test was done for 96 Omani children aged eight days to 10 years in a tertiary hospital in Oman for hypoglycemia investigation; 14 (15%) were confirmed to have hyperinsulinism. None of those cases confirmed to be due to insulinoma, suggesting it is a rare condition in Gulf countries.Among the Saudi population, there are few cases in adults; the youngest reported case was a 16-yearold girl.Our patient was only seven years old at presentation. Initially, our patient was evaluated for a seizure disorder, which is not an uncommon misleading workup for insulinoma,but not all neuroglycopenic symptoms, namely, deterioration of school performance and behavior, could be explained by focal epilepsy. Given low plasma glucose (≤ 2.5 mmol/l) in a patient appearing healthy and alleviation of neuroglycopenic symptoms after intake of carbohydrates, denoted as the Whipple triad, hypoglycemic hyperinsulinism may be assumed as the underlying cause, namely if elevated levels of insulin and C-peptide are measured, as in our case. Serum insulin levels should normally be low at the times of hypoglycemia to below the limits of detection of the assay. Thus, any detectable level of insulin at the critical time of hypoglycemia can be considered abnormal and must be viewed as evidence of inappropriate insulin secretion.evaluation of the glucose response to glucagon injection at the end of the fast or during a spontaneous hypoglycemia episode provides a quick and reliable measure of increased insulin effect and helps establish the diagnosis of hyperinsulinemia. An increase in serum glucose level of > 1.7 mmol/l (30 mg/dl) in response to glucagon administration (1 mg IV or intramuscular) suggests inappropriate preservation of liver glycogen at the time of hypoglycemia and indicates suppression of liver glycogenolysis by excess insulin effect.Insulinoma is poorly managed medically, and surgical options are always indicated. 14 It often presents a diagnostic challenge both clinically and radiologically, and even in the hands of an expert radiologist, the image could be reported as negative.Progressive resistance for anti-insulin medications and increase frequency of hypoglycemic attacks should put the diagnosis of insulinoma top of the list even with no radiological clues. The primary role of imaging is to localize the lesion preoperatively. The major challenge for the detection of insulinoma has been the small size of the lesion. using modern imaging techniques, such as 18-fluorine-DOPA positron emission tomography, is useful for patients with insulinoma, especially with negative CT and MRI, and can hold unnecessary investigations.unfortunately, it was not available at our institute.
## C o n c lu s i o n
We reported here the youngest insulinoma case described so far in Saudi Arabia. Our case emphasizes the importance of putting insulinoma at the top of the differential diagnosis list in any child after infancy who present with hypoglycemic hyperinsulinism, with a careful review of imaging studies and monitoring the response for anti-insulin medications and the frequency of hypoglycemic attacks. Accurate diagnosis and clear localizing of the tumor can save the patient from unnecessary total pancreatectomy and post-surgical consequences.
# Disclosure
The authors declared no conflict of interest.
|
Efficacy of a semirigid ankle brace in reducing mechanical ankle instability evaluated by 3D stress-MRI
Background: Novel imaging technologies like 3D stress-MRI of the ankle allow a quantification of the mechanical instability contributing to chronic ankle instability. In the present study, we have tested the efficacy of a semirigid ankle brace on joint congruency in a plantarflexion/supination position with and without load.Methods:In this controlled observational study of n = 25 patients suffering from mechanical ankle instability, a custom-built ankle arthrometer implementing a novel 3D-stress MRI technique was used to evaluate the stabilizing effect of an ankle brace. Three parameters of joint congruency (i.e., 3D cartilage contact area fibulotalar, tibiotalar horizontal and tibiotalar vertical) were measured. The loss of cartilage contact area from neutral position to a position combined of 40° of plantarflexion and 30° of supination without and with axial load of 200 N was calculated. A semirigid ankle brace was applied in plantarflexion/supination to evaluate its effect on joint congruence. Furthermore, the perceived stability of the brace during a hopping task was analyzed using visual analogue scale (VAS).Results:The application of a semirigid brace led to an increase in cartilage contact area (CCA) when the foot was placed in plantarflexion and supination. This effect was visible for all three compartments of the upper ankle joint (P < 0.001; η 2 = 0.54). The effect of axial loading did not result in significant differences. The subjective stability provided by the brace (VAS 7.6/10) did not correlate to the magnitude of the improvement of the overall joint congruency.Conclusions:The stabilizing effect of the semirigid ankle brace can be verified using 3D stress-MRI. Providing better joint congruency with an ankle brace may reduce peak loads at certain areas of the talus, which possibly cause osteochondral or degenerative lesions. However, the perceived stability provided by the brace does not seem to reflect into the mechanical effect of the brace.Trial registration The study protocol was prospectively registered at the German Clinical Trials Register (#DRKS00016356).
# Background
Chronic ankle instability (CAI) arises from the two etiologies of functional (FAI) and mechanical ankle instability (MAI), whose interaction is the subject of ongoing research. These insufficiencies may result in perceived instability, generally presented by recurrent sprains or feelings of "giving way".
While functional insufficiencies should be treated by functional, e.g., sensorimotor training, mechanical deficits may require a mechanical intervention like an external support using an ankle brace, taping or even surgical stabilization. Ankle brace or tape is widely used in athletic populations in order to reduce recurrence rate or severity of sprains. The effectiveness of external supports lies in the limitation of joint excursion which reduces maximum inversion angle and angular velocity. These effects are accompanied by functional adaptations, e.g., the preparatory muscle activation, which is increased when wearing an external support. Due to this interaction leading to a combined neuromechanical effect, it is still subject to debate whether the amount of stabilization provided by ankle braces under load exceeds the effect of the active, neuromuscular stabilization. A measurement of the isolated mechanical deficit and its potential improvement by an ankle brace will help to estimate the effect of this common treatment. A systematic review found that the reduction of recurrent sprains may be independent of the type of external support, taping or bracing. Furthermore, imaging studies have visualized the effect of ankle braces using different modalities like stress-roentgenology, arthrometric testing or, more recently, computed tomography.
However, diagnosing an unstable ankle is still challenging and has its limitations and flaws in practice, in particular when it comes to quantifying mechanical instability. A potentially useful parameter is the size of the contact area between the ankle's cartilage surfaces as a three-dimensional correlate of joint congruency. The higher their congruency, the more stable the joint. The measurable congruency changes during lateral opening, but has not been investigated much in connection with instabilities.
Novel imaging technologies like 3D stress-MRI of the ankle (3SAM) allow for a quantification of the mechanical instability contributing to CAI. The three-dimensional joint congruency between the distal fibula and the talus may be a decisive factor in the development of perceived and mechanical instability. This methodology also allows to investigate joint congruency with and without axial loading; thus, it may allow to estimate the stabilizing effect of an ankle brace on the different parts of the talocrural joint in vivo. Current evidence has shown that the main stabilizing effect will be in the vertical plane, while motion in the sagittal plane could be unrestricted.
In the present study, we have tested the efficacy of an ankle brace in reducing the previously established measures of mechanical ankle instability using 3D stress-MRI. The aim of this study was to quantify the improvement of joint congruency achievable by wearing a semi-rigid ankle brace with and without axial loading in vivo. We hypothesized that an effective brace will have its main effect on the fibulotalar articulation while having a smaller effect on the horizontal tibiotalar articulation.
# Methods
The study was approved by the ethics committee of the University Medical Center of Freiburg (protocol #118/19), and the study protocol was prospectively registered at the German Clinical Trials Register (#DRKS00016356). It was carried out according to the Declaration of Helsinki in its current form, and all patients declared informed consent prior to participation.
## Population
This is a separate study on a subgroup of MAI patients deducted from a previous investigation. The patients were recruited as a random community sample from the local university and outpatients of the university hospital's orthopedic department.
Selection criteria were defined according to the literature sing the Cumberland Ankle Instability Score (CAIT) for defining perceived instability. CAIT adds up a maximum score of 30 and any score < 24 is generally considered as a decisive criterion for CAI. Mechanical instability was assessed by physical examination (talar-tilt and anterior drawer test) by a blinded experienced orthopedic surgeon. For diagnosing mechanical ankle instability, both physical exams were rated in five steps (1 = stable, 2 = rather stable, 3 = intermediate, 4 = rather unstable, 5 = unstable) where there had to be a combined score > 8 in order to be rated as mechanically unstable. Moreover, an athletic background with an average sportive activity > 4 h per week was required. Exclusion criteria were previous surgery around the upper ankle joint, less than 3 months since the last ankle sprain and any contraindications to MRI diagnostics (tattoos, ferromagnetic implants) and acute illness.
Screening of n = 41 participants with subjective feelings of instability resulted in n = 25 included patients complying with the before-mentioned criteria. Reasons for the exclusion of screened participants were n = 6 presenting with perceived but no mechanical instability and n = 8 with intermediate scoring in CAIT or physical examination, n = 1 due to a novel tattoo in the region of interest and n = 1 due to a severe injury before final MRI examination. The final cohort of n = 25 patients showed average values of 24.6 ± 4.7 years of age, an average BMI (kg/ m 2 ) of 23 ± 3.5 and an average CAIT-Score of 18.8 ± 4.4.
## Mechanical testing
The mechanical stability testing was carried out using the previously described method of dynamic 3D stress ankle-MRI (3SAM). In this novel approach, the patient is placed supine in a custom-designed, non-ferromagnetic ankle arthrometer which allows free positioning of the foot in plantarflexion-dorsiflexion as well as pronation-supination. Furthermore, the device allows for the application of axial load up to 500 N using a pneumatic cylinder system. For axial loading, the patient is fixed to the table using a weightlifter's belt around the hip and adjustable straps tied to the table. In this study, the patients were measured under five different conditions as displayed in. The foot was therefore positioned neutral (NN) and in 40° of plantarflexion and 30° of supination (PS) without and with axial load of 200 N. The load was chosen from previous studies, where 200 N was the maximum load tolerated by the patients during PS-measurement without display of any adversities. To assess the effect of the ankle brace, it was worn in PS without and with axial load. The semirigid brace (Malle-oLoc ® , Bauerfeind AG, Zeulenroda, Germany) was composed of a plastic splint that was attached to the medial and lateral side of the ankle joint with two hook-and-loop straps assembled in a figure of eight.
All MRI experiments were performed on a Magnetom Trio 3 T system (Siemens Healthineers, Erlangen, Germany), using an 8-channel multipurpose coil (NORAS MRI Products, Germany) for signal reception. The protocol consisted of a 3D turbo-spin echo (TSE) sequence with GRAPPA parallel imaging acceleration by a factor of 2. The 3D imaging volume consisted of 128 sagittal slices with an in-plane resolution of 0.5 mm and a slice thickness of 0.6 mm.
In the post-processing, three different parameters of ankle joint congruity were calculated: cartilage contact area (CCA) in the fibulotalar (CCA FT ) as well as the horizontal (CCA TTH ) and the vertical (CCA TTV ) part of the tibiotalar joint. The outcome parameters consisted of the individual reduction of CCA during plantarflexionsupination as a percentage of CCA in neutral-null position (s., "Individual reference value"). Especially the loss of the CCA FT has been shown to be a potential measure of mechanical ankle instability and its diagnostic strength is comparable to stress-sonography. Thus, a significant reduction of the loss in CCA, so-tospeak improvement of the joint congruency can be interpreted as a positive protective effect of the brace.
For post-processing of the MRI data, a browser-based framework for medical image analysis (Nora Medical Imaging Platform, Freiburg, Germany) was used.
## Patient-reported outcomes
Apart from the CAIT-Score, we used visual analogue scales (VAS) ranging from 0 to 10 to evaluate the perceived stability and comfort experienced by the participants in order to compare the subjective with the objective mechanical stabilization effect of the brace. The patients performed 10 lateral skater hops with and without wearing the brace before answering the VAS questions.
## Statistics
Statistical analysis was conducted using the Statistical Package for the Social Sciences (SPSS) version 27 (IBM Corp., Armonk, NY, USA). Graphical display was performed using Veusz (v. 3.0.1 by Jeremy Sanders).
For statistical comparison, after checking for normal distribution using the Shapiro-Wilk test, a two-factor repeated-measures ANOVA was carried out with the factors load and brace. Furthermore, the effect of the interaction load * brace was analyzed. The level of significance was set at P < 0.05. In cases of statistical significance, pairwise comparison was performed using Bonferronicorrected t tests. Additionally, partial eta squared (η 2 ) was calculated as a measure of effect size. Effect sizes were interpreted following Cohen (small: 0.01, medium: 0.06 and large: 0.12.).
Moreover, bivariate two-tailed Spearman's correlation analyses were conducted to determine the strength of the linear relationship between the difference in CCA resulting from the brace condition and the visual analogue scales. Correlation strength was interpreted according to Cohen as follows: < 0.3: weak correlation, > 0.3-0.5:
# Results
The results are displayed in and . There was a significant effect of the brace (s. on all three measures of CCA with the largest effect size on the horizontal tibiotalar CCA TTH (P < 0.001; η 2 = 0.54) in rm-ANOVA. The average loss of CCA was 10.6% less for CCA FT , 8.0% less for CCA TTH and 18.7% for CCA TTV compared to the no-brace condition. The effect of load when analyzed using rm-ANOVA did not result in significant differences (s. . Neither did the interaction effect of load*brace show significant results with P = 0.06 for CCA FT , P = 0.2 for CCA TTH and P = 0.9 for CCA TTV .
Pairwise comparison revealed that there is a significant effect on CCA TTH (P < 0.001) and a significant effect on CCA TTV (P = 0.02) of the brace under load. However, there was no significant difference between the brace and no-brace condition under load in CCA FT (P = 0.24).
Correlation analysis revealed that there were no significant correlations between the improvement of CCA when wearing a brace and the perceived stability provided by the brace. There was a significant correlation between the perceived stability (VAS 7.9) and the . Furthermore, we evaluated the correlation between the perceived instability using VAS during the lateral Skater hop (VAS 7.9) and the quantitative improvement of the CCA FT by the orthosis (10.6% of CCA), which did not correlate significantly (Spearman's rho = 0.175, P = 0.4).
# Discussion
In this controlled observational study, we could demonstrate that semirigid ankle bracing has a significant effect on talocrural joint congruency in patients with mechanical ankle instability evaluated by 3D stress-MRI. More specifically, the application of a semirigid brace led to greater cartilage contact areas in a joint position close to maximum plantarflexion and supination. Stabilizing the position close to full plantarflexion and supination is challenging for the whole ankle joint complex, especially in patients with mechanical ankle instability, generally associated with an insufficiency of the lateral ligaments of the ankle. In MAI patients, these structures do not sufficiently restrict anterior talar movement and posterior fibular slide, leading to a reduction in chondral and osseous constraint. A recent systematic review and meta-analysis showed that wearing an ankle brace can positively influence the kinematics of the foot and the ankle. Mechanically, ankle braces have proven their ability to restrict vertical range of motion while allowing sagittal motion and this has been proposed to reduce the chance of further supination injury moments.
In the pilot study of Wenning et al., especially the reduction of CCA fibulotalar (corresponding to the lateral osseous constraint) in plantarflexion-supination was found to be decisive for the ankle's instability comparing healthy and mechanical instability patients. We therefore hypothesized that an effective brace will have its main effect on the fibulotalar articulation.
Interestingly, in this study pairwise comparison showed that there was a significant effect of the brace only without load for fibulotalar CCA. This might be related to the fact that loading itself already increases the stabilizing capacities of articular surfaces. As an example, it has been shown that the articular congruency may increase when weight bearing after a lateral ankle fracture, but it may be transferred similarly to a sprain condition. Additionally, early studies by Stormont et al. investigated the stabilizing capacity of the ligaments and articular surfaces in the ankle during supination. They demonstrated that the stabilizing capacity of the articular surface increases significantly during supination as the axial compression loading was raised from 0 to 670 N. These findings indicate that the articular surfaces display the individuals' predisposition and stability against supination with increase in axial compression. The data of Tohyama et al. additionally showed that the contribution of bracing to stabilization of the ankle is dependent on the axial compression load across the ankle, suggesting that axial compression should be applied in the controlled condition in evaluation of ankle stabilizing devices. This means that the contribution of bracing to stabilization of the ankle was smaller in the axial loading condition than in the no axial loading condition. When comparing these findings to our results, no significant effect of loading itself on cartilage contact areas was observed. Furthermore, the loss of articular surface during supination was less when wearing a brace; however, this beneficial effect was less pronounced when applying the axial load. This might be due to the custom-designed, non-ferromagnetic ankle arthrometer which only allows minimal motion and therefore also no significant change of contact areas during loading. Furthermore, it needs to be considered that the load applied in this study was a lot less (200 vs. 670 N), in order to enable the patient to keep the leg stable and without any adversities. This may also have reduced the effect of the loading in this study, which is why the results should be interpreted with care.
The findings match the assumption that external supports in preventing an ankle sprain may be pronounced mainly before ground contact because ankle bracing already reduces joint excursion during the swing phase of the gait cycle. Alterations during the gait cycle in CAI like a decreased foot clearance partially due to increased plantarflexion and inversion angle may be two of the factors that can be improved by ankle bracing as a secondary preventive measure, since they reduce recurrent sprains. In this particular aspect, the preventive measure on recurrent sprains of a brace would come into effect during the swing phase or just before ground contact while the brace is not loaded.
Nevertheless, this study found a significant effect of the semirigid ankle brace on CCA TTH and on CCA TTV in pairwise comparison under load. This leads to an increase in the stabilizing capacity of the horizontal and vertical tibiotalar articular surface and more symmetric load distribution. Consequently, peak loads at certain areas of the talus possibly causing degenerative or osteochondral lesions may be reduced by the brace. When looking at these findings in detail, the increase in CCA TTH will lead to a bigger surface for the distribution of weight and force during ground contact; subsequently, it would reduce the stress on the lateral and medial talar shoulder and prevent excessive impact on the cartilage and subchondral surfaces.
Moreover, the finding that the medial cartilage surface is also increased by the application of the brace may suggest that this increase joint congruency of the tibiotalar joint as a whole. However, the high variability of this parameter (CCA TTV ) requires a careful interpretation of this finding. Additionally, from a detailed biomechanical point of view it needs to be discussed, that the increase on the tibiotalar joints is not matched by an increase in the fibulotalar CCA, which can only be realized when there is a certain amount of joint play in the tibiofibular joint. Evidently, this is an early explorative aspect of this investigation and it requires further research including a dynamic analysis that can differentiate between tibiotalar, fibulotalar and tibiofibular biomechanics. Especially the latter is a challenge for research, which is why the role of the anterior inferior tibiofibular ligament in lateral ankle instability remains a mystery. Even though it is indicative, that braces have a tertiary preventive effect on the degeneration of the joint resulting from maldistribution of load, it cannot be concluded from these findings, but it should be the focus of future, longitudinal studies.
The interaction of the different components in CAI is part of an ongoing debate and we therefore included subjective measures as a potential correlate to mechanical performance of the brace. However, in our study there was no significant correlation between the improvement of CCA FT when wearing a brace and the perceived stability provided by the brace. This supports the hypothesis that mechanical and perceived insufficiencies are distinct entities contributing to chronic ankle instability. Furthermore, it promotes the theory that the functional, e.g., sensorimotor improvement, is a major factor in the effectiveness of ankle bracing. Again, additional research is necessary to further examine the relationship between functional and mechanical instabilities of the ankle to thoroughly evaluate the different adjustments effectuated by the brace.
Moreover, this underlines the recommendation that diagnosis of ankle instability with radiographic measurements alone is imprecise, because MAI can occur without subjective instability of the ankle and vice versa. In the future, it should be examined, which percentage in loss of CCA is clinically meaningful and these current findings of ankle efficacy will need to be compared to that. It may then also be of interest to compare conservative and operative procedures and their impact on CCA.
# Limitations
This study has a few limitations to consider, among which is the static testing of the cartilage contact areas. Therefore, transferring our results to the highly dynamic landing during sports performance should be made cautiously. Nevertheless, we were testing the benefit in of the brace in a close to accident position, which is of great scientific interest. Testing in even more different flexion and supination angles with MRI would have been excessively time consuming. Another limitation is the relatively small sample size as it was a subgroup of MAI patients deducted from a previous investigation. Furthermore, it has to be considered that all measures were performed by a single investigator and except from the pilot study no inter-rater reliability can be reported. However, this novel method has been shown to be practical and robust for analyzing CAI before.
# Conclusion
In conclusion, we found a significant effect of the semirigid ankle brace on joint congruency verified by using 3D stress-MRI. This effect may reduce peak loads at certain cartilage areas of the ankle and possibly delay degenerative or osteochondral lesions.
Abbreviations CAI: Chronic ankle instability; MAI: Mechanical ankle instability; FAI: Functional ankle instability; CAIT: Cumberland ankle instability tool; CCA : Cartilage contact area; ICC: Interclass coefficient; 3SAM: 3-Dimensional stress ankle-MRI.
|
Ultrasonographic features of cervical lymph node metastases from medullary thyroid cancer: a retrospective study
Background: To investigate sonographic features of cervical lymph node metastases from medullary thyroid cancer (LNM-MTC), as compared with lymph node metastases from papillary thyroid cancer (LNM-PTC).Methods: A total of 42 MTC patients with 52 metastatic LNs and 222 PTC patients with 234 metastatic LNs who were confirmed by fine needle aspiration and post-operative pathology, were enrolled in this study. The clinical characteristics and sonographic features of LNs were compared between the two groups. Univariate analysis and multivariate logistic regression analysis were performed on the sonographic features of LNs, including short and long-axis diameter, long-axis diameter/short-axis, shape, border, hilum, echogenicity, calcifications, cystic change and vascularity pattern. The discriminating performance was assessed with the area under the receiver operating characteristic curve (AUC).Results:The mean age of patients with LNM-MTC was older than that of patients with LNM-PTC (46.81 ± 13.05 vs 39.09 ± 12.05, P < 0.001). No differences were observed in gender, location, long-axis diameter/short-axis, shape, border, echogenicity, cystic change and vascularity pattern between LNM-MTC and LNM-PTC groups (P > 0.05, for all). However, long-axis and short-axis diameter, hilum and calcifications were statistically different between these two groups (P < 0.05, for all). The AUC of discriminate value between LNM-MTC and LNM-PTC was 0.808 (95% confidence interval 0.739-0.877).Conclusion:Compared with LNM-PTC, LNM-MTC tended to have the sonographic characteristics of larger size, absence of hilum, and less calcifications, and awareness of these features might be helpful to in the diagnosis of LNM-MTC.
# Background
As a neuroendocrine tumor, medullary thyroid carcinoma (MTC) is originated from parafollicular cells (C cells), which could secrete calcitonin [bib_ref] Clinical and ultrasonographic features of medullary thyroid microcarcinomas compared with papillary thyroid..., Li [/bib_ref]. MTC accounts for 5-8% of all thyroid malignant nodules [bib_ref] Diagnosis and pathologic characteristics of medullary thyroid carcinoma-review of current guidelines, Thomas [/bib_ref] , and it is prone to present lymph node metastasis (LNM), with a high recurrence rate and poor prognosis. Even after the initial surgery, the recurrence rate is still up to 40-66% [bib_ref] Biochemical cure after reoperations for medullary thyroid carcinoma: a meta-analysis, Rowland [/bib_ref]. Therefore, early diagnosis of regional lymph nodes is important for determining surgical strategy, which may be helpful to improve quality of life and reduce the mortality rate.
Ultrasound(US) is an effective method to distinguish benign from metastatic lymph nodes(LN), with a high Ni et al. specificity, however, the sensitivity is low for LNMs, especially for central LNs [bib_ref] Meta-analysis of ultrasound for cervical lymph nodes in papillary thyroid cancer: Diagnosis..., Zhao [/bib_ref]. Previous studies showed that specific sonographic features of metastatic LNs from thyroid carcinoma included microcalcifications, hyperechogenicity, partially cystic change, peripheral or diffusely increased vascularization [bib_ref] European thyroid association guidelines for cervical ultrasound scan and ultrasound-guided techniques in..., Leenhardt [/bib_ref] [bib_ref] Cervical lymph node imaging reporting and data system for ultrasound of cervical..., Ryu [/bib_ref]. However, the results of these studies were mainly based on lymph node metastases from papillary thyroid cancer(LNM-PTC), which might not be applicable to lymph node metastases from medullary thyroid cancer(LNM-MTC). To the best of our knowledge, only a few studies with limited cases had mentioned the sonographic features of LNM-MTC. The meta-analysis of US of metastatic lymph nodes in thyroid cancer only involved one case of MTC [bib_ref] Thyroid cancer neck lymph nodes metastasis: meta-analysis of US and CT diagnosis, Xing [/bib_ref]. Therefore, this study aimed to investigate the clinical and sonographic characteristics of LNM-MTC compared with LNM-PTC.
# Methods
## Patients
This retrospective study was approved by Ethics Committee of the Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, with waiver of informed consent. Consecutive patients with suspicious LNs who underwent US and fine needle aspiration (FNA) at our hospital from January 2014 to December 2020 were reviewed. Finally, 42 patients with 52 LNM-MTC and 222 patients with 234 LNM-PTC, who were confirmed by preoperative FNA and postoperative pathology, were enrolled in our study.
The inclusion criteria were as follows: (1) patients with suspicious cervical LNs which were evaluated by US examination and FNA; (2) US was performed by senior physicians with more than 10 years of experience in ultrasound operation; (3) LN dissection was performed, and LN was confirmed with post-operative pathology. Exclusion criteria were as follows: (1) patients without complete medical and US image information; (2) patients with history of neck irradiation; (3) cytological results were undiagnosable or unsatisfactory; (4) patients associated with other LN diseases, such as lymphoma, Kikuchi-Fujimoto disease, LN tuberculosis or metastasis from other diseases.
## Ultrasonography examination and fna
During the sonographic examination, the patient lay on a bed in the supine position, and the anterior area of the neck was fully exposed. US examinations were performed with a 4-to 13-MHz linear probe (Mindray Resona7, China; MyLab 90, EsaoteSpA, Genoa, Italy & iU22, Philips, Seattle,WA, USA) by 1 of 3 radiologists with more than 10 years of experience in thyroid and LN sonography (XFN, SYX, WZ). The gray scale and Doppler US were carefully adjusted to obtain the best image. Images of each suspicious LN were obtained in both transverse and longitudinal orientations, and images were recorded and uploaded for later retrospective analysis.
The sonographic appearances of LNs were assessed carefully, including location, long-axis diameter (L), short-axis diameter(S), Long-axis diameter/short-axis (L/S), shape, border, cortex echogenicity, hilum, calcifications, cystic change and vascularity pattern. Location was divided lateral region (level I-V) and central region (level VI or VII). If multiple LNMs were found in the same region, only the largest one was selected. Long-axis and short-axis diameters were obtained at the maximum of the longitudinal section of LN. L/S was classified into ≤ 2 and > 2. Shape was assessed as irregular/regular. Border was divided into sharp and unsharp. The cortex echogenicity was determined as hyperechoic, isoechoic, and hypoechoic, as compared with the surrounding muscles. Hilum, calcifications and cystic change were categorized into absent or present. Vascularity pattern was classified as abnormal flow, hilar flow or absent flow. If peripheral or diffusely increased flow was observed, it was regarded as abnormal flow [bib_ref] European thyroid association guidelines for cervical ultrasound scan and ultrasound-guided techniques in..., Leenhardt [/bib_ref] [bib_ref] Chapter 5 ultrasound characteristics of benign vs malignant cervical lymph nodes, Prativadi [/bib_ref].
FNA was performed by 1 of 3 radiologists with more than 5 years of experience in LN FNA (XFN, SYX, WZ), and the procedure was performed by freehand technique with 23G or 25G gauge needle and 5 ml syringe. Each LN was punctured for at least 3 times. Cytological smears were assessed by professional cytopathologists. LNs were included only if the FNA results were consistent with postoperative pathology.
# Statistical analysis
Mann-Whitney U test, χ2 and Fisher exact tests by univariate analysis were used to analyze clinical information, sonographic features on grey scale and color Doppler. In order to obtain the best cut-off points, receiver operating characteristic (ROC) curve analysis was used for long and short-axis diameter. A multiple logistic regression analysis by using binary logistic regression was applied to evaluate the malignancy risk for the independent features of sonographic features. We used the univariable analysis at a statistical significance level of P < 0.05 to determine of US features for the multivariate regression analysis. The B value, Wals χ, ratio (OR) and 95% CI were recorded. A P value < 0.05 was considered to indicate statistical significance. The area under the receiver operating characteristic curve (AUC) by logistic regression model was used to evaluate the predictive value of ultrasound in order to distinguish LNM-MTC and LNM-PTC. The statistical analysis was performed using SPSS version 25 (IBM Corporation, Armonk, NY, USA).
# Results
## Clinicopathological characteristics of lnm-mtc and lnm-ptc groups
The clinical characteristics of the patients with LNM-MTC and LNM-PTC were listed in [fig_ref] Table 1: Univariate analysis of clinical characteristics and ultrasound features of lymph nodes in... [/fig_ref]. There were no significant differences in gender and location between LNM-MTC and LNM-PTC groups (P > 0.05). The mean age of patients with LNM-MTC was older than that of patients with LNM-PTC (46.81 ± 13.05 vs 39.09 ± 12.05, P < 0.001). [fig_ref] Table 1: Univariate analysis of clinical characteristics and ultrasound features of lymph nodes in... [/fig_ref] showed the US features in LNM-MTC and LNM-PTC groups. LNM-MTCs were more likely to have the following US features [fig_ref] Table 1: Univariate analysis of clinical characteristics and ultrasound features of lymph nodes in... [/fig_ref] : L/S ≤ 2, regular shape, sharp border, isoechoic or hypoechoic, however, there were no significant differences in these features between LNM-MTC and LNM-PTC groups (P > 0.05). The ROC curve was used for the long and short-axis diameters. The cutoff points of long-axis diameter and short diameter were 8.95 and 7.85, and the areas under the curve were 0.641 and 0.664, respectively. Then, long and short-axis diameters were converted into binary variables. The mean sizes of LNM-MTC were larger than those of LNM-PTC [fig_ref] Figure 1: Ultrasonographic images of LNs [/fig_ref] in both long-axis diameter and short-axis diameter (P < 0.05). There were statistically significant differences in the features of hilum, calcifications and vascularity pattern between LNM-MTC and LNM-PTC (P < 0.05). By comparing the US characteristics of LNM-MTC in the lateral and central groups, it showed that the long axis diameter was significantly different between the two groups (P < 0.05), while the other characteristics had no differences [fig_ref] Table 2: Univariate analysis of ultrasound features of LNM-MTC in the lateral and central... [/fig_ref].
## Univariate analysis of us features in lnm-mtc and lnm-ptc groups
## Multivariate logistic regression analysis of us features in lnm-mtc and lnm-ptc groups
The results of multivariate logistic regression analysis were listed in [fig_ref] Table 3: Odds ratios for the selected sonographic features by multivariate logistic regression analysis [/fig_ref]. The ultrasonic indicators with statistical significance in the univariate analysis were included in the multivariate logistic regression analysis, and the malignancy risk was further evaluated for the independent features. Compared with LNM-PTC, long-axis diameter > 8.95 (OR = 3.543), shortaxis diameter > 7.85 (OR = 2.708), absence of hilum (OR = 3.031) and absence of calcification (OR = 0.140) were more common in LNM-MTC [fig_ref] Figure 1: Ultrasonographic images of LNs [/fig_ref] , P < 0.05), while there was no statistical significance in vascularity pattern (P > 0.05). AUC of discriminate ability between LNM-MTC and LNM-PTC was 0.808, with 95% confidence interval (CI) of 0.739-0.877 .
# Discussion
As previously reported, MTC is a rare thyroid malignant tumor with an incidence of 0.11/100,000 in the population, which can be divided into sporadic and genetic types [bib_ref] Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma, Jr [/bib_ref]. The C cells of MTC can secrete many kinds of substances, among which serum calcitonin is a valuable tumor marker of MTC. MTC is more aggressive than PTC, and it is prone to distant metastases including lung, liver, bone. Weber et al. reported that LNM-MTCs could be found in 75% patients who underwent total thyroidectomy and modified radical neck dissections [bib_ref] Impact of modified radical neck dissection on biochemical cure in medullary thyroid..., Weber [/bib_ref]. Cervical lymph node dissection had certain complications, such as thoracic duct leak, injury to the recurrent laryngeal nerve, and hypoparathyroidism, and repeated operations were very traumatic to patients. Therefore, accurate preoperative diagnosis of LNM is of great significance for surgical plans. In our study, there was no difference in gender between the two groups, however, patients with LNM-MTC were older than those with LNM-PTC. Similar results have also been reported in primary thyroid cancer. A previous study showed that patients with MTC were older than those with PTC [bib_ref] The KWAK TI-RADS and 2015 ATA guidelines for medullary thyroid carcinoma: combined..., Li [/bib_ref]. No difference was observed in the location between LNM-MTC and LNM-PTC groups. However, more LNMs were located at the lateral region in both groups. LNMs of thyroid cancer usually spread to the central region first, followed by the lateral region [bib_ref] Thyroid Association Management Guidelines for Adult Patients with thyroid nodules and differentiated..., Haugen [/bib_ref] , and central LNs dissection are routinely performed. Moreover, due to the poor US window and small size of LNs in the central region, the sensitivity of US in the detection of central LNMs was low [bib_ref] Meta-analysis of ultrasound for cervical lymph nodes in papillary thyroid cancer: Diagnosis..., Zhao [/bib_ref]. Thus, many LNs in the central region might not be recognized, and FNA was not performed. US is important for distinguishing benign and malignant LNs. The related ultrasonographic indicators of LN include size, L/S, shape, border, hilum, echogenicity, calcifications, cystic change and vascularization. LNMs of thyroid carcinoma are mainly manifested as microcalcifications, hyperechogenicity, cystic change and abnormal vascularization [bib_ref] European thyroid association guidelines for cervical ultrasound scan and ultrasound-guided techniques in..., Leenhardt [/bib_ref]. However, these features were commonly seen in LNM-PTC [bib_ref] Ultrasound-based clinical prediction rule model for detecting papillary thyroid cancer in cervical..., Patel [/bib_ref] [bib_ref] The application of ultrasound and fineneedle aspiration in low-volume lateral lymph nodes..., Chen [/bib_ref] , and they were rarely reported in LNM-MTC. To the best of our knowledge, the ultrasonic features of LNM-MTC have not been well documented in the literatures. In our study, no differences were observed in L/S, shape, border, echogenicity, cystic change and vascularity pattern between LNM-MTC and LNM-PTC groups. However, long-axis and short-axis diameter, hilum and calcifications were statistically different between these two groups. AUC showed good discrimination ability between LNM-MTC and LNM-PTC (AUC, 0.808; 95% CI 0.739-0.877). In our study, the mean long-axis and short-axis diameter in LNM-MTC were 16.43 ± 10.36 mm and 8.53 ± 5.79 mm, both larger than those of LNM-PTC. Compared with PTC, MTC had a higher degree of malignancy and invasion [bib_ref] The KWAK TI-RADS and 2015 ATA guidelines for medullary thyroid carcinoma: combined..., Li [/bib_ref] [bib_ref] Medullary thyroid carcinoma: comparison with papillary thyroid carcinoma and application of current..., Lee [/bib_ref] , which might be the reason for larger sizes of LNs. Normal LNs usually have an echogenic hilum, while metastatic LNs usually present with absence of hilum [bib_ref] Cervical lymph node imaging reporting and data system for ultrasound of cervical..., Ryu [/bib_ref] [bib_ref] Ultrasound-based clinical prediction rule model for detecting papillary thyroid cancer in cervical..., Patel [/bib_ref] [bib_ref] Value of contrast-enhanced ultrasound combined with elastography in evaluating cervical lymph node..., Jiang [/bib_ref]. The metastases of the tumor initially appear in the peripheral area of LN, but as the size of metastases become larger, the normal nodal tissue destroyed. When LNs were filled with tumor cells, hilum could not be recognized [bib_ref] Diagnosis of lymph node metastases of head and neck cancer and evaluation..., Furukawa [/bib_ref]. However, absence of hilum may be without significance in distinguishing LNM-PTC from normal LNs [bib_ref] European thyroid association guidelines for cervical ultrasound scan and ultrasound-guided techniques in..., Leenhardt [/bib_ref]. Due to its indolent nature, PTC usually develops slowly. When the medullary lymphatic sinuses have not been completely destroyed by tumor cells, hilum also exists in early LNMs [bib_ref] The linear echogenic hilus in cervical lymphadenopathy-a sign of benignity or malignancy?, Evans [/bib_ref]. The results in our study showed that absence of hilum was more commonly detected in LNM-MTC than in LNM-PTC. MTC is a moderate malignancy, and the tumor cells commonly grow with invasion. The difference might be due to the more aggressive biological behavior.
Nodal calcification plays an important role in differentiating benign and malignant LNs. It is also an important indicator of primary thyroid carcinoma [bib_ref] Evaluation of thyroid nodules by a scoring and categorizing method based on..., Xu [/bib_ref]. Li et al. revealed that macrocalcification was more frequent in MTC, while microcalcification was more common in PTC [bib_ref] Clinical and ultrasonographic features of medullary thyroid microcarcinomas compared with papillary thyroid..., Li [/bib_ref]. Microcalcifications are usually round or concentric under the light microscope, with a tiny diameter of 10 to 100 μm, which correspond to the psammoma bodies on pathology. In MTC, macrocalcifications are associated with amyloid deposits [bib_ref] Medullary thyroid carcinoma: role of high-resolution US, Gorman [/bib_ref]. Calcification in LNMs is generally rare, but it is common in LNM from thyroid cancer. found that there was no significant difference in calcifications between PTCs and MTCs [bib_ref] Ultrasonographic findings of medullary thyroid carcinoma: a comparison with papillary thyroid carcinoma, Kim [/bib_ref]. To the best of our knowledge, calcification in metastatic LNs has not been compared between the two tumors. Our results showed that the incidence of calcifications in LNM-MTC group was significantly lower than that in LNM-PTC group (P < 0.05). In the LNM-PTC group, calcification was found in 59.40% (139/234) LNs, with a relatively high incidence. However, in LNM-MTC group, the incidence of calcifications was not high in LNM-MTC group, and it was detected in only 21.2% (11/52) cases. Due to the low prevalence, calcifications were not further subdivided into coarse and fine calcifications. The difference between the two groups indicated that although calcification was regarded as an ultrasonic feature of LNM-PTC, this sign may not be applied to LNM-MTC.
Hyperechogenicity and cystic change are also the characteristic manifestations of LNM-PTCs, and there were no differences between LNM-MTC and LNM-PTC in our study. The typical echogenicity of LNM from thyroid cancer is hyperechoic in the cortex compared to skeletal muscle [bib_ref] Chapter 5 ultrasound characteristics of benign vs malignant cervical lymph nodes, Prativadi [/bib_ref]. This may be due to the deposition of thyroglobulin and clustering of the tumor cells in the LNs [bib_ref] Chapter 5 ultrasound characteristics of benign vs malignant cervical lymph nodes, Prativadi [/bib_ref] [bib_ref] Sonography of neck lymph nodes. Part II: abnormal lymph nodes, Ahuja [/bib_ref]. LNM should also be suspected in patients with suspicious thyroid nodules when cystic change existed. The mechanism of cystic change in LNM may be related to necrosis [bib_ref] Metastatic papillary thyroid cancer presenting with a recurrent necrotic cystic cervical lymph..., Clark [/bib_ref]. Our results suggested that hyperechogenicity and cystic change can also be applied to LNM-MTC. Normal LNs usually had hilar flow, while metastatic LNs often showed peripheral flow or mixed flow, which was usually considered as abnormal flow [bib_ref] Chapter 5 ultrasound characteristics of benign vs malignant cervical lymph nodes, Prativadi [/bib_ref]. Peripheral flow was a meaningful feature of LNM from thyroid cancer [bib_ref] Ultrasonographic evaluation of cervical lymph nodes in thyroid cancer, Machado [/bib_ref] , which was caused by tumor angiogenesis or recruitment of capsular vessels [bib_ref] Sonography of neck lymph nodes. Part II: abnormal lymph nodes, Ahuja [/bib_ref]. In our study, abnormal flow was more frequent in LNM-MTC than in LNM-PTC, but the difference was not significant (P > 0.05).
It was reported that the ultrasonographic characteristics of LNM-PTC in the central and cervical groups were different [bib_ref] Pre-operative ultrasound diagnosis of nodal metastasis in papillary thyroid carcinoma patients according..., Lee [/bib_ref]. The high frequency of malignant US characteristics in the lateral LNMs included partial cystic component, microcalcifications and diffusely increased echogenicity, however, these ultrasound findings may not apply to the central group. It was suggested that the taller-than-wide type was only malignant US feature in the central LNMs [bib_ref] Pre-operative ultrasound diagnosis of nodal metastasis in papillary thyroid carcinoma patients according..., Lee [/bib_ref]. Previous studies also suggested that the sensitivity of US in central LNMs was lower than that of lateral cervical LNM in PTC, because the lymph nodes in the central region lacked characteristic sonographic appearances [bib_ref] Meta-analysis of ultrasound for cervical lymph nodes in papillary thyroid cancer: Diagnosis..., Zhao [/bib_ref] [bib_ref] Pre-operative ultrasound diagnosis of nodal metastasis in papillary thyroid carcinoma patients according..., Lee [/bib_ref]. The comparison of ultrasound images between the central and lateral LNs in LNM-MTC group was performed in our study, however, there were no differences in other indicators except the long-axis diameter. According to the results of our study, we suggested that the US features of LNM-MTC could also be applicable to the central LNMs.
There were several limitations in this study: First, due to the retrospective nature of this study, it is likely to have selection bias and restricted quality of images and data. Second, the sample size of LNM-MTC group was small, and a further study with a larger sample size should be performed.
# Conclusions
Although some ultrasonic features were overlapped in LNM-MTC and LNM-PTC, LNM-MTC tended to have the sonographic characteristics of larger size, absence of hilum, and less calcifications. When the primary tumor was MTC, LNs with these features should be suspected of metastatic lymph nodes.
Abbreviations MTC: Medullary thyroid carcinoma; LNM: Lymph node metastasis; US: Ultrasound; LN: Lymph node; LNM-PTC: Lymph node metastases from papillary thyroid cancer; LNM-MTC: Lymph node metastases from medullary thyroid cancer; L: Long-axis diameter; S: Short-axis diameter; L/S: Long-axis diameter/ short-axis diameter; ROC: Receiver operating characteristic; OR: Odd ratio; CI: Confidence interval.
[fig] Figure 1: Ultrasonographic images of LNs. (A, B) A metastatic LN in a 53-year-old man with MTC. (C, D) A metastatic LN in a 29-year-old man with PTC. (A) The LN was located at left level IV, and the grey ultrasonographic images showed a LN measured 35.1 × 15.6 mm, with hyperechogenicity, L/S > 2 and absence of hilum. (B) Color Doppler showed that the LN had a mixed vascularity pattern (white arrow). (C) The LN was located at right level III. The grey ultrasonographic images showed a LN measured 15.7 × 9.0 mm, and it demonstrated absence of hilum, L/S ≤ 2, hyperechogenicity and microcalcifications (white arrow). (D) Color Doppler showed that the LN had a mix vascularity pattern (white arrow) Ni et al. BMC Medical Imaging (2022) 22:151 [/fig]
[table] Table 1: Univariate analysis of clinical characteristics and ultrasound features of lymph nodes in the MTC and PTC groups *using Fisher's exact tests, others using chi-square (χ2); # enrolled into Multivariate Logistic Regression [/table]
[table] Table 2: Univariate analysis of ultrasound features of LNM-MTC in the lateral and central groups * using Fisher's exact tests, others using chi-square (χ2) [/table]
[table] Table 3: Odds ratios for the selected sonographic features by multivariate logistic regression analysis [/table]
|
Uninstructed BIAT faking when ego depleted or in normal state: differential effect on brain and behavior
Background: Deception can distort psychological tests on socially sensitive topics. Understanding the cerebral processes that are involved in such faking can be useful in detection and prevention of deception. Previous research shows that faking a brief implicit association test (BIAT) evokes a characteristic ERP response. It is not yet known whether temporarily available self-control resources moderate this response. We randomly assigned 22 participants (15 females, 24.23 ± 2.91 years old) to a counterbalanced repeated-measurements design. Participants first completed a Brief-IAT (BIAT) on doping attitudes as a baseline measure and were then instructed to fake a negative doping attitude both when self-control resources were depleted and non-depleted. Cerebral activity during BIAT performance was assessed using high-density EEG.Results: Compared to the baseline BIAT, event-related potentials showed a first interaction at the parietal P1, while significant post hoc differences were found only at the later occurring late positive potential. Here, significantly decreased amplitudes were recorded for 'normal' faking, but not in the depletion condition. In source space, enhanced activity was found for 'normal' faking in the bilateral temporoparietal junction. Behaviorally, participants were successful in faking the BIAT successfully in both conditions.Conclusions: Results indicate that temporarily available self-control resources do not affect overt faking success on a BIAT. However, differences were found on an electrophysiological level. This indicates that while on a phenotypical level self-control resources play a negligible role in deliberate test faking the underlying cerebral processes are markedly different.
# Background
Test faking is a widespread problem especially when the content of human feelings and thought shall be explored. Socially sensitive topics (e.g., stereotyping, racism, doping) are particularly vulnerable to faking. For example, athletes' attitudes toward doping that were assessed via self-report measures have shown to be affected by deceptive responses [bib_ref] Does social desirability influence the relationship between doping attitudes and doping susceptibility..., Gucciardi [/bib_ref]. These authors illustrated that athletes' self-reported doping attitudes should be considered severely skewed towards the socially desired 'no to doping in sports' . Reaction-time based indirect tests like the Implicit Association Test [IAT have been introduced with the promise of being more robust towards deception attempts [bib_ref] Implicit measures in social cognition research: their meaning and use, Fazio [/bib_ref]. There is solid evidence that IAT's generally outperform traditional self-report measures when socially sensitive topics need to be assessed [bib_ref] Understanding and using the implicit association test: III. Meta-analysis of predictive validity, Greenwald [/bib_ref] [bib_ref] The effect of implicitly incentivized faking on explicit and implicit measures of..., Wolff [/bib_ref]. Recently a doping Brief IAT (BIAT) has been developed that showed to be a valid predictor of biochemical doping test results [bib_ref] Using response-time latencies to measure athletes' doping attitudes: the brief implicit attitude..., Brand [/bib_ref]. One reason for this might be that compared to direct tests IAT's are thought to be superior in hiding the true measurement goal (i.e., participants are not directly asked about their attitude towards a certain topic). Typically, these tests are presented as lexical sorting tasks on a computer, where two concepts (one target and one evaluative) are mapped on the same response key of the keyboard and participants are requested to respond as fast as possible. The task is easier and reaction times are faster when the two concepts sharing the same response key (e.g., doping and dislike) are closely associated than when they are not associated (e.g., doping and like).
However, even IAT's can be faked to some extent; even more if participants had the opportunity to familiarize themselves with the procedure at least once [bib_ref] Faking the IAT: aided and unaided response control on the implicit association..., Fiedler [/bib_ref].
Research indicates that participants who were provided with an explicit faking strategy were more successful at faking than participants who were simply instructed to find a way to 'trick the test' [bib_ref] Faking the IAT: aided and unaided response control on the implicit association..., Fiedler [/bib_ref] [bib_ref] Voluntary controllability of the implicit association test (IAT), Kim [/bib_ref] [bib_ref] Faking of the implicit association test is statistically detectable and partly correctable, Cvencek [/bib_ref] [bib_ref] Exaggeration is harder than understatement, but practice makes perfect! Faking success in..., Roehner [/bib_ref] [bib_ref] What do fakers actually do to fake the IAT? An investigation of..., Roehner [/bib_ref]. One major reason why faking IATs has sparked scientific interest is very practical: Test takers whose motivation to disguise their true attitude is high (for example towards doping in sports) will most likely begin to think about (and try) deception strategies. However, a recent study finds that in more realistic setting, i.e., when participants were only implicitly incentivized to fake doping attitude tests, faking occurred on self-reported measures but not on the BIAT [bib_ref] The effect of implicitly incentivized faking on explicit and implicit measures of..., Wolff [/bib_ref]. Research is needed to identify the cognitive processes that facilitate (or inhibit) participants' faking success. Studying the electrophysiological correlates underneath these processes helps to understand how (and drawing upon which resources) participants fake such response time-based tests. Using electroencephalography's (EEG) very high time resolution of eventrelated potentials (ERPs) enables to investigate very early cortical processes during faking.
## Cortical processes of test faking
So far, most ERP research on test faking focused on forced choice test formats [bib_ref] Differential effects of practice on the executive processes used for truthful and..., Johnson [/bib_ref] [bib_ref] The self in conflict: the role of executive processes during truthful and..., Johnson [/bib_ref] [bib_ref] The role of episodic memory in controlled evaluative judgments about attitudes: an..., Johnson [/bib_ref] [bib_ref] Temporal course of executive control when lying about self-and other-referential information: an..., Hu [/bib_ref]. Some studies found larger very early frontal negativity [bib_ref] Temporal course of executive control when lying about self-and other-referential information: an..., Hu [/bib_ref] and an occipital positivity [bib_ref] The self in conflict: the role of executive processes during truthful and..., Johnson [/bib_ref] for faking. Such early effects [bib_ref] Temporal course of executive control when lying about self-and other-referential information: an..., Hu [/bib_ref] have also been attributed partly to the blocked designs used in this research where participants had the opportunity to prepare themselves for giving an appropriate faked answer [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Consistently, for designs using an equal number of stimuli in the faking and non-faking condition, a reduced late positivity has been found, starting at the P300 . This is thought to reflect a form of cognitive control, where participants have to exert self-control [bib_ref] The self in conflict: the role of executive processes during truthful and..., Johnson [/bib_ref]. Recently ERP patterns were investigated while participants that were provided with an effective faking strategy faked a validated doping brief implicit association test [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Here, significant differences were found between faking and non-faking at early and late time points. An enlarged early frontal negativity and occipital positivity was found as well as a decreased P300/LPP component [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Source analyses revealed significantly more activity in the right inferior frontal gyrus and the bilateral temporo-parietal junction for faking compared to baseline blocks. Among other processes the right inferior frontal gyrus is involved in memory and motor inhibition [bib_ref] When 'go' and 'nogo' are equally frequent: ERP components and cortical tomography, Lavric [/bib_ref] [bib_ref] The functional neuroanatomical correlates of response variability: evidence from a response inhibition..., Bellgrove [/bib_ref] [bib_ref] Neural markers of inhibition in human memory retrieval, Wimber [/bib_ref] [bib_ref] Conflict control during sentence comprehension: fMRI evidence, Ye [/bib_ref] which suggests an inhibition of a prepotent motor response to the target stimulus. The enhanced activity for faking in the temporo-parietal junction was thought to reflect the monitoring on faking and faking success. For these middle temporal/ occipital regions enhanced activity is also found in an functional magnetic resonance imaging (fMRI) study on lying [bib_ref] Lying in the scanner: covert countermeasures disrupt deception detection by functional magnetic..., Ganis [/bib_ref]. To conclude, the act of suppressing a predominant response tendency (i.e., non-faking response) and substituting it with an experimentally required response (i.e., faking response) is an act of inhibitory cognitive control or self-control [bib_ref] The neuroscience of "ego depletion" or: how the brain can help us..., Inzlicht [/bib_ref]. To our current knowledge there seems to be no specific 'faking' component visible in ERPs [bib_ref] The self in conflict: the role of executive processes during truthful and..., Johnson [/bib_ref]. Individual differences in temporarily available self-control resources might affect how the cerebral response to such faking demands is and therefore partly explain the lack of a specific 'faking' component.
## Ego depletion as a cognitive moderator of the cerebral faking response
According to the strength model of self-control [bib_ref] Ego depletion: is the active self a limited resource?, Baumeister [/bib_ref] , the ability to volitionally inhibit predominant response tendencies is an act of self-control which is relies on the temporary availability of self-control strength. Baumeister and colleagues argue that all self-control acts e.g., emotion regulation [bib_ref] The role of self-control strength in the development of state anxiety in..., Englert [/bib_ref] and attention regulationare energized by one global resource, i.e., a metaphorical strength with limited capacity. Self-control strength can be depleted after a primary self-control act (i.e., ego depletion) and is not replenished immediately. Thus, in a state of ego depletion, less self-control strength is available for subsequent acts of self-control. This can negatively affect performance on tasks that require selfcontrol. Previous research has shown that the ability to suppress unwanted thoughts or attitudes (e.g., stereotypes) depends on self-control strength [bib_ref] Consequences of stereotype suppression and internal suppression motivation: a self-regulation approach, Gordijn [/bib_ref]. Therefore, we assume that faking also requires self-control strength. It has to be noted, that the strength models' propositions have been questioned lately [bib_ref] Publication bias and the limited strength model of self-control: has the evidence..., Carter [/bib_ref]. Specifically, it has been argued that the estimated ego-depletion effect size might have been overestimated [bib_ref] Publication bias and the limited strength model of self-control: has the evidence..., Carter [/bib_ref]. A Registered Replication Report is currently underway to investigate the size of the ego-depletion effect and to shed further light on this phenomenon (https://osf.io/jymhe/). Still, neuroscientific research has provided some support for the strength model of self-control. For instance, depleted participants displayed weaker error related negativity (ERN) signals while they performed a self-control task [bib_ref] Running on empty-neural signals for self-control failure, Inzlicht [/bib_ref]. Although neuroscientific findings of depletion effects have been found to be somewhat inconsistent [bib_ref] Motivation and cognitive control: from behavior to neural mechanism, Botvinick [/bib_ref] , one expectation of ego depletion effects is a reduction of activity within some parts of executive networks [bib_ref] Motivation and cognitive control: from behavior to neural mechanism, Botvinick [/bib_ref]. Specifically, Friese et al. [bib_ref] Suppressing emotions impairs subsequent stroop performance and reduces prefrontal brain activation, Friese [/bib_ref] found that ego depletion was associated with an activity decrease in the right lateral prefrontal cortex, an area which is responsible for the effortful implementation of control.
## The present research
Our study is motivated by the idea that in test faking, participants have to voluntarily inhibit a predominant or at least more 'automatic' response tendency (i.e., telling the truth). It is thus reasonable to assume that faking draws upon self-control resources. Faking research has not been linked with research on ego depletion yet and no study addressed how neurophysiological faking correlates differ as a function of the available self-control resources.
We hypothesize that ego depletion leads to a decreased ability to exert cognitive control [bib_ref] Ego-depletion and prejudice: separating automatic and controlled components, Govorun [/bib_ref]. We thus expect a distinctive ERP response for IAT faking under ego depletion compared to IAT faking when participants are not depleted. Further, we aim at extending previous findings on instructed BIAT faking by investigating the ERP response for non-depleted BIAT faking when participants have to search for an effective faking strategy themselves. For faking without an explicit strategy provided we expect similar ERP differences as for faking with an explicit strategy [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Namely, we expect an early frontal negativity and occipital positivity, and a subsequent decrease in the late components for faking. Further, in source space we expect more activity for faking in the right-inferior frontal gyrus and the bilateral temporoparietal junction (TPJ). Finally, we expect that for faking under ego depletion these differences should be reduced or even absent.
# Methods
In line with suggestions of Simmonset al."we report how we determined our sample size, all data exclusions, all manipulations, and all measures in the study. " All data are uploaded in Additional file 1, Additional file 2 and Additional file 3.
## Participants
Twenty-four participants were recruited at the University of Bielefeld. We aimed for the same sample size as in the previous study on IAT faking [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref] in order to increase comparability of results and because both studies are closely comparable in the experimental design. They gave written informed consent according to the Declaration of Helsinki and the study was approved by the ethics committee of the University of Bielefeld. Two participants were excluded due to excessive artifacts (both >50 % bad trials) leaving 22 participants for final analyses. These 22 participants (15 females) were 24.23 years old on average (SD = 2.91, Min = 21, Max = 31) and had normal or corrected to normal vision. Screenings with the German version of the Beck Depression Inventory and the State Trait Anxiety Inventory, revealed neither clinically relevant depression symptoms (M = 3.55, SD = 3.54) nor pathological anxiety scores (M = 34.73; SD = 5.12). No participant reported previous or current diagnosed mental or neurological disorders. One subject was left handed.
## Design/procedure
The participants' ability to fake a negative doping attitude in an ego depleted versus non-depleted state was tested in a counterbalanced within-group (repeated measures) experimental design [fig_ref] Figure 1: The experimental design [/fig_ref]. All participants completed a practice doping BIAT in order to familiarize themselves with the task [bib_ref] The Brief Implicit Association Test, Sriram [/bib_ref]. This BIAT was excluded from subsequent analyses. As a baseline measure of their doping attitude, participants then completed another doping BIAT. They were then randomized to either the 'ego The experimental design. Participants completed a total of four BIAT's. The first two BIAT's were completed with the standard instructions and the last two BIAT's were completed with the instruction to fake a negative doping attitude. Faking under ego depletion was operationalized by an incongruent Stroop preceding a BIAT (dark grey area) and 'normal' faking was operationalized by a congruent Stroop task preceding a BIAT (light grey area). a The practice BIAT consisted of 20 discrimination trials; and 20 trials in each of the doping + like and the doping + dislike blocks. b The discrimination block was removed for the subsequent BIAT's and the doping + like and doping + dislike blocks were increased to 40 trials each depletion' or to a 'non-depletion' condition and completed the respective tasks. Immediately after the manipulation both groups were asked to complete another doping BIAT and were instructed to fake the test in a way that they appeared strongly anti-doping. Participants were asked to find their own way to "trick" the test (An effective way to trick the test is to deliberately delay the response in the doping + like block and to respond with full effort in the doping + dislike block [bib_ref] Faking the IAT: aided and unaided response control on the implicit association..., Fiedler [/bib_ref] [bib_ref] Exaggeration is harder than understatement, but practice makes perfect! Faking success in..., Roehner [/bib_ref] [bib_ref] What do fakers actually do to fake the IAT? An investigation of..., Roehner [/bib_ref] [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Research has shown that significant resource replenishment occurs within 5 min [bib_ref] After depletion: the replenishment of the self's regulatory resources, Tyler [/bib_ref]. Therefore, after a break of 5 min the order was reversed: Participants who were depleted in the first sequence of the experiment were not depleted in the second sequence (and vice versa). Again, all participants were asked to fake the final doping BIAT in a way that they appeared strongly anti-doping. Manipulation checks whether or not participants perceived the depletion task as more depleting than the control condition were performed before both BIAT measurements. All experimental stimuli were programmed and run using Inquisit 3.0 experimental software (www.millisecond.com).
## Measures biat
Doping attitudes were assessed with a validated pictorial doping BIAT [bib_ref] Illegal performance enhancing drugs and doping in sport: a picture-based brief implicit..., Brand [/bib_ref]. The practice BIAT consisted of a discrimination block (20 trials) where participants learned to discriminate the doping stimuli from the nonfocal health food stimuli using the 'R' and 'I' keys on the computers keyboard. Then the compatible block (doping + dislike, 20 trials) was presented, followed by the incompatible block (doping + like, 20 trials). The order of compatible and incompatible blocks was counterbalanced between participants to avoid positioning effects. In the following BIAT's, the discrimination block was removed and compatible and incompatible blocks were expanded to 40 trials each. This BIAT is identical to the one introduced by Brand, Heck and Ziegler [bib_ref] Illegal performance enhancing drugs and doping in sport: a picture-based brief implicit..., Brand [/bib_ref] with the exception that (a) we expanded the blocks of interest to 40 trials to get an adequate number of trials per cell for ERP averaging and (b) set the inter-trial-interval to 1000 ms in order to avoid introducing artifacts into the EEG measure.
Pictures of syringes, ampules, and pills represented the focal 'doping' concept; pictures of apples, vegetables, and cereal stood for the 'health food' concept. The 'like' and 'dislike' concepts were represented by positive and negative Emoticons. D-scores are calculated according to the D4 algorithm such that negative scores represent a negative attitude towards doping [bib_ref] Understanding and using the implicit association test: I. An improved scoring algorithm, Greenwald [/bib_ref]. This means that reaction times of error trials, and those exceeding 10,000 ms are deleted and replaced by an error value (average reaction time of this participant in all correct trials of the block plus an error penalty of 600 ms; mere elimination of error trials would have a negative impact on the reliability of the test).
## Ego depletion manipulation
A computerized Stroop task [bib_ref] Studies of interference in serial verbal reactions, Stroop [/bib_ref] was used to induce ego depletion. This task has been frequently applied to experimentally manipulate self-control strength [bib_ref] Ego depletion and persistent performance in a cycling task, Englert [/bib_ref] [bib_ref] Can implementation intentions help to overcome ego-depletion?, Webb [/bib_ref]. In this task the participants see a color word on a computer screen. The font color in which the word is written either does match its semantic meaning (i.e., congruent trial; e.g., "blue" written in blue color) or it does not (i.e., incongruent trial; e.g., "blue" written in red color). Participants have to name the font color and not the semantic meaning of the color word. This requires the exertion of self-control as one has to volitionally override the automatic tendency to name the color word [bib_ref] Can implementation intentions help to overcome ego-depletion?, Webb [/bib_ref].
In our study participants indicated the font color of each word by pressing the respectively labeled key on a computer keyboard as fast and accurate as possible. All participants completed 30 practice trials (15 congruent, 15 incongruent) to get familiarized with the task first. Then, participants in the ego depletion condition performed the depleting Stroop task, which consisted of 40 congruent and 40 incongruent trials (order randomized). Participants in the non-depletion condition performed the non-depleting version of the Stroop task which consisted of 80 congruent trials (order randomized).
## Manipulation check and control variables
In order to assess the degree of self-control participants had to exert while working on the task, participants answered three items (e.g., "How mentally exhausting did you find the Stroop task?"; after depletion: α = 76; after non-depletion: α = .51). To control for differences in motivation participants answered three items (e.g., "How motivated were you to do well in the Stroop task?"; after depletion: α = 90; after non-depletion: α = .97). All items had to be answered on a 7-point Likert-type scale with answers ranging from 1 = not at all to 7 = very much.
As mood can affect performance on self-control tasks it is routinely controlled for in self-control research [e.g., [bib_ref] Ego depletion and persistent performance in a cycling task, Englert [/bib_ref] [bib_ref] Reduced self-control leads to disregard of an unfamiliar behavioral option: an experimental..., Wolff [/bib_ref]. We administered the German version of the Positive and Negative Affect Schedule [PANAS; 46]. Participants indicated their positive affect (10 items; e.g., "strong"; after depletion: α = 89; after non-depletion: α = .89) as well as their negative affect (10 items; e.g., "anxious"; after depletion: α = 75; after non-depletion: α = .63). Items had to be answered on a 5-point Likerttype scale with answers ranging from 1 = not at all to 5 = very much.
Indicating that our ego depletion manipulation worked, participants perceived the depletion task to be more depleting compared to the control task [M depletion-non depletion = 0.64; t [bib_ref] When 'go' and 'nogo' are equally frequent: ERP components and cortical tomography, Lavric [/bib_ref] = −2.54, p = .019, d = 0.56; 95 % CI +0.11 to +1.16], but did not differ in their motivation to perform well in it, p = .569. There were no differences in positive or negative affect, all p's > .20.
## Eeg recording and preprocessing
EEG signals were recorded from 128 BioSemi active electrodes (www.biosemi.com) with a sampling rate of 2048 Hz. Biosemi uses a Common Mode Sense active electrode (CMS) and a Driven Right Leg passive electrode (DRL) as two as ground electrodes. Four additional electrodes measured horizontal and vertical eye-movement.
Pre-processing and statistical analyses were done using SPM8 for EEG (http://www.fil.ion.ucl.ac.uk/spm/). SPM provides a unitary framework for the analysis of neuroscience data acquired with different technologies, including EEG [bib_ref] Hierarchical models, Penny [/bib_ref] [bib_ref] EEG and MEG data analysis in SPM8, Litvak [/bib_ref]. In a first step, data were offline re-referenced to average reference. To identify artifacts caused by saccades (horizontal) or eye blinks (vertical), virtual channels were created from the electrooculographic electrodes and then subtracted from the EEG. Data were then downsampled to 250 Hz and band-pass filtered from 0.166 to 30 Hz with a fifth-order zero phase Butterworth filter. Filtered data were segmented from 100 ms before stimulus onset until 1000 ms after stimulus presentation. 100 ms before stimulus onset were used for baseline-correction. Automatic artifact detection was used for trials exceeding a threshold of 150 µV. Data were averaged, using the robust averaging algorithm of SPM8, excluding possible further artifacts. The idea of robust averaging is that for each channel and each time point outliers are down-weighted. An advantage of this approach is that clean averages can be calculated without having clean trials as artifacts are supposed to not consistently overlap and distort only parts of the trials. We used the recommended offset of the weighting function, which preserves approximately 95 % of the data points drawn from a random Gaussian distribution [bib_ref] EEG and MEG data analysis in SPM8, Litvak [/bib_ref]. Overall, 8.10 % of all electrodes were interpolated by the recorded activity of all other channels by spherical spline interpolationand 18.85 % of all trials were rejected, leaving on average 32.46 trials per block.
Source reconstructions of the generators of significant ERP differences were generated and statistically assessed with SPM8 for EEG (http://www.fil.ion.ucl.ac.uk/spm) following recommended procedures. First, a realistic boundary element head model (BEM) was derived from the SPM's template head model based on the MNI brain. The standard coordinates for all electrode positions then were transformed to match the template head, which is thought to generate reasonable results even when individual subjects head differ from the template [bib_ref] EEG and MEG data analysis in SPM8, Litvak [/bib_ref]. To this aim, average electrode positions as provided by BioSemi were co-registered with the cortical mesh template for source reconstruction. Group inversion [bib_ref] Electromagnetic source reconstruction for group studies, Litvak [/bib_ref] was computed and the multiple sparse priors algorithm implemented in SPM8 was applied. This method allows activated sources to vary in the degree of activity, but restricts the activated sources to be the same in all subjects [bib_ref] Electromagnetic source reconstruction for group studies, Litvak [/bib_ref]. Compared to for single subjects, this is thought to result in superior source estimation. For source reconstruction, frequency contents between 0.166 and 30 Hz were analyzed [bib_ref] EEG and MEG data analysis in SPM8, Litvak [/bib_ref]. For each analyzed time window, three-dimensional source reconstructions were generated as NIFTI images. These images were smoothed using an 8 mm full-width half-maximum kernel (voxel size = 2 mm × 2 mm × 2 mm). Since we had to use all sensors for our source estimation [bib_ref] EEG source imaging, Michel [/bib_ref] , it has to be noted, that the number of interpolated electrodes might also have influenced our source estimation results.
## Statistical analyses biat results
In order to test whether participants differed in manifest faking success as a function of experimental condition, a repeated measures ANOVA (condition: baseline, 'normal' faking and ego depletion faking) was calculated to investigate main effects for the resulting D-scores. Planned contrasts (pairwise comparisons) were used to investigate significant effects' (p < .05) directions of differences.
## Eeg data analyses
EEG scalp-data were analyzed with EMEGS [http://www. emegs.org/; 53]. For statistical analyses 2 (block: doping + like vs. doping + dislike) × 3 (condition: baseline vs. 'normal' faking vs. ego depletion faking) repeated measures ANOVAs were set-up to investigate interaction effects in time windows and electrode clusters of interest. Whenever Mauchlys Test of Sphericity was violated, Greenhouse-Geisser corrections of the degrees of freedom were performed. Effect sizes for repeated measures were calculated for all effects. Time windows and electrode clusters were segmented similar to Schindler, Wolff and colleagues [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Time windows were segmented from 150 to 200 ms to N1/P2 effects, from 200 to 300 ms to investigate P2/N2 effects and from 250 to 400 ms to investigate P3 effects and from 500 to 700 ms to investigate LPP effects. However, the topographical information about significant interactions between block and condition were plotted (see Additional file 4: [fig_ref] Figure 1: The experimental design [/fig_ref] , and an additional time window from 120 to 150 ms was segmented to investigate parietal P1 effects. The selected electrode clusters for all investigated components are displayed in [fig_ref] Figure 2: Selected electrode clusters for all investigated components [/fig_ref].
Statistical tests for source estimations were calculated for significant scalp interaction time windows. In source space, within significant interactions, post hoc tests were calculated between the doping + like and doping + dislike block for each condition. A threshold of p < .005 [e.g., see [bib_ref] Anterobasal temporal lobe lesions alter recurrent functional connectivity within the ventral pathway..., Campo [/bib_ref] [bib_ref] Perceived communicative context and emotional content amplify visual word processing in the..., Schindler [/bib_ref] [bib_ref] People matter: perceived sender identity modulates cerebral processing of socio-emotional language feedback, Schindler [/bib_ref] with a minimum of 25 significant voxels [bib_ref] Perceived communicative context and emotional content amplify visual word processing in the..., Schindler [/bib_ref] [bib_ref] People matter: perceived sender identity modulates cerebral processing of socio-emotional language feedback, Schindler [/bib_ref] was applied. The identification of involved brain regions was performed using the AAL atlas [bib_ref] Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation..., Tzourio-Mazoyer [/bib_ref].
# Results
# Behavioral results
Faking in both the 'normal' and the depleted faking condition is most likely to occur by responses slowing on the doping + like block and normal performance in the doping + dislike block. However, since participants were not given an explicit strategy, we cannot be sure what strategy they used, or if they used a strategy at all. However, supporting that our faking instruction yielded the desired effect, participants' response times in the BIAT's 'doping + like' block were slower when they were asked to try faking a negative attitude to doping compared to the baseline BIAT (see [fig_ref] Figure 3: Mean reaction time in each block depicted for each condition [/fig_ref]. A repeated measures ANOVA on the participants' raw reaction times revealed a significant interaction between condition and block [F (1.46, 30.62) = 9.18, p = .002, η p 2 = 304]. Within baseline condition no differences between the doping + like and doping + dislike were found (p = .291). In contrast, slower responses were given in the doping + like block both for 'normal' faking (p = .011) and depleted faking (p = .002).
For the participants' D-scores, similarly a repeated measures ANOVA showed that participants were successful at faking a negative doping attitude on the doping BIAT [F (2, 42) = 13.04, p < .001, η p 2 = .383]. Planned simple contrast analyses showed that the D-scores in the depletion condition and in the nondepletion condition were significantly higher than baseline D-scores, [F (1, 21) = 18.40, p < .001, partial η 2 = .467], and [F (1, 21) = 8.57, p = .008, η p 2 = .290] respectively. This indicates that our manipulation worked and participants were able to fake an artificially negative attitude to doping. Interestingly, a comparison between the two faking conditions showed a more negative D-score under ego depletion compared to 'normal' faking [F [bib_ref] The functional neuroanatomical correlates of response variability: evidence from a response inhibition..., Bellgrove [/bib_ref] = 7.85, p = .011, partial η 2 = .272].
# Eeg results
## Centro-parietal sensor cluster
Over centro-parietal locations, a significant interaction between (block: doping + like vs. doping + dislike) and condition (condition: baseline, ego depletion faking, 'normal' faking) was found already at the P1 [F [bib_ref] Studies of interference in serial verbal reactions, Stroop [/bib_ref] Regarding the late stages, in the P3 time window (250-400 ms), a trend for a significant interaction between
## Frontal sensor cluster
Between 150 and 200 ms, over frontal regions, no interaction was found for the frontal P2 [F [bib_ref] Studies of interference in serial verbal reactions, Stroop [/bib_ref] = 0.53, p = .592, η p 2 = .025]. For the subsequently occurring N2, between 200 to 300 ms, again no significant block by condition interaction was found [F [bib_ref] Studies of interference in serial verbal reactions, Stroop [/bib_ref] = 0.51, p = .542, η p 2 = .029]. However, descriptively the expected negativity for the supposed faking block (doping + like) was most pronounced for the 'normal' faking condition (M doping+like-doping+dislike = −0.64 µV) compared to ego
## Occipital sensor cluster
Over occipital sensors, no interaction between block and condition occurred regarding the occipital N1
## Source analyses
Source analyses were calculated for the significant interaction effects in scalp space. However, no differences in differences for cortical generators could be observed for the P1 interaction. In contrast, in the LPP time window (500-700 ms) a significant interaction was also found in source space (see ; . Post-hoc tests showed that enhanced activity was found for the supposed faking block (doping + like) in the 'normal' faking condition. For the faking block, enhanced activity was found in the bilateral middle temporal gyri/temporoparietal junction. No differences were found for the baseline condition and the ego depletion faking condition.
# Discussion
Our studies goal was twofold: First, we were interested in the ERP correlates of IAT faking when participants were not provided with an effective faking strategy. Second, we assessed how the ERP's of faking differed when participants were high or low (i.e., ego depleted) in temporarily available self-control resources.
Behaviorally, participants were able to fake a more negative D-score in both faking conditions compared to the baseline. Somewhat surprisingly, the D-score under ego depletion was even more negative compared to 'normal' faking. Regarding ERPs, we found a significant interaction between block and condition already for the parietal P1 component. Post-hoc tests did not reveal significant differences within the respective conditions. However, for the supposed faking block in the 'normal' faking condition a descriptively smaller P1 was elicited, while for both the ego depletion faking condition and the baseline condition the reverse pattern was found. For such early components, differences are rarely found for perceptually identical stimuli. This might indicate that faking influences already very early time points, channeling later faking related cognitive processes. Of course one should be aware, that we did not expect this interaction and therefore such early effects need to be replicated.
Nevertheless, it is interesting that the pattern of results matches those found at the late time window (see [fig_ref] Figure 4: Results for the P1 and LPP time windows over parietal regions [/fig_ref].
In the P300 time window, there was only a trend for an interaction between block and condition. Importantly, in the later occurring LPP time window the interaction term became significant. Within the faking condition there was a significantly decreased amplitude for the supposed faking block (doping + like) in the 'normal' faking condition. The present data show that for uninstructed 'normal' BIAT faking the decrease in the late parietal amplitudes is similar to instructed BIAT faking [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Further, this pattern seems to be present already at the P1. We previously noted, that LPP modulations seem to Significant differences in source activity for 'normal' faking (t-contrasts). The faking block is displayed in italics. In the time window from 500 to 700 ms, more activity was be observed over the bilateral middle temporal gyri/temporoparietal junction for faking be preceded by earlier sensory differences, but are better predictors of IAT performance [bib_ref] Cerebral correlates of automatic associations towards performance enhancing substances, Schindler [/bib_ref]. In contrast, post hoc tests did not reveal differences between baseline or ego depletion blocks. For these conditions it seems that the doping + like block elicited larger amplitudes. In the LPP time window the ego depletion faking condition appears to be more similar to the baseline condition than compared to the 'normal' faking condition (see [fig_ref] Figure 4: Results for the P1 and LPP time windows over parietal regions [/fig_ref]. This finding could be related to a reduced ability to exert cognitive control in a state of ego depletion, as proposed by some researchers [bib_ref] Ego-depletion and prejudice: separating automatic and controlled components, Govorun [/bib_ref]. However, one should be aware that we found no differences in the faking success between both faking conditions. Since we regard shifts in the D-score as an indicator of faking success, we might even conclude that faking success was higher under ego depletion. To explain the difference between our behavioral and neuronal data, one might speculate that different brain activation patterns can lead to similar behavioral outcomes. This relates well to a recent review, which discusses evidence for compensatory brain activation under ego depletion [bib_ref] Motivation and cognitive control: from behavior to neural mechanism, Botvinick [/bib_ref]. However, another important reason for these differences is that ERPs and reaction times are essentially different measures. The ERP responses are direct measures of the processing of the presented visual stimuli. Here, next to the primary sensory processing, various other processing tasks modulate the ERP amplitudes, for example the response selection, response preparation and motor execution (and during faking an inhibition/delay of the response). On the other hand, reaction times are the very end of this process. Nevertheless, combining behavioral and neuronal measures might help to understand why sometimes differences are found under ego depletion, while sometimes no differences can be observed.
It should be mentioned, that in the absence of inhibitory tasks, slower responding has been found to affect late components like the P3 [bib_ref] Preparation for action: an ERP study about two tasks provoking variability in..., Wascher [/bib_ref]. Here, smaller amplitudes are reported for faster compared to slower subjects [bib_ref] Preparation for action: an ERP study about two tasks provoking variability in..., Wascher [/bib_ref]. Moreover, on the individual level, fast compared to slow responses were found to elicit a slightly earlier peaking P3 [bib_ref] Preparation for action: an ERP study about two tasks provoking variability in..., Wascher [/bib_ref]. This complicates direct comparisons of our supposed faking blocks to the respective baseline block. Since we have a systematical reaction time difference between the faking and baseline doping + like blocks (as induced by the manipulation), this issue cannot be eliminated by an Analysis of Covariance [bib_ref] Misunderstanding analysis of covariance, Miller [/bib_ref]. However, first we do not see a delayed peaking P3 for the doping + like block during faking (cf. [fig_ref] Figure 4: Results for the P1 and LPP time windows over parietal regions [/fig_ref]. Second, since the reaction times were comparably slow in the two faking blocks, the differential pattern between 'normal' and depleted faking cannot be explained by the slower response or the execution of the button press. On the other hand, despite differences in response speed, the ERPs seem quite similarly going in the ego depletion and baseline condition. Finally, an overall comparable pattern can be observed already at the P1, not considered to be affected by slower responses.
In source space, an enhanced activity for faking in the bilateral middle temporal gyri/TPJ could be found within the 'normal' faking condition. This region has been reported to be more active during forced-choice faking [bib_ref] Lying in the scanner: covert countermeasures disrupt deception detection by functional magnetic..., Ganis [/bib_ref]. However, as enhanced TPJ activity is also related to attention [bib_ref] Attentional control during the transient updating of cue information, Pessoa [/bib_ref] and intentional action execution [bib_ref] A region of right posterior superior temporal sulcus responds to observed intentional..., Saxe [/bib_ref] [bib_ref] Thinking about intentions, Den Ouden [/bib_ref] it has been previously interpreted to reflect the monitoring of the planned faking response [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. However, TPJ activity is not restricted to attention and executive control. It has shown to play a role in a wide variety of domains such as theory of mind [bib_ref] Fractionating theory of mind: a meta-analysis of functional brain imaging studies, Schurz [/bib_ref] [bib_ref] People thinking about thinking people-the role of the temporo-parietal junction in "theory..., Saxe [/bib_ref] , as well as in social cognition [bib_ref] Linking brain structure and activation in temporoparietal junction to explain the neurobiology..., Morishima [/bib_ref]. It has been suggested that the TPJ plays a role in both social cognitive specific as well as unspecific attention an memory processes [bib_ref] A nexus model of the temporal-parietal junction, Carter [/bib_ref]. No differences in activity were found for faking in right-inferior frontal regions [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref] , possibly due to a reduced signal to noise ratio for these estimations.
## Table 1 source analyses for significant scalp effects
Differences between the doping + like and doping + dislike block were calculated within significant interaction terms No. of sig. voxels per cluster = number of significant voxels for each cluster. unc = uncorrected p value. Each cluster may exhibit more than one peak, while only the largest peak is displayed. Peak coordinates are displayed in MNI space (x, y and z). The identification of area labels for each peak was performed using the AAL-atlas. R/L = right or left hemisphere. The faking block is displayed in italics Source estimation results might also indicate a differential cerebral processing in the two faking conditions [bib_ref] Motivation and cognitive control: from behavior to neural mechanism, Botvinick [/bib_ref]. However, while for example Friese et al. [bib_ref] Suppressing emotions impairs subsequent stroop performance and reduces prefrontal brain activation, Friese [/bib_ref] found decreased activity in the right lateral prefrontal cortex, we found no differences between the ego depletion faking and non-faking block. This might be due to the lower spatial resolution of the source estimations compared to fMRI measures. Depletion effects might lead to a reduction of activity within some parts of executive networks [bib_ref] Motivation and cognitive control: from behavior to neural mechanism, Botvinick [/bib_ref] , although present source estimations may not be able to capture these changes. Further some studies suggest that under ego depletion increased activations can be found which are supposed to reflect compensatory effort [bib_ref] Imaging fatigue of interference control reveals the neural basis of executive resource..., Persson [/bib_ref] [bib_ref] Selfregulatory depletion enhances neural responses to rewards and impairs top-down control, Wagner [/bib_ref]. Although we did not find differences in source space on a reasonable threshold, compensatory activations within some parts of executive network might account for the discrepancy between ERPs for 'normal' and ego depleted faking.
On the other hand, in contrast to previously reported modulations of early frontal and occipital components [bib_ref] Cerebral correlates of automatic associations towards performance enhancing substances, Schindler [/bib_ref] , we did not find substantial differences, neither for depleted nor for 'normal' faking. Descriptively, participants in the 'normal' faking condition showed a negativity over frontal and positivity over occipital sensors in the supposed faking block (doping + like). However, the insignificant interaction prohibited post hoc testing. An explanation might be the increased difficulty of the task. So far, studies reporting such early effects have used a very simple faking task for the participants, providing an instruction how and when to fake [bib_ref] Temporal course of executive control when lying about self-and other-referential information: an..., Hu [/bib_ref]. So, by using such blocked designs participants could prepare how to respond, which might have affected even such early sensory processing. In the current experiment participants were not provided with an explicit instruction and had to find a strategy on their own. This could have led to unsystematic faking strategies. Thus, not all participants might have faked using a set response strategy. Some might have waited for the stimulus onset and then planed how to respond. In line with this interpretation, standard deviations seem to be larger in the supposed faking blocks than in previous study on instructed faking (instructed faking SD = 348 ms; 'normal' uninstructed faking SD = 587 ms; ego depletion uninstructed faking SD = 685 ms) [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Another explanation for smaller interactions is based on methodological reasons: In the previous study on IAT faking, participants were instructed to fake negative as well as positive doping attitudes, thus both faking conditions should deviate in a different direction from baseline [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. In this experiment, the expected faking block and direction was the same for both faking conditions.
In principle, our results suggest that there is no specific ERP component for faking. In accordance with the literature [bib_ref] Differential effects of practice on the executive processes used for truthful and..., Johnson [/bib_ref] [bib_ref] The self in conflict: the role of executive processes during truthful and..., Johnson [/bib_ref] [bib_ref] The role of episodic memory in controlled evaluative judgments about attitudes: an..., Johnson [/bib_ref] [bib_ref] An event-related potential paradigm for identifying (rare negative) attitude stimuli that people..., Crites [/bib_ref] , we think that the cognitive processes involved in faking modulates the ERP responses. We have speculated that individual and temporal differences in the available self-control resources might affect the cerebral faking pattern. Supporting this, we found different EEG faking patterns for depleted and non-depleted participants. Future research on faking components might profit from monitoring both state and trait self-control as a possible moderator variable.
In regard to the BIAT, our study shows that it can be faked even if participants are not provided with an effective faking strategy and when their self-control resources are depleted (albeit to a lesser extent than when provided with an effective faking strategy; [bib_ref] Cerebral correlates of faking: evidence from a brief implicit association test on..., Schindler [/bib_ref]. Instructing participants to fake an (B)IAT and even providing them with a faking strategy is the standard setup of IAT faking research [e.g., [bib_ref] Faking the IAT: aided and unaided response control on the implicit association..., Fiedler [/bib_ref] [bib_ref] Voluntary controllability of the implicit association test (IAT), Kim [/bib_ref] [bib_ref] Faking of the implicit association test is statistically detectable and partly correctable, Cvencek [/bib_ref]. However, it has been argued that although this can be a powerful design to assess the principal fakeability of a test, it does not reflect real world faking [bib_ref] The effect of implicitly incentivized faking on explicit and implicit measures of..., Wolff [/bib_ref]. Consequently, Wolff and colleagues [bib_ref] The effect of implicitly incentivized faking on explicit and implicit measures of..., Wolff [/bib_ref] used an approach were participants were only implicitly incentivized to fake a BIAT: If their doping attitude-measured with a BIAT or a self-report measure-would exceed a certain (bogus) cut-off, participants would be subjected to a tedious anti-doping training program. This manipulation led to successfully faked scores only in the selfreport measure but not in the BIAT. As such a setting is more likely to mirror real-world faking, it will be interesting to investigate the associated ERP patterns. Even more so in light of our finding that effectively faked scores in the 'normal' faking and the ego depleted faking condition are associated with different electro-cortical correlates. Thus, even though participants are not successful at faking a BIAT in an implicitly incentivized scenario, the ERP's between the incentivized and the control condition may still differ. Finally, in cases were faking cannot be determined statistically from the behavioral data, ERPs and source reconstruction may help to distinguish between truth and lie.
# Conclusion
Although the effect is smaller, uninstructed BIAT faking in normal state resembles previous results from instructed BIAT faking, both behaviorally and neurophysiologically. However, our results also indicate that while participants succeed in faking a BIAT when self-control resources are temporarily depleted, the associated ERP patterns differ. When depleted, ERPs differences at early and late processing stages are similar to the baseline pattern but different from 'normal' faking results. In contrast to 'normal' faking, no reliable sources were found for faking under ego depletion. Taken together, our results underline the importance of further understanding the electrophysiological correlates of test
[fig] Figure 1: The experimental design. Participants completed a total of four BIAT's. The first two BIAT's were completed with the standard instructions and the last two BIAT's were completed with the instruction to fake a negative doping attitude. Faking under ego depletion was operationalized by an incongruent Stroop preceding a BIAT (dark grey area) and 'normal' faking was operationalized by a congruent Stroop task preceding a BIAT (light grey area). a The practice BIAT consisted of 20 discrimination trials; and 20 trials in each of the doping + like and the doping + dislike blocks. b The discrimination block was removed for the subsequent BIAT's and the doping + like and doping + dislike blocks were increased to 40 trials each [/fig]
[fig] Figure 2: Selected electrode clusters for all investigated components [/fig]
[fig] Figure 3: Mean reaction time in each block depicted for each condition. The D-score reflects the standardized mean difference between the doping + like and the doping + dislike blocks. Error Bars represent standard deviations block and condition was found [F (2,42) = 2.63, p = .084 η p 2 = .111]. This interaction became significant in the LPP time window (500-700 ms; see Fig. 4; F (2,42) = 5.06, p = .011 η p 2 = .194). Post-hoc tests showed, that at this time window a significant decreased LPP was observed for the doping + like block of the 'normal' faking condition compared to the doping + dislike block [M doping+like-doping+dislike = −0.65, t (1,21) = −2.16, p = .043, d = −0.39, 95 % CI −1.28 to −0.02]. In contrast, there were no differences between the two blocks in the ego depletion faking condition [M doping+like-doping+dislike = 0.35, t (1,21) = 1.32, p = .201, d = 0.19, 95 % CI −0.20 to +0.89]. Finally, no differences could be found between the two blocks of the baseline BIAT [M doping+like-doping+dislike = 0.27, t (1,21) = 1.14, p = .265, d = 0.17, 95 % CI −0.22 to +0.75]. [/fig]
[fig] Figure 4: Results for the P1 and LPP time windows over parietal regions. a Difference topographies (doping + like-doping + dislike) for the faking conditions: blue color indicates more negativity and red color more positivity for the supposed faking block. b Mean amplitudes in microvolt over the centro-parietal electrode cluster for the P1 and LPP. Error bars represent ±2 times the standard error of the mean. c Selected electrode PPOz displaying the time course over parietal sites depletion faking (M doping+like-doping+dislike = −0.28 µV) and baseline condition (M doping+like-doping+dislike = −0.25 µV). [/fig]
|
A feasibility study of accelerated polychemotherapy with cisplatin, epidoxorubicin and cyclophosphamide (PEC) in advanced ovarian cancer.
We have evaluated the feasibility of an increase in dose intensity of the cisplatin, epidoxorubicin and cyclophosphamide (PEC) regimen, with granulocyte colony-stimulating factor (G-CSF) support, in 22 patients with advanced ovarian cancer. Twenty-one patients (95.4%) received six cycles of treatment: of these, 13 (61.9%) were also able to repeat cycles every 14 days as planned. Marrow toxicity was similar to that observed during conventional treatments. No severe mucositis or thrombocytopenia was observed. A clinical complete response was observed in 9 out of 16 evaluable patients (56.2%).
The importance of dose intensity in chemotherapy, i.e. the amount of drug delivered per unit of time, has been stressed in several experimental and clinical contributions after early retrospective analyses by and [bib_ref] The impact of dose-intensity on the design of clinical trials, Hryniuk W [/bib_ref] underlined the relationship between planned dose intensity and response rate in breast cancer. Based on another retrospective study [bib_ref] The application of dose intensity in chemotherapy of ovarian and endometrial cancer, Levine L And [/bib_ref] , ovarian cancer appears to be a particularly suitable model for intensification of chemotherapy: however, prospective data about the use of intensified regimens in this disease are still scarce.
In the present study we have evaluated the feasibility of an increase of the dose intensity of a polychemotherapy regimen commonly used in ovarian cancer (PEC; cisplatin, epidoxorubicin and cyclophosphamide) . Such increase was obtained reducing the intervals between cycles, with the use of granulocyte colony-stimulating factor (filgrastim, G-CSF) as prophylactic supportive treatment.
# Patients and methods
To be eligible for the trial, patients had to have histological diagnosis of epithelial ovarian carcinoma; previous adequate cytoreductive surgery was required, with FIGO III -IV staging; serum creatinine, serum bilirubin and haemogram had to be within normal limits. Informed consent was obtained from all patients.
Chemotherapy consisted of cisplatin 50 mg m-2, epidoxorubicin 60 mg m-2 and cyclophosphamide 600 mg m-2, all administered intravenously on day 1 every 14 days for six cycles. Filgrastim was administered subcutaneously from day 4 to day 9 between cycles.
In the case of incomplete marrow recovery (WBC < 3000 mm-' and/or platelets < 100.000 mm-3) on day 1 of the cycle, chemotherapy was delayed for 1 week or until complete marrow recovery. In the case of haemoglobin levels of 9 g dl-' or less a blood transfusion was supplied and chemotherapy was not postponed. In the case of a platelet count of 50 000 mm-3 or less, detected at any time during therapy, a 50% reduction of the doses of all drugs was planned for the remaining courses.
The primary aim of the study was to evaluate the feasibility of accelerated PEC treatment: the activity of the treatment was also assessed according to WHO response criteria. Toxicity was evaluated according to ECOG criteria.
Planned and total delivered dose intensity were calculated as the amount of drug (mg m-2) administered per unit of time (week), according to the indications of and [bib_ref] The description of chemotherapy delivery: options and pitfalls, Coppin Cml [/bib_ref]. The planned dose intensity was 25 mg m-2 per week for cisplatin, 30 mg m-2 per week for epidoxorubicin and 300 mg m-2 per week for cyclophosphamide. For each patient, the actually delivered dose intensity (received dose intensity, RDI) was calculated as a percentage of the planned one
# Results
Twenty-two patients entered the trial. The main characteristics of patients are shown in. One patient refused to continue the trial after two cycles. Twenty-one patients (95.4%) received all six cycles of PEC without any reduction in doses: of these, 13 (59.1%) also completed the treatment without delays between cycles. Two patients (9.5%) completed the treatment with a delay of 1 week (between the fifth and the sixth cycle), and four (19%) with a delay of 2 weeks (1 week between the fourth and the fifth cycle and 1 week between the fifth and the sixth cycle). A delay greater than 2 weeks (respectively of 4 and 5 weeks) occurred in two patients (9.5%), one of which was suffering grade III acute emesis and delayed emesis that caused a poor compliance to chemotherapy, while the second patient delayed the cycles because of psychological distress.
The average RDI in the 21 evaluable patients was 93.9% of the planned one (range 70.3% -100%): in 19 cases (90.5%; 95% confidence limits, 77.9% -100%) RDI was at least 85% of planned, and in 13 cases (61.9%; 95% confidence limits, 41.1%-82.7%) it was 100% of the planned intensity.
A clinical complete response, as confirmed by computerised tomography (CT) scan, serum markers and pelvic examination, was observed in nine patients (56.2%) out of the 16 who entered the trial having evaluable disease after surgery.
Toxicities are shown in. Values of WBC, platelets and haemoglobin during treatment are shown in. An overall decline of platelets and leucocytes was observed. This decline reached levels lower than those required for recycle in eight cases (in the last two cycles). In spite of the general decline observed in blood cells, only six patients had haemoglobin levels below 9 g dl-' and needed a red cell transfusion. No case of febrile neutropenia was observed.
# Discussion
Since the introduction of platinum-based combinations, the prognosis of advanced ovarian carcinoma has improved . To assess the possibility of further progress, researchers in recent years have explored fields such as the association of platinum and anthracyclines [bib_ref] Cyclophosphamide plus cisplatin versus cyclophosphamide, doxorubicin and cisplatin chemotherapy of ovarian carcinoma:..., Ovarian Cancer [/bib_ref] and the issue of dose intensity: retrospective analyses, in fact, showed a direct relationship between clinical results and average relative dose intensity, i.e. a mean of the dose intensities of each drug in different regimens (Levine and [bib_ref] The impact of dose-intensity on the design of clinical trials, Hryniuk W [/bib_ref]. There are basically two ways to increase the dose intensity of a chemotherapy regimen: the first is to increase the dose of drugs in each cycle, while the second is to shorten the intervals between standard-dosed cycles. Based on the characteristics of available growth factors (G-CSF and GM-CSF), which allow a more rapid marrow recovery from previous chemotherapy [bib_ref] Reduction by granulocyte colony-stimulating factor of fever and neutropenia induced by chemotherapy..., Ozer H [/bib_ref] [bib_ref] Effect of granulocyte colony stimulating factor on neutropenia and associated morbidity due..., Jl [/bib_ref] [bib_ref] The use of granulocyte colony stimulating factor to increase the intensity of..., Bronchud Mh [/bib_ref] , we have chosen in the present study to accelerate a combination of cisplatin, epidoxorubicin and cyclophosphamide (PEC) that is often used in our country . Cycles were to be repeated every 2 weeks instead of every 3 or 4 weeks as is usual when PEC is administered without growth factor support. The results show the feasibility of such an accelerated regimen, with 90% of patients being able to receive at least 85% of the planned dose intensity and more than 60% receiving 100%. Since the addition of anthracyclines generally results in a reduction of dose intensity of cyclophosphamide and cisplatin, our study suggests that G-CSF support is not only able to avoid this, but is also able to obtain a consistent increase of dose intensity in a very manageable way. No life-threatening toxicity was observed and marrow toxicity remained similar to that reported during conventional treatments. Interestingly, in spite of the increased dose intensity, no severe mucositis or thrombocytopenia were observed. The issue of the possible clinical advantages associated with an increase of dose intensity, however, cannot be resolved by this study. Other trials have reported contrasting data. One studycompared weekly cisplatin vs standard cisplatin plus cyclophosphamide given every 3 weeks: as the two regimens had superimposable results, at least in patients with residual tumour >2 cm after surgery, it was suggested that the intensification of cisplatin is able to counterbalance the possible disadvantage of using singleagent chemotherapy.
The Gynecology Oncology Study Group (1987) has specifically tested the hypothesis of dose intensity in a randomised study comparing two planned doses of cisplatin and cyclophosphamide: in spite of the achievement of a significantly higher actually delivered dose intensity, better clinical results were not obtained with respect to conventional doses [bib_ref] A phase III tnral of dose intensity versus standard dose cisplatin and..., Mcguire [/bib_ref]. More encouraging results were reported by Kaye et al. in a comparative trial of two different dose intensities of cisplatin and cyclophosphamide, which was closed because of a significant survival advantage for the higher doses; however, long-term follow-up showed a reduced survival benefit [bib_ref] Long-term follow up of a randomized trial of cisplatin dose in advanced..., Sb [/bib_ref]. The Italian cooperative group GONO has compared two regimens of PEC using cisplatin at the dose of 50 mg 2 and 100 mg-2 respectively . The high-dose cisplatin regimen seemed significantly more toxic but not more active, although definitive results are not available.
On the whole, the possibility of obtaining clinical benefits from an increase of chemotherapy dose intensity in ovarian cancer still seems controversial. Future studies should also examine aspects such as quality of life and cost/benefit considerations: other interesting data may derive from the exploration of much higher dose intensity increases, obtainable for example with peripheral blood stem cell reinfusion.
|
Progressive Tumefactive Demyelination as the Only Result of Extensive Diagnostic Work-Up: A Case Report
Tumefactive demyelinating lesions belong to the rare variants of multiple sclerosis, posing a diagnostic challenge since it is difficult to distinguish them from a neoplasm or other brain lesions and they require a careful differential diagnosis. This contribution presents the case report of a young female with progressive tumefactive demyelinating brain and spinal cord lesions. An extensive diagnostic process including two brain biopsies and an autopsy did not reveal any explanatory diagnosis other than multiple sclerosis. The patient was treated by various disease-modifying treatments without significant effect and died from ascendent infection via ventriculoperitoneal shunt resulting in Staphylococcus aureus meningitis.
# Introduction
Tumefactive demyelination belongs to the rare variants of multiple sclerosis (MS), posing a diagnostic challenge and therapeutic enigma as they can mimic other pathologies such as brain neoplasm, abscess, vasculitis, or granulomatous disease.
Atypical features of MS plaques on MRI include size >2 cm, mass effect, edema, and/or the presence of ringlike or open-ring enhancement. Lesions with these characteristics are often described as tumefactive demyelinating lesions (TDLs).
The prevalence of TDLs is estimated to be one to three per 1,000 cases of MS (2), although Sánchez et al. report the prevalence of 21 per 1,000 cases of MS. Neuroimaging is necessary to confirm the diagnosis, and a biopsy may be warranted if imaging is not precise. The clinical presentation of patients with TDLs is variable and could be atypical for the demyelinating disease. The mass effect is usually the cause of symptoms due to the displacement of the surrounding tissue, is present in about half of TDL cases, and may lead to increased intracranial pressure and cerebral herniation. This contribution presents the case report of a young female with TDLs. Such a case report of a patient exhibiting similar TDLs has not been reported before.
## Case report
A 23-year-old female was admitted to the neurological department of a major university hospital presenting with a mild central paraparesis of the lower extremities; an MRI indicated T2-hyperintense lesions in a periventricular, infratentorial, and intramedullary localization; both contrast-enhancing and non-enhancing lesions were found. Five oligoclonal bands (OCBs) appeared in the cerebrospinal fluid (CSF), but no OCB appeared in the serum; no signs of neuro-infection were found in the CSF. The patient's medical and family history was unremarkable, without any chronic disease, neoplasm, or autoimmune disease.
The MS diagnosis was determined according to the McDonald 2010 criteria. The patient was treated with high-dose steroids resulting in a slight reduction of complaints. Chronic treatment with interferon beta-1b commenced, and the patient was relapsefree for 4 years; no MRI progression appeared.
At the age of 27, the patient exhibited a mild central paraparesis of the lower extremities (treated with high-dose steroids), and the chronic treatment began with dimethylfumarate. An MRI of the brain and spinal cord showed multiple TDLs. An extensive diagnostic process was made, including positron emission tomography (PET) demonstrating high accumulation of 18F-fluoroethyl-L-thyrosine (FLT) within the lesions and MRI spectroscopy revealing elevation of choline peak and choline/creatine ratio. The histopathological findings from stereotactic brain biopsy of the lesion in the left occipital lobe confirmed demyelination; no neoplasm signs were found.
A broad range of tests was undertaken during follow-up, all with negative results. The patient did not exhibit any other autoimmune disease or any other non-neurological manifestation over the whole follow-up period. She underwent a second CSF examination: no signs of infection or neoplasm were found, one OCB appeared in the CSF, and no OCB appeared in the serum.
At the age of 28, the patient exhibited left-sided negative sensitive symptoms and was treated with high-dose steroids. An MRI of the brain and cervical spinal cord showed significant progression of TDLs, and the treatment with natalizumab commenced.
At the age of 29, cognitive and gait problems together with a headache and papilledema occurred, and a diagnosis of obstructive hydrocephaluswas established. Thus, a ventriculoperitoneal shunt was inserted. Brain MRI revealed further progression of TDLs, and the patient underwent another brain lesion biopsy. Histopathological findings revealed demyelination.
At the age of 30, left-sided hemiparesis together with further progression of TDLs appeared. The patient was treated with highdose steroids and an immunomodulatory dose of intravenous immunoglobulins resulting in a slight reduction of complaints to mild central paraparesis of the lower extremities and mild left-sided hemiparesis. From a chronic treatment point of view, hematopoietic stem-cell transplantation was considered. However, the patient died from circumscribed peritonitis complicated by ascendent infection via ventriculoperitoneal shunt resulting in Staphylococcus aureus meningitis. The autopsy
## Laboratory test categories
# Detailed description result
Autoimmune antibodies
Anti-aquaporin-4 antibodies, anti-myelin oligodendrocyte glycoprotein antibodies, anti-nuclear antibodies, antibodies against extractable nuclear antigens, anti-double-and single-stranded DNA, anti-neutrophil cytoplasmic antibodies, anticardiolipin antibodies, rheumatoid factor, anti-cyclic citrullinated peptide antibodies LGI, leucine-rich glioma-inactivated; GABAR, gamma-aminobutyric acid receptor.
# Discussion
This contribution presents the case report of a young female with TDLs. An extensive diagnostic work-up, including two biopsies and an autopsy, did not reveal any other explanatory diagnosis other than tumefactive MS. The report is unique due to subacute progressive TDLs of the brain and spinal cord, nonresponsiveness to the high-efficacy drugs, and high diagnostic certainty due to multiple brain biopsies and an autopsy. Such a case report of a patient exhibiting similar TDLs has not been reported before. In distinguishing between TDLs and other pathologies on MRI, the features listed above can be helpful along with followup imaging since TDLs tend to resolve in response to steroid therapy (6). However, the differentiation between TDLs and brain tumors by MRI alone may be difficult. One of the pathologies to be considered as a differential diagnosis of TDLs is primary central nervous system lymphoma (PCNSL) as cases of either concurrence of MS and PCNSL or demyelinating lesions preceding the development of PCNSL have been reported.
The value of advanced MRI techniques such as diffusion, perfusion imaging, or magnetic resonance spectroscopy has been studied in distinguishing between TDLs and brain tumors; however, it appears that those techniques still cannot provide definite diagnosis, and further studies are required to determine their additional value.
PET may also play some role in distinguishing TDLs from brain tumors. FLT tracer, which has been used in the diagnostic work-up in our patient, is generally referred to as a marker of proliferation in brain tumors; however, it should not be considered entirely specific as its increased uptake had also been observed in demyelinating lesions.
In the presented case, we observed several imaging features considerably atypical for TDLs like mostly homogenous enhancement, spectroscopy, and FLT-PET findings. Also, considering the constant progression of the mass lesions during the therapy leading finally to the development of obstructive hydrocephalus, the coincidence of tumor infiltration was suspected, and the brain biopsy was required to provide a definite diagnosis.
There is little comment on the effect of disease-modifying therapy on the evolution of TDLs. Some evidence supports that fingolimod should be avoided in MS patients with TDLs due to possible exacerbation (4). Patients with TDLs may have a better prognosis compared to MS patients without such lesions, especially when there is a good recovery from a tumefactive lesion; however, there is a dearth of information on untreated TDLs in the literature.
# Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.
# Ethics statement
Ethical approval was not provided for this study on human participants because The manuscript presents a case report study. Written informed consent was not provided because According to the local ethics committees, the written informed consent is not necessary for a post-mortem case report study. Written informed consent was not obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
|
Estrogen Receptor-Alpha 36 Mediates Mitogenic Antiestrogen Signaling in ER-Negative Breast Cancer Cells
It is prevailingly thought that the antiestrogens tamoxifen and ICI 182, 780 are competitive antagonists of the estrogenbinding site of the estrogen receptor-alpha (ER-a). However, a plethora of evidence demonstrated both antiestrogens exhibit agonist activities in different systems such as activation of the membrane-initiated signaling pathways. The mechanisms by which antiestrogens mediate estrogen-like activities have not been fully established. Previously, a variant of ER-a, EP-a36, has been cloned and showed to mediate membrane-initiated estrogen and antiestrogen signaling in cells only expressing ER-a36. Here, we investigated the molecular mechanisms underlying the antiestrogen signaling in ERnegative breast cancer MDA-MB-231 and MDA-MB-436 cells that express high levels of endogenous ER-a36. We found that the effects of both 4-hydoxytamoxifen (4-OHT) and ICI 182, 780 (ICI) exhibited a non-monotonic, or biphasic dose response curve; antiestrogens at low concentrations, elicited a mitogenic signaling pathway to stimulate cell proliferation while at high concentrations, antiestrogens inhibited cell growth. Antiestrogens at l nM induced the phosphorylation of the Src-Y416 residue, an event to activate Src, while at 5 mM induced Src-Y527 phosphorylation that inactivates Src. Antiestrogens at 1 nM also induced phosphorylation of the MAPK/ERK and activated the Cyclin D1 promoter activity through the Src/EGFR/ STAT5 pathways but not at 5 mM. Knock-down of ER-a36 abrogated the biphasic antiestrogen signaling in these cells. Our results thus indicated that ER-a36 mediates biphasic antiestrogen signaling in the ER-negative breast cancer cells and Src functions as a switch of antiestrogen signaling dependent on concentrations of antiestrogens through the EGFR/STAT5 pathway.
# Introduction
The diverse physiological functions of estrogens are mediated by estrogen receptors ER-a and ER-b, both of which are ligandactivated transcription factors that stimulate target gene transcription [bib_ref] Mechanisms of estrogen action, Nilsson [/bib_ref]. Estrogen-induced transcription regulation has been prevailingly thought as the only mechanism of estrogen action. However, it became apparent now that not all of the physiological effects mediated by estrogens are accomplished through a direct effect on gene transcription. Another signaling pathway (also known as a 'non-classic,' 'non-genomic' or 'membrane-initiated' signaling pathway) exists that involves cytoplasmic signaling proteins, growth factor receptors and components of other membrane-initiated signaling pathways [bib_ref] Rapid actions of plasma membrane estrogen receptors, Kelly [/bib_ref] [bib_ref] Estrogen action and cytoplasmic signaling cascades. Part I: membrane-associated signaling complexes, Segars [/bib_ref].
Since mitogenic estrogen signaling plays a pivotal role in development and progression of ER-positive breast cancer, treatment with antiestrogens such as tamoxifen (TAM) has become a first-line therapy for advanced ER-positive breast cancer. However, laboratory and clinical evidence indicated that TAM and its metabolites such as 4-hydroxytamoxifen (4-OHT) have mixed agonist/antagonist or estrogenic/anti-estrogenic actions depending on cell and tissue context, and the agonist activity of tamoxifen may contribute to tamoxifen resistance observed in almost all patients treated with tamoxifen [bib_ref] Acquired tamoxifen resistance: correlation with reduced breast tumor levels of tamoxifen and..., Osborne [/bib_ref] [bib_ref] Tamoxifen and the isomers of 4-hydroxytamoxifen in tamoxifen-resistant tumors from breast cancer..., Osborne [/bib_ref] [bib_ref] New endocrine approaches to breast cancer, Howell [/bib_ref]. As a consequence, a more potent and ''pure'' antiestrogen, ICI 182, 780 (Fulvestrant, Faslodex) has been developed [bib_ref] A potent specific pure antiestrogen with clinical potential, Wakeling [/bib_ref].
TAM and 4-OHT are thought to function as antagonists by competing with 17-b-estradiol (E2b) and other estrogens for binding to ERs. Further structural studies revealed that TAM induces an ER-a conformation that does not recruit coactivators to trans-activate target genes but recruits co-repressors [bib_ref] Cofactor dynamics and sufficiency in estrogen receptor-regulated transcription, Shang [/bib_ref] , suggesting that TAM-and 4-OHT-bounded ER-a is unable to effectively activate genes involved in cell growth and breast cancer development. On the other hand, ICI 182, 780, a 'pure' antiestrogen, works in a different mechanism. ICI 182, 780 binds to ERs, impairs receptor dimerization and inhibits nuclear localization of receptor [bib_ref] Inhibition of estrogen receptor-DNA binding by the ''pure'' antiestrogen ICI 164,384 appears..., Fawell [/bib_ref] [bib_ref] Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover, Dauvois [/bib_ref]. Furthermore, ICI 182, 780 also accelerates degradation of the ER-a protein without a reduction of ER-a mRNA [bib_ref] Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover, Dauvois [/bib_ref] [bib_ref] Responses to pure antiestrogens (ICI 164384, ICI 182780) in estrogen-sensitive and -resistant..., Nicholson [/bib_ref]. Thus, ICI 182, 780 binds ER-a and accelerates degradation of ER-a protein, resulting in a complete inhibition of estrogen signaling mediated by ER-a.
Although ICI 182, 780 has been depicted as a non-agonist or 'full' or 'pure' antiestrogen, a number of laboratories reported estrogenic agonist activities of ICI 182, 780 in different systems. Estrogenic agonist activity of ICI 182, 780 has been reported in hippocampal neurons and in bone cells where ICI 182, 780 promoted bone growth [bib_ref] Estrogenic agonist activity of ICI 182,780 (Faslodex) in hippocampal neurons: implications for..., Zhao [/bib_ref] [bib_ref] Effect of the highaffinity estrogen receptor ligand ICI 182,780 on the rat..., Sibonga [/bib_ref]. Agonist-like activities of ICI 182, 780 have also been reported in human breast cancer cells [bib_ref] Complex agonist-like properties of ICI 182,780 (Faslodex) in human breast cancer cells..., Wu [/bib_ref] , sheep uterus [bib_ref] ICI 182,780 acts as a partial agonist and antagonist of estradiol effects..., Robertson [/bib_ref] and yeast [bib_ref] Activation of the human estrogen receptor by the antiestrogens ICI 182,780 and..., Dudley [/bib_ref]. The molecular mechanisms by which ICI 182, 780 acts as an estrogenic agonist have never been elucidated. Studies from several laboratories suggested that a membrane-associated estrogen-binding receptor mediates the agonist actions of ICI 182, 780 in neurons [bib_ref] Cellular and molecular mechanisms of estrogen regulation of memory function and neuroprotection..., Brinton [/bib_ref] [bib_ref] Estrogen actions throughout the brain, Mcewen [/bib_ref] [bib_ref] Selective estrogen receptor modulators (SERMs) for the brain: current status and remaining..., Zhao [/bib_ref] [bib_ref] Estrogens and ICI182,780 (Faslodex) modulate mitosis and cell death in immature cerebellar..., Wong [/bib_ref].
Previously, we identified and cloned a 36-kDa variant of ER-a, ER-a36 [bib_ref] Identification, cloning, and expression of human estrogen receptor-alpha36, a novel variant of..., Wang [/bib_ref]. ER-a36 lacks both transcription activation domains AF-1 and AF-2 of the 66 kDa ER-a (ER-a66), consistent with the fact that ER-a36 has no intrinsic transcriptional activity [bib_ref] Identification, cloning, and expression of human estrogen receptor-alpha36, a novel variant of..., Wang [/bib_ref] [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref] ER-a36 transcripts are generated from a promoter located in the first intron of the ER-a66 gene [bib_ref] Estrogen receptor-alpha (ER-alpha) suppresses expression of its variant ER-alpha 36, Zou [/bib_ref] , indicating that ER-a36 expression is regulated differently from ER-a66. Indeed, ER-a36 is expressed in specimens from ER-negative patients and ERnegative breast cancer cells that lack ER-a66 expression [bib_ref] ER-alpha36, a novel variant of ER-alpha, is expressed in ER-positive and -negative..., Lee [/bib_ref] [bib_ref] Expression of ER-{alpha}36, a novel variant of estrogen receptor {alpha}, and resistance..., Shi [/bib_ref] [bib_ref] ER-alpha36, a novel isoform of ER-alpha66, is commonly over-expressed in apocrine and..., Vranic [/bib_ref]. ER-a36 is mainly expressed on the plasma membrane and mediates membrane-initiated estrogen signaling [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref] [bib_ref] Involvement of estrogen receptor variant ER-alpha36, not GPR30, in nongenomic estrogen signaling, Kang [/bib_ref]. Antiestrogens such as TAM and ICI 182, 780 at 10 nM induced phosphorylation of the MAPK/ERK in HEK/293 cells expressing recombinant ER-a36 [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref]. ER-a36 also mediates agonist activity of tamoxifen in endometrial cancer cells [bib_ref] ER-alpha36, a variant of ER-alpha, promotes tamoxifen agonist action in endometrial cancer..., Lin [/bib_ref]. These results suggested that ER-a36-mediated non-genomic signaling pathway is involved in agonist activities of antiestrogens.
Recently, we reported that ER-a36 mediated mitogenic estrogen signaling in ER-negative breast cancer cells such as MDA-MB-231 and MDA-MB-436 cells that lack expression of ER-a66 but highly express ER-a36 [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref]. To exclude the involvement of ER-a66, we used these cells to study the effects and the underlying mechanisms of pharmacological high concentrations and clinical relevant low concentrations of antiestrogens. In addition, although MDA-MB-231 cells express the full-length ER-b, MDA-MB-436 cells express undetectable levels of fulllength ER-b [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref] , which will then exclude the possible involvement of ER-b.
In the current study, we examined the agonist activities of antiestrogens ICI 182, 780 and 4-OHT in the ER-negative breast cancer MDA-MB-231 and MDA-MB-436 cells and found that the ER-negative breast cancer cells exhibited biphasic growth response curves in response to these antiestrogens. We also found that ER-a36-mediated Src/EGFR/STAT5 signaling pathway plays an important role in the biphasic antiestrogen signaling.
# Results
## Antiestrogens stimulates proliferation of er-negative breast cancer cells
To test if antiestrogens such as ICI 182, 780 (ICI) and 4-OHT act as agonists in the ER-negative breast cancer cells, the growth rate of each cell line was determined by counting the number of cells cultured in different concentrations of ICI and 4-OHT. As shown in [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref] , the ER-negative breast cancer cells treated with low concentrations (,1 nM) antiestrogens exhibited an increased growth rate compared with cells treated with vehicle. The dose-response curves of these cells to antiestrogens exhibited a non-monotonic or biphasic pattern; increasing concentrations of antiestrogens that initially stimulated cell growth but inhibited cell growth at higher concentrations [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref]. Our data indicated that antiestrogens induced proliferation of ER-negative breast cancer cells in a biphasic pattern.
## Antiestrogens induces biphasic activation of the mapk/ erk and cyclin d1 expression in er-negative breast cancer cells
To determine whether antiestrogens induced phosphorylation of the MAPK/ERK1/2, a typical non-genomic estrogen-signaling event, in these two cell lines, we treated cells with ICI and 4-OHT at different concentrations (1 nM and 5 mM) for 10 min. These concentrations were chosen to reflect physiological estrogen concentration and pharmacological antiestrogens concentration, respectively. Western blot analysis with a phospho-specific ERK1/ 2 antibody was performed to assess the phosphorylation levels of the ERK1/2. As shown in [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref] , we found that both ICI and 4-OHT were able to induce the activation of the MAPK/ERK at a low concentration (1 nM) in both cell lines. However, the activation of the MAPK/ERK was not observed in cells treated with a high concentration (5 mM) of ICI and 4-OHT [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref] , consistent with the biphasic pattern of the dose-response curves of these cells to ICI and 4-OHT. To determine whether high concentrations of antiestrogens failed to activate the MAPK/ERK or inhibited the ERK activation, we examined the effects of high concentrations of ICI and 4-OHT on the ERK activation induced by EGF. We found that antiestrogens failed to inhibit ERK activation induced by EGF (data not shown), suggesting that high concentrations of antiestrogens may fail to activate the MAPK/ ERK.
It is well known that induction of the growth-promoting gene Cyclin D1 by estrogen contributes to estrogen-stimulated proliferation of ER-positive breast cancer cells. Previously, we reported that E2b was also able to induce expression of c-Myc and Cyclin D1 in the ER-negative breast cancer cells [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref]. To assess whether antiestrogens were also able to induce expression of Cyclin D1, we treated cells with two concentrations (1 nM or 5 mM) of antiestrogens for six hours, and Western blot analysis was performed to examine Cyclin D1 expression. We found that at 1 nM, both ICI and 4-OHT up-regulated expression levels of Cyclin D1 [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref] whereas at 5 mM, both antiestrogens failed to induce Cyclin D1 expression [fig_ref] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling [/fig_ref]. Thus, antiestrogens elicited a biphasic induction of Cyclin D1 expression in these ERnegative breast cancer cells.
## Src/egfr/stat5 are involved in biphasic antiestrogen signaling in er-negative breast cancer cells
Recently, we reported that E2b induced phosphorylation of Src-Tyr-416 and activated Src activity, which then induced phosphorylation of EGFR-Tyr-845 in these ER-negative breast cancer cells [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref]. We then examined the phosphorylation status of Src-Tyr-416 and EGFR-Tyr-845 in the cells treated with different concentrations of antiestrogens. [fig_ref] Figure 2: Different concentrations of antiestrogens induce Src phosphorylation at distinct residues [/fig_ref] shows that in both cell lines, 1 nM of ICI and 4-OHT elicited phosphorylation of Src-Tyr-416 and EGFR-Tyr-845 while failed to do so at 5 mM. Intriguingly, 5 mM of ICI and 4-OHT strongly induced phosphorylation of Src-Tyr-527, an event associated with inactivation of Src activity, which was not observed in the cells treated with 1 nM of antiestrogens. These results suggested that antiestrogens at low concentrations induced phosphorylation of Src-Y-416 and activated Src whereas at high concentrations, antiestrogens induced Src-Y-527 phosphorylation and inactivated Src activity.
It was reported that signal transducer and activator of transcription 5 (STAT5), Src and EGFR play important roles in estrogen-stimulated proliferation of ER-positive breast cancer cells [bib_ref] Signal transducer and activator of transcription 5b, c-Src, and epidermal growth factor..., Fox [/bib_ref] ; estrogen-induced Src activation and Src-dependent phosphorylation of EGFR-Tyr-845 recruit STAT5 as a downstream effector of phosphorylated EGFR-Tyr-845 [bib_ref] Signal transducer and activator of transcription 5b, c-Src, and epidermal growth factor..., Fox [/bib_ref]. To examine whether STAT5 is involved in the observed biphasic antiestrogen signaling, we transfected MDA-MB-231 and MDA-MB-436 cells with the 4 X M67 pTATA-TK-luciferase reporter plasmid that contains four copies of STAT-binding site and treated with antiestrogens at 1 nM and 5 mM. We found that 1 nM of antiestrogens activated the promoter activity of the reporter plasmid while 5 mM of antiestrogens failed to do so [fig_ref] Figure 3: Antiestrogens induce biphasic STAT5 activities in ER-negative breast cancer cells [/fig_ref] , suggesting that antiestrogens at low concentrations were able to activate STAT-mediated transcription. To confirm the involvement of STAT5, we included two dominant-negative mutants of STAT5a (STAT5aD713 and STAT5aD740) that inhibit transcription activation mediated by STAT5a/b [bib_ref] Naturally occurring dominant-negative Stat5 suppresses transcriptional activity of estrogen receptors and induces..., Yamashita [/bib_ref]. We found that both dominant-negative mutants of STAT5a potently inhibited 1 nM of both ICI and 4-OHT induced promoter activity of the 4 X M67 pTATA-TK-luciferase reporter plasmid [fig_ref] Figure 3: Antiestrogens induce biphasic STAT5 activities in ER-negative breast cancer cells [/fig_ref] & C), indicating that STAT5 is involved in the biphasic antiestrogen signaling.
## Src is involved in biphasic cyclin d1 expression induced by different concentrations of antiestrogens
In the experiments described above, we observed that the cells treated with different concentrations of antiestrogens also exhibited biphasic patterns of Cyclin D1 expression. We decided to examine whether the Src signaling pathway is involved in the induction of Cyclin D1 expression by low concentrations of antiestrogens. We first tested if the Src inhibitors PP2 and dasatinib were able to inhibit Cyclin D1 induction by 1 nM of antiestrogens. Cells were treated with 1 nM of either ICI or 4-OHT and together with the Src inhibitors PP2 and dasatinib, the EGFR inhibitor AG1478 or the PI3K inhibitor LY294002, and Western blot analysis was performed to examine Cyclin D1 expression. [fig_ref] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression [/fig_ref] shows that 1 nM antiestrogen-induced Cyclin D1 expression was blocked by the Src inhibitors but not by AG1478 and LY294002, suggesting that Src is involved in Cyclin D1 expression induced by low concentrations of antiestrogens in these ER-negative breast cancer cells. To confirm Src function in Cyclin D1 induction by antiestrogens, we transiently transfected both cell lines with a human Cyclin D1 promoter-luciferase construct and then treated transfected cells with 1 nM or 5 mM antiestrogens. We found that 1 nM of both antiestrogens was able to induce Cyclin D1 promoter activity whereas at 5 mM, both antiestrogens failed to induce Cyclin D1 promoter activity [fig_ref] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression [/fig_ref] , indicating the biphasic effects of antiestrogens on induction of Cyclin D1 expression is through regulation of its promoter activity. The Cyclin D1 promoter activity induced by 1 nM of antiestrogens was inhibited by the Src inhibitors PP2 and dasatinib but not by AG1478 [fig_ref] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression [/fig_ref]. To further confirm the involvement of Src in the antiestrogen-induced Cyclin D1 expression, these ER-negative breast cancer cells were transiently co-transfected with the Cyclin D1 promoter reporter plasmid and pCMV5/SrcK295M, a dominant-negative mutant of Src, or pCMV5/SrcY527F, a constitutively active mutant of Src, respectively. We found that co-transfection of the dominantnegative mutant of Src abrogated the Cyclin D1 promoter activity induced by 1 nM estrogen while had no effects in cells treated with 5 mM antiestrogens [fig_ref] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression [/fig_ref]. On the contrary, the constitutively active mutant of Src (SrcY527F) released the Cyclin D1 promoter activity suppressed by 5 mM antiestrogens [fig_ref] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression [/fig_ref]. These results indicated Src plays an integral role in biphasic response of Cyclin D1 to different concentrations of antiestrogens.
## Stat5 is involved in antiestrogen induced cyclin d1 promoter activity
Previously, it was reported that prolactin induces Cyclin D1 promoter activity through activation of STAT proteins and their interaction with the consensus gamma-interferon-activation sites (GAS) located in the Cyclin D1 promoter [bib_ref] PRL activates the cyclin D1 promoter via the Jak2/Stat pathway, Brockman [/bib_ref]. We decided to examine whether antiestrogens function the same as prolactin in these ER-negative breast cancer cells. Two dominant-negative mutants of STAT5a were co-transfected with the Cyclin D1 promoter reporter plasmid, and the transfected cells were treated with 1 nM of ICI or 4-OHT. We found that inclusion of the two mutants of STAT5a strongly suppressed the Cyclin D1 promoter activity induced by 1 nM of antiestrogens , indicating that 1 nM of antiestrogens induced the Cyclin D1 promoter activity through STAT5 in ER-negative breast cancer cells.
In human cyclin D1 promoter, there are two GAS consensus sequences at 2457 (GAS1) and 2224 (GAS2) (relative to the transcription initiation site) that have been previously shown to be sites for STAT protein binding induced by prolactin [bib_ref] PRL activates the cyclin D1 promoter via the Jak2/Stat pathway, Brockman [/bib_ref]. To assess involvement of the two GAS sequences in antiestrogeninduced Cyclin D1 promoter activity, we transiently transfected these ER-negative breast cancer cells with two mutants of the Cyclin D1 promoter/reporter constructs, GAS1mut and GAS2mut that mutated the two GAS sequences respectively to prevent STAT protein binding. The Cyclin D1 promoter containing the GAS1 mutation failed to respond to 1 nM of ICI or 4-OHT while GAS2 mutant retained the ability to response to 1 nM of both antiestrogens , indicating that the STAT-binding GAS1 site is involved in the increase of Cyclin D1 promoter activity induced by low-concentrations of antiestrogens.
## Er-a36 mediates mitogenic antiestrogen signaling in ernegative breast cancer cells
Previously, we reported that ER-a36 mediates mitogenic estrogen signaling in ER-negative breast cancer MDA-MB-231 and MDA-MB-436 cells using shRNA method [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref]. To determine the involvement of ER-a36 in the antiestrogen signaling of these breast cancer cells, we tested antiestrogen signaling in the cell lines derived from MDA-MB-231 and MDA-MB-436 that carrying knocked-down levels of ER-(36 expression by the shRNA method. Cells derived from both cell lines with ER-(36 expression knockeddown by shRNA failed to increase cell proliferation in response to low-concentrations of antiestrogens [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref] , suggesting that ER-(36 mediates mitogenic antiestrogen signaling in these ERnegative breast cancer cells. However, at 5 mM, both antiestrogens potently inhibited proliferation of MDA-MB-231 and MDA-MB-436 cells with knocked-down levels of ER-a36 expression [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref].
We found that 1 nM antiestrogens failed to induce Src-Tyr-416 and EGFR-Tyr-845 phosphorylation in MDA-MB-231 and MDA-MB-436 cells with knocked-down level of ER-a36 expression [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref]. However, the basal levels of Src-Tyr-527 phosphorylation were dramatically increased in MDA-MB-231 and -436 cells transfected with ER-a36 shRNA expression vector compared to control cells transfected with the empty expression vector [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref] , which was not further induced by 5 mM of antiestrogens [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref]. We also tested whether antiestrogensare able to induce Cyclin D1 promoter activity in the cells with ER-a36 knock-down. Both antiestrogens at 1 nM failed to induce Cyclin D1 promoter activity [fig_ref] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells [/fig_ref].
## Different concentrations of antiestrogens affect the association of er-a36 and src differently
To elucidate the molecular mechanism by which different concentrations of antiestrogens influence Src phosphorylation in ER-negative breast cancer cells, we examined the effects of different concentrations of antiestrogens on the association of ER-a36 with Src as we reported before [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref]. MDA-MB-231 cells were transiently transfected with an expression vector for HA-tagged ER-a36 and treated with different concentrations of antiestrogens for 10 min. Cell lysates were immunoprecipitated with preimmune and anti-HA antibodies, and blotted by anti-HA and anti-Src antibodies. [fig_ref] Figure 7: Different concentrations of antiestrogens affect the association of ER-a36 and Src differently [/fig_ref] shows that at l nM, antiestrogens induced association of ER-a36 and Src, which was decreased when treated with 5 mM of antiestrogens [fig_ref] Figure 7: Different concentrations of antiestrogens affect the association of ER-a36 and Src differently [/fig_ref]. This result demonstrated that antiestrogens at 1 nM induced interaction between ER-a36 and Src but failed to do so at 5 mM.
# Discussion
Here, we used ER-negative breast cancer MDA-MB-231 and MDA-MB-436 cells as models to study the effects and the underlying mechanisms of the rapid, non-genomic antiestrogen signaling mediated by ER-a36. We found that these ER-negative breast cancer cells exhibited a biphasic growth response curve to antiestrogens ICI 182, 780 and 4-OHT; antiestrogens stimulated cell proliferation at sub-nM range while inhibited cell growth at mM range.
Antiestrogens such as tamoxifen and ICI 182, 780 are widely used for the treatment of advanced breast cancer, especially ERpositive breast cancer. It is prevailingly thought that ER-negative breast cancer is less or no responsive to antiestrogen therapy. However, it has also been reported that about 45% ER2/PR+ breast tumor patients and 10% ER2/PR2 tumor patients responded to tamoxifen treatment (Reviewed in [bib_ref] Activation of the p38 mitogen-activated protein kinase pathway by estrogen or by..., Zhang [/bib_ref] , suggesting that a subset of ER-negative breast cancer still responses to antiestrogen therapy. Previously, several in vitro studies showed that tamoxifen and 4-OHT can cross-talk with other signaling pathways such as the p38/MAPK and the SAPK/JNK pathways, and induce apoptosis in both ER-positive breast cancer cells such as MCF7 and ER-negative cells such as MDA-MB-231 [bib_ref] Activation of the p38 mitogen-activated protein kinase pathway by estrogen or by..., Zhang [/bib_ref] [bib_ref] Activation of caspase-3 and c-Jun NH2-terminal kinase-1 signaling pathways in tamoxifen-induced apoptosis..., Mandlekar [/bib_ref] [bib_ref] Estrogen receptor-dependent and estrogen receptor-independent pathways for tamoxifen and 4-hydroxytamoxifen-induced programmed cell..., Obrero [/bib_ref] [bib_ref] Mechanisms of tamoxifen-induced apoptosis, Mandlekar [/bib_ref]. However, we did not observe significant apoptosis in antiestrogen treated ER-negative breast cancer cells at 5 mM, the highest concentration we used, presumably because much higher concentrations of 4-OHT ($10-20 mM) were required to induce apoptosis in ER-negative breast cancer cells [bib_ref] Activation of the p38 mitogen-activated protein kinase pathway by estrogen or by..., Zhang [/bib_ref].
It is well-known that tamoxifen and its metabolite 4-OHT act as both agonists and antagonists in a tissue specific fashion. However, the mechanisms underlying the paradox effects of tamoxfen have never been fully elucidated. Different patterns of co-regulators expression and different post-translation modifications of ER have been proposed to be involved in tissue specific agonist/antagonist properties of tamoxifen (Reviewed in [bib_ref] Antiestrogen resistance in breast cancer and the role of estrogen receptor signaling, Clarke [/bib_ref]. We recently reported that ER-a36 mediates the agonist activity of tamoxifen by activation of the MAPK/ERK and PI3K/AKT signaling pathways in endometrial cancer cells that lack expression of ER-a66 [bib_ref] ER-alpha36, a variant of ER-alpha, promotes tamoxifen agonist action in endometrial cancer..., Lin [/bib_ref]. Our current results thus suggested that the agonist/ antagonist activities of tamoxifen are concentration dependent and ER-a36 is involved in the agonist/antagonist activities of tamoxifen.
ICI 182, 780 has been portrayed as a ''pure'' antiestrogen without any estrogenic activity [bib_ref] A potent specific pure antiestrogen with clinical potential, Wakeling [/bib_ref]. Here, we found that ICI 182, 780 worked just like 4-OHT and exhibited concentrationdependent agonist/antagonist activities in cells expressing ER-a36. Recently, we reported that ICI 182, 780 failed to induce degradation of ER-a36 [bib_ref] Breast cancer cell growth inhibition by phenethyl isothiocyanate is associated with down-regulation..., Kang [/bib_ref] , presumably because ER-a36 has a truncated ligand-binding domain that lacks the last 4 helixes (helix 9-12) of ER-a66 [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref]. The helix-12 domain is critical in protein degradation induced by ICI 182, 780 and different positioning of the helix 12 and the F domain of ER-a66 regulates functional differences between agonists and antagonists [bib_ref] Specific mutations in the estrogen receptor change the properties of antiestrogens to..., Mahfoudi [/bib_ref] [bib_ref] Modulation of estrogen receptor alpha function and stability by tamoxifen and a..., Pearce [/bib_ref] [bib_ref] Different positioning of the ligandbinding domain helix 12 and the F domain..., Nichols [/bib_ref]. This may provide a molecular explanation for the failure of ICI 182, 780 to induce ER-a36 degradation and inhibited ER-a36 activity. Previously, low concentrations of ICI 182, 780 were found to stimulate the growth of tamoxifen-resistant KPL-1 human breast cancer cells [bib_ref] A pure antiestrogen, ICI 182,780, stimulates the growth of tamoxifen-resistant KPL-1 human..., Kurebayashi [/bib_ref] and to induce phosphorylation of the MAPK/ ERK in neonatal rat cerebellar neurons [bib_ref] Estrogenic agonist activity of ICI 182,780 (Faslodex) in hippocampal neurons: implications for..., Zhao [/bib_ref]. Thus, like tamoxifen, ICI 182, 780 also has concentration-dependent agonist/antagonist activities.
It is worth noting that the earlier version of ER-a66/knockout mice that was generated by insertion of a Neo cassette into the first coding exon of the mouse ER-a gene [bib_ref] Alteration of reproductive function but not prenatal sexual development after insertional disruption..., Lubahn [/bib_ref] (the exon that is sipped in the generation of the transcript encoding ER-a36) retains ER-a36 expression (Elliot Sharon, personal communication). This version of ER-a66 deficient mice also retained ICI 182, 780 insensitive non-genomic estrogen-signaling in different tissues [bib_ref] Rapid action of 17beta-estradiol on kainateinduced currents in hippocampal neurons lacking intracellular..., Gu [/bib_ref] [bib_ref] Estrogenic responses in estrogen receptor-alpha deficient mice reveal a distinct estrogen signaling..., Das [/bib_ref] [bib_ref] The AF-1 activation-function of ERalpha may be dispensable to mediate the effect..., Pendaries [/bib_ref]. Our current results suggested that ER-a36 may be involved in the ICI 182, 780 resistant non-genomic estrogen signaling observed in the early version of ER-a66 deficient mice.
The involvement of Src in rapid estrogen signaling has been reported in the mouse neocortex, ER-positive breast cancer cells, and prostate cancer cells [bib_ref] Estradiol (E2) elicits SRC phosphorylation in the mouse neocortex: the initial event..., Nethrapalli [/bib_ref] [bib_ref] The role of Shc and insulin-like growth factor 1 receptor in mediating..., Song [/bib_ref] [bib_ref] Steroid-induced androgen receptor-oestradiol receptor beta-Src complex triggers prostate cancer cell proliferation, Migliaccio [/bib_ref]. It has also been reported that antiestrogen tamoxifen promotes phosphorylation of the adhesion molecules, p130Cas/BCAR1, FAK and Src [bib_ref] Tamoxifen treatment promotes phosphorylation of the adhesion molecules, p130Cas/BCAR1, FAK and Src,..., Cowell [/bib_ref]. In the present study, we found that at 1 nM, antiestrogens induced phosphorylation of Src at Tyr-416 and the downstream MAPK/ERK1/2. Intriguingly, we found that 5 mM antiestrogenstriggered phosphorylation of Src-Tyr-527 and failed to induce phosphorylation of Src-Tyr-416 and the MAPK/ERK1/2. Src can be switched from an inactive to an active state through control of its phosphorylation state [bib_ref] Structure-function relationships in Src family and related protein tyrosine kinases, Superti-Furga [/bib_ref]. Src-Tyr-416 can be auto-phosphorylated, which activates Src by displacing the P-Tyr-416 from the binding pocket, allowing the substrate to gain access. However, phosphorylation of Tyr-527 inactivates Src through the interaction of P-Tyr-527 with a SH2 domain, which effectively folds Src up into a closed, inactive state. Our results thus demonstrated, for the first time, that phosphorylation state of Src-Y-416 and-Y-527 acts as a switch of concentration dependent agonist/antagonist activities of antiestrogens.
Previously, we reported that E2b induced the physical interaction of ER-a36 and Src, and consequently the autophosphorylation of Src-Y-416 in the ER-negative breast cancer cells [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref]. Here, we found that ER-a36 knock-down diminished the Src-Y-416 phosphorylation induced by 1 nM of antiestrogens, indicating ER-a36 is involved in the auto-phosphorylation of Src-Y-416 induced by low-concentrations of antiestrogens. However, the cells with ER-a36 knock-down exhibited high basal levels of Src-Tyr-527 phosphorylation, which was not further induced by antiestrogens at 5 mM, indicating that abrogation of ER-a36 activity increased basal levels of Src-Tyr-527 phosphorylation and silenced Src activity, consistent with our previous findings that the ER-negative breast cancer cells with ER-a36 knock-down failed to form xenograft tumors [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref]. Furthermore, in the co-immunoprecipitation assays, we found that antiestrogens at low concentration (1 nM) induced interaction between ER-a36 and Src, suggesting that like estrogen, both antiestrogens are able to induce association of ER-a36 and Src as well as auto-phosphorylation of Src. At high concentration (5 mM), antiestrogens failed to induce the interaction of ER-a36 and Src. It is possible that different concentrations of antiestrogens may trigger different conformations of ER-a36, which regulates ER-a36 accessibility for Src binding. The failure of the interaction between ER-a36 and Src may increase the basal levels of Src-Tyr-527 phosphorylation to silence Src as we observed in the cells with ER-a36 expression knocked-down.
The present study demonstrated that Cyclin D1 expression also exhibited biphasic response to antiestrogens in these ER-negative breast cancer cells through the Src/EGFR/STAT5 pathway. The low concentrations of antiestrogens induced Src-mediated phosphorylation of the EGFR-Tyr-845 residue, which then recruits STAT5 as a downstream effector to induce Cyclin D1 expression through the GAS site located in the Cyclin D1 promoter. Srcdependent phosphorylation of EGFR-Tyr-845 is required for DNA synthesis induced by transactivating agonists of EGFR, such as endothelin, lysophosphatidic acid, cytokines and growth hormones [bib_ref] The EGF receptor as central transducer of heterologous signalling systems, Zwick [/bib_ref]. It was reported that STAT5b, c-Src and EGFR play important roles in estrogen-stimulated proliferation of ERpositive breast cancer cells [bib_ref] Signal transducer and activator of transcription 5b, c-Src, and epidermal growth factor..., Fox [/bib_ref]. Introduction of dominantnegative STAT5a mutants into ER-positive T47D breast cancer cells inhibits estrogen-stimulated cell growth and induces apoptosis . STAT5 is involved in antiestrogen-induced Cyclin D1 promoter activity. (A). The involvement of STAT5 in antiestrogens-induced Cyclin D1 promoter activity. Cells were transfected with the luciferase reported plasmid Cyclin D1 pl-963 together with an empty expression vector or two dominant-negative STAT5a mutants, STAT5aD713 and STAT5aD740, respectively. Transfected cells were treated with vehicle (ethanol), 1 nM or 5 mM of antiestrogens. Columns: means of the relative luciferase activity from four independent experiments. Luciferase activity in the cells transfected with an empty expression vector and treated with vehicle is arbitrarily set as 1.0; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. (B). The GAS1 is involved in induction of the Cyclin D1 promoter activity by antiestrogens. Cells were transiently transfected with either the wild-type Cyclin D1 promoter (CycD1) or the same promoter construct containing mutated GAS1 (GAS1mut) or GAS2 (GAS2mut) sequence, respectively. Transfected cells were treated with vehicle or 1 nM of antiestrogens, and the luciferase activity was presented relative to the wild-type Cyclin D1 promoter-transfected cells treated with vehicle that is arbitrarily set as 1.0. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. doi:10.1371/journal.pone.0030174.g005 [bib_ref] Naturally occurring dominant-negative Stat5 suppresses transcriptional activity of estrogen receptors and induces..., Yamashita [/bib_ref]. Thus, our results indicated that the EGFR/Src/STAT5 pathway is also involved in the biphasic antiestrogen signaling in ER-negative breast cancer cells.
In summary, we have shown that ER-a36 expressing ERnegative breast cancer cells exhibited biphasic response to antiestrogens, which further confirm that ER-a36 mediates nongenomic antiestrogen signaling. Our results also provided a possible explanation to the previous findings of the existence of two non-genomic estrogen-signaling pathways, one sensitive to antiestrogens and the other insensitive. The finding that antiestrogens at higher concentrations inhibit proliferation of ER-negative breast cancer cells through suppressing the EGFR/ Src/STAT5 signaling pathway provided a rational for development of more effective therapeutic approaches for ER-negative breast cancer using combinations of antiestrogens with EGFR or Src inhibitors.
# Materials and methods
## Chemicals and antibodies
The Src inhibitors PP2, the PI3K inhibitor LY294002, 4hydoxy tamoxifen (4-OHT) and ICI 182, 780 were from Tocris Bioscience (Ellisville, MO). The Src inhibitor dasatinib was obtained from LC Laboratories (Woburn, MA). Phospho-EGFR and -Src antibodies, EGFR and Src antibodies, anti-phospho-p44/42 ERK (Thr202/Tyr204) (197G2) mouse monoclonal antibody (mAb) and anti-p44/42 ERK (137F5) rabbit mAb were all purchased from Cell Signaling Technology (Boston, MA). Polyclonal anti-ER-a36 antibody was generated and characterized as described before [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref]. Antibody for Cyclin D1 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
## Cell culture, treatment and growth assay
MDA-MB-231 and MDA-MB-436 cells were obtained from American Type Culture Collection (ATCC, Manassas, VA). All parental and derivative cells were maintained at 37uC in a 10% CO 2 atmosphere in DMEM and 10% fetal calf serum in a humidified incubator. For antiestrogen treatment, cells were maintained in phenol red-free media with 2.5% charcoal-stripped fetal calf serum for two to three days, and then in serum-free medium for 24 hours before experimentation. For ERK activation assays, cells were treated with vehicle (ethanol) and indicated concentrations of 4-OHT and ICI 182, 780. To test the effects of different inhibitors, all inhibitors were added 10 min. before addition of antiestrogens.
Since these ER-negative breast cancer cells express high levels of EGFR, which make cells proliferate at a near-maximal rate in serum-supplemented medium, the effects of antiestrogen signaling on proliferation of these cells are too subtle to detect most time. To alleviate this problem, we reduced charcoal-stripped fetal calf serum concentration in culture medium from 10% to 2.5% and increased estrogen treatment time to 12 days in our cell growth assays.
To examine cell growth in the presence or absence of antiestrogens, cells maintained for three days in phenol red-free DMEM plus 2.5% dextran-charcoal-stripped fetal calf serum (HyClone, Logan, UT) were treated with different concentrations of 4-OHT, ICI 182, 780 or ethanol vehicle as a control. The cells were seeded at 1610 4 cells per dish in 60 mm dishes and the cell numbers were determined using the ADAM automatic cell counter (Digital Bio., Korea) after 12 days. Five dishes were used for each treatment and experiments were repeated more than three times.
Cell lines with ER-a36 expression knocked down by the shRNA method in MDA-MB-231 and MDA-MB-436 cells were generated and described before [bib_ref] A variant of estrogen receptor-{alpha}, hER-{alpha}36: transduction of estrogen-and antiestrogen-dependent membrane-initiated mitogenic..., Wang [/bib_ref].
## Plasmids, dna transfection and luciferase assay
The expression vectors for a dominant-negative mutant of Src (pCMV5/SrcK295) and a constitutively active mutant of Src (pCMV5/SrcY527F) were obtained from Dr. Yun Qiu at the Department of Pharmacology and Experimental Therapeutics, University of Maryland School of Medicine. Dr. Linda Schuler at Department of Comparative Biosciences, University of Wisconsin-Madison kindly provided the luciferase reporter plasmids of the Cyclin D1 promoter (pl-963) carrying GAS1 and 2 mutations. Two dominant-negative STAT5 mutants, Stat5aD713 and Stat5aD740 were provided by Dr. H Yamashita at Department of Surgery II, Nagoya City University. The wild-type luciferase reporter plasmid of the Cyclin D1 promoter, Cyclin D1 pl-963 was obtained from Dr. Chris Albanese at Departments of Oncology and Pathology, Georgetown University Medical Center. The 4XM67 pTATA-TK-luciferase reporter plasmid was purchased from Addgene (Cambridge, MA). Cells were cotransfected with a cytomegalovirus-driven Renilla luciferase plasmid, pRL-CMV (Promega, Madison, WI) to establish transfection efficiency. Twenty-four hours after transfection, cells were treated with vehicle, 10 mM of dasatinib, PP2, or LY294002 for twenty-four hours. Forty-eight hours after transfection, cell extracts were prepared and luciferase activities were determined and normalized using the Dual-Luciferase Assay System (Promega, Madison, WI) and a TD 20/20 Luminometer (Turner BioSystems, Inc., Sunnyvale, CA) as instructed by the manufacturer.
# Western blot analysis
For Western blot analysis, cells washed with cold PBS were lysed with the lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.25 mM EDTA pH 8.0, 0.1% SDS, 1% Triton X-100, 50 mM NaF) supplemented with protease and phosphatase inhibitors from Sigma. The protein amounts were measured using the DC protein assay kit (BIO-RAD Laboratories, Hercules, CA). The same amounts of the cell lysates were boiled for 5 minutes in loading buffer and separated on a SDS-PAGE gel. After electrophoresis, the proteins were transferred to a PVDF membrane. The membranes were probed with various primary antibodies, HRP-conjugated secondary antibodies, and visualized with enhanced chemiluminescence (ECL) detection reagents (GE Healthcare Bio-Sciences Corp. Piscataway, NJ).
## Immunoprecipitation and immunoblot analysis
For imunoprecipitation assays, cells were washed twice with ice-cold PBS and lysed with the lysis buffer (150 mM NaCl, 20 mM TrisHCl, pH 7.4, 0.1% NP-40) supplemented with protease and phosphatase inhibitors (Sigma, St. Louis, MO). Cell lysates were then incubated with indicated anti-HA antibodies, or pre-immune serum and immunoprecipitated with protein A/G plus agarose. The precipitates were then washed, separated on SDS-PAGE and analyzed with Western blot analysis as described before [bib_ref] Involvement of estrogen receptor variant ER-alpha36, not GPR30, in nongenomic estrogen signaling, Kang [/bib_ref] [bib_ref] A positive feedback loop of ER-alpha36/EGFR promotes malignant growth of ERnegative breast..., Zhang [/bib_ref].
# Statistical analysis
Data were summarized as the mean 6 standard error (SE) using the GraphPad InStat software program (GraphPad Software, La Jolla, CA, USA). Tukey-Kramer Multiple Comparisons Test was also used, and the significance was accepted for P,0.05.
[fig] Figure 1: ER-negative breast cancer cells exhibit biphasic antistrogen signaling. (A). The effects of 4-OHT and ICI 182, 780 on the proliferation rate of MDA-MB-231 and MDA-MB-436 cells. Cells maintained for three days in phenol red-free DMEM plus 2.5% dextran-charcoalstripped fetal calf serum were treated with indicated concentrations of 4-OHT, ICI or ethanol vehicle as a control. The cell numbers were determined using an automatic cell counter after 12 days. Five dishes were used for each concentration and experiments were repeated more than three times. The mean cell number 6 SE are shown. (B). The dose-dependent phosphorylation pattern of the MAPK/ERK1/2 in MDA-MB-231 and MDA-MB-436 cells treated with different concentrations of antiestrogens. Starved cells were treated with indicated doses of 4-OHT or ICI 182, 780 (ICI) for 10 min. Western blot analysis was performed to assess induction of ERK1/2 phosphorylation. The experiment was repeated more than three times. The representative results are shown. (C). The dose dependent induction Cyclin D1 by antiestrogens in MDA-MB-231 and MDA-MB-436 cells. The experiment was repeated more than three times. The representative results are shown. doi:10.1371/journal.pone.0030174.g001 [/fig]
[fig] Figure 2: Different concentrations of antiestrogens induce Src phosphorylation at distinct residues. Western blot analysis of the effects of different concentrations of antiestrogens on the phosphorylation levels of EGFR-Y845, Src-Y416 and Src-Y527 in MDA-MB-231 and MDA-MB-436 cells. doi:10.1371/journal.pone.0030174.g002 [/fig]
[fig] Figure 3: Antiestrogens induce biphasic STAT5 activities in ER-negative breast cancer cells. (A). ER-negative breast cancer cells were transfected with the luciferase reported plasmid 4XM67 TATA-TK-Luc that containing four copies of STAT-binding sites upstream of the minimal TK promoter. Transfected cells were treated with vehicle (ethanol), 1 nM or 5 mM of 4-OHT or ICI 182, 780. The luciferase activities were assayed and normalized using a cytomegalovirus-driven Renilla luciferase plasmid. Columns: means of the relative luciferase activity from four independent experiments. Luciferase activity in transfected cells treated with vehicle is arbitrarily set as 1.0; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. #, p,0.05, for cells treated with 5 mM vs 1 nM of antiestrogens. (B&C). Cells were transfected with the 4XM67 TATA-TK-Luc reporter together with an empty expression vector (vector) and the expression vectors of two dominant-negative STAT5a mutants carrying truncations at their C-terminal (STAT5aD713 and STAT5aD740) before treated with vehicle (ethanol), 1 nM or 5 mM of antiestrogens. Columns: means of the relative luciferase activity from three independent experiments. Luciferase activity of cells co-transfected with an empty expression vector and treated with vehicle is arbitrarily set as 1.0; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. doi:10.1371/journal.pone.0030174.g003 [/fig]
[fig] Figure 4: Src is involved in antiestrogen-induced Cyclin D1 expression. (A). Western blot analysis of Cyclin D1 expression in MDA-MB-231 and -436 cells. Cells were treated with vehicle (ethanol) and antiestrogens alone or together with the Src inhibitors PP2 and dasatinib, the EGFR inhibitor AG1478 and PI3K inhibitor LY294002. Cell lysates were analyzed with anti-Cyclin D1 antibody and anti-Acin antibody was used to ensure equal loading. The experiment was repeated three times, and the representative results are shown. (B). Src inhibitors inhibit antiestrogen-induced Cyclin D1 promoter activity. ER-negative breast cancer cells were transfected with the luciferase reported plasmid Cyclin D1 pl-963 that containing a luciferase gene driven by the Cyclin D1 promoter. Transfected cells were treated with vehicle (ethanol), 1 nM or 5 mM of antiestrogens, and 1 nM of antiestrogens together with different inhibitors. The luciferase activities were assayed and normalized using a cytomegalovirus promoter-driven Renilla luciferase plasmid. Columns: means of the relative luciferase activity in cells treated with vehicle that is arbitrarily set as 1.0 from four independent experiments; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens, or vehicle (V) vs 1 nM of antiestrogens plus AG1478. (C). The involvement of Src in antiestrogen-induced Cyclin D1 promoter activity. Cells were transfected with the luciferase reported plasmid Cyclin D1 pl-963 together with an empty expression vector or Src mutants, a dominant-negative mutant (SrcK295R) and a constitutively active mutant (SrcY527F). Transfected cells were treated with vehicle (ethanol), 1 nM or 5 mM of antiestrogens. The luciferase activities were assayed and normalized using a cytomegalovirus-driven Renilla luciferase plasmid. Columns: means of the relative luciferase activity from four independent experiments. Luciferase activity in transfected cells treated with vehicle is arbitrarily set as 1.0; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. #, p,0.05, for cells treated with vehicle (V) vs 5 mM of antiestrogens. doi:10.1371/journal.pone.0030174.g004 [/fig]
[fig] Figure 6: ER-a36 mediates biphasic antiestrogen signaling in ER-negative breast cancer cells. (A). The effects of antiestrogens on the proliferation rate of MDA-MB-231 and MDA-MB-436 cells with or without ER-a36 expression knocked-down. Cells maintained for three days in phenol red-free DMEM plus 2.5% dextran-charcoal-stripped fetal calf serum were treated with indicated concentrations of antiestrogens or ethanol vehicle as a control. The cell numbers were determined using an automatic cell counter after 12 days. Five dishes were used for each concentration and experiments were repeated three times. The mean cell number 6 SE are shown. (B). Western blot analysis of the effects of 1 nM or 5 mM of antiestrogens on the phosphorylation levels of the SrcY416, SrcY527 and EGFRY845 in MDA-MB-231 and MDA-MB-436 cells. (C). Cells were transfected with the luciferase reporter plasmid driven by the wild-type Cyclin D1 promoter and transfected cells were treated with vehicle (ethanol), 1 nM or 5 mM of antiestrogens. Columns: means of the relative luciferase activity in transfected cells treated with vehicle that is arbitrarily set as 1.0 from three independent experiments; bars, SE. *, p,0.05, for cells treated with vehicle (V) vs 1 nM of antiestrogens. doi:10.1371/journal.pone.0030174.g006 [/fig]
[fig] Figure 7: Different concentrations of antiestrogens affect the association of ER-a36 and Src differently. Co-immunoprecipitation and Western blot analysis of HA-ER-a36 and Src in MDA-MB-231 cells. Cells transiently transfected with an expression of HA-tagged ER-a36 and treated with different concentrations of antiestrogens for 10 min were lysised and the cell lysates were immunoprecipitated with pre-immune and anti-HA antibodies. The immunoprecipitates were blotted by anti-HA and anti-Src antibodies. doi:10.1371/journal.pone.0030174.g007 [/fig]
|
Duration of obesity exposure between ages 10 and 40 years and its relationship with cardiometabolic disease risk factors: A cohort study
BackgroundIndividuals with obesity do not represent a homogeneous group in terms of cardiometabolic risk. Using 3 nationally representative British birth cohorts, we investigated whether the duration of obesity was related to heterogeneity in cardiometabolic risk.Methods and findingsWe used harmonised body mass index (BMI) and cardiometabolic disease risk factor data from 20,746 participants (49.1% male and 97.2% white British) enrolled in 3 British birth cohort studies: the 1946 National Survey of Health and Development (NSHD), the 1958 National Child Development Study (NCDS), and the 1970 British Cohort Study (BCS70). Within each cohort, individual life course BMI trajectories were created between 10 and 40 years of age, and from these, age of obesity onset, duration spent obese (range 0 to 30 years), and cumulative obesity severity were derived. Obesity duration was examined in relation to a number of cardiometabolic disease risk factors collected in mid-adulthood: systolic (SBP) and diastolic blood pressure (DBP), high-density-lipoprotein cholesterol (HDL-C), and glycated haemoglobin (HbA1c).A greater obesity duration was associated with worse values for all cardiometabolic disease risk factors. The strongest association with obesity duration was for HbA1c: HbA1c levels in those with obesity for <5 years were relatively higher by 5% (95% CI: 4, 6), compared with never obese, increasing to 20% (95% CI: 17, 23) higher in those with obesity for 20 to 30 years. When adjustment was made for obesity severity, the association with obesity duration was largely attenuated for SBP, DBP, and HDL-C. For HbA1c, however, the association with obesity duration persisted, independent of obesity severity. Due to pooling of 3 PLOS MEDICINE
# Introduction
Obesity is a global public health concern. Worldwide prevalence of child and adolescent obesity (defined according to a body mass index [BMI] of >2 standard deviations above age-specific World Health Organization cut-offs) has increased from 0.9% and 0.7% in boys and girls, respectively, in 1975 to 7.8% and 5.6%, respectively, in 2016. These increases in child obesity accompany significant increases in global adult obesity, with prevalence increasing from 3% and 6.6% of males and females, respectively, in 1975 to 11.6% and 15.7%, respectively, in 2016 (defined according to a BMI of >30 kg/m 2 ). While this epidemic is associated with many adverse health outcomes, particularly cardiovascular disease-related morbidity and mortality, individuals with obesity do not represent a homogeneous group in terms of cardiometabolic risk. Indeed, there exists a group of individuals who, while exceeding the standard BMI cut-off for obesity (�30 kg/m 2 ), are regarded as metabolically healthy because they have an absence of other major cardiovascular risk factors. The life course traits contributing to this heterogeneity in cardiometabolic risk have received little attention, but it is likely that a large proportion of the heterogeneity is related, in particular, to the length of time a person spends obese. It has been demonstrated that younger individuals are now accumulating greater exposure to overweight or obesity throughout their lives, so a comprehensive understanding of the influence of the duration of obesity on the development of cardiometabolic risk factors is critical.
Abraham and colleaguespublished one of the first studies investigating this heterogeneity in cardiometabolic risk for a given weight, observing that rates of some cardiovascular diseases were highest among individuals who were most overweight in adulthood but below the average weight in childhood. As this study, and others which have replicated that analysis, are based on weight status at just 2 time points however, obesity duration can be estimated only crudely. More frequent longitudinal measurements of weight are required for a fuller picture. Furthermore, a detailed measurement schedule is also required in order to differentiate between obesity duration, the age of obesity onset, and the severity of obesity, as these, although correlated, may confer different health risks. For example, due to the changes in insulin sensitivity that occur during pubertal development, an obesity onset in adolescence may be more deleterious for insulin resistance and diabetes than an onset during another period of the life course.
A handful of studies with such data have observed positive associations between obesity duration and several cardiometabolic disease risk factors including metabolic syndrome, hypertriglyceridemia, dyslipidaemia, and blood pressure. Most evidence relate to the association with type 2 diabetes, however, with numerous studies observing a positive relationship with obesity duration. The largest of these studies (n = 61,821)observed that for each 2-year increment in obesity duration, the risk of type 2 diabetes increased by 14%, although, as observed in other studies, estimates were attenuated upon adjustment for current weight (representing obesity severity). However, these studies have important limitations, including retrospective designs, categorising the outcome variable (thus ignoring the observed distribution), an a priori assumption of a linear relationship between obesity duration and outcomes, and assuming that once a person becomes obese they remain obese, thus removing the possibility for weight cycling. Another important limitation is the adjustment of the obesity duration-outcome relationship for current BMI (i.e., at outcome assessment) in order to separate the effects of obesity duration and severity. BMI at outcome assessment does not capture the true extent of obesity severity as it ignores (potentially greater) severity occurring at earlier time points. For example, consider 2 adults, adult A and adult B, who both have a BMI of 35 kg/m 2 at follow-up and who have both been obese for 20 years. Adult A has had a constant BMI of 35 kg/m 2 , while adult B has had a BMI as high as 45 kg/m 2 during this period. It is unlikely that the cardiometabolic health risks associated with these 2 profiles are homogeneous.
Recently, the concept of "obese-years" has been proposed, which combines the degree and duration of obesity into a single variable. In the study by Araujo and colleagues, an area under the curve (AUC) of BMI (BMI AUC ) was used to summarize duration and severity of BMI. It is possible however to obtain a mean obesity severity over any period by dividing this AUC by obesity duration, thus separating the effects of severity and duration. To our knowledge, no study has done this, and thus, robust evidence of the association between obesity duration and cardiometabolic risk factors, which is truly independent of obesity severity, is lacking.
Using data from 3 British birth cohort studies, the aim of the present study was to model serial measurements of BMI obtained across the life course in order to derive, for each individual, the following obesity traits: duration of obesity exposure between ages 10 and 40 years, age of obesity onset, and obesity severity. These parameters were then used to relate obesity duration, with and without adjustment for obesity severity, to systolic (SBP) and diastolic blood pressure (DBP), high-density lipoprotein cholesterol (HDL-C), and glycated haemoglobin (HbA1c) in mid-adulthood.British birth cohort studies used in these analyses have been previously described in detail elsewhereand were designed to be nationally representative when initiated. The MRC National Survey of Health and Development (NSHD) was initiated in 1946 and recruited 5,362 participants. The National Child Development Study (NCDS) was initiated in 1958 and recruited 17,416 participants.British Cohort Study (BCS70) was initiated in 1970 and recruited 16,571 participants.
# Methods
## Samples
# Ethics statement
All of the studies have received ethical approval and obtained informed parental and/or participant consent, both of which cover the secondary analyses reported here. .
For this analysis, we identified a target sample of 20,746 (NSHD: n = 2,968; NCDS: n = 9,302; BCS: n = 8,476) participants who attended the biomedical sweep where cardiometabolic disease risk factor data were collected (see below) and contributed BMI data for the derivation of our exposure variable: obesity duration (S1 Fig).
## Serial bmi data
As described elsewhere, serial BMI (kg/m 2 ) was derived and harmonised in each study from measured or self-reported weight and height collected at the target ages 11, 15, 20 (selfreport), 26 (self-report), 36, and 43 years in the 1946 NSHD; 11, 16, 23 (self-report), 33, and 42 (self-report) years in the 1958 NCDS; and 10, 16 (one-third self-report), 26 (self-report), 30 (self-report), 34 (self-report), and 42 (self-report) years in the 1970 BCS.
There were 21,009 observations of BMI from 4,702 participants in the 1946 cohort, with 74% of the sample having 4 or more observations. There were 57,545 observations of BMI from 16.274 participants in the 1958 cohort, with 80% of the sample having 3 or more observations. Finally, there were 56,275 observations of BMI from 15,437 participants in the 1970 cohort, with 72% of the sample having 3 or more observations.
## Cardiometabolic disease risk factors in adulthood
In each cohort, a biomedical sweep, with venous blood sampling was conducted in adulthood, at 53 years in the 1946 cohort (n = 3 053), 44 years in the 1958 cohort (n = 9 377), and 46 years in the 1970 cohort (n = 8 581). Measurements of SBP and DBP were obtained as well as blood cardiometabolic biomarkers (HbA1c and HDL-C). More information about the measurement protocols can be found in S1 Text.
# Statistical analysis
TN and WJ determined which analyses to perform and include in the present paper in January 2019 after discussing options with all coauthors. The analysis plan was revised in May (modelling obesity duration as a categorical variable rather than a continuous variable) and October 2019 (removing LDL-cholesterol as an outcome due to high amount of missing data) when further exposure and outcome data were obtained and explored. Further analyses were added in June 2020 in response to reviewer comments (adjusting for further putative confounding variables in the regression models, adding sexXduration interaction models in the supplementary analyses).
Obesity duration parameters. In order to identify obesity and derive obesity parameters throughout the life course, we modelled individual child-adulthood trajectories of BMI from 10 to 40 years of age. These life course BMI trajectories were modelled within each cohort separately, due to the previously described between-cohort heterogeneity in the age-related progression of obesity from childhood to adulthood. Models included all participants who contributed at least 1 measurement of BMI during the studied age range (NSHD: 11 to 43 years; NCDS: 11 to 42 years; BCS: 10 to 42 years). The BMI trajectories were modelled using restricted cubic splines with mixed effects, with measurement occasion at level 1 and individuals at level 2. The restricted cubic splines split the trajectories into piecewise functions of age separated by "knots." Between the adjacent knots, separate cubic polynomials were fitted, with the spline terms constrained to be linear in the 2 tails. The number of knots (using the default knot positions as proposed by Harrellwas chosen based on the Bayesian information criterion (BIC), with a lower BIC indicating a better fitting model. Once the best fitting model was identified, sex was added as a fixed effect and as interaction terms with the age terms identified in the previous step. Finally, an adjustment for level 1 variation was included to allow for differing error associated with measured versus self-reported BMI. From these models, fitted annual BMI values between 10 and 40 years were obtained for each individual.
Using these fitted BMI values, z-scores were created relative to the International Obesity Task Force (IOTF) reference. Obesity was defined as a z-score of >2.288 in males and >2.192 in females, which corresponds to a BMI value of 30 kg/m 2 at 18 years. Using the sexspecific obesity cut-off, several obesity parameters were derived for each individual. Firstly, the presence of obesity at any timepoint was identified, representing any BMI z-score, which exceeded the obesity threshold. Secondly, the "number of times obese" was calculated as the number of times an individual's BMI z-score crossed upwards through the obesity threshold. Thirdly, "age first obese" was derived, representing the age, in years, when BMI z-score first crossed upwards through the obesity threshold. "Total duration of obesity" was calculated as the length of time, in years, that a person's BMI z-score exceeded the obesity threshold; these values were categorised as 0: never obese; 1: obesity 0.01 to <5 years; 2: obesity 5 to <10 years; 3: obesity 10 to <15 years; 4: obesity 15 to <20 years; 5: obesity 20+ years. Finally, we used the composite trapezoid rule to derive a cumulative obesity severity variable, represented inby the AUC and above the obesity threshold. Severity here is expressed in BMI-years above the obesity threshold, reflecting the fact that it incorporates both duration of obesity and the extent to which BMI exceeded the age-specific obesity threshold. If this is then divided by obesity duration, it can be interpreted as the "average obesity severity," i.e., the mean excess BMI above the obesity cut-off.
Linking obesity parameters to cardiometabolic disease risk factors. Prespecified constants were added to the cardiometabolic disease risk factors to adjust for being on medication, which has been found to reduce bias. The constants were +10 mmHg and +5 mmHg for SBP and DBP, respectively, −5% for HDL-C and +1% (absolute) for HbA1c, obtained from meta-analyses of the effect of blood pressure lowering, lipid-regulating, and diabetesmedications on the respective cardiometabolic risk factors.
Multiple linear regression was used to relate obesity parameters to the continuous cardiometabolic risk factors. As uncertainty in estimated obesity parameters are not taken into account in the confidence intervals for their associations with these continuous cardiometabolic risk factors, standard errors may be underestimated. To correct for this, robust standard errors in these subsequent models were estimated. Data were pooled across cohorts and sexes, thus enabling adjustment of the association between obesity duration and cardiometabolic risk factors for cohort and sex. As HDL-C and HbA1c required transformation to achieve normal distributions, for consistency, we transformed all continuous cardiometabolic risk factors to the 100 log e scale, so that the regression coefficients are in units of percentage difference in cardiometabolic risk factor per unit difference in covariate. In a first set of models, the binary variable ever (versus never) obese (between 10 and 40 years) was tested for association with each cardiometabolic risk factor. In a subsequent set of models, we related the categorical obesity duration variable to each cardiometabolic risk factor, with never obese the referent group. The above steps were unadjusted for covariates. A subsequent model included adjustments for sex, cohort, birth weight (kg), ethnicity (white versus non-white), social class in childhood (father's social class reported when the child was 10 to 11 years and according to the Registrar General's Social Classes schema; see S2 Text for more details), and age at follow-up. A final model also included an adjustment for average obesity severity. In order to address missingness in covariate data, we used multiple imputation by chained equations (MICE)combining estimates using Rubin's rules. The number of imputations required to achieve convergence of parameter estimates was determined as 100 × fraction missing information (FMI).
In addition, in order to aid presentation, we repeated the above steps for a number of derived dichotomous cardiometabolic disease risk factor variables, using generalised linear models (Poisson distribution with robust error variances) to estimate relative risks (RRs) for each outcome. The derived cardiometabolic disease risk factor variables were hypertension (SBP > 140mmHg and/or DBP > 90mmHg or reported use of BP-lowering medication), low HDL-cholesterol (<1.03mmol/L in males and <1.29mmol/L in femalesor reported use of lipid-regulating medication) and elevated HbA1c (>5.7%or reported use of diabetes medication).
Beta coefficients from these regression models, i.e., percentage change for continuous variables and RRs for binary variables, were plotted. Each figure is split into 2, with the left-hand side (model 1) showing the estimates from the regression of ever obese (versus never) and the right-hand side (model 2) showing estimates of the categorical obesity duration variable (versus never).
Sensitivity analyses. First, we repeated the analyses excluding the NSHD cohort as the biomedical sweep occurred much later in this cohort compared with NCDS and BCS70 cohorts, which may have resulted in an underestimation of the association between obesity duration and cardiometabolic disease risk factors. In a related sensitivity analyses, we also replaced the NSHD blood pressure variables to those collected at the 43-year sweep in order to align with the timing of blood pressure measurements in the NCDS and BCS70. No other outcome data were available at that age in NSHD however. To identify the extent to which relationships were sex-specific, we also repeated the analyses including a "sex X obesity duration" interaction. We also performed an analysis, which was restricted to those who remained obese, assuming that relationships would strengthen when limited to those with persistent obesity and not cycles of obesity.
Analyses were performed in Stata version 15 (Stata Corp, College Station, Texas) and R version 3.5.3 (R Core Team 2019, Vienna, Austria).
This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist).
## Code availability
The statistical code for the analyses in this paper has been placed in GitHub, the open-access online repository (repository URL: https://github.com/tomnorris1988/Obesity-duration-andcardiometabolic-outcomes).
# Results
Descriptive statistics of the cohorts are shown in. A total of 49.1% of the sample were male, and 97.2% were white British. As shown in, the prevalence of "ever obese" between 10 and 40 years was approximately 3 times higher in the most recent cohort BCS70 (19.7%, n = 1673), compared with the oldest cohort NSHD (6.6%, n = 197). The average age of first obesity onset was less in more recent cohorts, with a median of 30. The average BMI at the biomedical sweep was in the overweight category (>25 kg/m 2 ) in all 3 cohorts but was highest in the BCS70 cohort (27.6 kg/m 2 ; IQR: 24.6, 31.5). For all 5 cardiometabolic risk factors, the mean was highest in the NSHD cohort, reflecting the older age at follow-up. This was most notable for SBP, with a mean of 136.0 mmHg (SD: 20.1) in the NSHD compared with 126.6 mmHg (16.5) and 124.6 mmHg (15.2) in the NCDS and BCS70, respectively. This translated to a much higher prevalence of hypertension in the NSHD cohort (68.1%) compared with the NCDS (27.8%) and BCS70 cohorts (23.9%). The presence of elevated HbA1c was also considerably higher in the NSHD cohort compared with the NCDS and BCS70 (35.8% versus 15.0% and 16.5%, respectively).
## Relationship of obesity parameters with cardiometabolic disease risk factors
Results from the unadjusted analysis are included in S1 and S2 Tables. Here, we report estimates from the adjusted analyses, presented in Figs 2-4 and with corresponding estimates in S3-S6 Tables.
HbA1c. Being ever obese at any age between 10 and 40 years (versus never obese) was associated with a 9.0% (95% CI: 8.2, 9.9) higher HbA1c , which reduced to 4.5% higher (95% CI: 3.5, 5.6) when adjusted for obesity severity. HbA1c increased linearly with obesity duration, from 5% excess for <5 years duration up to 19.9% (95% CI: 16.5, 23.3) for 20 to 30 years duration (p(trend) < 0.001). Upon adjustment for obesity severity, the trend remained (p(trend) = 0.007) but was attenuated, particularly for 20 to 30 years, which reduced from 19.9% to 11.6% (95% CI: 5.9, 17.2), a relative reduction of 42%.
There was also a linear trend between obesity duration and risk for elevated HbA1c, with those obese for <5 years having a 2.1 (95% CI: 1.8, 2.4) times higher risk of elevated HbA1c compared with never obese, which more than doubled in those obese for 20 to 30 years (RR 4.6; 95% CI: 3.9, 5.5, p(trend) < 0.001) . However, upon adjustment for obesity severity, this graded relationship was attenuated (p(trend) = 0.006).
## Sbp and dbp.
There was a positive relationship between ever being obese between 10 and 40 years and both SBP and DBP. For example, ever obese was associated with a 6.1% (95% CI: 5.6, 6.6) higher SBP and 7.1% (95% CI: 6.6, 7.7) higher in DBP at follow-up (versus never obese). Obesity duration was also positively associated with both SBP and DBP, such that SBP was 5.0% higher in those who were obese for <5 years compared with those never obese, increasing to 9.0% higher for 20 to 30 years (p(trend) < 0.001). However, upon adjustment for obesity severity, evidence for this dose-response association was greatly reduced (SBP: p(trend) = 0.975; DBP: p(trend) = 0.294).
## Fig 2. association between ever obese and categories of obesity duration (versus never obese) and hba1c (left panel) and risk for elevated hba1c (right panel).
https://doi.org/10.1371/journal.pmed.1003387.g002 Consistent with these findings, ever being obese between 10 and 40 years (versus never) was associated with an RR for hypertension of 1.6 (95% CI: 1.5, 1.7), independent of obesity severity . For obesity duration, a similar pattern was observed to that seen for SBP and DBP, i.e., a gradually increasing risk for hypertension with increasing time spent obese (p(trend) < 0.001), evidence for which weakened when adjusted for obesity severity (p(trend) = 0.456).
HDL-C. A negative relationship was observed between obesity and HDL-C, such that obesity at any point between 10 and 40 years was associated with a 16.4% (95% CI: 17.6, 15.2) lower HDL-C at follow-up, attenuating to 12.3% lower when adjusted for severity. There was a linear trend in the effect of obesity duration on HDL-C, such that HDL-C levels in those with obesity for <5 years were 12.4% (95% CI: 10.4, 14.4) lower than those never obese, which increased to 24.8% (95% CI: 20.5, 29.1) lower in those who had been obese for 20 to 30 years (p(trend) < 0.001). Upon adjustment for obesity severity, evidence for the trend attenuated (p(trend) = 0.117).
This resulted in an RR for low-HDL-C of 2.0 (95% CI: 1.8, 2.2) in those who were ever obese between 10 and 40 years (versus never), independent of obesity severity . For obesity duration, there was a linear trend of increasing risk (p(trend) < 0.001), which remained on adjustment for severity, although evidence for this was attenuated (p(trend) = 0.037).
Sensitivity analysis. Similar results were found when the analysis was limited to the NCDS and BCS70 cohorts (S7 and S8 Tables), thus accounting for the difference in the age at follow-up in the NSHD. Similarly, replacing the blood pressure variables in the NSHD cohort with those collected at the age 43-year sweep, in order to be more aligned with the age at follow-up in the NCDS and BCS70, did not change results (S9 and S10 Tables). When stratified by sex, associations were consistently stronger in females (S11 and S12 Tables) and especially for the dichotomous cardiometabolic disease risk factor variables. Finally, estimates were largely unchanged when the analysis was limited to those with persistent obesity (i.e., staying obese after first onset) (S13 and S14 Tables).
# Discussion
This study utilised longitudinal BMI data from 3 British birth cohort studies to model each person's obesity history and derive individual obesity parameters. Ever being obese between 10 and 40 years of age, compared with never being obese, was associated with less favourable levels of all cardiometabolic disease risk factors. More time spent obese was associated with worse profiles for all cardiometabolic disease risk factors, although greatest for HbA1c. When adjustment was made for obesity severity, the strength of the evidence in support of an association between obesity duration and SBP, DBP, and HDL-C was weak (p > 0.1). For HbA1c however, although the association with obesity duration also attenuated when adjusting for obesity severity, the strength of evidence remained strong. The study design, in particular the fact that most individuals who became obese remained obese, has meant that age of obesity onset and obesity duration are very highly negatively correlated. Our results also therefore mean that after accounting for obesity severity, an earlier age of obesity onset was only associated with HbA1C. These key findings were robust to a range of sensitivity analyses.
In attempting to separate the effects of obesity duration and severity on cardiometabolic health, previous studies have simply adjusted for BMI (or waist circumference) at the time of outcome assessment. This, however, only provides an indication of obesity severity at that particular point in time. Our study represents an advance over these studies however, as we have been able to measure obesity severity accumulated over the life course, and by adjusting this for the time spent obese, we have been able to appropriately separate the effects of obesity duration and severity. As such, these findings provide novel, robust evidence regarding the independent association of obesity duration with cardiometabolic disease risk factors.
Our findings are in line with other studies, which have observed an attenuated, but persisting, effect of obesity duration on diabetes risk or impaired glucose metabolism, once obesity severity is accounted for. In another NCDS analysis (n = 7855), Power and colleaguesobserved that compared with those never obese, those with the greatest duration of obesity (i.e., onset < 16 years) had an almost 24-fold increased risk of having HbA1c of >7% (and/or a diagnosis of diabetes) at 45 years. While this risk was substantially attenuated upon adjustment for current BMI, it still remained over 4 times higher compared with those never obese. In addition we have observed, in line with Pontirolli and colleagues, a specific effect of obesity duration on glucose metabolism. In their study of 760 obese adults (average age 51 years), obesity duration was a risk factor for glucose intolerance and type 2 diabetes but not for hypertension or hyperlipidaemia. Evidence in support of our finding of no independent association of obesity duration with HDL-C is lacking. To our knowledge only 1 other study has investigated this and observed an association in females only, although the strength of evidence was modest (p = 0.05).
In addition to the cited empirical studies, there is also a plausible biological mechanism supporting the observed association between obesity duration and HbA1c (reflecting impaired glucose metabolism). Obesity is characterised by enlarged fat stores, which results in enhanced lipolysis and an increase in circulating free fatty acids. This state leads to peripheral and hepatic insulin resistance, resulting in a compensatory insulin hypersecretion by the pancreatic β-cells in order to preserve normoglycemia. Prolonged obesity leads to β-cell exhaustion, culminating in a reduced insulin response and an inability to maintain normoglycemia. In addition, prolonged obesity may represent a state in which subcutaneous adipose stores have been exhausted, with the consequence being a deposition of adipose tissue around the visceral organs (e.g., liver and pancreas), with fat stored in these areas (i.e., "ectopic fat") being strongly related to insulin resistance.
Despite the persisting independent effect of obesity duration on HbA1c levels, a substantial reduction in the effect was observed once the severity of obesity had been accounted for. This suggests that in those who have been exposed to obesity for a prolonged period, there is still opportunity to return to more favourable HbA1c levels if a degree of weight loss is achieved. For example, upon adjustment for severity, the risk of elevated HbA1c in those who had been obese for 20 to 30 years reduced from more than a 4-fold increased risk (relative to never obese), to a level similar to those obese for half as long, i.e., 10 to 15 years (RR: 3.0; 95%CI: 2.3, 4.0).
There was some evidence that the association between obesity duration and the dichotomous cardiometabolic outcomes was stronger in females than in males (S11 and S12 Tables). Sex-specific associations have been observed in other studies. Janssen and colleagues, for example, observed an independent effect of overweight/obesity duration on risk for insulin resistance and type 2 diabetes (and also hypertension, hypertriglyceridemia, low-HDL-C, and metabolic syndrome) in females, but not in males except for hypertriglyceridemia. A sex difference was also observed in the Framingham Heart Study. Tanamas and colleagues observed an association between obesity duration and risk for hypertension in females but not in males (ages 30 to 62 years). As summarised in the review by Jarvis, there are fundamental differences in the control of metabolic homeostasis between males and females. Females are more likely to gain fat, and although abdominal obesity more commonly affects males than females, the prevalence of abdominal obesity has increased more in females than in males . Furthermore, the prevalence of visceral obesity associated with metabolic syndrome is 2 to 10 times higher in women throughout the world. It may be therefore that compared with males, females are more exposed to this metabolically volatile adipose tissue and thus at increased risk of its deleterious outcomes.
# Strengths
The key strength of our study is the derivation, using over 130,000 serial BMI observations across the life course, of individualised obesity parameters, which enabled us to distinguish between obesity severity and duration. In addition, the pooling of data from 3 nationally representative cohorts means the observed associations are based on a far larger sample than most previous studies and are likely to be generalisable to the underlying population.
# Limitations
Our definition of obesity was based on BMI, which despite exhibiting a strong positive correlation with direct estimates of fat mass, is only an indicator of total body adiposity. However, it remains the most commonly used, widely accepted and practical measure of obesity in both children and adults. Our trajectories were dependent on the frequency of BMI measurements across the life course, with some intervals spanning 10 years. As such, we may not have captured instances of weight cycling between measurement occasions. Measurement protocols for weight and height were not consistent within and between studies, which may have introduced bias if self-reported measurements were systemically under or overreported. It has been shown that people with greater BMIs tend to underreport their weight, suggesting that estimates of obesity duration (and severity) may be conservative in our study. Our regression models included adjustment for only a small number of covariates, which means there is a possibility of residual confounding. As we have combined 3 cohorts, any included variable must be harmonised across each cohort so that the variable conveys the same thing in each cohort. This is only the case for a small number of variables in the cohorts we have used. As all of the included studies suffered from attrition, which is more extensive in those from lower socioeconomic position (SEP) groups and/or with higher BMI, we may have inadvertently selected a more socioeconomically advantaged and thinner sample which, in addition to a loss of power, may have introduced bias into the observed associations. In addition, as the NSHD, NCDS, and BCS70 cohorts are either exclusively (NSHD), or predominantly white British, we are unable to generalise the results to other ethnic groups. Finally, the biomedical sweep in the NSHD cohort was conducted 9 and 7 years later than the NCDS and BCS cohorts, respectively, which may impair cross-cohort comparability (underpinning the decision to pool cohorts). However, supplementary analyses limited to the NCDS and BCS cohorts only and replacing the NSHD blood pressure variables with those collected at 43 years produced similar estimates (S7-S10 Tables).
Associations observed in this study suggest that there are benefits in delaying the onset of obesity, as risks of elevated HbA1c were positively associated with time spent with obesity, independent of the degree of severity. Interventions aiming to prevent childhood obesity therefore have the potential to reduce the long-term risk of developing diabetes. However, we also observed an amelioration of HbA1c profiles in those who had been exposed to obesity for a prolonged period, once severity of obesity is accounted for. As such, people with obesity should be encouraged to lose weight in order to return their HbA1c levels to more favourable values. Firstly, however, more research using different epidemiological approaches underpinned by different assumptions and sources of bias (e.g., mendelian randomization and negative control studies) are needed to test the robustness of these findings.
# Conclusion
We found a dose-response relationship between the duration of obesity and HbA1c, independent of obesity severity. Given that the obesity epidemic is characterised by trends towards earlier onset and consequently greater lifetime exposure, our findings are important as they suggest that health policy recommendations aimed at preventing early onset obesity, and therefore reducing lifetime obesity exposure, may help reduce the risk for diabetes. For those who are already obese, reducing obesity severity can also improve their metabolic profile. Accordingly, prevention strategies could consider both the duration and severity of obesity. Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors � † (imputed, unadjusted). (DOCX) S2 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, unadjusted). Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors � † (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, and childhood social class). (DOCX) S4 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, and childhood social class). (DOCX) S5 Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors � † (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity). (DOCX) S6 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity). (DOCX) S7 Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): excluding NSHD.
## Supporting information
[formula] (DOCX) S3 [/formula]
(DOCX) S8 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): excluding NSHD. (DOCX) S9 Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): using blood pressure at 43 years in NSHD. (DOCX) S10 Association between ever obese and categories of obesity duration (versus never obese) and categorical cardiometabolic outcomes (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): using blood pressure at 43 years in NSHD. (DOCX) S11 Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors (imputed, adjusted for cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): sex interaction. (DOCX) S12 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, adjusted for cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): sex interaction. (DOCX) S13 Association between ever obese and categories of obesity duration (versus never obese) and cardiometabolic disease risk factors (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): limited to those who once obese were always obese. (DOCX) S14 Association between ever obese and categories of obesity duration (versus never obese) and dichotomous cardiometabolic outcomes (imputed, adjusted for sex, cohort, age at follow-up, ethnicity, birth weight, childhood social class, and obesity severity): limited to those who once obese were always obese.
|
Effects of intrapulmonary viral tropism and cytokine expression on the histological patterns of cytomegalovirus pneumonia
Pulmonary cytomegalovirus (CMV) infection causes fatal CMV pneumonia (CMVp) in immunocompromised patients; however, the mechanisms underlying CMV-infectioninduced pulmonary lesion development remain largely unknown. We examined the relationship between CMVp patterns and intrapulmonary viral tropism, including expression of inflammatory cytokines and related molecules.Double immunohistochemistry of CMV antigen and cellular markers showed that epithelial tropism was associated with a diffuse alveolar damage (DAD) pattern (CMVp-DAD) while stromal tropism was associated with a predominantly interstitial inflammation/fibrosis (IIF) (CMVp-IIF) or a combination of DAD and IIF (CMVp-complex). Transforming growth factor (TGF)-b1 expression was relevant to CMV-induced tissue injury, and its expression was higher in CMVpcomplex and CMVp-IIF than in CMVp-DAD.Expression of integrin b6 (ITGB6), an adhesion molecule and important activator of TGF-b1 in interstitial pneumonia, was lost in CMV-infected pneumocytes, especially CMVp-DAD, whereas CMV-negative pneumocytes in CMVp-complex and CMVp-IIF showed overexpression. Diffuse interleukin (IL)-8 up-regulation and strong expression were present in both CMV-infected pneumocytes and stromal cells only in CMVp-IIF cases with marked interstitial neutrophilic infiltration. On the basis of viral tropism and the expression of TGF-b1, ITGB6, and IL-8, we conclude that CMV-infected pulmonary cells play an important role in the development of diverse CMVp patterns.Key words: chromogenic in situ hybridization, cytomegalovirus pneumonia, double immunostain, integrin b6, interleukin-8, transforming growth factor-b1Cytomegalovirus (CMV) is a major pathogenic microbe in immunocompromised individuals, and CMV pneumonia (CMVp) is a critical complication because of the high fatality. 1,2 The clinical findings of CMVp have been well documented; 3,4 however, how CMV infection causes pulmonary lesions is not yet well understood.Cytomegalovirus pneumonia, a secondary interstitial pneumonia (IP), exhibits various histopathological characteristics, i.e., focal or diffuse interstitial lesions with/without hemorrhage, hyaline membrane, and necrobiosis. 5-7 It has been reported that CMV infects a wide variety of cell types, including pneumocytes, fibroblasts, macrophages, and endothelial cells in the lung. 8 Although cytomegalic cells are a wellknown hallmark of CMV-infected cells, only a few reports have immunohistochemically confirmed these cell types. 8-10 Moreover, to our knowledge, the proportions of CMV-infected cell types among the diverse patterns of CMVp have not been described in previous reports.Many growth factors and cytokines have been identified as pathogenetic factors for the initiation or progression of both idiopathic and secondary IP. 11,12 Transforming growth factor (TGF)-b1 has been implicated as a pivotal molecule in the development of acute and chronic IP 13,14 and CMVp. 15 CMV infection was reported to induce production of TGF-b1,16and this protein also enhanced viral replication in some CMV-infected cells. 17 Inflammatory cytokines, such as tumor necrosis factor-a, interleukin (IL)-6, and IL-8, were highly expressed in IP, 18-20 and were up-regulated by CMV infection.21,22Furthermore, high expression of IL-8 and its receptor in CMV-infected human lung fibroblasts enhanced its
## Pulmonary cytomegalovirus (cmv) infection causes fatal cmv pneumonia (cmvp) in immunocompromised patients; however, the mechanisms underlying cmv-infectioninduced pulmonary lesion development remain largely unknown. we examined the relationship between cmvp patterns and intrapulmonary viral tropism, including expression of inflammatory cytokines and related molecules. double immunohistochemistry of cmv antigen and cellular markers showed that epithelial tropism was associated with a diffuse alveolar damage (dad) pattern (cmvp-dad) while stromal tropism was associated with a predominantly interstitial inflammation/fibrosis (iif) (cmvp-iif) or a combination of dad and iif (cmvp-complex). transforming growth factor (tgf)-b1 expression was relevant to cmv-induced tissue injury, and its expression was higher in cmvpcomplex and cmvp-iif than in cmvp-dad. expression of integrin b6 (itgb6), an adhesion molecule and important activator of tgf-b1 in interstitial pneumonia, was lost in cmv-infected pneumocytes, especially cmvp-dad, whereas cmv-negative pneumocytes in cmvp-complex and cmvp-iif showed overexpression. diffuse interleukin (il)-8 up-regulation and strong expression were present in both cmv-infected pneumocytes and stromal cells only in cmvp-iif cases with marked interstitial neutrophilic infiltration. on the basis of viral tropism and the expression of tgf-b1, itgb6, and il-8, we conclude that cmv-infected pulmonary cells play an important role in the development of diverse cmvp patterns.
Key words: chromogenic in situ hybridization, cytomegalovirus pneumonia, double immunostain, integrin b6, interleukin-8, transforming growth factor-b1
Cytomegalovirus (CMV) is a major pathogenic microbe in immunocompromised individuals, and CMV pneumonia (CMVp) is a critical complication because of the high fatality. [bib_ref] Pulmonary involvement during cytomegalovirus infection in immunosuppressed patients, De Maar [/bib_ref] [bib_ref] Pneumonitis in human cytomegalovirus infection, Langhoff [/bib_ref] The clinical findings of CMVp have been well documented; 3,4 however, how CMV infection causes pulmonary lesions is not yet well understood.
Cytomegalovirus pneumonia, a secondary interstitial pneumonia (IP), exhibits various histopathological characteristics, i.e., focal or diffuse interstitial lesions with/without hemorrhage, hyaline membrane, and necrobiosis. [bib_ref] Cytomegalovirus pneumonia in bone marrow transplant recipients: Miliary and diffuse patterns, Beschorner [/bib_ref] [bib_ref] Pulmonary cytomegalovirus infection in the adult, Craighead [/bib_ref] [bib_ref] Cytomegaloviral infection presenting as a solitary pulmonary nodule, Ravin [/bib_ref] It has been reported that CMV infects a wide variety of cell types, including pneumocytes, fibroblasts, macrophages, and endothelial cells in the lung. [bib_ref] Fibroblasts, epithelial cells, endothelialcells and smooth muscle cells are major targets of..., Sinzger [/bib_ref] Although cytomegalic cells are a wellknown hallmark of CMV-infected cells, only a few reports have immunohistochemically confirmed these cell types. [bib_ref] Fibroblasts, epithelial cells, endothelialcells and smooth muscle cells are major targets of..., Sinzger [/bib_ref] [bib_ref] Cytomegalovirus infection is associated with absence of alveolar epithelial cell HLA class..., Ng-Bautista [/bib_ref] [bib_ref] In situ characterization of human cytomegalovirus infection of bronchiolar cells in human..., Morbini [/bib_ref] Moreover, to our knowledge, the proportions of CMV-infected cell types among the diverse patterns of CMVp have not been described in previous reports.
Many growth factors and cytokines have been identified as pathogenetic factors for the initiation or progression of both idiopathic and secondary IP. [bib_ref] Chemokine/cytokine cocktail in idiopathic pulmonary fibrosis, Agostini [/bib_ref] [bib_ref] Molecular targets in pulmonary fibrosis: The myofibroblast in focus, Scotton [/bib_ref] Transforming growth factor (TGF)-b1 has been implicated as a pivotal molecule in the development of acute and chronic IP [bib_ref] The role of transforming growth factor b in lung development and disease, Bartram [/bib_ref] [bib_ref] Transforming growth factor b: A central modulator of pulmonary and airway inflammation..., Sheppard [/bib_ref] and CMVp. [bib_ref] Balance between alveolar macrophage IL-6 and TGF-b in lung-transplant recipients, Magnan [/bib_ref] CMV infection was reported to induce production of TGF-b1, [bib_ref] Human cytomegalovirus infection induces transcription and secretion of transforming growth factor b1, Michelson [/bib_ref] and this protein also enhanced viral replication in some CMV-infected cells. [bib_ref] Antagonistic modulation of human cytomegalovirus replication by transforming growth factor b and..., Alcami [/bib_ref] Inflammatory cytokines, such as tumor necrosis factor-a, interleukin (IL)-6, and IL-8, were highly expressed in IP, [bib_ref] Enhanced IL-1b and tumor necrosis factor-a release and messenger RNA expression in..., Zhang [/bib_ref] [bib_ref] Inverse effects of interleukin-6 on apoptosis of fibroblasts from pulmonary fibrosis and..., Moodley [/bib_ref] [bib_ref] Serum level of interleukin 8 is elevated in idiopathic pulmonary fibrosis and..., Ziegenhagen [/bib_ref] and were up-regulated by CMV infection. [bib_ref] Monitoring of alveolar macrophage production of tumor necrosis factor-alpha and interleukin-6 in..., Magnan [/bib_ref] [bib_ref] Profiling of inflammatory cytokines produced by gingival fibroblasts after human cytomegalovirus infection, Botero [/bib_ref] Furthermore, high expression of IL-8 and its receptor in CMV-infected human lung fibroblasts enhanced its function in an autocrine manner and promoted CMV replication in vitro. [bib_ref] Enhancement of human cytomegalovirus replication in a human lung fibroblast cell line..., Murayama [/bib_ref] Therefore, CMV-infected cells may play an important role in the formation of IP lesions in CMVp.
However, these findings may not be sufficient to account for the diverse histology of CMVp. Very few reports have analyzed CMVp pathogenesis from a histopathological point of view. [bib_ref] Fibroblasts, epithelial cells, endothelialcells and smooth muscle cells are major targets of..., Sinzger [/bib_ref] [bib_ref] Induction of an epithelial integrin avb6 in human cytomegalovirus-infected endothelial cells leads..., Tabata [/bib_ref] In this study, we focus on the types of CMVinfected cells and the expression of inflammatory cytokines and related molecules in the various histological types of CMVp.
# Materials and methods
## Case selection
From the autopsy files of the Department of Pathology, Hamamatsu University School of Medicine, 12 cases of severe CMVp, with an average of at least 100 CMV antigenpositive cells in each immunohistochemistry (IHC) section, were selected [fig_ref] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/fig_ref] to examine the characteristic CMVinduced changes. Cases of mild CMVp with fewer CMVpositive cells, as well as cases concomitant with other opportunistic infections such as pneumocystis, fungi, severe bronchopneumonia, or intrapulmonary neoplasms were excluded. Two cases of idiopathic pulmonary fibrosis (IPF) (case nos. 9 and 12) were included; however, the remarkably CMV-infected lesions were carefully compared with the non-CMV-infected areas.
All 12 patients had underlying primary diseases causing immunosuppression during medical treatment. Systemic CMV infection in multiple organs was observed in all cases except case no. 10. Case no. 12 had regional lung necropsy, although CMV infection in the visceral organs was not determined. Case nos. 1 and 2 had systemic lupus erythematosus with no apparent pulmonary involvement. The pulmonary lesion in case no. 7 was clinically diagnosed as non-specific IP, but was diagnosed as CMVp at autopsy. After deterioration or the appearance of respiratory disease, all patients died of respiratory failure or hypoxia-induced cardiovascular collapse within 1 to 8 weeks, except case nos. 4, 5, 7, and 12 (the durations to death were not reported in the autopsy records) [fig_ref] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/fig_ref].
## Specimens
All autopsied lung materials were fixed in 10% formalin for more than 7 days. At least one tissue block was prepared from the gross lesions in each lobe, and the tissue blocks were embedded in paraffin. In every case, 1 or 2 paraffinembedded blocks containing the most CMV-infected cells, as shown by CMV IHC, were selected. Additional blocks without CMV-infected cells were prepared in case nos. 9 and 12. A set of 4 mm thick sections was cut for histopathological analyses, including hematoxylin and eosin (HE) and aniline blue staining. §Post-operation. ¶Post-transplantation. † †CMV-induced lesions with a sign of inequality or equality, according to those predominances, and primary or complicated lesions, shown in parentheses.
‡ ‡The period after deterioration or the appearance of respiratory symptoms to death. Adr, adrenal gland; BP, bronchopneumonia; CRF, chronic renal failure; DAD, diffuse alveolar damage; Gbl, gallbladder; Hrt, heart; IIF, interstitial inflammation/fibrosis; Kdn, kidney; LGI, lower gastrointestinal tract, including small intestine and colon; Liv, liver; MI, myocardial infarction; N.D., not determined; NSIP, non-specific interstitial pneumonia; OP, organizing pneumonia; Ovr, ovary; PE, pulmonary edema; PH, pulmonary hemorrhage; Pnc, pancreas; Prs, prostate; Skn, skin; Spl, spleen; Thy, thyroid gland; Ubl, urinary bladder; UGI, upper gastrointestinal tract, including esophagus, stomach and duodenum; UIP, usual interstitial pneumonia; Utr, uterus; wk, weeks; yr, years.
## Primary antibodies
Primary antibodies to cytokeratin (CK) (clone AE1/AE3; Dako, Glostrup, Denmark), CK7 (clone OV-TL 12/30; Dako), surfactant apoprotein A (SP-A) (clone PE10; Dako), vimentin (clone V9; Dako), a-smooth muscle actin (SMA) (clone 1A4; Dako), CD45 (LCA) (clone 2B11 + PD7/26; Dako), CD68 (clone PG-M1; Dako), CMV (clone CCH2 + DDG9; Dako), TGF-b1 (clone TB21; Chemicon, Temecula, CA, USA) and integrin b6 (ITGB6) (clone 442.5C4; Calbiochem, Darmstadt, Germany) were used in this study.
## Ihc
Double IHC was performed by combining antibodies for CMV with those for cell type markers or cytokines, using peroxidase (POD)-conjugated universal immuno-enzyme polymer (UIP), anti-mouse solution (Nichirei Biosciences, Tokyo, Japan) and alkaline-phosphatase (ALP)-conjugated UIP, anti-mouse solution. Dewaxed sections were initially incubated in 3% hydrogen peroxide solution at room temperature for 20 min. Antigen retrieval was done according to the manufacturer's instructions. Each antibody reaction was done at 37°C for 1 h. The first round of immunostaining was colored bluish purple with fast blue BB salt (Sigma, St. Louis, MO, USA). [bib_ref] Murine cytomegalovirus induces apoptosis in non-infected cells of the developing mouse brain..., Kosugi [/bib_ref] To block cross-reactivity in sequential rounds of immunostaining, the microwave oven heating (MW) method was performed. [bib_ref] A novel, simple, reliable, and sensitive method for multiple immunoenzyme staining: Use..., Lan [/bib_ref] The second round of immunostaining was colored red with 3-amino-9-ethyl carbazole (AEC) + substrate-chromogen (DakoCytomation, Carpinteria, CA, USA). The specificity and sensitivity of each marker was verified by single IHC prior to double IHC using POD-conjugated UIP and colored brown with a liquid 3,3′-diaminobenzidine tetrahydrochloride (DAB) substrate chromogen system (DakoCytomation), then counter-stained with hematoxylin.
## Chromogenic in situ hybridization (cish) for whole cmv genome (cish-cmv) combined with immunostaining of itgb6
A DNA probe for CISH-CMV, which was derived from a bacterial artificial chromosome (BAC) and encoded 230 kb of the whole genome of human CMV Towne strain (a gift from Dr Fenyong Liu, University of California, Berkeley, USA), was labeled with digoxigenin (DIG)-11-dUTP (Roche Diagnostics, Penzberg, Germany) using a nick translation kit (Roche Diagnostics). The hybridization and washing procedures have been described previously. [bib_ref] Initial intermittent microwave irradiation for fluorescence in situ hybridization analysis in paraffin-embedded..., Kitayama [/bib_ref] Sections were subsequently incubated with POD-conjugated anti-DIG Fab fragments (1:100, Roche Diagnostics) and colored red with AEC+, followed by hematoxylin counter staining.
Combined ITGB6 IHC and CISH-CMV were also performed. The former was preceded with POD-conjugated UIP, and then colored brown with DAB substrate. After MW, CISH was done and colored red with AEC+, and then counterstained with hematoxylin.
## Semi-quantitative reverse transcriptase-polymerase chain reaction (rt-pcr) for tgf-b1
Ten slices of 10 mm thick paraffin-embedded tissue sections, which were cut from the same samples used for histology and IHC, were placed into a 1.5 mL tube. Total RNA was extracted using ISOGEN (Nippon gene, Tokyo, Japan), according to the manufacturer's protocol. Following treatment with RNase-free DNase I (Roche Diagnostics) for 15 min at 37°C, RNA was reverse transcribed using the SuperScript II First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA). Isolated reverse transcribed product (0.5 mg) was used as the RT-PCR template. The primer pairs for RT-PCR were as follows: 5′-AAGATATCGAATTCTC CGAGAAGCGGTAC-3′ and 5′-CGCGGATCCTCCGGTGAC ATCAAAAGATA-3′ for TGF-b1; 5′-GAAGGTGAAGGTCG GAGTC-3′ and 5′-GAAGATGGTGATGGGATTTC-3′ for glyceraldehyde-3-phosphate dehydrogenase (GAPDH); and 5′-ATGAAGTGTATTGGGCTAACTATGC-3′ and 5′-TTCTC CTAAGTTCATCCTTTTTAGC-3′ for CMV. An initial denaturation step at 95°C for 10 min was followed by 40 cycles of denaturation at 95°C for 30 s, and annealing at 56°C, 58°C, and 60°C for 30 s for TGF-b1, GAPDH, and CMV, respectively. The final elongation step was at 72°C for 10 min. Amplified aliquots were separated on a 2% agarose gel and visualized by ethidium bromide staining.
## In situ hybridization (ish) of il-8 messenger rna (mrna) and ihc for cmv
To detect in situ expression of IL-8 mRNA, a fragment of IL-8 complementary DNA, corresponding to nucleotides 1134-1245 (GenBank Accession No. NM_000584), was cloned into the pGEM-T vector (Promega, Madison, WI, USA). Antisense and sense IL-8 riboprobes were prepared with a DIG RNA labeling kit (Roche Diagnostics) using pGEM-T/ IL-81134-1245 as the template. Hybridization and washing procedures have been previously described. [bib_ref] Down-regulation of human X-box binding protein 1 (hXBP-1) expression correlates with tumor..., Takahashi [/bib_ref] For signal detection, POD-conjugated anti-DIG Fab fragments were applied, and then colored red with AEC+. The specificity of ISH was demonstrated by parallel hybridization of the sections with sense riboprobes. For double staining by IL-8 ISH and CMV IHC, completion of the former was followed by MW. CMV IHC was performed using ALP-conjugated UIP and colored bluish purple with fast blue BB salt.
## Cell counting
The number of cells positive for CMV as well as the number of cells double positive for CMV and cellular markers were counted in more than three non-overlapping 5 ¥ 5 mm 2 fields. The ratio of double positive cells to CMV-positive cells was calculated. Since endothelial cell markers, such as CD31 and factor VIII, tend to lose their immunoreactivity in CMVinfected endothelial cells, [bib_ref] Human cytomegalovirus infection of endothelial cells triggers platelet adhesion and aggregation, Rahbar [/bib_ref] [bib_ref] Circulating cytomegalovirus (CMV)-infected endothelial cells in marrow transplant patients with CMV disease..., Salzberger [/bib_ref] CMV-infected endothelial cells were morphologically identified in double vimentin-and CMVstained sections and were counted. The average number of CMV-positive cells and the frequency of CMV infection, calculated by the average number of CMV-positive cells per counted area (mm 2 ), were also examined.
## Histomorphological evaluation of interstitial fibrosis
Aniline blue stains collagen-deposition areas dark blue compared to the light blue-stained reticulin fibers or basement membrane. [bib_ref] Selective demonstration of elastin, reticulum and collagen by silver, orcein and aniline..., Humason [/bib_ref] Images of several non-overlapping regions, excluding those with large vessels or bronchi, were taken with a digital camera system (DP70; Olympus, Tokyo, Japan). Each dark blue-stained area was extracted as a two-tone image and the ratio of the stained area to whole image was determined using Adobe Photoshop 7.0 (Adobe Systems Inc., San Jose, CA, USA). The mean aniline blue-positive ratio in each case was used as an indicator of interstitial fibrosis.
# Results
## Histomorphological characteristics of severe cmvp
Pulmonary lesions are shown in [fig_ref] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/fig_ref]. CMV-induced pulmonary lesions exhibited diffuse alveolar damage (DAD) and/or interstitial inflammation and fibrosis (IIF). In DAD lesions, hyaline membranes, detached swollen pneumocytes, intra-alveolar exudation, and alveolar wall edema were commonly observed. Cytomegalic cells were frequently seen on the alveolar surface or in the alveolar spaces [fig_ref] Figure 1 11: and 12 [/fig_ref]. In contrast, CMV-induced IIF lesions presented various degrees of interstitial inflammatory infiltrates and fibrous thickening as well as reactive alveolar epithelial proliferation. Cytomegalic cells were observed not only on the alveolar surface, but also in the stroma [fig_ref] Figure 1 11: and 12 [/fig_ref].
Three cases (case nos. 1-3) had DAD with negligible IIF (CMVp-DAD) [fig_ref] Figure 1 11: and 12 [/fig_ref] and six cases (case nos. 4-9) had DAD and IIF with variable severity in a complex pattern (CMVp-complex) [fig_ref] Figure 1 11: and 12 [/fig_ref]. In the remaining three cases (case nos. 10-12), IIF was predominant (CMVp-IIF) [fig_ref] Figure 1 11: and 12 [/fig_ref] , although minor DAD patterns were also seen in case nos. of intra-alveolar fibrosis, regarded as an organizing stage of DAD, were also observed.
## Cmv-infected cell types and cmvp patterns
Numerous CMV antigen-positive cells, many more than the number of cytomegalic cells, were found by IHC. These CMV-positive cells were doubly labeled with epithelial, mesenchymal, or leukocyte markers by IHC without crossreactivity [fig_ref] Figure 2: Double immunohistochemistry [/fig_ref] , except for endothelial cell markers CD31 and factor VIII, which only stained the CMV-negative endothelial cells well (data not shown).
The results of intrapulmonary CMV tropism by double IHC are shown in [fig_ref] Figure 3: Cytomegalovirus [/fig_ref]. Although the proportion of CMVinfected cells was generally very small in specific stromal cell types, such as smooth muscle cells, myofibroblasts, leukocytes, macrophages, and endothelial cells, the proportion in vimentin-positive whole stromal cells, including fibroblasts, and cytokeratin-positive epithelial cells, including pneu-mocytes, was larger and varied among the cases. The major histological patterns of CMVp depended on the proportion of CMV-infected pneumocytes and total stromal cells. In cases with CMVp-DAD, CMV infection was more frequent in pneumocytes than in stromal cells, whereas in CMVp-complex or CMVp-IIF cases, CMV infection was more prevalent in stromal cells, except in case no. 5. Collagen deposition, quantified by aniline blue staining, was less than 3% in CMVp-DAD cases and as high as 14% in CMVp-IIF cases [fig_ref] Figure 3: Cytomegalovirus [/fig_ref] ; however, the percentage of staining was not strictly proportional to the duration of respiratory symptoms [fig_ref] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/fig_ref] or the ratio of CMV-infected stromal cells [fig_ref] Figure 3: Cytomegalovirus [/fig_ref].
## Expression of tgf-b1 and itgb6 and presence of the cmv genome associated with cmv propagation
Transforming growth factor-b1 immunoreactivity was observed in various cell types, such as fibroblasts, endothelial [fig_ref] Figure 1 11: and 12 [/fig_ref] , CMVp-complex (case nos. [bib_ref] Diagnosis and treatment approaches to CMV infections in adult patients, De La Hoz [/bib_ref] [bib_ref] Cytomegalovirus pneumonia in bone marrow transplant recipients: Miliary and diffuse patterns, Beschorner [/bib_ref] [bib_ref] Pulmonary cytomegalovirus infection in the adult, Craighead [/bib_ref] [bib_ref] Cytomegaloviral infection presenting as a solitary pulmonary nodule, Ravin [/bib_ref] [bib_ref] Fibroblasts, epithelial cells, endothelialcells and smooth muscle cells are major targets of..., Sinzger [/bib_ref] [bib_ref] Cytomegalovirus infection is associated with absence of alveolar epithelial cell HLA class..., Ng-Bautista [/bib_ref] , and CMVp-interstitial inflammation/fibrosis (IIF) (case nos. [bib_ref] In situ characterization of human cytomegalovirus infection of bronchiolar cells in human..., Morbini [/bib_ref] [bib_ref] Chemokine/cytokine cocktail in idiopathic pulmonary fibrosis, Agostini [/bib_ref] [bib_ref] Molecular targets in pulmonary fibrosis: The myofibroblast in focus, Scotton [/bib_ref] are indicated by pink, yellow, and green, respectively. cells, reactive pneumocytes, and intra-alveolar mononuclear cells [fig_ref] Figure 4: Transforming growth factor [/fig_ref]. Not all CMV-infected cells were positive for TGF-b1. The percentage of TGF-b1 expressing, CMVinfected cells ranged from 5 to 40% [fig_ref] Figure 3: Cytomegalovirus [/fig_ref] , and had no apparent relationship to CMVp pattern. The TGF-b1 staining intensity and distribution also varied among the cases [fig_ref] Figure 4: Transforming growth factor [/fig_ref]. In CMVp-DAD cases, weak to moderate TGF-b1 staining was restricted to swollen pneumocytes and intraalveolar mononuclear cells [fig_ref] Figure 4: Transforming growth factor [/fig_ref]. In CMVp-complex and CMVp-IIF cases, TGF-b1 immunoreactivity increased in intensity and was found in both pneumocytes and stromal cells, and stromal cells exhibited heterogeneous and collective staining patterns [fig_ref] Figure 4: Transforming growth factor [/fig_ref]. Strong TGF-b1 expression was noted in areas of active tissue injury accompanied by CMV infection and inflammatory infiltration.
The expression of TGF-b1 mRNA, determined by semiquantitative RT-PCR, was low in representative cases of CMVp-DAD, while expression was relatively high in many CMVp-complex and CMVp-IIF cases [fig_ref] Figure 4: Transforming growth factor [/fig_ref] , which was similar to the TGF-b1 protein expression intensities shown by IHC [fig_ref] Figure 4: Transforming growth factor [/fig_ref]. Expression in CMVp-complex and CMVp-IIF cases was correlated with the CMV infection frequency, except in case no. 6.
Immunohistochemical staining for TGF-b1 and for both ITGB6 and CMV in the serial sections from CMVp-complex or CMVp-IIF cases showed CMV antigens in regions with TGF-b1 expression [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref] , and overexpression of ITGB6 in CMV-negative pneumocytes, but not in CMVpositive pneumocytes [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref]. In cases of CMVp-DAD, ITGB6 expression was rare, while co-localization of TGF-b1 and CMV was observed in pneumocytes (data not shown).
With CISH-CMV, positive signals were seen, not only in the cytomegalic cells, but also in various types of infected cells with indistinct cytomegaly, since the sensitivity of CISH-CMV is much higher than that of conventional CMV ISH. A large number of CISH-positive cells were often associated with severe DAD and/or IIF lesions. Overexpression of ITGB6 was observed in pneumocytes and accompanied by remarkable proliferative inflammation in the stroma where the number of CISH-positive cells was increased [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref]. However, as demonstrated by double staining, CISH-CMV-positive pneumocytes significantly lost ITGB6 expression [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref].
## Il-8 expression and iif severity
Expression of IL-8 mRNA by ISH was barely detected in CMVp-DAD cases [fig_ref] Figure 6: Double staining of interleukin [/fig_ref] , while focal expression was observed in several cases with CMVp-complex [fig_ref] Figure 6: Double staining of interleukin [/fig_ref]. Interestingly, all CMVp-IIF cases showed diffuse IL-8 up-regulation, and case no. 12 exhibited strong IL-8 expression [fig_ref] Figure 6: Double staining of interleukin [/fig_ref]. Expression of IL-8 was observed both in pneumocytes and stromal cells, including fibroblastic cells and endothelial cells. Double staining for IL-8 ISH and CMV IHC suggested that some CMV-infected pneumocytes and stromal cells induced IL-8 expression [fig_ref] Figure 6: Double staining of interleukin [/fig_ref] , arrows), particularly in CMVp-IIF, which exhibited remarkable interstitial neutrophilic infiltration.
# Discussion
Here we demonstrated a possible relationship between major CMVp histology and pulmonary viral tropism, namely, CMVp-DAD with epithelial tropism and CMVp-complex or CMVp-IIF with stromal tropism. The desquamative reactions in CMVp-DAD might be caused by CMV-induced alveolar epithelial injury. Previous reports suggested that CMV-infected pneumocytes tended to lose expression of their functional molecules. [bib_ref] Fibroblasts, epithelial cells, endothelialcells and smooth muscle cells are major targets of..., Sinzger [/bib_ref] [bib_ref] Cytomegalovirus infection is associated with absence of alveolar epithelial cell HLA class..., Ng-Bautista [/bib_ref] [bib_ref] In situ characterization of human cytomegalovirus infection of bronchiolar cells in human..., Morbini [/bib_ref] Compared with the rates of CK-(AE1/AE3) or CK7-positive CMVinfected cells, the rates of SP-A-positive CMV-infected cells were lower in almost all cases [fig_ref] Figure 3: Cytomegalovirus [/fig_ref]. Cell-cell or cellmatrix adhesion molecules, including ITGB6, were also reported to be down-regulated by CMV infection in some permissively infective cells. [bib_ref] Down-regulation of integrin a1/b1 expression and association with cell rounding in human..., Warren [/bib_ref] [bib_ref] Cytotrophoblasts infected with a pathogenic human cytomegalovirus strain dysregulate cell-matrix and cell-cell..., Tabata [/bib_ref] Our IHC and CISH analyses confirmed that ITGB6 expression was significantly lost in CMV-infected pneumocytes [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref]. The decreased adhesion between penumocytes and the basement membrane might result in frequent epithelial detachment, which could be a subsequent source of CMV propagation to other pneumocytes. Finally, plasma exudation, which forms a hyaline membrane on the alveolar surface, causes DAD lesions .
In addition to its function as an adhesion molecule, ITGB6 has been shown to play an important role in the development of TGF-b1-mediated IP lesions in combination with integrin av (as a heterodimer). [bib_ref] The integrin avb6 binds and activates latent TGFb1: A mechanism for regulating..., Munger [/bib_ref] Transforming growth factor-b1, the majority of which is produced in a latent form, 35 requires proteolytic enzyme reactions or a conformational change mediated by membrane-bound integrins for activation. [bib_ref] Integrin avb6-mediated activation of latent TGF-b requires the latent TGF-b binding protein-1, Annes [/bib_ref] In the absence of ITGB6 expression in CMV-positive pneumocytes, TGF-b1 signal transduction may be insufficient for the development of interstitial lesions, thereby resulting solely in a CMVp-DAD histological pattern, even if increased TGF-b1 expression is present. In case no. 3, which had 4 weeks of symptoms of respiratory disease, the organizing stage of DAD with intra-alveolar fibrous reaction was observed, while IIF lesions were underdeveloped.
Conversely, in CMVp-complex and CMVp-IIF cases, ITGB6 was overexpressed in CMV-negative pneumocytes along with TGF-b1 overexpression [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref]. CISH-CMV revealed a number of CMV-positive cells in the stroma of such ITGB6-overexpressing alveoli [fig_ref] Figure 5: The relationship between cytomegalovirus [/fig_ref] stromal trophic CMV propagation might promote functional TGF-b1 expression, primarily according to the frequency of CMV infection intensity, and cause persistent interstitial injury, leading to prolonged wound healing and exacerbation of IIF . The variation in DAD and IIF severity in CMVpcomplex may reflect, not a series of DAD lesions, but the reciprocal tissue injury induced by transitional CMV propagation between pneumocytes and stromal cells . It is essential to consider host immune status to understand CMV pathology and viral propagation. [bib_ref] Clinical relevance of cytomegalovirus infection in patients with disorders of the immune..., Steininger [/bib_ref] Marked systemic CMV involvement in CMVp-DAD cases [fig_ref] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/fig_ref] suggests a profound loss of the immune response against CMV infection. The severe immune-refractoriness may bring about CMV reactivation and rapid propagation in the lung with epithelial tropism, thus leading to CMVp-DAD with slight inflammatory infiltrates despite the alveolar injury.
In contrast, chronic, excessive inflammation, accompanied by IL-8 overexpression, was noted in cases of heavily infected CMVp-IIF. The CMV genome harbors several functional homologs of cytokine, chemokine, and chemokine receptors.A transcript of UL146, an open reading frame in the CMV genome, is known to act as an IL-8 receptor agonist, [bib_ref] The cytomegalovirus UL146 gene product vCXCL1 targets both CXCR1 and CXCR2 as..., Lüttichau [/bib_ref] promoting transendothelial migration of the neutrophils and viral transmission through migratory cell-cell microfusion events. [bib_ref] Cytomegalovirusinfected endothelial cells recruit neutrophils by the secretion of C-X-C chemokines and..., Grundy [/bib_ref] Furthermore, CMV-infected neutrophils are protected from apoptosis and remain in the tissue with enhanced function, 42 especially in the presence of IL-8. [bib_ref] Immediate-early antigen expression and modulation of apoptosis after invitro infection of polymorphonuclear..., Saez-Lopez [/bib_ref] Hence, CMV can propagate and modulate various host cellular immune functions, [bib_ref] Cytomegalovirus MHC class I homologues and natural killer cells: An overview, Farrell [/bib_ref] [bib_ref] Immunobiology of human cytomegalovirus: From bench to bedside, Crough [/bib_ref] leading to regulatory failure of the host immune system.
In the underlying pathology of IPF in case nos. 9 and 12, IL-8 overexpression may be a factor that exacerbates CMVinduced IIF [fig_ref] Figure 3: Cytomegalovirus [/fig_ref] ; however, the mechanism underlying induction of in vivo IL-8 overexpression remains controversial. Stromal components, such as vascular endothelial cells and myofibroblasts, have been proposed to play a critical role in regulating IPF [bib_ref] New mechanisms of pulmonary fibrosis, Strieter [/bib_ref] [bib_ref] Pulmonary hypertention and idiopathic pulmonary fibrosis: A tale of angiogenesis, apoptosis, and..., Farkas [/bib_ref] and CMVp. [bib_ref] Induction of an epithelial integrin avb6 in human cytomegalovirus-infected endothelial cells leads..., Tabata [/bib_ref] Endothelial cells infected with CMV were found in the present CMVp cases, as in the previous reports; 6,48 however, the ratio of positive cells was typically lower [fig_ref] Figure 3: Cytomegalovirus [/fig_ref] , and are not likely to be a major regulator of local CMVp lesion formation. Infection with CMV in SMA-positive cells was mostly not found in myofibroblasts, but was observed in pre-existing peribronchiolar or perivascular smooth muscle cells [fig_ref] Figure 2: Double immunohistochemistry [/fig_ref]. As to the concept of epithelial mesenchymal transition, i.e., TGF-b1-mediated cellular transformation from epithelial cells to myofibroblasts, [bib_ref] Induction of epithelial-mesenchymal transition in alveolar epithelial cells by transforming growth factor-b1;..., Willis [/bib_ref] further evidence is required to elucidate its significance in IIF pathogenesis.
In conclusion, the present histopathological and pathogenetic analysis of severe CMVp suggested that CMV tropism between pneumocytes and stromal cells, CMV-induced cytokine expressions, and host immune conditions were crucial in the formation of a variety of CMVp types.
[fig] Figure 1 11: and 12. All DAD lesions had acute stage characteristics with hyaline membrane formation, except case no. 3, in which foci Three major histological patterns of cytomegalovirus (CMV) pneumonia (CMVp). (a) Diffuse alveolar damage (DAD) pattern of CMVp (CMVp-DAD) (case no. 1, x20), (b) intermixed pattern of DAD and interstitial inflammation/fibrosis (IIF) (CMVp-complex) (case no. 7, x10), and (c) predominantly IIF pattern (CMVp-IIF) (case no. 12, x20). Cytomegalic cells (arrows) are observed on the alveolar surface, in the alveolar space, and in the alveolar wall. [/fig]
[fig] Figure 2: Double immunohistochemistry (IHC) of cytomegalovirus (CMV, red) and each cellular marker (blue) in representative cases of CMV pneumonia (CMVp)-diffuse alveolar damage (DAD) (case no. 2), CMVp-complex (case no. 4), and CMVp-interstitial inflammation/fibrosis (IIF) (case no. 12). [/fig]
[fig] Figure 3: Cytomegalovirus (CMV) tropism and severity of fibrosis among cases. (a) The ratio of doubly positive number for CMV and cellular marker to CMV-positive number in every section with the average number of CMV-positive cells and the frequency of CMV infection shown in parentheses; (b) representative aniline blue stained images and extracted two-tone images with the value of the positive ratio (left panels), and the aniline blue-positive ratio in all cases (right panel). The cases of CMVp-diffuse alveolar damage (DAD) (case nos [/fig]
[fig] Figure 4: Transforming growth factor (TGF)-b1 expression in immunohistochemistry (IHC) and semi-quantitative reverse transcriptasepolymerase chain reaction (RT-PCR). (a) Representative cases of TGF-b1 IHC of cytomegalovirus pneumonia (CMVp)-diffuse alveolar damage (DAD) (case no. 2) (left panel), CMVp-complex (case no. 5) (middle panel), and CMVp-interstitial inflammation/fibrosis (IIF) (case no. 11) (right panel), the positive findings in the reactive pneumocytes (solid arrowheads), intra-alveolar mononuclear cells (open arrowheads), fibroblasts (arrows), and vascular endothelial cells (asterisks); (b) the findings of semi-quantitative RT-PCR for TGF-b1, CMV, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in representative cases of CMVp-DAD (case nos. 2 and 3), CMVp-complex (case nos. 5, 6, and 7), and CMVp-IIF (case nos. 11 and 12) with the frequency of CMV infection (CMV freq.) as described in Figure 3a. [/fig]
[fig] Figure 5: The relationship between cytomegalovirus (CMV) infection/replication and functional expression of transforming growth factor (TGF)-b1, associated with integrin b6 (ITGB6) overexpression. (a) Double immunohistochemistry (IHC) of CMV (red) and ITGB6 (blue) and (b) single IHC of TGF-b1 (brown), examined at the same region of the serial sections (case no. 5). Up-regulation of ITGB6 in the pneumocytes accompanied by TGF-b1 expression (solid arrowheads) as well as colocalization of TGF-b1 expression and CMV antigens (circles) in both CMV pneumonia (CMVp)-complex and CMVp-interstitial inflammation/fibrosis (IIF) cases. (c and d) In double staining of chromogenic in situ hybridization for whole CMV genome (CISH-CMV) (red) and ITGB6 IHC (brown) (case no. 12), ITGB6 overexpression area is surrounded by remarkable IIF with increased CISH-positive stromal cells (c, x10). CISH-positive pneumocytes are lacking ITGB6 expression (d, arrows, x40). [/fig]
[fig] Figure 6: Double staining of interleukin (IL)-8 ISH (red) and cytomegalovirus (CMV) IHC (blue) in representative cases of CMVp-DAD (a, case no. 1), CMVp-complex (b, case no. 8), and CMVp-interstitial inflammation/fibrosis (IIF) (c, case no. 12); arrows, IL-8 overexpression in CMV infected cells; al, alveolar space; cap, capillary endothelial cells. [/fig]
[table] Table 1: Clinical data and histopathological findings of 12 autopsy cases [/table]
|
Strengthening the science of addressing antimicrobial resistance: a framework for planning, conducting and disseminating antimicrobial resistance intervention research
Antimicrobial resistance (AMR) has the potential to threaten tens of millions of lives and poses major global economic and development challenges. As the AMR threat grows, it is increasingly important to strengthen the scientific evidence base on AMR policy interventions, to learn from existing policies and programmes, and to integrate scientific evidence into the global AMR response. While rigorous evaluations of AMR policy interventions are the ideal, they are far from the current reality. To strengthen this evidence base, we describe a framework for planning, conducting and disseminating research on AMR policy interventions. The framework identifies challenges in AMR research, areas for enhanced coordination and cooperation with decision-makers, and best practices in the design of impact evaluations for AMR policies. This framework offers a path forward, enabling increased local and global cooperation, and overcoming common limitations in existing research on AMR policy interventions.
# Introduction
Antimicrobial resistance (AMR)the process by which microbes acquire resistance to antimicrobial medicines is widely recognised as a serious threat to global public health. The likelihood of drug resistance increases when microbes are exposed to antimicrobials and, unlike previous generations, we can no longer count on the development of new drugs to overcome this threat. The development of resistance has been accelerated by overuse of antimicrobials for medical and agricultural purposes. AMR now threatens tens of millions of lives, in addition to posing major global economic and development challenges.
AMR is politically, economically and microbially difficult to tackle from a policy perspective. Efforts to evaluate AMR interventions would be significantly improved by increasing investments in monitoring and surveillance for antimicrobial resistance. Controlling AMR will require a suite of effective antimicrobial stewardship and conservation strategies to ensure the appropriate use of antimicrobials, in addition to substantial international cooperation on the regulation and surveillance of antimicrobials and their use. Substantial research is needed to generate evidence on the effects and effectiveness of various possible AMR policies and to ensure that health system investments in AMR are evidence informed. Existing research has created little clarity about what interventions are best suited to achieve AMR goals across contexts, cultures and health systems. Many efforts to reduce AMR are designed as policies to reduce the use of antimicrobials; in this manuscript, when we refer to 'AMR policy', we refer also to these antimicrobial use policies. Policy recommendations for AMR have changed little since the late 1990s. Worldwide, millions of dollars are invested annually in public programmes to raise awareness about AMR, educate health professionals on appropriate prescribing, and decrease antimicrobial consumption in the health and agricultural sectors. Despite major financial and political investments, it has been difficult to link these programmes to concrete improvements in antimicrobial use, resistance or health outcomes more generally, particularly as major surveillance and information gaps impede the global response to AMR(Box 1).
As the threat posed by AMR grows, it is increasingly important to strengthen the scientific evidence base on AMR policy interventions, to learn from existing policies and programmes, and integrate scientific evidence into the global AMR response. The goal of this paper is to develop a framework that facilitates the strengthening of this evidence base. This paper is not intended as a formal research prioritisation process but, rather, builds upon the findings from recent systematic reviews of interventions to reduce antimicrobial consumptionand efforts by others to strengthen research on AMR and public health, and aims to draw insights for improving the planning, conduct and dissemination of research to evaluate AMR policy interventions. The framework identifies challenges in AMR evaluation research, areas for enhanced coordination and cooperation with policy-makers, and best- Around the world, 129 governments are currently in the process of developing or implementing a National Action Plan to address antimicrobial resistanceGlobal capacity for AMR surveillance is lacking; discrepancies between methods and monitoring systems, data quality concerns and lack of representativeness make it challenging to compare AMR data between countriesMany evaluations of AMR policy interventions are conducted retrospectively by academics who were not involved in the design or implementation of the interventionA systematic review of experimental and quasi-experimental studies evaluating government policy interventions to reduce the use of antimicrobialsfound that 30 of the 69 studies used low-quality study designs, such as uncontrolled before-after designs, which severely limits the validity of their findings. Among these 69 studies, only 4 used a randomised controlled design which is considered the gold standard for evaluating interventions. Another systematic review of 221 interventions for improving antibiotic prescribing among hospital inpatients found the quality of the reporting for the 163 non-randomised studies was so poor that it was difficult for professionals to use the research findings or to implement interventions that were shown to be useful; further, this systematic review found that no useful evidence could be gleaned from studies using controlled before-after and non-randomised trial designsReporting of AMR policy intervention studies is weak; studies often fail to describe the intervention in sufficient detail for replication and many do not report the reason the intervention is expected to workIn the broader field of public health, researchers have estimated that at least 50% of published research is not sufficiently clear, complete or accurate for others to interpret or useThere are no standardised measures and metrics for AMR research; many AMR intervention studies report antimicrobial use in defined daily dose per 1000 population or a simple prescribing ratepractices for overcoming common limitations in evaluating AMR policies. Some of these challenges are specific to AMR, while others are shared with other areas of health research.
## Research planning
Prioritising research AMR researchers need to prioritise the study of 'what' works, 'when' it works, 'why' it works, and 'what' elements are necessary for its success.
AMR needs better mechanisms for prioritising important research questions that can shape effective action. There has already been substantial research on the root social and microbial causes of AMRand we argue that attention now needs to be focused on determining which interventions are effective at addressing the underlying root causes of AMR, why these interventions work, what elements are necessary to their success, and in what contexts and circumstances these interventions work. More evidence on all four questions would be invaluable for policy planning. As the majority of existing research evidence focuses on interventions in high-income settings, additional research on these questions would particularly benefit low-and middle-income countries (LMICs) and other resource limited settings to identify policy interventions that can be adapted to meet local needs and priorities.
More formally, research prioritisation can be improved by undertaking structured prioritisation and consensus processes in collaboration with stakeholder groups, including policy-makers at different levels of government, civil society, health professionals and research funders. Research funders can support this work through opportunities for strategically funded research to address AMR rather than relying on researcher-led operating grants. For example, the Joint Programming Initiative on Antimicrobial Resistance (JPIAMR) recently funded an international workshop and formal consensus process to develop 10 research priorities for behavioural approaches to develop more impactful hospital antimicrobial stewardship programmes. The James Lind Alliance in the United Kingdom, which provides a platform for priority-setting partnerships between clinicians, patients and carers, has produced a guidebook that outlines their method for identifying research uncertainties and producing an agreed list of research priorities. Research prioritisation processes can also highlight the differences between research priorities in different contexts. Economic, political and cultural differences between countries and regions are likely to introduce new priorities. A recent prioritisation exercise looking at global health trial methodology found different research priorities in LMICs compared with the United Kingdom. The Cochrane Collaboration has previously published a special series on priority-setting that offers guidance on topics such as applying an equity lens to priority-settingand effective stakeholder participation in priority-setting. Formal prioritisation processes would help drive research agendas at the international level, such as those of JPIAMR, WHO, the Food and Agriculture Organization of the United Nations, and the World Organisation for Animal Health. Recently, the WHO has engaged in formal prioritysetting exercises for AMR research and development, first through a multi-criteria decision analysis exercise to develop its list of priority pathogens for research and development of new antimicrobials, and more recently through the Global Antibiotic Research & Development Partnership (GARDP). At the country level, such prioritisation processes can help drive national research funding, prompt updates to AMR national action plans, and support maximal learning from national AMR actions.
## Systematic reviews
Researchers need to use rigorous systematic reviews to inform research prioritisation and to summarise the effectiveness of policy interventions.
Rigorous systematic reviews and evidence gap maps can support the planning of policy interventions by ensuring that they are adequately supported by evidence. However, to be useful, these reviews must be high quality and regularly updated. Health Systems Evidence has appraised and catalogued more than 50 systematic reviews related to health systems and antimicrobial use dating back to the year 2000, and this database shows that the rigour and quality of these reviews is mixed .
Conducting systematic reviews to summarise what we currently know is an essential input for research prioritisation. Reviews can collate empirical evidence to answer a specific research questionbut they can also map the availability of evidence to identify evidence gaps. The Campbell Collaboration and others have recently developed methodological guidance for creating evidence gap maps. Systematic reviews need to be regularly updated to include new evidence as it arises, to determine whether research gaps have been addressed, and to determine whether the research question has been satisfactorily answered or whether future, more refined research is needed. Given that the current evidence base on AMR policy is weak, policy decisions in the near future will be informed by relatively weak signals from the research base, which reinforces the need for further evaluation.
Systematic reviews can also ensure that research efforts are not wasted on questions that have already been definitively answered. Although replication is key to science, there is a point at which additional replication holds little additional value. For example, it has been definitively shown that hospital antimicrobial stewardship interventions are effective at reducing antimicrobial use compared to control and should be a standard part of the AMR response. Future research can, instead, focus on optimising stewardship for different contexts in order to maximise effects. This includes opportunities to embrace a philosophy of radical incrementalism, where a series of small evaluated changes one after the other result in radical cumulative change. Finally, employing advanced analytical techniques, such as network meta-analysis, in systematic reviews can improve research prioritisation by enabling better exploration of heterogeneity in reviews of complex interventions.
## More planned evaluations
Researchers need to work with stakeholders to ensure that rigorous evaluations of all new AMR programmes are the norm.
Researchers can actively advance progress on AMR by ensuring that evaluations of policies become the norm. While rigorous evaluations of all AMR interventions would be ideal, at present, we are very far from this reality. Policy responses to AMRfrom legislation and government regulation to public awareness campaignshave played a major role in responding to AMR, yet these policies are rarely conceptualised as population health interventions and, as a result, are rarely pilot-tested, reviewed or evaluated with sufficient rigour to expand the AMR evidence base.
Without a culture of evaluation, we risk implementing, maintaining and even spreading ineffective or inefficient AMR policies, at great financial and opportunity costs. Additionally, though policy-makers do not always see their added value, there are political advantages in conducting good impact evaluations; evaluation puts policymakers in the politically attractive position of continuous policy improvement, enables them to ensure that research assessing their initiative is appropriate, and reduces political risk because they can acknowledge that they are operating with imperfect information. Where possible, researchers should advise and partner with policy-makers to raise the rigour of evaluations, simultaneously making progress on scientific questions. In particular, researchers should advocate for the development of protocols and evaluation plans a priori, which will help minimise waste of public resources in ineffective programmes by ensuring that the data collected is appropriate to answer key policy-maker questions, while also supporting implementation and future improvements in practice by ensuring that data is internationally comparable, and can feed into future evidence syntheses of similar policy interventions.
Research that is responsive to stakeholder needs AMR research needs to be planned to address policymakers' questions about effectiveness, implementation, costs and equity.
Moving from evidence to policy inevitably involves consideration of pragmatic and ideological factors beyond evidence of effectiveness. Researchers can support evidence-informed policy-making by considering, in advance, the likely information needs of policy-makers. In addition to effectiveness, decisions to pursue or pass over various AMR policy interventions are likely to be informed by their perceived cost-effectiveness, equity and differential impacts based on gender, race and socioeconomic status, implementation challenges, and acceptability to diverse social groups. This is particularly true in the case of LMICs, where resources are scarce and where the level of evidence required before investing in policy action may be substantially higher.
One simple and intuitive tool for enabling policymaker engagement in research planning is to use the APEASE criteria(Box 2). Originally created as a framework for evaluating ideas for interventions, APEASE offers a useful structure for framing research questions and evidence needs in partnership with stakeholders and for communicating policy-relevant research findings. APEASE addresses many common stakeholder concerns, including equality and equity considerations, acceptability across a wide range of groups, and the feasibility and practicality of an intervention in a given context, while recognising important trade-offs. Consider, for example, that both the clinical effectiveness and cost-effectiveness of an intervention is irrelevant if the intervention is unaffordable or infeasible given funding and resource constraints in a specific context. In addition to the feasibility of implementation, it is useful to consider whether the intervention should be, or can be, equally applied across the whole population, and whether it will reach its target population and intended beneficiaries. Again, the inclusion of these additional research questions in the planning phase will particularly benefit policy-makers in lower-resource settings, who must consider whether interventions from high-income settings could translate effectively to their setting, given the staff and resource limitations particular to their context.
## Box 2
The APEASE CriteriaAffordability Practicability Effectiveness/cost-effectiveness Acceptabilitypublic, professional, political Side-effects/safety Equality
## Research conduct better design of amr interventions
Researchers need to use theory, frameworks and logic models to design more coherent AMR policy interventions.
Unfortunately, to date, the approach to designing AMR policy interventions has been ad hoc and seems to be guided by the 'it seemed like a good idea at the time' principle, rather than by an explicit process that considers the determinants of the problem, relevant theory and available empirical evidence. This strategy has often resulted in poorly considered AMR policies that, although designed to change attitudes, beliefs and practices around antimicrobial use, cannot clearly articulate how their intervention will successfully bring about this change. Poorly designed policies may also fail to recognise and address key AMR determinants, leading to ineffective or sub-optimally effective interventions.
Researchers can improve the design of AMR policy interventions by employing and advocating for the inclusion of theory, frameworks and logic models in the early stages of intervention design to describe how and why an intervention is expected to work. These steps can substantially address the common tendency in AMR to re-invent the 'square' wheel rather than build on existing evidence from behavioural and implementation science. When planned without the use of theory, interventions are more likely to be unclear about the behaviours and outcomes targeted, and the means by which the intervention will achieve its intended effect. The process of building such models can encourage researchers to consider all aspects of the intervention and the existing AMR evidence base. The United Kingdom Medical Research Council has published a useful framework for developing and evaluating complex interventions. Other useful frameworks include the Behaviour Change Wheelto guide intervention development, the Theoretical Domains Framework to assess factors that impact behaviour, and the Behaviour Change Techniques taxonomy, which considers individual component strategies employed to change behaviour (e.g. feedback on behaviour, goal setting, prompts and cues) under the umbrella of a larger intervention. Finally, the use of theories, frameworks and logic models also facilitate research communication and dissemination, making clear the considerations and circumstances that drove the initial hypothesis.
## Better design of evaluations
Researchers need to ensure that AMR policy interventions are evaluated using the most rigorous designs feasible in the given circumstances.
When promoting a culture of AMR policy evaluation, improving and strengthening the design of evaluations should also be considered. One challenge in AMR research has been a lack of differentiation between programme evaluation (i.e. evaluating whether the local programme achieved its goals) and research evaluation (i.e. addressing generalisable concerns about what, when, how and why an intervention works). While programme evaluation is important, policy-making needs to be guided by robust evidence generated using rigorous study designs. AMR policy intervention research has been plagued by poor quality and inappropriate evaluation methods. In a recent systematic review, 30 of 69 included studies used uncontrolled before-after designs and simple descriptive methods that cannot control for important design concerns, including bias, confounding and secular trends, and are generally uninterpretable.
Overall, interventions should be evaluated using the most rigorous study designs feasible in the given circumstances in order to minimise bias and maximise the generalisability of findings. The choice of evaluation design will be guided by numerous factors, such as whether it is feasible to randomise intervention sites, whether it is necessary to introduce the intervention at all sites simultaneously, the acceptability of having a no-intervention control group, and the number of available intervention sites.
The focus on education, attitudes and behaviour change in AMR research, and the associated high likelihood of contamination, will likely preclude the use of randomised controlled trials (RCTs) with individuals (e.g. citizens, patients) as the units of randomisation in much of AMR policy intervention research. In most cases, a cluster RCT should be considered the gold standard for AMR policy intervention evaluations. Cluster randomisation can be done at numerous levels, although trade-offs exist between the number of available units (e.g. regional level) and the risk of contaminations (e.g. provider level). There are many possible cluster randomised trial designs, including two arm trials, multi-arm trials and factorial designs where two or more interventions can be implemented simultaneously. Another cluster design, the stepped wedge trialwhere all sites start in the control arm and end in the intervention arm, crossing over to intervention sequentially and in random ordermay have many practical benefits for large policy evaluations where rollout of an intervention to all sites within a health system or community is a requirement.illustrates these recommended evaluation designs.
Where random allocation is not feasible (either due to an inadequate number of randomisation units being available or because simultaneous implementation across the health system is required) interrupted time series (ITS) methods are the strongest study design for AMR research. However, adding a control groupor additional sites with staggered implementation of interventions (multiple baselines ITS designcan strengthen this design. Researchers should pay particular attention to whether an ITS design is appropriate for their study. ITS designs are best used for evaluating AMR interventions that have been implemented at a clearly defined point in time; many AMR studies have inappropriately used this design to measure complex interventions rolled out in stages across several months or years. de Kraker et al. have carefully considered the validity and bias concerns associated with these study designs and have published two excellent guides for evaluating antimicrobial stewardship interventions, which are also highly relevant to other AMR policy interventions. Uncontrolled before-after study designs, which historically have been common in AMR research, should be avoided if at all possible. The apparent effects of an intervention using these designs are completely confounded by secular trends and concurrent events. Even controlled before-after studies, which mitigate this threat of confounding to a limited extent, should be avoided unless there are at least two intervention sites and two control sites; however, given more sites, other more rigorous study designs may be feasible and preferable. Controlled before-after studies and non-randomised trials are an option when randomisation is not possible and there is an insufficient number of time intervals to conduct an ITS; however, these designs should only be considered as hypothesis generating.outlines some of the methodological considerations for choosing a prospective evaluation design. Iterative improvement on existing trials AMR policy research needs to include head-to-head comparisons of different intervention variations and promote radical incrementalism to enhance the effectiveness of extant policies. Researchers can advocate for the use of more rigorous study designs and partner with policy-makers to ensure that evaluations are appropriately conducted. To move beyond effectiveness, research evaluations need to incorporate opportunities for addressing the other important AMR policy questions described earlier, namely why interventions work and what elements are necessary for their success. Once the initial effectiveness of an AMR policy intervention has been shown, there is an opportunity to conduct controlled head-to-head comparisons of different intervention iterations in the interest of optimisation, either through sequential trials comparing variations of the intervention, or using factorial designs to evaluate whether the addition of cointerventions meaningfully changes the intervention's effectiveness. Evaluations can be facilitated by the development of implementation laboratories, which involve close collaboration between research teams and health systems delivering implementation strategies at scale. Trials can also be designed to enhance the generalisability of evaluation by including fidelity sub-studies to determine whether an intervention was delivered and received as designed, mechanistic sub-studies to determine whether the intervention acted through the hypothesised pathways, qualitative process evaluations to investigate experience and acceptability, and economic evaluations.
## A set of standard measures and metrics
Researchers need to define a set of core outcome measures for AMR research that address appropriate antimicrobial prescribing, development of antimicrobial resistance, and cost-effectiveness of policy interventions.
AMR lacks both an agreed-upon metric for evaluating progress and a common system for measuring the scope of the problem. Systematic reviews of interventions to reduce antimicrobial use have shown that researchers use a wide range of prescribing and dispensing-focused metrics. The Tripartite Monitoring and Evaluation Frameworkrecently released by WHO, the Food and Agriculture Organization and the World Organisation for Animal Health, is a useful starting point that suggests many important One Health indicators for AMR; however, this framework lacks detail on operationalising these measures. Researchers can have a significant impact in these ongoing conversations by developing measures that facilitate data sharing; for example, through an agreed minimum dataset for collecting intervention data and a core outcome set of measures to facilitate evaluations. Consideration should also be given as to whether a core outcome set can feed into AMR and antimicrobial use surveillance to facilitate the use of routinely collected data in impact evaluations.
A harmonised set of measures for conducting and evaluating interventions serves three purposes. First, it creates consistency between evaluations that facilitates systematic reviews to inform evidence-informed policymaking. Lack of a shared outcome measure is a common challenge in public health systematic reviews and one which limits the amount of evidence that can be rigorously synthesised. Second, creating a harmonised set of measures offers the opportunity to include, as a requirement, metrics beyond impact that are relevant to stakeholders such as equity considerations and costeffectiveness. Finally, developing a set of harmonised measures offers an opportunity to consider common barriers to AMR policy implementation and to develop standard measures and indicators that also address these needs.
While there are substantial political and economic barriers to improving data collection and evaluation, from a metrics viewpoint, there are numerous successful initiatives that have adopted research governance principles to improve data collection and comparability of research studies, including the Core Outcome Measures in Effectiveness Trials (COMET) groupand ESSENCE for Health Research. Outcome Measures in Rheumatology (OMERACT), for example, uses a data-driven, iterative process to choose shared outcome measures across four domains, and each selected measure must meet three criteriatruth, discrimination and feasibility.
## Research dissemination better reporting of interventions
Researchers need to register research protocols and evaluation trials and use reporting guidelines and checklists to improve reporting quality.
An effective and coherent global response to AMR requires full and transparent reporting of all aspects of AMR intervention studies. Unfortunately, public health research as a whole, and AMR research specifically, faces challenges in this domain. Inaccessible research, partial reporting, and publication bias are all common in AMR researchand limit opportunities for researchers and policy-makers to learn from experiences in other contexts.
Researchers should register protocols, primary evaluations of policy interventions (both RCTs and quasiexperiments) and systematic reviews in trial registers. Pre-registration of an evaluation helps address the tendency to avoid publishing the results of research with neutral or negative findings, which is common to much of scientific research, and may be particularly common in policy research where government partners may feel that they lose credibility if a policy intervention is found to be less effective than expected. The fields of clinical Researchers should also embrace the use of reporting guidelines and checklists to overcome common reporting challenges such as not describing the intervention specifically enough to allow replication, using the same description to represent different types of interventions, using different terminology to represent the same content, and repetition without improvement. The TIDieR toolfor describing and replicating interventions, and its extension TIDieR-PHPfor population health and policy studies, are both useful tools for researchers to ensure that their intervention has been thoroughly described. The SQUIRE checklistfor quality improvement interventions, the CONSORT statementfor RCTs and its many extensions to cluster RCTs, stepped wedge trials, pilot studies, pragmatic trials, and the Equator Networkare all useful tools to ensure the full reporting of methods and findings within a study report.
## Shared learning opportunities
Researchers need to embrace open data and open access opportunities to widely disseminate AMR research findings.
The principle of shared learning is familiar to researchers and embedded in much of health research. A key question facing the field is how best to promote this ideal when working with stakeholders to develop effective and efficient AMR policy interventions. However, as many governments embrace nudge units, innovation hubs and radical incrementalism, there is an opportunity for researchers to reiterate the substantial benefits of shared learning. Many new methodological tools can support these effortsdata sharing platforms (e.g. the World Wide Antimalarial Resistance Network, WWARNand open-access information repositories will both go a long way to ensuring the evidence generated from policy experiments and intervention evaluations can support shared learning. Likewise, living systematic reviews, which are continually updated as new evidence is generated, will also help ensure that new policies are planned based on current evidence.
# Conclusions
Although the threat posed by AMR has been wellknown for many decades, recent escalation in multidrug-resistant and extensively drug-resistant bacterial infections has elevated AMR to a more prominent position on the international political agenda. Substantial work is ongoing at the national level to address domestic AMR concerns, and there are robust conversations at the international level about pursuing more largescale, coordinated efforts to mitigate the AMR threat such as an international legal treaty on AMR. Innovative approaches have been taken to research funding (e.g. JPIAMR, health systems strengthening (e.g. Accreditation Canada, and global policy monitoring (e.g. WHO.
These national and international efforts require substantially more evidence for the effectiveness and feasibility of AMR policy interventions than is currently available. However, this gap provides an opportunity for researchers to engage in a meaningful conversation about the importance of evidenceinformed policy-making for AMR. Mitigating the threat posed by AMR will also require substantial collaboration among researchers and policy-makers.describes many of the ways in which research funders, publishers and policy-makers can jointly support researchers and facilitate action across the priorities identified in this framework. With their ability to strategically fund innovative research, encourage researchers to use rigorous study designs and reporting checklists, and facilitate shared learning, these partners have an opportunity to amplify researchers' calls for better AMR practice.
As we have highlighted, it is increasingly important to strengthen the scientific evidence base on AMR policy interventions, to learn from existing policies and programmes, and integrate scientific evidence into the global AMR response. This framework offers a path forward, increasing local and global cooperation, and overcoming common limitations in existing research on AMR policy interventions.
Abbreviations AMR: antimicrobial resistance; ITS: interrupted time series; JPIAMR: Joint Programming Initiative on AMR; LMICs: low-and middle-income countries; RCT: randomised controlled trial Funding This work was completed as part of the International Collaboration for Capitalizing on Cost-Effective and Life-Saving Commodities (i4C) that is funded through the Research Council of Norway's Global Health & Vaccination Programme (GLOBVAC Project #234608). SRVK is supported by an Ontario Graduate Scholarship and SJH is additionally supported by the Canadian Institutes of Health Research. JMG holds a Canada Research Chair in Health Knowledge Transfer and Uptake. None of the funders had a role in the design of the study, the preparation of this manuscript, or the decision to publish it.
|
Antimicrobial Susceptibility of Standard Strains of Nontuberculous Mycobacteria by Microplate Alamar Blue Assay
In this study, 24 standard nontuberculous mycobacteria (NTM) species strains including 12 slowly growing mycobacteria strains and 12 rapidly growing mycobacteria strains were subjected to drug susceptibility testing using microplate Alamar Blue assay-based 7H9 broth. The most active antimicrobial agents against the 24 NTM strains were streptomycin, amikacin, the fluoroquinolones, and the tetracyclines. Mycobacterium chelonae, Mycobacterium abscessus, Mycobacterium bolletii, and Mycobacterium simiae are resistant to most antimicrobial agents. The susceptibility results of this study from 24 NTM standard strains can be referenced by clinicians before susceptibility testing for clinical isolates is performed or when conditions do not allow for susceptibility testing. The application of broth-based methods is recommended by the Clinical and Laboratory Standards Institute, and the documentation of the susceptibility patterns of standard strains of mycobacteria can improve the international standardization of susceptibility testing methods.
# Introduction
Although the prevalence of tuberculosis is decreasing globally, increased numbers of nontuberculous mycobacteria (NTM) have been reported in human infections in recent years because of the growing number of immunosuppressed patients coupled with better diagnostic techniques. Classified into rapidly growing mycobacteria (RGM) and slowly growing mycobacteria (SGM), NTM are opportunistic pathogens that can cause a wide variety of disseminated or localized diseases, particularly pulmonary, skin, and soft tissue infections. Due to the differences between even individual NTM strains, these organisms require individualized treatment that must be selected on the basis of results obtained from in vitro drug susceptibility tests (DST).
With the evolution of assay techniques, especially the wide application of a new commercially available DNA strip assay (GenoType Mycobacterium, Hain Lifescience, Nehren, Germany) [bib_ref] Evaluation of the GenoType Mycobacterium Assay for identification of mycobacterial species from..., Richter [/bib_ref] [bib_ref] Molecular detection and identification of mycobacterium tuberculosis complex and four clinically important..., Bicmen [/bib_ref] , Mycobacteria can be easily identified to the species level; however, our knowledge about the susceptibility patterns of NTM is limited. Our results presented here about the susceptibilities of 15 antibiotic agents against standard NTM strains could be highly valuable for clinicians.
The methods for antimycobacterial susceptibility testing include the Clinical and Laboratory Standards Institute (CLSI) brothbased methodology, E-test [bib_ref] Assessment of in vitro susceptibility to antimicrobials of rapidly growing mycobacteria by..., Garcia-Agudo [/bib_ref] , agar-based testing methods [bib_ref] In vitro drug susceptibility of 2275 clinical non-tuberculous Mycobacterium isolates of 49..., Van Ingen [/bib_ref] , and the disk elution and diffusion method [bib_ref] Antibiotic susceptibility pattern of rapidly growing mycobacteria, Gayathri [/bib_ref] [bib_ref] Antibiotic susceptibility pattern of Mycobacterium marinum, Aubry [/bib_ref]. The CLSI currently recommends Mueller-Hinton broth-based methods for RGM and Mueller-Hinton or 7H9 broth-based methods supplemented with OADC or acid-albumin-dextrose-catalase (ADC) for SGM. Another broth-based method, the microplate Alamar Blue assay, has been used for years with favorable results for Mycobacterium tuberculosis complex isolates [bib_ref] Interand intra-assay reproducibility of microplate Alamar blue assay results for isoniazid, rifampicin,..., Leonard [/bib_ref] [bib_ref] Evaluation of the accuracy of the microplate Alamar Blue assay for rapid..., Chauca [/bib_ref]. In this study, we used the microplate Alamar Blue assay to test the activities of 15 drugs against 24 standard NTM strains in China.
# Methods
## Strains
Twenty-four NTM standard strains including 12 RGM and 12 SGM were included in this study [fig_ref] Table 1: MIC [/fig_ref].
## Antibiotics and chemicals
Middlebrook 7H9 broth and ADC supplement were purchased from Difco (Detroit, MI, USA). Powders of 15 antimicrobial agents including rifampicin, isoniazid, streptomycin, amikacin, kanamycin, ciprofloxacin, ofloxacin, levofloxacin, capreomycin, cefoxitin, doxycycline, minocycline, ethionamide, and p-aminosalicylic acid were purchased from Sigma-Aldrich Company (St. Louis, USA), while dipasic was purchased from Chongqing Huapont Pharmaceutical Company (Chongqing, China). Alamar Blue was purchased from AbD Serotec (Oxford, UK). All of the antibiotic solutions were prepared before the day of the experiment and stored at 270uC.
## Antimicrobial susceptibility testing
Susceptibility testing was performed using the Middlebrook 7H9 broth microdilution method. All tests for each strain were carried out at least in duplicate. The isolates were grown on microplates. The inocula were prepared from actively growing bacteria collected from Lowenstein-Jensen slants. The strains were then adjusted with saline to a bacterial cell density of 3.0610 8 (McFarland 1.0 standard), and then adjusted to a 1:20 dilution with Middlebrook 7H9 Supplement (7H9-S) (7H9 broth +10% ADC + 0.5% glycerol). Antibiotics were serially diluted twofold in 100 mL of 7H9-S. The range of antibiotic concentrations was 256-0.5 mg/mL except for ciprofloxacin and ofloxacin, which were 128-0.25 mg/mL, and minocycline, which was 42.6-0.04 mg/mL. The final reaction volume was 200 mL (100 mL of antibiotic solution and 100 mL of bacterial suspension). Three negative controls were set in this study. Drug free control well(7H9-S+inoculum) was used to decide the time of adding alamar blue. The medium(7H9-S) without inoculum control well was used to decide the interference of 7H9-S to alamar blue and a series control wells of rifampicin concentration gradients of rifampicin and 7H9-S mixture were also used to decide the interference of the color of rifampicin-7H9-S mixture to alamar blue. The plates were sealed in individual Ziploc bags and then incubated at 37uC.
After 24 h, the first drug-free growth control wells were examined using indicator (20 mL of Alamar Blue and 50 mL of sterile 5% Tween-80). The plates were then re-incubated for 8 h. If the control well turned pink, all of the other wells received the indicator. After a further 24 h of incubation, the colors of all wells were recorded. If the first drug-free growth control well did not change to pink, the second drug-free control well received the indicator and the above steps were repeated. Each minimum inhibitory concentration (MIC) was read on the 3 rd to 6 th days. The MIC was defined as the lowest drug concentration that prevented a change in color. The final result of MIC of each drug for each strain was the mean value from two tests. The MIC breakpoints of the drugs indicating sensitivity, moderate susceptibility, and resistance were interpreted according to the approved guidelines established by the National Committee for Clinical Laboratory Standardsand World Health Organization guidance(Table 2) except that minocycle was according to Vanitha et al [bib_ref] Evaluation of microplate Alamar blue assay for drug susceptibility testing of Mycobacterium..., Vanitha [/bib_ref].
# Results
The results of the antimicrobial susceptibility testing of the 12 standard RGM and 12 standard SGM strains are shown in [fig_ref] Table 1: MIC [/fig_ref]. All 24 strains were highly resistant to isoniazid (all MIC . 256 mg/mL). Nine and six of 12 RGM strains were resistant to rifampicin and streptomycin, respectively, especially Mycobacterium abscessus and Mycobacterium bolletii, which showed very high resistance, while seven and two of 12 SGM strains were resistant to rifampicin and streptomycin, respectively. Among members of the Mycobacterium chelonae-abscessus complex, M. abscessus was more resistant to rifampicin than M. chelonae.
All of the NTM strains were sensitive or moderately susceptible to amikacin. However, most of the standard NTM strains (19/23) were resistant to the aminoglycoside kanamycin. The 12 SGM strains were all susceptible to amikacin, whereas only one SGM (Mycobacterium senegalense) was susceptible to kanamycin.
The fluoroquinolones were active against most of the NTM. Nine RGM strains (Mycobacterium doricum, Mycobacterium fortuitum, Mycobacterium peregrinum, Mycobacterium obuense, Mycobacterium phlei, Mycobacterium duvalii, Mycobacterium parafortuitum, Mycobacterium gilvum, and Mycobacterium flavescens) and 5 SGM strains (Mycobacterium xenopi, Mycobacterium kansasii, Mycobacterium scrofulaceum, Mycobacterium malmoense, and Mycobacterium terrae) were all susceptible to ciprofloxacin, ofloxacin, and levofloxacin.
All of the standard NTM strains were highly resistant to capreomycin (MIC . 256 mg/mL) except for M. scrofulaceum and Mycobacterium gordonae, which were low-level resistant (MIC = 4 mg/mL).
Twelve NTM strains (6/12 RGM strains and 6/12 SGM strains) were sensitive or moderately susceptible to the b-lactam antibiotic cefoxitin, while the other strains were highly resistant. Cefoxitin was the only antibiotic agent that M. chelonae was more resistant to than M. abscessus (.256 mg/mL vs. 64 mg/mL). Dipasic is a mixture of p-aminosalicylic acid and isoniazid, and its MIC breakpoints (mg/mL) corresponded to those of paminosalicylic acid in this study (the MIC of dipasic for these 24 strains were all . 256 mg/mL). As such, all 24 NTM strains were resistant to p-aminosalicylic acid and dipasic.
# Discussion
Analyzing the susceptibility results of 24 standard NTM strains using the microplate Alamar Blue assay, it became evident that these findings yielded important clues for the optimization of NTM species-specific therapy. The results showed that streptomycin, amikacin, the fluoroquinolones, and the tetracyclines were the most active antimicrobial agents against the 12 RGM and 12 SGM strains. This is the first report of susceptibility patterns of standard NTM strains.
It is well know that most NTM strains are resistant to conventional anti-tuberculous agents [bib_ref] An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Griffith [/bib_ref] [bib_ref] High prevalence of antimicrobial resistance in rapidly growing mycobacteria in Taiwan, Yang [/bib_ref] [bib_ref] Nontuberculous mycobacteria isolated from pulmonary specimens between 2004 and 2009: causative agent..., Bicmen [/bib_ref] [bib_ref] Clinical and taxonomic status of pathogenic nonpigmented or late-pigmenting rapidly growing mycobacteria, Brown-Elliott [/bib_ref] , a fact that was further proven by the current findings. However, more SGM strains were susceptible to the drugs than RGM stains. Most of the SGM strains(10/12) were also susceptible to streptomycin. Several reports have shown that amikacin has good activity against RGM [bib_ref] Antibiotic susceptibility pattern of rapidly growing mycobacteria, Gayathri [/bib_ref] [bib_ref] In vitro activities of tigecycline and 10 other antimicrobials against nonpigmented rapidly..., Fernandez-Roblas [/bib_ref] [bib_ref] In Vitro activities of isepamicin, other aminoglycosides, and capreomycin against clinical isolates..., Shen [/bib_ref]. In this study, 22 NTM strains were susceptible to and 2 NTM strains (M. abscessus and M. chelonae) were moderately susceptible to amikacin.
M. abscessus is naturally sensitive to amikacin, cefoxitin, and imipenem [bib_ref] Mycobacterium abscessus: an emerging rapid-growing potential pathogen, Petrini [/bib_ref] and very resistant to many other chemotherapeutic agents [bib_ref] Diagnosis and treatment of infections caused by rapidly growing mycobacteria, Colombo [/bib_ref]. In our study, M. abscessus was resistant to 13 antibiotics and moderately susceptible to amikacin and cefoxitin. These findings are comparable to those described in the other studies. A total of 38 (95%) isolates in a Taiwanese study (40 isolates of M. abscessus isolates obtained from January 2006 to December 2008) and 73 (99%) isolates in a Korean study (74 isolates of M. abscessus isolates obtained from July 2005 to December 2006) of in vitro antimicrobial susceptibility were sensitive to amikacin [bib_ref] Clinical outcome of Mycobacterium abscessus infection and antimicrobial susceptibility testing, Huang [/bib_ref] [bib_ref] In vitro antimicrobial susceptibility of Mycobacterium abscessus in Korea, Park [/bib_ref]. M. abscessus and M. chelonae are members of the M. chelonae-abscessus complex, the susceptibility patterns of which are similar, although M. abscessus was more resistant than M. chelonae to rifampicin, ciprofloxacin, ofloxacin, and levofloxacin. Cefoxitin was the only antibiotic agent tested here that M. chelonae was more resistant to than M. abscessus (.256 mg/mL vs. 64 mg/mL), suggesting that cefoxitin resistance could be a way to distinguish between M. chelonae and M. abscessus.
M. fortuitum and M. peregrinum are members of the M. fortuitum complex. However, M. peregrinum was more sensitive to rifampicin, streptomycin, amikacin, cefoxitin, doxycycline, minocycline, and ethionamide than was M. fortuitum. Studies have reported that the M. fortuitum complex was much less drug-resistant than M. abscessus and M. chelonae [bib_ref] Clinical and taxonomic status of pathogenic nonpigmented or late-pigmenting rapidly growing mycobacteria, Brown-Elliott [/bib_ref] [bib_ref] Comparison of the in vitro activity of the glycylcycline tigecycline (formerly GAR-936)..., Wallace [/bib_ref] [bib_ref] Activities of ciprofloxacin and ofloxacin against rapidly growing mycobacteria with demonstration of..., Wallace [/bib_ref] [bib_ref] Susceptibilities of Mycobacterium fortuitum biovar. fortuitum and the two subgroups of Mycobacterium..., Wallace [/bib_ref] ; in our study, the M. fortuitum complex was more susceptible to amikacin, kanamycin, ciprofloxacin, ofloxacin, and levofloxacin than M. abscessus and M. chelonae. M. bolletii and M. simiae were resistant to 14 of 15 antibiotic agents and susceptible to amikacin, so more agents should be included in future tests. One study showed that 38% (11/29), 25% (7/29), 100% (29/29), 90% (26/29), and 66% (19/29)of M. simiae clinical isolates were susceptible to ciprofloxacin, clarithromycin, cycloserine, clofazimine, and prothionamide, respectively [bib_ref] Clinical outcome of Mycobacterium abscessus infection and antimicrobial susceptibility testing, Huang [/bib_ref]. However, the efficacy of these drugs in NTM treatment has not been sufficiently proven and they are limited by their toxicity [bib_ref] An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Griffith [/bib_ref] [bib_ref] Synergic activity of D-cycloserine and beta-chloro-D-alanine against Mycobacterium tuberculosis, David [/bib_ref].
In summary, M. abscessus, M. chelonae, and M. bolletii were resistant to almost all 15 antimicrobial agents, while the other nine standard RGM strains were resistant to 4-11 drugs [fig_ref] Figure 1: The susceptibility distributions to 15 antimicrobial agents of 12 standard rapidly growing... [/fig_ref]. A total of 11 SGM standard strains were resistant to 6-11 drugs, while M. simiae was resistant to 14 drugs [fig_ref] Figure 2: The susceptibility distributions to 15 antimicrobial agents of 12 standard slowly growing... [/fig_ref].
The American Thoracic Society advocated the use of macrolide-based multidrug regimens for NTM treatment [bib_ref] An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Griffith [/bib_ref] , and some studies have reported that most NTM strains are sensitive to the macrolide clarithromycin [bib_ref] In vitro drug susceptibility of 2275 clinical non-tuberculous Mycobacterium isolates of 49..., Van Ingen [/bib_ref] [bib_ref] Evaluation of antimicrobial susceptibilities of rapidly growing mycobacteria by Sensititre RAPMYCO panel, Cavusoglu [/bib_ref] and that rifabutin and tigecycline also showed high activity against many NTM strains [bib_ref] In vitro drug susceptibility of 2275 clinical non-tuberculous Mycobacterium isolates of 49..., Van Ingen [/bib_ref] [bib_ref] In vitro activities of tigecycline and 10 other antimicrobials against nonpigmented rapidly..., Fernandez-Roblas [/bib_ref]. As such, streptomycin, amikacin, the fluoroquinones, the tetracyclines, and the above three antibiotics were the alternative choices for the treatment of NTM infection. Studies have shown that the susceptibilities of clinical NTM isolates of a species were also quite different [bib_ref] Clinical outcome of Mycobacterium abscessus infection and antimicrobial susceptibility testing, Huang [/bib_ref] [bib_ref] Evaluation of antimicrobial susceptibilities of rapidly growing mycobacteria by Sensititre RAPMYCO panel, Cavusoglu [/bib_ref] , so our results from NTM standard strains can only be referenced by clinicians before susceptibility testing for clinical isolates is performed or when conditions do not allow for susceptibility testing. Susceptibility testing for clinical isolates should always be performed prior to treatment unless conditions do not permit such. The technique described here can offer the MIC of antimicrobial agents within 6 days. The microplate Alamar Blue assay is inexpensive and reliable for the DST of NTM. Alternatively, the application of broth-based methods is recommended by the CLSI and the susceptibility patterns of standard strains of Mycobacterium can improve the international standardization of susceptibility testing methods.
[fig] Figure 1: The susceptibility distributions to 15 antimicrobial agents of 12 standard rapidly growing mycobacteria strains. doi:10.1371/journal.pone.0084065.g001 [/fig]
[fig] Figure 2: The susceptibility distributions to 15 antimicrobial agents of 12 standard slowly growing mycobacteria strains. doi:10.1371/journal.pone.0084065.g002 [/fig]
[table] Table 1: MIC (mg/mL) of the 15 antimicrobial agents* to the 24 standard NTM strains. [/table]
[table] Table 2: The MIC breakpoints (mg/mL) of the 15 antimicrobial agents. MIC, minimum inhibitory concentration. doi:10.1371/journal.pone.0084065.t002 [/table]
|
Asymmetric optic disc edema in a young patient with POEMS: A rare presentation of a rare disease
A B S T R A C TPurpose: To describe a case of asymmetric optic disc edema presenting as the initial ocular feature of POEMS (Polyneuropathy, Organomegaly, Endocrinopathy, Monoclonal gammopathy, Skin changes) syndrome. Observations: A 29-year-old female patient presented with 3 weeks history of blurred vision, proptosis, and peripheral neuropathy as well as hypothyroidism. Fundoscopy revealed optic disc edema associated with visual loss in the left eye. Following a computed tomography (CT) scan and a positron emission tomography/CT (PET/CT) scan which respectively revealed hepatomegaly and multiple osteosclerotic lesions, as well as laboratory findings of monoclonal gammopathy and elevated vascular endothelial growth factor (VEGF) levels, she was diagnosed with POEMS syndrome. After treatment with an autologous stem cell transplant, the optic disc edema and blurred vision resolved. Conclusions and importance: The most reported ocular manifestation of POEMS syndrome, a rare and complex multisystem disorder, is bilateral optic disc edema that typically occurs in older males. Therefore, this report presents an uncommon case of asymmetric optic disc edema in a younger, female patient.
# Introduction
POEMS syndrome, also known as Crow-Fukase syndrome, osteosclerotic myeloma, or Takatsuki syndrome, is a rare, multisystem paraneoplastic disorder associated with peripheral neuropathy (P), organomegaly (O), endocrinopathy (E), monoclonal gammopathy (M), and skin changes (S), although not all features must be present for diagnosis. [bib_ref] POEMS syndrome, Dispenzieri [/bib_ref] Other major criteria include osteosclerotic lesions and elevation of serum vascular endothelial growth factor (VEGF) levels. [bib_ref] Update on diagnosis, risk-stratification, and management, Dispenzieri [/bib_ref] Ocular manifestations are also important to consider, as papilledema is one of six minor criteria for POEMS diagnosis according to the most recent assessment by Dispenzieri2. Specifically, optic disc edema [bib_ref] Infiltrative orbitopathy, optic disk edema, and POEMS, Bourdette [/bib_ref] [bib_ref] Optic disk swelling with peripheral neuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Bolling [/bib_ref] [bib_ref] POEMS syndrome: an unusual cause of bilateral optic disk swelling, Wong [/bib_ref] [bib_ref] Optic disk drusen, peripapillary choroidal neovascularization, and POEMS syndrome, Diduszyn [/bib_ref] [bib_ref] Unusual cause of bilateral optic disc swelling: POEMS syndrome, Wiaux [/bib_ref] [bib_ref] Optic disc edema, cystoid macular edema, and elevated vascular endothelial growth factor..., Chong [/bib_ref] [bib_ref] Papilledema as an indicator of POEMS syndrome, Barnés [/bib_ref] [bib_ref] Ocular findings in patients with polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Kaushik [/bib_ref] [bib_ref] Ocular manifestations and treatment outcomes in Chinese patients with POEMS syndrome, Zhang [/bib_ref] is the most frequent ocular finding in POEMS patients, with recent reports citing an incidence of 52%, 10 67.5% 11 and 79%, [bib_ref] Correlation between peripapillary retinal thickness and serum level of vascular endothelial growth..., Yokouchi [/bib_ref] and has presented bilaterally at the initial visit in all but one 13 individual case report. Furthermore, the syndrome is typically diagnosed in male patients ranging from 40 to 65 years of age. Therefore, we present a rather unique POEMS case in a 29-year-old female patient with asymmetric optic disc edema.
## Case report
A 29-year-old female patient was referred for a neuroophthalmological consultation prompted by symptoms of blurry vision and photopsias for 3 weeks. Her past medical history was notable for chronic migraines. Her recent medical history included menstrual dysfunction, hand and feet paresthesias, fatigue, bilateral arm rashes, joint pain and swelling, abdominal distention, and hair loss throughout the past six months. She had noted a 20-pound weight gain for the past three-to-four months despite a consistent workout regimen and was recently started on synthroid for newly diagnosed hypothyroidism.
Ophthalmologic examination showed an uncorrected visual acuity of 20/20 in both eyes and 1 mm of proptosis in the left eye with diffuse conjunctival injection. Intraocular pressures were 14 mmHg in both eyes. AOHRR color plates, sensorimotor testing and slit-lamp examination did not demonstrate any abnormalities. She was emmetropic in both eyes and upon testing, there was an afferent pupillary defect in the left eye. Visual field testing showed fluctuations nasally and in the blind spot of the left eye and was normal in the right eye [fig_ref] Figure 1: Visual fields of right and left eyes before and after treatment with... [/fig_ref]. Dilated fundus examination showed normal right optic nerve and optic disc edema without vitritis in the left eye [fig_ref] Figure 2: Fundus images of right and left eyes before and after treatment with... [/fig_ref]. Optical coherence tomography of the optic nerve corroborated the clinically-observed disc edema in the left eye and did not exhibit subclinical edema in the right eye [fig_ref] Figure 3: Optical coherence tomography [/fig_ref].
Further work-up included a brain and orbit MRI showing bilateral prominence of the lacrimal gland and extra-ocular muscles without optic nerve enhancement [fig_ref] Figure 4: Axial section of orbital-cerebral magnetic resonance imaging [/fig_ref]. Lumbar puncture showed an opening pressure of 20.5 mmHg and CSF protein of 70. She was discharged on Diamox, pending further workup.
Follow-up visits after discharge showed increasing disc edema (Figs. 1B, 2B and 3B), hand and feet parasthesias, and monoclonal gammopathy with IgG lambda levels of 900 mg/dL. Consultation with neurology corroborated a suspected diagnosis of POEMS. Subsequent work-up indicated VEGF levels of 409 (normal: 9-86), mild erythrocytosis and thrombocytosis (platelet count of 407 × 10(3)/uL), presence of monoclonal IgG gammopathy (1494 mg/dL) and a kappa light chain level of 1.96 mg/dL (range: 0.33-1.94 mg/dL) with a kappa to lambda ratio of 0.32 (range: 0. . An electromyogram also demonstrated chronic multifocal demyelinating sensorimotor polyneuropathy. CT imaging of chest, abdomen and pelvis revealed hepatomegaly (18.1 cm) as well as multiple spine and pelvic osteolytic lesions with central sclerosis. A bone marrow biopsy failed to show a clonal plasma cell population, therefore a sacral lesion biopsy was performed and revealed sheets of lambda-restricted plasma cell neoplasm [fig_ref] Figure 5: Sacral lesion biopsy at one month after initial visit [/fig_ref]. Finally, a PET/CT scan evidenced multiple osteosclerotic lesions along the spine and pelvic bone, demonstrating multifocal hypermetabolic bone involvement [fig_ref] Figure 6: Positron emission tomography [/fig_ref].
After a series of hematology and oncology consultations, a treatment course of Melphalan200 (200 mg/m 2 ) conditioning and autologous stem cell transplantation (ASCT) was selected. Prior to initiating treatment, the patient chose to cryopreserve her eggs. Stem cell steady state mobilization was induced with 10 mg/kg GCSF divided twice a day for four days, with stem cells collected by apheresis on the fifth day. The patient exhibited typical post-transplant complications such as mucositis, as well as engraftment syndrome which was treated with a prolonged steroid taper. Ophthalmologic examination 12 days after the transplant also showed optic disc hemorrhage and worsening edema [fig_ref] Figure 1: Visual fields of right and left eyes before and after treatment with... [/fig_ref]. The patient was discharged after 68 days once all complications had resolved. Follow-up visit seven months posttransplant showed improvement in rashes, blurred vision and the optic disc edema [fig_ref] Figure 1: Visual fields of right and left eyes before and after treatment with... [/fig_ref] , in addition to significant improvement in the serum M protein level. VEGF levels normalized approximately eleven weeks following the transplant.
# Discussion
Due to its variety of clinical features, POEMS syndrome is not thoroughly understood and can be quite challenging to diagnose. While our patient was a 29-year-old female, a Mayo Clinic study involving 291 POEMS patients, two-thirds of which were male, cited a median age of 52 years at baseline 14 and a recent nationwide survey of Japanese POEMS patients noted a 1.5 ratio of diagnosed men to women and a median age of 54 years at onset. [bib_ref] Prevalence, clinical profiles, and prognosis of POEMS syndrome in Japanese nationwide survey, Suichi [/bib_ref] Few case reports involving younger women have been published. [bib_ref] POEMS syndrome with Guillan-Barre syndromelike acute onset: a case report and review..., Isose [/bib_ref] [bib_ref] POEMS syndrome: clinical, pathological and immunological study of a case, Orefice [/bib_ref] [bib_ref] Peripheral neuropathy: clinical diagnosis of POEMS syndrome and treatment with radiotherapy, Ng [/bib_ref] Current theories link the manifestation of POEMS to elevated cytokine levels, primarily vascular endothelial growth factor (VEGF), interleukin-6 (IL-6), and interleukin-12 (IL-12). 1,2 In particular, elevated serum and plasma VEGF levels have become an important element in distinguishing POEMS from similar disorders such as other plasma cell dyscrasias, monoclonal gammopathy of undetermined significance (MGUS), and/or peripheral neuropathy. [bib_ref] Update on diagnosis, risk-stratification, and management, Dispenzieri [/bib_ref] [bib_ref] The utility of plasma vascular endothelial growth factor levels in the diagnosis..., D'souza [/bib_ref] [bib_ref] Overproduction of vascular endothelial growth factor/vascular permeability factor is causative in Crow-Fukase..., Watanabe [/bib_ref] Optic disc edema (ODE) is considered the most common ocular manifestation in POEMS patients. In addition, findings of macular edema, retinal hemorrhages, retinal serous detachments, and increased choroidal thickness have also been reported so far. [bib_ref] Optic disk drusen, peripapillary choroidal neovascularization, and POEMS syndrome, Diduszyn [/bib_ref] [bib_ref] Optic disc edema, cystoid macular edema, and elevated vascular endothelial growth factor..., Chong [/bib_ref] [bib_ref] Ocular findings in patients with polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Kaushik [/bib_ref] [bib_ref] Ocular manifestations and treatment outcomes in Chinese patients with POEMS syndrome, Zhang [/bib_ref] [bib_ref] A case of POEMS syndrome with cystoid macular edema, Imai [/bib_ref] Most cases of ODE in POEMS tend be either asymptomatic or present with blurred vision. [bib_ref] Ocular findings in patients with polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Kaushik [/bib_ref] [bib_ref] Ocular manifestations and treatment outcomes in Chinese patients with POEMS syndrome, Zhang [/bib_ref] The term papilledema has been preferentially used in studies that describe the diagnostic features of POEMS, 2,9 however, whether or
## Not intracranial hypertension (ih) is related to the presence of ode in
POEMS has yet to be confirmed, especially since a co-morbidity of IH and ODE has not been consistently reported in POEMS patients. [bib_ref] POEMS syndrome: an unusual cause of bilateral optic disk swelling, Wong [/bib_ref] [bib_ref] Optic disc edema, cystoid macular edema, and elevated vascular endothelial growth factor..., Chong [/bib_ref] Literature exploring the relationship between ODE and increased serum VEGF levels have alternatively proposed vascular hyperpermeability due to angiogenesis as the mechanism behind ODE in POEMS, [bib_ref] Optic disc edema, cystoid macular edema, and elevated vascular endothelial growth factor..., Chong [/bib_ref] [bib_ref] Ocular findings in patients with polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Kaushik [/bib_ref] [bib_ref] Ocular manifestations and treatment outcomes in Chinese patients with POEMS syndrome, Zhang [/bib_ref] especially as edema and ascites often present in other parts of the body. [bib_ref] Update on diagnosis, risk-stratification, and management, Dispenzieri [/bib_ref] While the majority of existing POEMS case studies have corroborated this particular theory, other proposed possibilities include infiltrative orbitopathy 3 and ischemia. [bib_ref] POEMS syndrome, Dispenzieri [/bib_ref] [bib_ref] POEMS syndrome with venous sinus thrombosis and visual failure: a case report, Witoonpanich [/bib_ref] It is important to consider, however, that nearly all previous studies exploring the mechanism behind ODE in POEMS have specifically investigated ODE that presented bilaterally in the initial assessment. In two case series reports exploring this relationship, 10,11 the symmetry of the ODE at the time of the initial visit was not explicitly stated, and most of the ODE cases were generally described as bilateral. Only one individual case report thoroughly examines unilateral ODE. [bib_ref] Proptosis with orbital soft tissue and bone changes and unilateral papilloedema: unusual..., Gandhi [/bib_ref] Given that our patient exhibited bilateral ODE five weeks after the first visit [fig_ref] Figure 3: Optical coherence tomography [/fig_ref] , we believe this is the first clearly documented case of asymmetric ODE in POEMS, notably in a young woman additionally presenting with lacrimal gland and extra-ocular muscle enlargement. This is especially important as ODE can be an early sign of POEMS [bib_ref] Optic disk swelling with peripheral neuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin..., Bolling [/bib_ref] [bib_ref] Proptosis with orbital soft tissue and bone changes and unilateral papilloedema: unusual..., Gandhi [/bib_ref] and thus ophthalmologists should be aware that it can manifest asymmetrically.
The initially-unilateral presentation of ocular manifestations in this particular case may be due to multiple factors. However, the rare occurrence of asymmetric papilledema-especially in POEMS-, the systemic spread of vascular hyperpermeability, and the lack of evidence for local venous or arterial thromboses point to infiltrative orbitopathy as a potential driving force of ODE in our patient, which reflects the findings of the individual unilateral ODE case report. [bib_ref] Proptosis with orbital soft tissue and bone changes and unilateral papilloedema: unusual..., Gandhi [/bib_ref] This is further corroborated by the concurrent presence of extra-ocular muscle and lacrimal gland prominence and proptosis, which have been strongly linked to plasma cell infiltration in the context of multiple myeloma. [bib_ref] Multiple myeloma and its ocular manifestations, Knapp [/bib_ref] [bib_ref] Ocular manifestations of multiple myeloma: three cases and a review of the..., Chin [/bib_ref] Further studies should confirm if cellular infiltration could indeed relate to the development of asymmetric ODE in POEMS.
Regarding POEMS treatment, Melphalan200 conditioning and ASCT was deemed the appropriate treatment for this case. An alternative such as radiation is considered more suitable for patients with limited or more localized lesions, [bib_ref] Outcomes of patients with POEMS syndrome treated initially with radiation, Humeniuk [/bib_ref] and recent literature has demonstrated more reliable outcomes with ASCT relative to other systemic treatments such as proteasome inhibitor-based therapy or immunomodulatory imide drugs. [bib_ref] Update on diagnosis, risk-stratification, and management, Dispenzieri [/bib_ref] [bib_ref] Long-term outcome of patients with POEMS syndrome: an update of the Mayo..., Kourelis [/bib_ref] [bib_ref] Long-term outcomes after autologous stem cell transplantation for patients with POEMS syndrome..., D'souza [/bib_ref]
# Conclusions
In summary, this case illustrates the rare, multisystem disease of POEMS syndrome as an uncommon cause of asymmetric optic disc edema. While it is often diagnosed in older males, ophthalmologists should consider POEMS in any patient initially presenting with bilateral or unilateral optic disc edema in conjunction with multiple systemic symptoms.
## Patient consent
This report does not contain any personal identifying information. Patient consent is available.
# Funding
No funding or grant support was received for this work.
## Authorship
All authors attest that they meet the current ICMJE criteria for Authorship.
## Declaration of competing interest
Dr. Amer Assal wishes to disclose advisory board participation with Incyte Corporation and Boston Biomedical, as well as research funding received from Incyte Corporation.
The following authors have no financial disclosures: HM, JKR, DB, DCP, GM.
[fig] Figure 1: Visual fields of right and left eyes before and after treatment with autologous stem cell transplant. (A) At initial visit, visual field of right eye is normal and the left eye shows fluctuations nasally and in the blind spot. (B) One week after baseline assessment, visual field of right eye is normal and the left eye shows fluctuations in the blind spot. (C) 12 days post-treatment, visual field of right eye shows superior and central defects and the left eye shows an enlarged blind spot with extension into the superior field. (D) Seven months post-treatment, visual field of right eye is normal and the left eye shows a minimally enlarged blind spot. [/fig]
[fig] Figure 2: Fundus images of right and left eyes before and after treatment with autologous stem cell transplant. (A) Posterior poles of wide-field fundus photos at initial visit show a normal optic disc in the right eye and mild optic disc edema in the left eye. (B) Fundus photos one week after baseline assessment show no frank optic disc edema in the right eye and worsening optic disc edema in the left eye. (C) Fundus photos 12 days post-treatment show optic disc hemorrhage and worsening edema in both eyes. (D) Fundus photos seven months post-treatment show resolution of the optic disc edema in both eyes. [/fig]
[fig] Figure 3: Optical coherence tomography (OCT) findings before and after treatment with autologous stem cell transplant. (A) OCT at initial visit shows a normal retinal nerve fiber layer in the right eye and elevation in the left eye. (B) OCT one week after baseline assessment continues to show a normal retinal nerve fiber layer in the right eye and disc edema in the left eye. (C) OCT 12 days post-treatment shows elevation of the retinal nerve fiber layer in both eyes. (D) OCT seven months posttreatment shows a normal retinal nerve fiber layer in the right eye and possible retinal nerve fiber layer thinning in the left eye. [/fig]
[fig] Figure 4: Axial section of orbital-cerebral magnetic resonance imaging (MRI) at initial visit. Images show bilateral prominence of the (A; arrows) extra-ocular muscles and (B; arrows) lacrimal gland without optic nerve enhancement. [/fig]
[fig] Figure 5: Sacral lesion biopsy at one month after initial visit. (A; 10x) showing osteosclerotic bone and (B; 40x) sheets of mature-appearing plasma cells that are (C; 40x) positive for CD138 and (D; 40x) cytoplasmic lambda light chain restricted. [/fig]
[fig] Figure 6: Positron emission tomography (PET)/computed tomography (CT) scan at one month after initial visit featuring multiple spine and pelvic bone lytic lesions (orange circles). [/fig]
|
Multisource feedback analysis of pediatric outpatient teaching
Background: This study aims to evaluate the outpatient communication skills of medical students via multisource feedback, which may be useful to map future directions in improving physician-patient communication.Methods: Family respondents of patients, a nurse, a clinical teacher, and a research assistant evaluated video-recorded medical students' interactions with outpatients by using multisource feedback questionnaires; students also assessed their own skills. The questionnaire was answered based on the video-recorded interactions between outpatients and the medical students. Results: A total of 60 family respondents of the 60 patients completed the questionnaires, 58 (96.7%) of them agreed with the video recording. Two reasons for reluctance were "personal privacy" issues and "simply disagree" with the video recording. The average satisfaction score of the 58 students was 85.1 points, indicating students' performance was in the category between satisfied and very satisfied. The family respondents were most satisfied with the "teacher"s attitude," followed by "teaching quality". In contrast, the family respondents were least satisfied with "being open to questions". Among the 6 assessment domains of communication skills, the students scored highest on "explaining" and lowest on "giving recommendations". In the detailed assessment by family respondents, the students scored lowest on "asking about life/school burden". In the multisource analysis, the nurses' mean score was much higher and the students' mean self-assessment score was lower than the average scores on all domains. Conclusion: The willingness and satisfaction of family respondents were high in this study. Students scored the lowest on giving recommendations to patients. Multisource feedback with video recording is useful in providing more accurate evaluation of students' communication competence and in identifying the areas of communication that require enhancement.
# Background
Communication is an important component of patient care. The physician-patient interview is the key component of all health care, particularly of primary medical care [bib_ref] Physician-patient communication in the primary care office: a systematic review, Beck [/bib_ref] [bib_ref] Patient-doctor communication, Teutsch [/bib_ref]. Outpatient clinics offer trainees one of the most varied clinical experiences within the hospital setting, but they are often chaotic and over-stretched, with limited time for teaching [bib_ref] Teaching and learning in out-patient clinics, Williamson [/bib_ref]. When patients are informed and involved in decision making, they adhere to medical recommendations (e.g., vaccination and dietary modification) [bib_ref] Physician-patient communication in the primary care office: a systematic review, Beck [/bib_ref] [bib_ref] Clinical guides to preventing ethical conflicts between pregnant women and their physicians, Chervenak [/bib_ref]. Such joint decision making requires patients to be fully informed about alternatives and potential risks of treatment [bib_ref] Physician-patient communication in the primary care office: a systematic review, Beck [/bib_ref] [bib_ref] Patients' preferences for risk disclosure and role in decision making for invasive..., Mazur [/bib_ref]. Clinical practice within outpatient clinics can strengthen the collective knowledge of trainees [bib_ref] Teaching and learning in out-patient clinics, Williamson [/bib_ref]. This establishment of practice may also validate the role of trainees in the management of patients and facilitate social learning [bib_ref] Teaching and learning in out-patient clinics, Williamson [/bib_ref].
Assessment is an essential step in the curricular development process [bib_ref] Teaching skills for students: our future educators, Burgess [/bib_ref]. An evaluation method is important for the improvement of the quality of learning among medical students; however, such method is rare [bib_ref] Teaching skills for students: our future educators, Burgess [/bib_ref]. Concern about the inability of monitored examinations to assess the full spectrum of clinical competence, including humanistic quality, knowledge, and communication skills, stimulated the introduction of the "patient and peer assessment module" [bib_ref] The value of patient and peer ratings in recertification, Lipner [/bib_ref]. On the other hand, supervision features observation and sharing of clinical feedback, which can improve clinical performance [bib_ref] Teaching and learning in out-patient clinics, Williamson [/bib_ref]. Effective needs-assessment strategies include Multisource feedback (MSF) from educators and learners [bib_ref] Multi-method needs assessment optimises learning, Keister [/bib_ref]. Thus, the patient and peer assessment, and self-evaluation modules are better methods to evaluate the communication and clinical performance of medical students [bib_ref] The value of patient and peer ratings in recertification, Lipner [/bib_ref].
We aimed to evaluate the outpatient communication skills of medical students by using MSF from family respondents of patients, nurses, a clinical teacher, and a research assistant. This may be an effective evaluation method in the future to improve physician-patient communication skills in the outpatient setting.
# Methods
Settings: undergraduate students of Chang Gung Memorial Hospital (CGMH)
We employed a multi-respondent evaluation method using a structured paper questionnaire to investigate the communication skills of our students in the pediatric outpatient clinic. All seventh-year medical students trained at CGMH were enrolled in this study. There were 32 males and 28 females, with mean age of 25 years (24-27 y). MSF was obtained from a nurse, a clinical teacher, and a research assistant to assess the medical students' communication competence; a self-assessment of skills was also administered by the students. The satisfaction score of the medical students was evaluated by 4 respondent groups which included the family, research assistant, nurse and students. The same teacher and the same research assistant were involved in the whole study. The patients, 36 males and 24 females, were one month to 16 years old (median: 3.3 y). The patients who were suitable for this teaching clinic were classified by nurses before the patients went into the outpatient clinic. The study was approved by the Ethics and Clinical Research Committee of CGMH (No. 98-2202B).
## Instruments
Each outpatient interaction led by the medical students was video-recorded. The students, a nurse, a teacher, and a research assistant, watched the video together in a room at the same time and then filled out the corresponding paper questionnaire. The observers (students, nurse, teacher, research assistant) discussed the videotaped interactions after completing the assessments. The family respondents completed reasons for agreeing/disagreeing to the questionnaires before video recording and completed their assessments after the outpatient clinic at a room next to the clinic.
## Assessment and evaluation
The completed questionnaires were validated by 3 professional teachers. The questionnaire included items such as family respondent's reasons for agreeing/disagreeing with the video recording, multisource satisfaction with the outpatient interaction, and evaluation of skills in 6 domains (i.e., giving recommendations, listening, explaining, acknowledging, negotiation, and patient-centered communication). These skills were subdivided accordingly as shown in [fig_ref] Table 1: Subdivided outpatient skills evaluation by the different groups [/fig_ref] and [fig_ref] Figure 1 8: vs [/fig_ref] [bib_ref] Physician-patient communication in the primary care office: a systematic review, Beck [/bib_ref] [bib_ref] Using the objective structured clinical examinations in undergraduate midwifery students, Delavar [/bib_ref]. Every question can be answered with the following options and corresponding scores: very satisfied (100 points), satisfied (80 points), no opinion (60 points), dissatisfied (40 points), and very dissatisfied (20 points).
Our patients included children with respiratory tract infection, liver cirrhosis, and abdominal pain, and those requiring post-operative care. These patients were classified into difficult or common cases evaluated by students after the clinics. The liver cirrhosis patients included hepatitis or biliary atresia with routine followup and without complications.
The questionnaires in this study were tested with Cronbach's α for reliability. Statistical methods employed were descriptive statistics and student's t-test (one-sided) in comparison between difficult and common cases. Our hypothesis is that the student will have lower scores with difficult cases. Comparisons in teaching satisfaction and six domains between different respondent groups were analyzed by the ANOVA with Bonferroni's correction, when multiple comparisons were evaluated. The correlation in the overall satisfaction with number of questions asked was also analyzed by Pearson correlation test. A p-value < 0.05 was considered statistically significant. The statistical analyses were performed using the Statistical Package for Social Science (SPSS, version 12) software package.
# Results
## Reliability
Cronbach's α coefficient was used to assess reliability [bib_ref] Using the objective structured clinical examinations in undergraduate midwifery students, Delavar [/bib_ref] [bib_ref] Assessment of resident physicians in professionalism, interpersonal and communication skills: a multisource..., Qu [/bib_ref]. The Cronbach's α values for the questionnaires of the 4 respondent groups were as follows: 0.696 for the students, 0.974 for the nurses, 0.914 for the teacher and 0.918 for the research assistant. The Cronbachs' α values for the students' scores would be 0.857 had the item "asking about chief complaints" been deleted. The item "asking about chief complaints" is of low reliability for the students' assessing themselves in this study. The overall Cronbachs' α value in this study was 0.867.
## Reasons for agreeing/disagreeing with video recording
A total of 60 family respondents of the 60 patients completed the questionnaires, 58 (96.7%) of them agreed with the video recording and is the source of data included for statistical analysis. Two reasons for reluctance were "personal privacy" issues and they "simply disagree" with video recording. Among the family respondents who agreed with the video recording, the following reasons were obtained: to contribute to medical education (41.8%), to get a more accurate diagnosis (13.9%), and to cooperate on the study (12.7%) [fig_ref] Figure 1 8: vs [/fig_ref].
## Satisfaction
The average satisfaction score of the 58 medical students was 85.1 points (references 60 points), indicating students' performance was in the category between satisfied and very satisfied, which was derived from the following group scores: 88.9 points (family), 85.6 (research assistant), 84.0 (nurses), 82.2 (students) . All participants were most satisfied with the "teacher's attitude" (92.9 points) and "teaching quality" (88.4 points). The family respondents were likewise most satisfied with the" "teacher's attitude" (90.7 points), followed by the "clinic quality" (90.7 points); students were most satisfied with the "teacher's attitude" (97.5 points), followed by "being open to questions" (92.5 points). In contrast, the family respondents were least satisfied with "being open to questions" (86.3 points), while the students were least satisfied with the "student's attitude" (74.7 points) . We discarded the students' self-ratings as input in the t-test for differences between difficult and common cases. There was no significant difference in the overall satisfaction by the other 3 groups' (family respondents, nurses, research assistant) evaluation between difficult cases and common cases (82.4 ± 13.8 vs. 85.6 ± 13.3, t = 1.481, degrees of freedom (df ) = 172, P = 0.070). Lower scores were obtained in the difficult cases than in the common cases with regard to "student's attitude" (80.0 ± 14. Teaching satisfaction. Average satisfaction score was 85.1 points. Students scored highest on "teacher's attitude," followed by "teaching quality." Family respondents were most satisfied with "teacher's attitude" followed by "teaching quality"; students were most satisfied with "teacher's attitude" and "being open to questions." The family respondents of patients were least satisfied with "being open to questions," whereas students were least satisfied with "student's attitude." The results represented were mean ± SD. * P < 0.05 compared to the average scores.
the overall satisfaction with the number of questions asked (r = 0.021, P = 0.898).
## Evaluation of outpatient skill in 6 domains
Outpatient skills were assessed in 6 domains. Overall, students obtained the highest score on "explaining" (82.6 points) and the lowest score on "giving recommendations" (73.6 points).
The teacher assessed the students to have the highest score on the "explaining" outpatient skill (83.8 points), while the lowest was on "giving recommendations" (71.8 points). In the self-assessment of students, they had the highest score on the "explaining" skill (80.2 points) and the lowest on "providing recommendations" (72.5 points) [fig_ref] Figure 3: Outpatient skills in six assessment domains [/fig_ref]. The teacher's and the research assistant's scores were comparable to the 4 groups' average score. The nurses' score was the highest and the students' selfassessed score was lower than the average score of the 4 groups [fig_ref] Table 2: The total scores of outpatient skill in 6 domains between the different... [/fig_ref].
## Subdivided outpatient skills evaluation by the different groups
In the subdivided outpatient teaching skills, the students performed best in "asking patients about their chief complaints" (90.5 points), while worst in "asking about school/life burden" (65.6 points). Teachers believed that the students were best in "consensus decision maker" (94.5 points), followed by "giving self-introduction" (90.0 points). In contrast, they believed that the students were worst in "asking about school/life burden" (52.3 points), as did the students (66.4 points), nurses (62.3 points), and the research assistant (74.0 points). The second worst skill of the students was "Get a consensus with patient in different views" (average 69.2 points) [fig_ref] Table 1: Subdivided outpatient skills evaluation by the different groups [/fig_ref].
# Discussion
The agreement and satisfaction rate was high in this pediatric outpatient teaching evaluation. Different respondents will give much different scores in the evaluation. We noticed that the four respondents assessments could measure out more accurately the strengths and weaknesses of the students outpatient skills than the single one respondent assessment.
Why were their differences in satisfaction and skill scores between assessor groups? Were senior doctors more critical and strict in their assessments [bib_ref] An analysis of multisource feedback within the foundation programme, Ellul [/bib_ref] [bib_ref] Use of a multisource feedback tool to develop pharmacists in a postgraduate..., Davies [/bib_ref] [bib_ref] Self-other agreement in multisource feedback: the influence of doctor and rater group..., Roberts [/bib_ref] ?. In our study, we found that students gave more strict scores to themselves than the rest of the groups. Did the nurses give more favorable assessments because they understand the trainees better and thus make allowances for weakness [bib_ref] An analysis of multisource feedback within the foundation programme, Ellul [/bib_ref] [bib_ref] Self-other agreement in multisource feedback: the influence of doctor and rater group..., Roberts [/bib_ref] ?. This favorable bias could be found in the nurses' scores in our study, but not in the research assistant and the teacher. Alternatively, were their assessments more reliable because they know the students better [bib_ref] An analysis of multisource feedback within the foundation programme, Ellul [/bib_ref] ?. We found that the teacher's and the research assistant's scores were comparable to that of the 4 groups' average score in [fig_ref] Table 2: The total scores of outpatient skill in 6 domains between the different... [/fig_ref]. The nurses' mean score was much higher and the students' mean self-evaluation score was much lower than the average score. We believe that MSF containing four respondents assessments used in this study can give a more accurate assessment to distinguish students' strong and weak points [bib_ref] Multisource feedback for residents: how high must the stakes be?, Cate [/bib_ref] [bib_ref] Feasibility and reliability of a multisource feedback tool for emergency medicine residents, Garra [/bib_ref]. The quality of the results may also be influenced by personal relationships, stakes and equivalence [bib_ref] The value of patient and peer ratings in recertification, Lipner [/bib_ref]. Students' own scores tended to be the lowest in our study. This is common in our teaching environment, and is probably due to the named questionnaires. Physician-patient communication has frequently been judged to be inadequate and imperfect [bib_ref] Interview with eve j, Higginbotham [/bib_ref]. It is thus important that communication skills of physicians be assessed periodically with a confidential peer evaluation survey [bib_ref] The value of patient and peer ratings in recertification, Lipner [/bib_ref] , as such surveys may provide unbiased and factual information from respondents. However, in our study, preserving the anonymity and confidentiality of students and patients was difficult to achieve. First, selecting patients who were willing to participate in the study was not easy; second, the students who participated all completed the peer assessment in one outpatient room. However, this may be improved in the future by using a one-way mirror room or by ensuring that students complete the questionnaires in separate rooms.
Published findings showed that the number of questions asked about a patient's illness was inversely related to patient satisfaction [bib_ref] Communication patterns of primary care physicians, Roter [/bib_ref]. However, this inverse relation was not obtained in our study. We found that students assigned to difficult cases scored lower on "student's attitude" and "being open to questions." Counseling about unhealthy or risky behaviors is an important communication skill that should be a vital part of health care visits [bib_ref] Patient-doctor communication, Teutsch [/bib_ref]. Physicians' attitudes towards the physicianpatient relationship may contribute to the diagnostic value of the patient history [bib_ref] The diagnostic value of the medical history. Perceptions of internal medicine physicians, Rich [/bib_ref]. To this end, we should further educate the students about the appropriate attitude and approach in answering questions of patients with diseases that are difficult to treat. This may be an important future direction of research on physicianpatient communication.
A patient-centered medical interview is essential to create good interpersonal relationships and information exchange; it may also contribute to the diagnostic value of the patient history and facilitate informed decision making [bib_ref] The diagnostic value of the medical history. Perceptions of internal medicine physicians, Rich [/bib_ref] [bib_ref] Effective physician-patient communication and health outcomes: a review, Stewart [/bib_ref]. Medical educators should focus on teaching and reinforcing behaviors that are known to enhance favorable patient outcomes and satisfaction [bib_ref] Physician-patient communication in the primary care office: a systematic review, Beck [/bib_ref]. Patient health outcomes can be improved with good physician-patient communication [bib_ref] Effective physician-patient communication and health outcomes: a review, Stewart [/bib_ref]. As Aspergren noted, communication skills can be taught in courses but are easily forgotten if not maintained by practice [bib_ref] Teaching and learning communication skills in medicine-a review with quality grading of..., Aspegren [/bib_ref]. We believe that this pediatric-patient communication project will be helpful for the development and promotion of clinical skills of medical students in an outpatient practice. One limitation of this study was that this was a single site study using a specifically designed instrument and therefore might not be generalised. Therefore, these findings need to be tested in a large-scale study.
# Conclusions
MSF with video-recorded is important in providing a more accurate evaluation of students' communication competence and in identifying the areas of communication that require enhancement.
Abbreviations CGMH: Chang Gung Memorial Hospital; MSF: Multisource feedback.
[fig] Figure 1 8: vs. 87.2 ± 13.5, t = 3.289, df = 172, P = 0.001), and "being open to questions" (79.3 ± 15.2 vs. 86.8 ± 12.8, t = 3.451, df = 172, P = 0.001). The number of questions asked was about 10-15 in each case. There was no significant correlation in Agree reasons. Patients agreed with the video recording to contribute to medical education, to get a more accurate diagnosis, and to cooperate on the study. [/fig]
[fig] Figure 3: Outpatient skills in six assessment domains. Students scored highest on "explaining" and lowest on "giving recommendations" as in the average scores. The results represented were mean ± SD. * P < 0.05 compared to the average scores. [/fig]
[table] Table 1: Subdivided outpatient skills evaluation by the different groups [/table]
[table] Table 2: The total scores of outpatient skill in 6 domains between the different respondents [/table]
|
Interleukin-23 Facilitates Thyroid Cancer Cell Migration and Invasion by Inhibiting SOCS4 Expression via MicroRNA-25
Interleukin-23 (IL-23) is a conventional proinflammatory cytokine that plays a role in tumor progression by inducing inflammation in the tumor microenvironment. However, the role of IL-23 in thyroid cancer migration and invasion remains unclear. In the present study, we observed that the treatment with IL-23, induced migration and invasion in human thyroid cancer cells. Additional data demonstrate that SOCS4 negatively regulates IL-23-mediated migration and invasion. On investigating the mechanisms involved in IL-23 mediated migration and invasion, we observed that miR-25 promotes the migration and invasion of thyroid cancer cells by directly binding to the 3 0 -UTR of SOCS4 that leads to the inhibition of SOCS4. In addition, we also demonstrated that IL-23 increases miR-25 expression levels, and overexpressed miR-25 is involved in IL-23-associated SOCS4 inhibition and cell migration and invasion. Together, our data suggest that IL-23 induces migration and invasion in thyroid cancer cells by mediating the miR-25/SOCS4 signaling pathway.
# Introduction
Thyroid cancer is a common type of endocrine malignancy that has shown a rapid increase in worldwide incidence during the past few decades [bib_ref] Increasing incidence of differentiated thyroid cancer in the United States, Chen [/bib_ref]. Despite improvements in therapeutic strategies, some patients are difficult to treat and develop invasion and metastasis [bib_ref] Increasing incidence of thyroid cancer in the United States, Davies [/bib_ref]. Therefore, it is essential to identify the molecular mechanisms underlying thyroid cancer invasion and metastasis. Recent reports demonstrated that inflammation is a strong promoter of carcinogenesis and malignancy in many forms of cancer [bib_ref] Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion, Schreiber [/bib_ref] [bib_ref] Crosstalk between cancer and immune cells: role of STAT3 in the tumour..., Yu [/bib_ref]. Inflammation seems to be an important mediator for the development of cancer and provides the cancer cells a hospitable microenvironment [bib_ref] Chemokine/chemokine receptor interactions contribute to the accumulation of Th17 cells in patients..., Chen [/bib_ref]. Interleukin-23 (IL-23), a heterodimeric type 1 cytokine composed of the IL-12/p40 subunit and the specific p19 subunit, belongs to the interleukin-6 superfamily [bib_ref] Novel p19 protein engages IL-12p40 to form a cytokine, IL-23, with biological..., Oppmann [/bib_ref]. Previous studies have shown that IL-23 is associated with carcinogenesis as well as inflammation. High levels of IL-23 were found in human hepatocellular carcinoma, colorectal carcinoma, squamous carcinoma, and esophageal carcinoma [bib_ref] Interleukin-23 promotes the epithelial-mesenchymal transition of oesophageal carcinoma cells via the Wnt/beta-catenin..., Chen [/bib_ref] [bib_ref] Astrocytes facilitate melanoma brain metastasis via secretion of IL-23, Klein [/bib_ref] [bib_ref] Interleukin 23 promotes hepatocellular carcinoma metastasis via NF-kappa B induced matrix metalloproteinase..., Li [/bib_ref] [bib_ref] IL-23 directly enhances the proliferative and invasive activities of colorectal carcinoma, Suzuki [/bib_ref]. Evidence suggests that IL-23 overexpression can induce metastasis in colorectal, lung, and oral cancer [bib_ref] Interleukin-23 promotes the epithelial-mesenchymal transition of oesophageal carcinoma cells via the Wnt/beta-catenin..., Chen [/bib_ref] [bib_ref] Astrocytes facilitate melanoma brain metastasis via secretion of IL-23, Klein [/bib_ref] [bib_ref] Interleukin 23 promotes hepatocellular carcinoma metastasis via NF-kappa B induced matrix metalloproteinase..., Li [/bib_ref] [bib_ref] IL-23 directly enhances the proliferative and invasive activities of colorectal carcinoma, Suzuki [/bib_ref].
The suppressors of cytokine signaling (SOCS) are important negative feedback regulators of cytokine signaling [bib_ref] Negative regulators of cytokine signal transduction, Hilton [/bib_ref]. The SOCS proteins are a family of 8 proteins (SOCS1-7 and a cytokine-inducible SH2-containing protein or CIS). Each SOCS protein contains a central SH2 domain that interacts with phosphorylated tyrosines [bib_ref] SOCS/CIS protein inhibition of growth hormone-stimulated STAT5 signaling by multiple mechanisms, Ram [/bib_ref]. SOCS proteins have been recently investigated for their role in the development of different cancers [bib_ref] Aberrant methylation of SOCS-1 was observed in younger colorectal cancer patients, Fujitake [/bib_ref] [bib_ref] Constitutional activation of IL-6-mediated JAK/STAT pathway through hypermethylation of SOCS-1 in human..., To [/bib_ref] [bib_ref] SOCS-3 is frequently silenced by hypermethylation and suppresses cell growth in human..., He [/bib_ref] [bib_ref] SOCS-3 is frequently methylated in head and neck squamous cell carcinoma and..., Weber [/bib_ref] [bib_ref] SOCS6, down-regulated in gastric cancer, inhibits cell proliferation and colony formation, Lai [/bib_ref] [bib_ref] Prognostic and predictive relevance of DNAM-1, SOCS6 and CADH-7 genes on chromosome..., Storojeva [/bib_ref]. However, little is known about the role of SOCS4 in carcinoma, and their possible influence on tumor growth and malignancy.
MicroRNAs (miRNAs) are a species of small noncoding single stranded RNAs that play an important role in the development of different cancers by binding the 3 0 -untranslated region (3 0 -UTR) of targeted genes [bib_ref] Conserved seed pairing, often flanked by adenosines, indicates that thousands of human..., Lewis [/bib_ref]. Aberrant miRNA expression has also been frequently reported in numerous tumors [bib_ref] MicroRNAs: target recognition and regulatory functions, Bartel [/bib_ref]. In recent years, multiple evidence point to a role for miRNAs in tumor cell biological processes, including cell proliferation, differentiation, migration, and invasion [bib_ref] ) microRNAs as oncogenes and tumor suppressors, Zhang [/bib_ref]. MicroRNA-25 belongs to the miR-106b-25 cluster that includes miR-106b, miR-93, and miR-25. MicroRNA-25 has been reported to be aberrantly overexpressed in several tumors, such as ovarian cancer, lung cancer, gastric cancer, and colorectal cancer [bib_ref] Emerging role of miR-106b-25/miR-17-92 clusters in the control of transforming growth factor..., Petrocca [/bib_ref] [bib_ref] MiR-25 promotes ovarian cancer proliferation and motility by targeting LATS2, Feng [/bib_ref] [bib_ref] MicroRNA-25 functions as a potential tumor suppressor in colon cancer by targeting..., Li [/bib_ref] [bib_ref] Downregulation of miR-25 modulates non-small cell lung cancer cells by targeting CDC42, Yang [/bib_ref] [bib_ref] MiR-25 promotes gastric cancer cells growth and motility by targeting RECK, Zhao [/bib_ref]. Although the expression of miR-25 in different tumors has been described, a clear role for miR-25 in thyroid carcinoma remains unclear.
In this study, we demonstrate that IL-23 promotes thyroid cancer cell migration and invasion. We further demonstrate that IL-23 regulates the migration and invasion of thyroid cancer cells via a miR-25/SOCS4 signaling pathway.
# Materials and methods
# Ethics statement
All participants gave written informed consent to participate in the study. The study was conducted according to the principles of the Declaration of Helsinki and approved by the Institutional Review Board of the Remin Hospital of Wuhan University, in accordance with its guidelines for the protection of human subjects.
## Samples and cases
Thyroid tissues were collected at the Remin Hospital of Wuhan University from February 2010 to February 2014. Tissue samples were cut into two parts, one was reviewed by two expert pathologists to verify the histologic diagnosis, the other immediately snap-frozen in liquid nitrogen, and stored in liquid nitrogen until RNA extraction. None of the patients had received any preoperative treatment. Tumors were staged according to the American Joint Committee on Cancer (AJCC) pathologic tumor-node-metastasis (TNM) classiication. The characteristics of patients are described in S1 and S2 Tables.
## Cell culture
Human thyroid cancer cell lines K1 (papillary) and WRO (follicular) were cultured in Dulbecco's modified Eagle's medium (DMEM) (Gibco BRL, Grand Island, NY) supplemented with 10% fetal bovine serum (Gibco BRL), 100 units/ml penicillin, and 100 μg/ml streptomycin sulfate. Cells were maintained at 37°C in a 5% CO 2 incubator. All these cell lines were original purchased from cell bank of the Chinese Academy of Science, Shanghai.
## Reagents
Recombinant human IL-23 (rhIL-23) was purchased from R&D Systems (Minneapolis, MN). Monoclonal antibodies (Abs) against human SOCS4 and GAPDH were purchased from Sigma (St Louis, MO). Chemically synthesized miRNA mimics and miRNA inhibitors were purchased from Ambion (Austin, TX, USA). TRIzol, Lipfectamine-2000, Enzyme MIX were purchased from Invitrogen (Basel, Switzerland).
## Quantitative real time pcr
Quantitative real time PCR analysis was performed to determine mature miRNA and mRNA levels. For quantitative mature microRNAs detection, Total miRNAs were isolated using a mir-Vana miRNA isolation kit (Ambion), according to the manufacturer's instructions. Total RNA (2 μg) was reversetranscribed with Bulge-Loop mature miRNA-specific reverse transcription primers (Ambion) and Moloney murine leukemia virus reverse transcriptase (Promega). Quantitative Real-time PCR-based quantification of miRNAs was performed using the miRNA analysis kits (Ambion), according to the manufacturer's instructions. The levels of miRNAs were normalized to those of the internal control U6 snRNA. To detect cellular mRNAs, total RNA was isolated using TRIzol (Invitrogen, Basel, Switzerland). Cellular RNA samples were reverse-transcribed using random primers. Real-time PCR was performed using a LightCycler 480 (Roche) and the SYBR Green system (Applied Biosystems). GAPDH was amplified as an internal control. Primers used this study are listed in S3 Table and the SYBR green products verified by sequencing.
## Transwell assay
Cell migration and invasion assay were performed as described previously [bib_ref] Interleukin 17A promotes hepatocellular carcinoma metastasis via NF-kB induced matrix metalloproteinases 2..., Li [/bib_ref]. Briefly, cells were treated either with 50 ng/mL rhIL-23 or BSA buffer control for described time and dose and observed accordingly. The mean number of migrating and invading cells was expressed as a percentage relative to the control, which was designated as 100%.
## Wound closure assay
A wound was introduced on the confluent monolayer cells using a micropipette tip. Photographs were taken at 40 X magnification using phase-contrast microscopy immediately after wound incision and at selected timepoints. Wound closure was measured by calculating pixel densities in the wound area by Cella software (Olympus Biosystem Gmb, Hamburg, Germany) and expressed as percentage of wound closure of triplicate areas ± SD.
## Mtt assay
The 3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used in the evaluation of cells proliferation. Cells were seeded into 96-well plates at 5×10 3 cells/well. Twenty-four hours later, MTT assay was conducted. Finally, the optical density was determined at 570nm using the ELISA plate reader (Model 550; Bio-Rad). At least three independent experiments were ensured.
## Transfection and luciferase reporter assay
Cells were seeded on 24-well or 6-well dishes, depending on the experiment, and were grown to the confluence reaching approximately 80-90% at the time of transfection. Cells were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the protocol provided by the manufacturer. A Renilla luciferase reporter vector pRL-TK was used as internal control.
Luciferase assays were performed with a dual-specific luciferse assay kit (Promega, Madison, WI). Firefly luciferase activities were normalized on the basis of Renilla luciferase activities. All reporter assays were repeated for at least three times. Data shown were average values ± SD from one representative experiment.
## Rna interference
SOCS4 shRNA and irrelevant shRNA control (shRNA-control) were purchased from Gene-Pharma (GenePharma, Shanghai) and prepared by ligation of the corresponding pairs of oligonucleotides to PGPU6/GFP/Neo. The target sequence can be found in S4 [fig_ref] Table: Correlation of IL-23, miR-25 and SOCS4 expression with clinicopathologic features in papillary... [/fig_ref] Western blot analysis Whole-cell lysates were prepared by lysing cells with PBS pH 7.4 containing 0.01% Triton X-100, 0.01% EDTA, and 10% protease inhibitor cocktail (Roche, Applied Science). Protein concentration was determined by the Bradford assay (Bio-Rad). The polypeptides from cell lysates were separated on SDS 12% polyacrylamide gels cross-linked with N,N -methylenebisacylamide, and transferred electrically to nitrocellulose membranes (Millipore, Billerica, MA). Nonspecific binding was blocked with 5% milk in PBST before adding primary antibodies used in this study. Horseradish peroxidase-linked anti-rabbit and anti-mouse antibodies (Sigma, St Louis, MO) were used as secondary antibodies. Protein levels were quantified by scanning blots on a Gel Doc EZ imager (Bio-Rad) and analysis with Quantity One 1D image analysis software 4.4.0 (Bio-Rad)
# Statistical analysis
Statistical analyses were performed using the GraphPad Prism 5 software (GraphPad Software, La Jolla, CA, USA). Parametric and nonparametric data were analyzed using a two-tailed t-test and the Mann-Whitney U test respectively. A value of P < 0.05 was considered statistically significant. Data are presented as mean±S.D. or mean±S.E.M.
# Results
## Il-23 promotes the migration and invasion of thyroid cancer cells
We wanted to test the effect of IL-23 on the cell proliferation of thyroid cancer cells in order to observe if cell proliferation disturbs the migration and invasion capacity of the cells. The MTT assay showed that the proliferation of K1 cells and WRO cells were hardly affected by any dose of IL-23 (S1 Fig). We next investigated whether IL-23 could affect the migration and the invasion of thyroid cancer cells. Enhanced movement and invasion of K1 cells were detected in the presence of 50 and 100 ng/ml of IL-23 . Similarly, increased migration and invasion were also detected as early as 48 hours (h) following treatment with 50 ng/ml IL-23 protein . We also demonstrated that IL-23 increased the migration and invasion of WRO cells in a dose-and time-dependent manner (S2 In addition, wound healing assay also show that IL-23 increased the migration of K1 cells . Together, these results confirm the dose-and time-dependent promigratory and proinvasive effect of IL-23.
## Il-23 induces the migration and invasion of thyroid cancer cells through socs4
To identify the role of IL-23 in the expression of SOCS4, K1 cells were stimulated with IL-23 protein for 24 h at the concentration of 50 ng/ml. The suppressors of cytokine signaling (SOCS4) mRNA and protein were detected by real-time PCR and western blot, respectively. Results showed that IL-23 suppressed SOCS4 mRNA and protein expression in K1 cells IL-23 promotes K1 cell migration and invasion. (A) K1 cells were treated with rhIL-23 for 48 h at the indicated concentrations, followed by culturing in a transwell system for 24 hours. (B) Cell invasion assay experiments were performed with K1 cells that were treated as in (A). (C) K1 cells were treated with a concentration of 50 ng/ml rhIL-23 for the indicated time points, followed by culturing in a transwell system for 24 h. (D) The experiments were performed as in (C) for the cell invasion assay. (E) K1 cells were treated with rhIL-23 for 48 h at the indicated concentrations, followed by introducing a wound. Cell migration into the wound was monitored at 24 hours. (F) K1 cells were treated with a concentration of 50 ng/ml rhIL-23 for the indicated time points, followed by introducing a wound. Cell migration into the wound was monitored at 24 hours. Wound closure was measured by calculating pixel densities in the wound 2A). Similar results were also obtained in the WRO cell line [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. To determine whether SOCS4 participates in the IL-23-mediated migration and invasion of thyroid cancer cells, we constructed 4 human SOCS4-specific short hairpin RNA (shRNA). Real-time PCR results showed that the #2 shRNA plasmids could markedly inhibit the expression of SOCS4 in K1 cells, whereas the other shRNA plasmids had little effect on the expression of SOCS4 [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. SOCS4 protein levels showed a similar trend as determined by Western blot [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. In cell migration experiments, the overexpression of SOCS4 inhibited IL-23-induced migration in K1 cells [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Conversely, the knockdown of SOCS4 expression enhanced IL-23-induced migration in K1 cells [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. The degrees of induction were correlated with the efficiencies of SOCS4 knockdown by each shRNA plasmid [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Similar results were obtained with transwell invasion experiments [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Since SOCS4 negatively regulates migration, we next analyzed the role of SOCS4 in the cell proliferation of thyroid cancer cells. Results from MTT assay indicated that the proliferation of K1 cells were hardly affected by the expression of SOCS4. In addition, overexpression or Knockdown of SOCS4 and treatment with IL-23 could not affect the cell proliferation of thyroid cancer cells (S3 These results suggest that the activation of IL-23-regulated migration and invasion may require SOCS4.
## Microrna-25 downregulates socs4 expression by directly targeting its 3 0 -utr
To analyze the miRNAs that may target the 3 0 -UTR of SOCS4, we used 3 online databases, Tar-getScanHuman, miRDB, and miRWalk2.0, to search for potential candidates. Fourteen of the common potential miRNAs were found in the 3 databases. To determine the effect of the predicted miRNAs on the expression of SOCS4, the SOCS4 3 0 -UTR was cloned into a firefly luciferase reporter plasmid. The luciferase activity assays indicated that 6 miRNAs inhibited the SOCS4 luciferase activity below 50% as compared with the miR-Ctrl (S4A We next investigated which miRNAs were activated by IL-23-induced migration and invasion. K1 cells were treated with IL-23 protein at 50 ng/ml for 48 h. Results from real-time PCR demonstrated that among the miRNAs that could attenuate SOCS4 luciferase activity, the expression of miR-25 was significantly induced (S4B .
From our data we hypothesized that miR-25 may play a role in the IL-23/SOCS4 signaling pathway. We then investigated whether miR-25 regulates SOCS4 expression by post-transcriptional targeting of its 3 0 -UTR. A distinct mutation was generated in the 3 0 -UTR of SOCS4 at predicted seed-matching sites to test the interaction between miR-25 and SOCS4 3 0 -UTR [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. Transient transfection of K1 cells with wild type (WT) SOCS4 3 0 -UTR, and the miR-25 mimic, leads to a significant decrease in the reporter activity compared with that of the miR-control [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. This phenomenon was disrupted when the same cell lines were transfected with SOCS4 3 0 -UTR mutants [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. The miR-25 inhibitor (miR-25-Inh) significantly stimulated the luciferase activity of the WT SOCS4 3 0 -UTR, without any effect on Mut SOCS4 3 0 -UTR in K1 cells [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. MicroRNA-25 overexpression in K1 cells significantly suppressed both the mRNA and protein levels of SOCS4 [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. Conversely, miR-25 inhibitor-mediated knockdown of endogenous miR-25 increased SOCS4 mRNA and protein expression in K1 cells [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. To determine whether the miR-25-mediated downregulation of SOCS4 expression is a common feature in thyroid cancer cells, similar experiments were performed in WRO cells [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. Together, these results suggest that SOCS4 is a target of miR-25 in K1 cells and WRO cells. area and expressed as percentage of wound closure of triplicate areas ± standard deviations In the transwell migration, invasion and wound healing assay, the data are expressed as a percentage of the control. Data represent mean ± SD, n = 3 (**P < 0.01; *P < 0.05). Since miR-25 can suppress SOCS4 expression by sequence-specific binding to its 3 0 -UTR, we hypothesized that miR-25 may affect the migration and invasion of thyroid cancer via SOCS4. To test the hypothesis, K1 cells were co-transfected with the SOCS4 expression vector (or empty vector) and the miR-25 mimics (or miR-Ctrl). Transwell migration assays demonstrate that miR-25 promotes the migration of K1 cells and cell migration induced by miR-25 is reversed by SOCS4 overexpression [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. In addition, we observed that the knockdown of SOCS4 may significantly enhance the effect of miR-25 on cell migration [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. Similar results were obtained in the invasion assays and wound healing assay [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. The effect of the miR-25/SOCS4 signaling pathway on the migration of thyroid cancer cells was further evaluated using the miR-25 inhibitor. As shown in [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref] inhibitor significantly inhibits the migration of K1 cells. In addition, the miR-25 inhibitor and the SOCS4 expression vector synergistically inhibit the migration of K1 cells [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. Conversely, high levels of migration were present in K1 cells when SOCS4 was knocked down by shRNA-SOCS4#2 [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. Similar results were also obtained in invasion assays and wound healing assay [fig_ref] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid... [/fig_ref]. These results suggest that the inhibition of SOCS4 expression is responsible for the ability of miR-25 to promote cell invasion and migration.
## Il-23 stimulates mir-25 expression in thyroid cancer cells
To identify the role of IL-23 in the expression of miR-25, K1 cells were stimulated with human IL-23 protein at different concentrations for 48 h. MicroRNA-25 levels were detected by realtime PCR. Results show that miR-25 levels are upregulated by the IL-23 protein in a dosedependent manner [fig_ref] Fig 5: IL-23 stimulates miR-25 expression [/fig_ref]. Consistently, K1 cells were treated with IL-23 protein at different time points, at a concentration of 50 ng/ml. Results from real-time PCR analyses demonstrate that miR-25 levels increase as the time increases [fig_ref] Fig 5: IL-23 stimulates miR-25 expression [/fig_ref]. The role of IL-23 on miR-25 expression was confirmed by repeating the experiments using WRO cells [fig_ref] Fig 5: IL-23 stimulates miR-25 expression [/fig_ref]. These results suggest that IL-23 activates miR-25 expression.
## Microrna-25 plays an important role in il-23-mediated socs4 inhibition and cell migration and invasion
To define the role of miR-25 in the downregulation of IL-23-mediated SOCS4 expression, K1 cells were transfected with either miR-25 mimics or miR-Ctrl and treated with or without IL-23 protein for 48 h. Results from real-time PCR and western blot analyses show that transfection with miR-25 mimics increases IL-23-mediated inhibition of SOCS4 mRNA and protein expression [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref]. In contrast, high levels of SOCS4 mRNA and protein are present in K1 cells, when the expression of miR-25 is inhibited by the miR-25 inhibitor [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref]. We also examined whether miR-25 is involved in IL-23-mediated thyroid cancer cell line motility. Using transwell migration and invasion assays, we showed that miR-25 overexpression stimulates IL-23-mediated activation of migration and invasion [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref] , and knockdown of miR-25 expression inhibits IL-23-mediated activation of migration and invasion [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref]. Similar results were also obtained in wound healing assay [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref].Taken together, these data suggest that miR-25 is the key component involved in IL-23-mediated thyroid cancer cell migration and invasion through the inhibition of SOCS4 expression.
# Discussion
In this study, we defined a novel signaling pathway implicated in the control of thyroid cancer cell migration and invasion. First, we demonstrated that IL-23 upregulated miR-25 expression as well as downregulated SOCS4 expression in thyroid cancer cell migration and invasion. Further, we also showed that miR-25 promotes thyroid cancer cell migration and invasion by targeting SOCS4. Finally, our data suggest that miR-25 is involved in IL-23-associated SOCS4 expression and cell migration and invasion.
Tumor cell migration and invasion is a very complicated process in which cancer cells spread from the primary tumor, survive in the circulation, and grow in distant locations in the body [bib_ref] In vivo monitoring the process of tumor growth, metastasis and bacterial infection..., Xiong [/bib_ref]. Each process is determined by the migration and invasion ability of the tumor cells and the local tumor microenvironment that provide a favorable environment for tumor cells to survive and metastasize [bib_ref] Tumor microenvironment: a main actor in the metastasis process, Spano [/bib_ref]. Recent investigations have demonstrated that high expression of levels of IL-23, which can be detected in the microenvironment, could help to facilitate tumor metastasis. For example, IL-23 promotes hepatocellular carcinoma metastasis by NF-κB-upregulated MMP9 expression [bib_ref] Interleukin 23 promotes hepatocellular carcinoma metastasis via NF-kappa B induced matrix metalloproteinase..., Li [/bib_ref]. IL-23 is highly expressed in metastases-associated astrocytes, and IL-23 induces the progression of melanoma brain metastasis [bib_ref] Astrocytes facilitate melanoma brain metastasis via secretion of IL-23, Klein [/bib_ref]. IL-23 plays a pivotal role in the development of esophageal cancer via an epithelial-mesenchymal transition [bib_ref] Interleukin-23 promotes the epithelial-mesenchymal transition of oesophageal carcinoma cells via the Wnt/beta-catenin..., Chen [/bib_ref]. IL-23 can enhance the proliferation and invasion of colorectal carcinoma cells [bib_ref] IL-23 directly enhances the proliferative and invasive activities of colorectal carcinoma, Suzuki [/bib_ref]. However, the role of IL-23 in thyroid cancer cell migration and invasion is still unknown. To our knowledge, this is the first study to demonstrate the direct effects of IL-23 on the migration and invasion of thyroid cancer cells. Interestingly, a recent study demonstrated that IL-23 regulates the proliferation of lung cancer cells [bib_ref] Interleukin 23 regulates proliferation of lung cancer cells in a concentration-dependent way..., Li [/bib_ref]. Nonetheless, unlike lung cancer cells, we did not find evidence to support that IL-23 induces the proliferation of thyroid cancer cells (S1 We speculate that there may be some intrinsic differences between thyroid cancer and lung cancer. On the other hand, IL-23 promotes the migration and invasion of thyroid cancer cells.
Currently, two reports have implied the potential role of SOCS4 in cancer. One study used a double combination array analysis to prove that SOCS4 is a novel gastric cancer suppressor gene [bib_ref] Suppressor of cytokine signaling 4 detected as a novel gastric cancer suppressor..., Kobayashi [/bib_ref]. The other study compared the expression differences of SOCS1-7 between breast cancer tissue and background breast tissue [bib_ref] Higher expression levels of SOCS 1,3,4,7 are associated with earlier tumour stage..., Sasi [/bib_ref]. High expression of SOCS4 is significantly associated with an earlier tumor stage and a better clinical outcome in human breast cancer [bib_ref] Higher expression levels of SOCS 1,3,4,7 are associated with earlier tumour stage..., Sasi [/bib_ref]. In this study, we observed that the levels of SOCS4 are decreased in IL-23-induced migration and invasion of thyroid cancer cells [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Treatment with IL-23 resulted in reduced expression levels of SOCS4 in thyroid cancer cells [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Through overexpression and knockdown experiments, we demonstrate that SOCS4 negatively regulates IL-23-induced migration and invasion [fig_ref] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer... [/fig_ref]. Recently, several studies have shown evidence that the SOCS family has a strong tumor suppressing role in several types of solid and hematological tumors [bib_ref] Higher expression levels of SOCS 1,3,4,7 are associated with earlier tumour stage..., Sasi [/bib_ref] [bib_ref] Differential hypermethylation of SOCS genes in ovarian and breast carcinomas, Sutherland [/bib_ref]. When considering the next step, exploring the relationship between the SOCS family and IL-23 could be of great help in further clarifying the role that is played by the SOCS family in the development of thyroid tumors.
The role of miR-25 in cancer is not consistent and is occasionally controversial. In some studies, miR-25 may suppress the proliferation and the migration of colon cancer cells as a tumor suppressor gene in vitro and in vivo [bib_ref] MicroRNA-25 functions as a potential tumor suppressor in colon cancer by targeting..., Li [/bib_ref] [bib_ref] MiR-25 promotes gastric cancer cells growth and motility by targeting RECK, Zhao [/bib_ref] , whereas in others, miR-25 could remarkably promote cell proliferation and suppress apoptosis in gastric cancer [bib_ref] MiR-25 promotes gastric cancer cells growth and motility by targeting RECK, Zhao [/bib_ref]. In this study, we found that miR-25 promotes the migration and invasion of thyroid cancer cells by targeting SOCS4 [fig_ref] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR [/fig_ref]. Additional data reveal that miR-25 is involved in IL-23-regulated cell migration and invasion [fig_ref] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression... [/fig_ref]. These results are consistent with the above studies showing that miR-25 promotes tumor cell migration and invasion. Besides, we also analyzed the role of miR-25 in proliferation, apoptosis, and cell cycle in thyroid cancer cell lines. However, there was no association between miR-25 and cell proliferation, apoptosis, and cell cycle in thyroid cancer cells (data not shown). These results indicate that deregulated miR-25 expression plays different roles in different types of cancers.
In summary, our present study provides a novel evidence indicating that miR-25 and SOCS4 play a functional role in regulating IL-23-mediated migration and invasion in thyroid cancer cells. Although additional studies are required to understand the intricate regulatory mechanisms of thyroid cancer cell migration and invasion, the results obtained may shed new light in understanding cancer resistance to therapy.
## Supporting information
[fig] Fig 2: Analysis of the role of So4 in the regulation of thyroid cancer cell migration and invasion mediated by IL-23. (A) K1 cells were treated with 50 ng/ml rhIL-23 for 48 h. SOCS4 RNA levels were quantified by qRT-PCR (left panel) and the protein levels of SOCS4 were detected by western blot (right MicroRNA-25 promotes the motility of thyroid cancer cells by targeting SOCS4 [/fig]
[fig] Fig 3: MicroRNA-25 downregulates SOCS4 expression by directly targeting its 3 0 UTR. (A) The sequences of miR-25 binding sites within the human SOCS4 3 0 UTRs and the schematic reporter constructs. In this panel, WT represents the reporter constructs containing the entire 3 0 UTR sequences of SOCS4. SOCS4-MUT represents the reporter constructs containing mutated nucleotides. (B, C) K1 cells were co-transfected with either WT or MUT SOCS4 3 0 UTR reporter plasmids and either miR-25 mimic (B) or miR-25 inhibitor (C). Luciferase activities were measured after 48 h using a dual-luciferase assay kit and normalized to Renilla luciferase. (D) K1 cells were transfected with miR-25 mimic or controls for 48 h. SOCS4 mRNA (left panel) and protein (right panel) were detected by qRT-PCR and western blot, respectively. (E) Cells were transfected with miR-25 inhibitor or control and the experiments were performed as in D. (F, G) WRO cells were utilized and experiments were performed as in D and E. All the experiments were repeated at least 3 times with similar results. Bar graphs represent mean ± SD, n = 3 (**P < 0.01; *P < 0.05). doi:10.1371/journal.pone.0139456.g003 [/fig]
[fig] Fig 4: Determination of the effect of miR-25 on the regulation of SOCS4-suppressed thyroid cancer cell migration and invasion. (A, B) K1 cells were co-transfected with indicated miR-RNA mimic and plasmid (A) or shRNA (B) for 48 h and allowed to migrate towards serum for 24 h. (C, D) Cell invasion assay experiments were performed with the same conditions as in A and B. (E, F) Wound healing assay experiments were performed with the same conditions as in A and B. (G, H) Cells were transfected with miR-25 inhibitor or control and the experiments were performed as in A and B. (I, J) Cell invasion assay experiments were performed with the same conditions as in G and H. (K, L) Wound healing assay experiments were performed with the same conditions as in G and H. Bar graphs represent mean ± SD, n = 3 (**P < 0.01; *P < 0.05). doi:10.1371/journal.pone.0139456.g004 The expression of IL-23, miR-25 and SOCS4 in thyroid cancer tissues To validate the role of IL-23, miR-25 and SOCS4 in thyroid cancer, the expression of IL-23, miR-25 and SOCS4 were analyzed in 35 pairs of clinical PTC, 26 pairs of clinical FTC, and 22 normal thyroid samples. As shown in S5A-S5C Fig, the expression of SOCS4 was lower and the expression level of IL-23 and miR-25 was higher in PTC and FTC specimens than normal thyroid samples. In addition, high levels of IL-23 were correlated with high levels of miR-25 (S5D Fig). Interestingly, low levels of SOCS4 expression were correlated with high levels of IL-23 and miR-25 expression in PTC and FTC specimens (S5E and S5F Fig). These observations strongly suggest that alterations of IL-23, miR-25 and SOCS4 expression could be involved in thyroid cancer progression. [/fig]
[fig] Fig 5: IL-23 stimulates miR-25 expression. MicroRNA-25 levels were quantified by real-time PCR and the experiments were performed as described inFig 1.The expression levels were normalized to U6 snRNA. Data represent mean ± SD, n = 3 (**P < 0.01; *P < 0.05).doi:10.1371/journal.pone.0139456.g005 [/fig]
[fig] Fig 6: Analysis of the role of miR-25 in the regulation of SOCS4 expression and IL-23-mediated thyroid cancer cell migration and invasion. (A) K1 cells were transfected with the indicated plasmid, then treated with rhIL-23 (50 ng/ml) for 48 h. SOCS4 mRNA (left panel) and protein (right panel) were detected by qRT-PCR and western blot, respectively. (B) Cells were transfected with miR-25 inhibitor or control, and the experiments were performed as in A. (C, D) K1 cells were transfected with miR-25 mimic (C) or inhibitor (D) and controls, then treated with rhIL-23 (50 ng/ml) for 48 h and allowed to migrate towards serum for 24 h. (E, F) Cell invasion assay experiments were performed with the same conditions as in C and D. (G, H) Wound healing assay experiments were performed with the same conditions as in C and D. All the experiments were repeated at least 3 times with similar results. Bar graphs represent mean ± SD, n = 3 (**P < 0.01; *P < 0.05). doi:10.1371/journal.pone.0139456.g006 [/fig]
[fig] S1: Fig. IL-23 can not affect the proliferation of these thyroid cancer cells. (DOC) S2 Fig. IL-23 promotes the migration and invasion of WRO cell. (DOC) S3 Fig. SOCS4 can not affect the proliferation of these thyroid cancer cells. (DOC) S4 Fig. Screening for miRNAs that are involved in the IL-23-induced signaling pathway and that target the 3'UTR of SOCS4. (DOC) S5 Fig. Expression IL-23, miR-25 and SOCS4 in thyroid cancer tissues. (DOC) [/fig]
[table] Table: Correlation of IL-23, miR-25 and SOCS4 expression with clinicopathologic features in papillary thyroid cancers (PTC). (DOC) S2Table. Correlation of IL-23, miR-25 and SOCS4 expression with clinicopathologic features in follicular thyroid cancers (FTC). (DOC) S3 Table. Primers Used in Real-time PCR. (DOC) S4 Table. The target sequence of shRNAs. (DOC) [/table]
|
Can Controlling Endoplasmic Reticulum Dysfunction Treat Allergic Inflammation in Severe Asthma With Fungal Sensitization?
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Severe asthma is a heterogeneous disease entity to which diverse cellular components and pathogenetic mechanisms contribute. Current asthma therapies, including new biologic agents, are mainly targeting T helper type 2 cell-dominant inflammation, so that they are often unsatisfactory in the treatment of severe asthma. Respiratory fungal exposure has long been regarded as a precipitating factor for severe asthma phenotype. Moreover, as seen in clinical definitions of allergic bronchopulmonary aspergillosis (ABPA) and severe asthma with fungal sensitization (SAFS), fungal allergy-associated severe asthma phenotype is increasingly thought to have distinct pathobiologic mechanisms requiring different therapeutic approaches other than conventional treatment. However, there are still many unanswered questions on the direct causality of fungal sensitization in inducing severe allergic inflammation in SAFS. Recently, growing evidence suggests that stress response from the largest organelle, endoplasmic reticulum (ER), is closely interconnected to diverse cellular immune/inflammatory platforms, thereby being implicated in severe allergic lung inflammation. Interestingly, a recent study on this issue has suggested that ER stress responses and several associated molecular platforms, including phosphoinositide 3-kinase-δ and mitochondria, may be crucial players in the development of severe allergic inflammation in the SAFS. Defining emerging roles of ER and associated cellular platforms in SAFS may offer promising therapeutic options in the near future.
# Introduction
Bronchial asthma is a common chronic inflammatory disorder of the airways that is increasingly recognized as a heterogeneous clinical syndrome. Through a vast amount of research, current knowledge on the pathogenesis of asthma has shifted from a single disease due to airway obstruction to a complex disorder consisting of diverse phenotypes. [bib_ref] Evolving concepts of asthma, Gauthier [/bib_ref] The heterogeneous nature of bronchial asthma necessitates advancement in treatment, particularly for severe asthmatic patients refractory to conventional therapies, including inhaled or systemic corticosteroids (CS) and bronchodilators. Currently, most of the asthma treatment guidelines are based on the strategy which mainly targets T helper type 2 (TH2) cells-dominant airway inflammation and these are efficacious in many patients. However, 5%-10% asthmatic patients remain poorly managed by these modalities and account for about 50% of the health care expenditure for asthma care. [bib_ref] Severe asthma: advances in current management and future therapy, Barnes [/bib_ref] Notably, severe asthma is also heterogeneous in nature, in which different cellular components and pathogenetic mechanisms contribute to the disease, requiring different therapeutic approaches. [bib_ref] Clinical heterogeneity in the severe asthma research program, Moore [/bib_ref] Therefore, defining etiolo-gies and underlying pathobiologic mechanisms (i.e. endotypes) in various types of severe asthma is a major field of respiratory research. [bib_ref] Asthma endotypes: a new approach to classification of disease entities within the..., Lötvall [/bib_ref] [bib_ref] International ERS/ATS guidelines on definition, evaluation and treatment of severe asthma, Chung [/bib_ref] Respiratory fungal exposure is constant in humans, and fungal spores constitute the largest proportion of aerobiological particles in usual air environment. [bib_ref] Mold allergens in respiratory allergy: from structure to therapy, Twaroch [/bib_ref] Thus, the impact of respiratory fungal exposure on clinical courses of bronchial asthma has been widely reported in the literature for a long time. [bib_ref] The link between fungi and severe asthma: a summary of the evidence, Denning [/bib_ref] [bib_ref] Fungi and allergic lower respiratory tract diseases, Knutsen [/bib_ref] In particular, fungal exposure has been reported to be associated with the development of asthma. [bib_ref] Fungi and pollen exposure in the first months of life and risk..., Harley [/bib_ref] Furthermore, exposure to environmental fungal spores also leads to the exacerbation of allergic symptoms and asthma. [bib_ref] The effect of outdoor fungal spore concentrations on daily asthma severity, Delfino [/bib_ref] [bib_ref] Does spore count matter in fungal allergy?: the role of allergenic fungal..., Lin [/bib_ref] These results suggest the crucial role of fungi in allergic lung inflammation, especially in the more severe phenotype of the disease. Importantly, this idea is further supported by the fact that fungi can colonize, actively germinate, and infect the human respiratory tract, thereby potently sensitizing and inducing the host immune response. [bib_ref] The link between fungi and severe asthma: a summary of the evidence, Denning [/bib_ref] [bib_ref] Severe asthma with fungal sensitization, Agarwal [/bib_ref] Moreover, they can produce a wide array of enzymes and toxins closely implicated in allergic inflammation. [bib_ref] Cleavage of fibrinogen by proteinases elicits allergic responses through Toll-like receptor 4, Millien [/bib_ref] In this context, epidemiologic studies have also shown that fungal sensitivity is a possible precipitating factor for life-threatening asthma. [bib_ref] Sensitisation to airborne moulds and severity of asthma: cross sectional study from..., Zureik [/bib_ref] [bib_ref] Mold sensitization is common amongst patients with severe asthma requiring multiple hospital..., O'driscoll [/bib_ref] Based on this knowledge, severe asthma with fungal sensitization (SAFS) has been proposed to be investigated as a particular phenotype of severe asthma with therapeutic implications in clinical trials. [bib_ref] The link between fungi and severe asthma: a summary of the evidence, Denning [/bib_ref] [bib_ref] Randomized controlled trial of oral antifungal treatment for severe asthma with fungal..., Denning [/bib_ref] [bib_ref] Effectiveness of voriconazole in the treatment of Aspergillus fumigatus-associated asthma (EVITA3 study), Agbetile [/bib_ref] However, treatment with antifungal agents in SAFS patients failed to show a consistent beneficial effect on the clinical course of the disease in 2 randomized clinical trials. [bib_ref] Randomized controlled trial of oral antifungal treatment for severe asthma with fungal..., Denning [/bib_ref] [bib_ref] Effectiveness of voriconazole in the treatment of Aspergillus fumigatus-associated asthma (EVITA3 study), Agbetile [/bib_ref] In fact, little is known about the exact role of fungi in severe allergic lung inflammation.
Recently, increasing evidence suggests that cellular stresses due to impaired biological processes can be integrated into inflammatory response, thereby being implicated in various human inflammatory disorders. Endoplasmic reticulum (ER) is the largest intracellular organelle which manipulates crucial soluble and membrane proteins, and it is abundantly equipped by numerous enzymes that facilitate proper folding of the client proteins. [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] In this respect, stress response originates from the ER related to the accumulation of unfolded and/or misfolded proteins in the ER lumen (i.e. ER stress), [bib_ref] Endoplasmic reticulum stress in disease: mechanisms and therapeutic opportunities, Hosoi [/bib_ref] has been increasingly reported to be involved in the pathogenesis of diverse inflammatory disorders. [bib_ref] From endoplasmic-reticulum stress to the inflammatory response, Zhang [/bib_ref] Furthermore, involvement of ER stress in the pathogenesis of numerous pulmonary disorders, including bronchial asthma, chronic obstructive pulmonary disease (COPD), fibrotic lung disorders, and acute lung injury (ALI)/ acute respiratory distress syndrome (ARDS), has been widely investigated. [bib_ref] Endoplasmic reticulum stress influences bronchial asthma pathogenesis by modulating nuclear factor κB..., Kim [/bib_ref] [bib_ref] The UPR and lung disease, Osorio [/bib_ref] [bib_ref] Endoplasmic reticulum stress in chronic obstructive lung diseases, Ribeiro [/bib_ref] [bib_ref] Endoplasmic reticulum stress enhances fibrotic remodeling in the lungs, Lawson [/bib_ref] [bib_ref] Inhibition of endoplasmic reticulum stress alleviates lipopolysaccharide-induced lung inflammation through modulation of..., Kim [/bib_ref] In this review, we focus on the emerging role of ER and associated cellular immune/inflammatory molecular platforms in the lung, particularly in the pathogenesis of fungal allergy and SAFS. Additionally, we present our recent data on these issues, which further highlight the therapeutic potential of ER stress in fungi-induced severe allergic inflammatory disorders in the lung.
## Safs and other fungal sensitization/allergy-associated clinical conditions
Fungal sensitization and allergy development refer to an exaggerated immune response typically against non-pathogenic fungi. Fungal sensitization and allergy are further distinguished by the presence of immune-mediated tissue damage because sensitization itself does not always cause inflammatory response. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] In contrast, the term fungal infection can be applied when there is evidence of tissue dysfunction associated with the growth and invasion of pathogenic fungi in the host. Whereas viable fungi are implicated in both allergy and infection, causative agents for allergy do not need to be viable to induce an allergic response (e.g. structural components and byproducts). There are several important disease entities representing the severe end of fungal sensitization/allergy development-associated conditions. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] These include allergic bronchopulmonary aspergillosis/mycosis (ABPA/ABPM) and SAFS.
ABPA is a complex hypersensitivity reaction that often occurs in patients with asthma or cystic fibrosis when bronchi become colonized by Aspergillus species, mostly Aspergillus fumigatus. [bib_ref] Allergic bronchopulmonary aspergillosis, Agarwal [/bib_ref] In general, the diagnosis of ABPA is a composite of clinical, radiological, and immunologic features . In the later courses of ABPA, repeated episodes of bronchial obstruction, inflammation, and mucoid impaction can lead to irreversible structural changes. Many patients with ABPA respond well to treatment with systemic CS, whereas some patients are poorly controlled and can be complicated by progression to bronchiectasis and pulmonary fibrosis. Similarly, fungi other than Aspergillus species, such as Candida albicans, can induce an AB-PA-like disease process, which refers to ABPM.
Whereas ABPA was first reported in 1952, the definition of SAFS was introduced in 2006 7 and has been used in clinical trial settings to demonstrate the possible role of antifungal therapy in treating a particular phenotype of severe asthma associated with fungi. [bib_ref] Randomized controlled trial of oral antifungal treatment for severe asthma with fungal..., Denning [/bib_ref] [bib_ref] Effectiveness of voriconazole in the treatment of Aspergillus fumigatus-associated asthma (EVITA3 study), Agbetile [/bib_ref] In fact, ABPA/ABPM is an extreme spectrum of allergic inflammatory response against fungi, thus most patients sensitized to fungi without convincing evidence of lung damage had not been incorporated into a specific disease entity. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] Therefore, researchers have proposed SAFS that can be defined as patients having both severe asthma and evidence for fungal sensitization (i.e. positive skin prick test, positive fungal specific IgE in blood) without satisfying the criteria of ABPA. [bib_ref] The link between fungi and severe asthma: a summary of the evidence, Denning [/bib_ref] Thereafter, several clinical studies suggested the role of antifungal agents in the treatment of the SAFS patient group. [bib_ref] Randomized controlled trial of oral antifungal treatment for severe asthma with fungal..., Denning [/bib_ref] [bib_ref] Voriconazole and posaconazole improve asthma severity in allergic bronchopulmonary aspergillosis and severe..., Chishimba [/bib_ref] Meanwhile, most of the current diagnostic criteria for ABPA/ ABPM have been in practice since the 1970s and do not pre- cisely reflect the natural history of the disease, especially in its early phase. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] Consequently, these diagnostic criteria of ABPA/ ABPM may not be effective for the prevention of permanent lung damage, such as bronchiectasis. Additionally, the definition of SAFS does not represent the direct causality of fungal sensitization in inducing severe asthma, even though it is convenient for patient inclusion in clinical trials. In this respect, SAFS was not proposed as an asthma endotype having a distinct pathobiological mechanism. [bib_ref] Asthma endotypes: a new approach to classification of disease entities within the..., Lötvall [/bib_ref] Conflicting results from previous clinical trials, in which efficacies of antifungal agents in the treatment of SAFS were investigated, 16,17 may be partly due to the limitation in SAFS definition and unstandardized testing tools demonstrating fungal allergy. More precise definitions that represent the diverse spectrum of fungal allergy-associated clinical conditions are needed in the near future.
## Interactions between fungi and host
Research on the mechanisms of fungi-associated disorders, including fungal allergy and infection, has revealed intricate molecular networks between the fungi and host. In particular, they mainly focused on what kind of fungi-derived molecules can be effectively recognized by the host immune system through what kind of innate receptors and how these receptorligand interactions can be interpreted and linked to appropriate host inflammatory responses.
Fungi contain various antigens in their cell wall. Among various components, polysaccharides, including αand β-glucans, galactomannan, and chitin, are major constituents of fungal cell wall. [bib_ref] C-type lectin receptors orchestrate antifungal immunity, Hardison [/bib_ref] In addition, fungi can produce large amounts of secreted enzymes, including proteases and glycosidases, which can directly damage host tissues. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] Both fungal cell wall and secreted components can induce protective host immune responses against fungi. At the same time, they also act as major fungal allergens in the lung.
## Initial recognition of fungi
The complex interaction between the host and fungi begins with recognition of the fungal components, called pathogenassociated molecular patterns (PAMPs), through several evolutionarily conserved pattern recognition receptors (PRRs) of the innate immunity. [bib_ref] Pattern recognition receptors and inflammation, Takeuchi [/bib_ref] The C-type lectin receptor (CLR), Toll-like receptor (TLR), and NOD-like receptor (NLR) families of PRRs are known to be involved in anti-fungal host immunity.
## Clrs
Since about 60% of the dry weight of the fungal cell wall is β-glucans, glucan concentration in dust can be used as a surrogate marker for estimating fungal exposure. [bib_ref] Fungi and allergic lower respiratory tract diseases, Knutsen [/bib_ref] [bib_ref] Use of (1-3)-beta-d-glucan concentrations in dust as a surrogate method for estimating..., Iossifova [/bib_ref] CLRs consist of a transmembrane receptor family having a carbohydrate-binding domain, thereby recognizing carbohydrates on the fungi. [bib_ref] Pattern recognition receptors and inflammation, Takeuchi [/bib_ref] Among them, dectin-1 and dectin-2 expressed on macrophages, neutrophils, and dendritic cells (DCs) sense β-glucans of fungi and generate anti-fungal effector responses with phagocytosis, oxidative burst, and production of inflammatory mediators, including interleukin (IL)-6, IL-8, and tumor necrosis factor (TNF)-α. [bib_ref] Fungi and allergic lower respiratory tract diseases, Knutsen [/bib_ref] In CLR signaling, activation of downstream Syk tyrosine kinase is known to be integral in the activation of mitogen-activated protein (MAP) kinases and important transcription factors, including nuclear factor (NF)-κB through CARD9. [bib_ref] C-type lectin receptors orchestrate antifungal immunity, Hardison [/bib_ref] The clinical importance of the CLR-associated pathway is further emphasized by the fact that mutations in dectin-1 and CARD 9 lead to chronic fungal infection. [bib_ref] Human dectin-1 deficiency and mucocutaneous fungal infections, Ferwerda [/bib_ref] [bib_ref] A homozygous CARD9 mutation in a family with susceptibility to fungal infections, Glocker [/bib_ref] Moreover, considering that aerosol exposure of β-glucan resulted in pulmonary eosinophilia in an animal model [bib_ref] Inhalation of (1→3)-beta-Dglucan causes airway eosinophilia, Fogelmark [/bib_ref] and that the levels of β-glucan were positively associated with peak expiratory flow variability in children with asthma, [bib_ref] 1→3)-beta-D-glucan and endotoxin in house dust and peak flow variability in children, Douwes [/bib_ref] the CLR pathway may play an important role in the pathogenesis of fungal allergic lung inflammation.
## Tlrs
TLRs are broadly expressed on nonprofessional innate cells, such as epithelial cells and fibroblasts, as well as innate immune cells. A wide range of fungal PAMPs (e.g. phospholipomannan, glucuronoxylomannan, O-and N-linked mannans, DNA, and RNA) are recognized by numerous TLRs. [bib_ref] Fungal pathogens-a sweet and sour treat for Toll-like receptors, Bourgeois [/bib_ref] Their engagements with cognate TLRs activate intracellular pathways via association with the myeloid differentiation primary response gene 88 (Myd88) (except for the TLR3 pathway), a key downstream adaptor for most TLRs and IL-1 receptors, [bib_ref] Pattern recognition receptors and inflammation, Takeuchi [/bib_ref] and intersect with cellular adaptor proteins and important transcription factors, such as NF-κB. [bib_ref] Fungal pathogens-a sweet and sour treat for Toll-like receptors, Bourgeois [/bib_ref] Among them, TLR2 and TLR4 have been thought to be crucial for the recognition of common pathogenic fungi, including Candida and Aspergillus species. In particular, macrophages lacking a functional TLR4 or both TLR2 and TLR4 result in decreased nuclear translocation of NF-κB and release of various pro-inflammatory cytokines in response to viable conidia of Aspergillus fumigatus. [bib_ref] Toll-like receptor (TLR) 2 and TLR4 are essential for Aspergillusinduced activation of..., Meier [/bib_ref] In addition, the lack of TLR2 and/or TLR4 significantly impairs recruitment and effector function of neutrophils in response to Aspergillus fumigatus. Furthermore decreased recruitment of neutrophils and macrophages to the site of Candida albicans infection was observed in TLR2 -/mice, [bib_ref] Absence of TLR2 influences survival of neutrophils after infection with Candida albicans, Tessarolli [/bib_ref] and mice expressing nonfunctional TLR4 showed decreased phagocytosis and production of pro-inflammatory mediators in macrophages against Candida albicans infection. [bib_ref] Absence of functional TLR4 impairs response of macrophages after Candida albicans infection, Gasparoto [/bib_ref]
## Nlrs
NLRs, a cytoplasmic PRR family, are also involved in the recognition of fungi and induction of anti-fungal immune response. [bib_ref] Syk kinase signalling couples to the Nlrp3 inflammasome for anti-fungal host defence, Gross [/bib_ref] [bib_ref] Sensing and reacting to microbes through the inflammasomes, Franchi [/bib_ref] Two NLRs, namely, NLRP3 and NLRC4, are reported to sense fungi to form the proteolytic multiprotein complex termed inflammasomes, processing and activating IL-1β and IL-18 which are essential cytokines for protection against fungi. [bib_ref] C-type lectin receptors orchestrate antifungal immunity, Hardison [/bib_ref] [bib_ref] Sensing and reacting to microbes through the inflammasomes, Franchi [/bib_ref] On sensing certain fungi, these NLRs oligomerize to form caspase-1-activating scaffold through recruiting and nucleating numerous apoptosis-associated speck-like proteins containing a CARD (ASCs) and pro-caspase-1. Then, active caspase-1 cleaves pro-IL-1β and pro-IL-18 into their biologically active forms IL-1β and IL-18, respectively. Interestingly, activation of NLRP3 inflammasome against Aspergillus fumigatus and Candida albicans may require Syk tyrosine kinase, which is located in the downstream of CLR signaling. [bib_ref] Syk kinase signalling couples to the Nlrp3 inflammasome for anti-fungal host defence, Gross [/bib_ref] Furthermore, the CLRs-Syk pathway is also known to be related to non-NLR-associated inflammasome to produce IL-1β against fungi. [bib_ref] C-type lectin receptors orchestrate antifungal immunity, Hardison [/bib_ref] Therefore, the CLRs-Syk pathway may be a potent upstream modulator in the activation of various inflammasomes against fungi.
## Action of fungal proteases
Many of the fungal allergens are known as proteases (www.allergen.org. Accessed July, 2017). Similarly, increased expression of protease-activated receptor-2 (PAR-2) is observed in asthmatic airway epithelium and diverse proteases from many aeroallergens, including fungi, activate PAR-2. [bib_ref] Protease-activated receptors in human airways: upregulation of PAR-2 in respiratory epithelium from..., Knight [/bib_ref] Therefore, elucidating the action mechanism of various fungal proteases in allergic lung disease is an important issue to unveil the fundamental molecular basis of the disease. In particular, Homma et al. [bib_ref] Role of Aspergillus fumigatus in triggering protease-activated receptor-2 in airway epithelial cells..., Homma [/bib_ref] recently demonstrated that PAR-2 activation by Aspergillus fumigatus may weaken TH1 cells-mediated antiviral/interferon response without affecting Th2 cells-mediated response in airway epithelial cells, and thus this may impact on airway cell populations toward TH2 cells. Additionally, highly purified protease derived from Aspergillus fumigatus was shown to be sufficient to convert type I allergen ovalbumin (which typically require aluminum-based adjuvants before respiratory challenge to induce pulmonary allergic inflammation in mice) to type II allergen (which can readily induce allergic lung inflammation via airway challenge without additional extra-pulmonary adjuvants) when primary respiratory exposure to ovalbumin occurs in the presence of fungal protease, which implies that active proteases are the essential adjuvant factor responsible for overcoming tolerogenic responses and inducing allergic reaction. [bib_ref] A protease-activated pathway underlying Th cell type 2 activation and allergic lung..., Kheradmand [/bib_ref] Furthermore, fibrinogen cleavage product by fungal protease in airway epithelium has been reported to activate PRRs, such as TLR4, also known as one of the important upstream regulators of T helper responses to allergens, 45,46 thereby associating antifungal innate immunity with adaptive allergic inflammation. [bib_ref] Cleavage of fibrinogen by proteinases elicits allergic responses through Toll-like receptor 4, Millien [/bib_ref] Importantly, fungal proteases directly influence the integrity of asthmatic bronchial epithelium and induce morphologic changes and pro-inflammatory cytokine production, thereby possibly enhancing the allergenic potential of the other fungal allergens at the same time. [bib_ref] The link between fungi and severe asthma: a summary of the evidence, Denning [/bib_ref] Given that epithelial exposure to various fungal allergens occurs simultaneously, these aspects of action may highlight the vital role of fungal proteases in initiating allergic airway disorders.
## Interpretation of fungal exposure and induction of host immune responses
Initial fungal recognition leads to release of endogenous danger molecules, including uric acid, ATP, and various epithelium-derived cytokines, such as IL-33, IL-25, and thymic stromal lymphopoietin (TSLP) from airway epithelium. [bib_ref] The airway epithelium in asthma, Lambrecht [/bib_ref] These molecules are critically implicated in the pathogenesis of fungal allergy. For example, IL-33 is a member of the IL-1 family of cytokines and normally localized in the nucleus of airway epithelial cells. However, IL-33 can be released and processed to a potent mature form in response to epithelial stimulation from protease action and PRR activation. Then, through engagement of IL-33 with IL-1 receptor family member suppressor of tumorigenicity 2 (ST2), group 2 innate lymphoid cells (ILCs) can be activated and rapidly induce type 2 inflammation in the airways. In this context, IL-33/ST2 signaling has reported to be prerequisite for airway eosinophilia following exposure to Alternaria alternata, which is mainly orchestrated by IL-5/IL-13-producing group 2 ILCs in the lung. [bib_ref] IL-33-responsive lineage-CD25+ CD44(hi) lymphoid cells mediate innate type 2 immunity and allergic..., Bartemes [/bib_ref] Moreover, type 2 airway inflammation induced by chitin (a polysaccharide constituent of fungal cell wall) and protease allergen papain also involves epithelial production of IL-33, IL-25, and TSLP, and subsequent activation of group 2 ILCs, DCs, and other innate immune cells. [bib_ref] Chitin activates parallel immune modules that direct distinct inflammatory responses via innate..., Van Dyken [/bib_ref] [bib_ref] Group 2 innate lymphoid cells are critical for the initiation of adaptive..., Halim [/bib_ref] Importantly, epithelial release of IL-33 leads to exacerbation and airway remodeling of allergic airway disease in murine models, 50,51 which highlights the possible implication of IL-33 in fungi-induced severe allergic inflammation. Furthermore, a recent study has shown that children with SAFS is associated with higher airway IL-33 levels in bronchoalveolar lavage fluid and endobronchial biopsy specimens compared to children having severe asthma without fungal sensitization. Similarly, pulmonary IL-33 levels and IL-13+ group 2 ILCs numbers are increased in the Alternaria alternata-induced SAFS animal model. Notably, Alternaria alternata-induced elevations in lung IL-33 levels and IL-13+ group 2 ILCs numbers remain increased with CS treatment, demonstrating that anti-fungal host immune response involving epithelial IL-33/group 2 ILCs may contribute to CS resistance of fungi-induced allergic lung inflammation. [bib_ref] Pediatric severe asthma with fungal sensitization is mediated by steroid-resistant IL-33, Castanhinha [/bib_ref]
## Genetic background of the host
Several genetic problems affecting the phagocytic activity of cell-mediated immunity have been well known to be associated with increased risk of fungal infection. [bib_ref] Immune interactions with pathogenic and commensal fungi: a two-way street, Underhill [/bib_ref] For instance, genetic defects in several proteins comprising the NADPH oxidase complex, which is essential for the generation of reactive oxygen species (ROS) in phagocytes, are related to increased incidence of invasive fungal infection. However, scarce information is available on genetic associations concerning SAFS and other fungal allergic diseases. Most knowledge on these subjects came from small studies on ABPA. Several implicated genes have been identified, which includes IL-4Rα, cystic fibrosis transmembrane conductance regulator, and HLA-DR. [bib_ref] Fungal allergy in asthma-state of the art and research needs, Denning [/bib_ref] In regards to SAFS, HLA class II restriction involving HLA-DR and HLA-DQ was reported to be implicated in Alternaria-sensitive moderate-to-severe asthma. [bib_ref] Mold-sensitivity in children with moderate-severe asthma is associated with HLA-DR and HLA-DQ, Knutsen [/bib_ref] In particular, there is possibility that genetic loci encoding IL-33 and its receptor could be linked to SAFS development, since these have consistently been reported to be a critical susceptibility factor for human asthma. [bib_ref] A large-scale, consortium-based genomewide association study of asthma, Moffatt [/bib_ref] Further research is warranted regarding this issue.
## Implication of er stress in fungal allergy and safs er stress and unfolded-protein response (upr)
ER is located adjacent to nuclear envelope and structurally consists of continuous membrane network that expands throughout the cytoplasm. ER participates in a variety of cellular functions. In addition to its well-known role as a major cellular storage for calcium, ER plays an essential role in the synthesis, correct folding, assembly, modification, and transport of soluble and membrane proteins. To efficiently accomplish its roles as a 'protein folding factory, ' ER is abundantly equipped with chaperones, oxidases, and isomerases that facilitate proper folding of proteins and thus manipulates nearly one-third of the cellular proteome. [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] Meanwhile, optimal milieus within ER, such as adequate concentrations of ATP and calcium or oxidizing environments, are essential to retain the maximal activities of many enzymes for protein folding. [bib_ref] Functional and morphological impact of ER stress on mitochondria, Vannuvel [/bib_ref] Therefore, ER is highly sensitive to various stresses that perturb cellular energy levels, calcium concentration, and redox state, all of which can lead to the accumulation of misfolded and/or unfolded proteins in the ER lumen. Accumulation of misfolded/unfolded proteins are toxic to cells and can cause an imbalance between the ER folding capacity and the folding load of nascent protein entering ER, which is referred to as ER stress. To maintain cellular homeostasis against this crisis, eukaryotic cells have evolved adaptive response, UPR. UPR signaling is mediated by 3 ER-localized sensors in mammalian cells. These consist of inositol-requiring 1α (IRE1α), double-stranded RNA-dependent protein kinase (PKR)-like ER kinase (PERK), and activating transcription factor 6 (ATF6); each of these ER transmembrane proteins senses unfolded proteins in ER through ER-luminal domain. According to the proposed model of UPR activation, all these sensors remain in the inactive state through association with abundant ER chaperone glucose-regulated protein 78 (GRP78) in resting cells. In a condition of ER stress, however, GRP78 preferentially associates with accumulated unfolded proteins in ER and dissociates from the 3 sensor proteins. Then, ER stress sensor proteins become activated (as for IRE1α, direct activation through engagement with misfolded proteins has been also demonstrated) [bib_ref] Unfolded proteins are Ire1-activating ligands that directly induce the unfolded protein response, Gardner [/bib_ref] and transmit signals about the folding status of ER to the cytosol and nucleus, inducing UPR-target genes to restore the ER folding capacity. [bib_ref] The unfolded protein response: controlling cell fate decisions under ER stress and..., Hetz [/bib_ref] Although the 3 UPR signaling pathways are simultaneously activated in response to the accumu-lation of unfolded/misfolded proteins, contribution of each pathway substantially varies depending on the severity and chronicity of ER stress.
Activation of the PERK pathway is known to be a relatively immediate response to ER stress. PERK is a serine/threonine protein kinase and, on its activation, PERK becomes dimerized and autophosphorylated. This process allows PERK to phosphorylate the α-subunit of eukaryotic translation-initiation factor 2α (eIF2α), which inhibits the assembly of the 80S ribosome in eukaryotic cells, therefore the synthesis of proteins. Considering that blockade of PERK-induced eIF2α phosphorylation leads to abnormally elevated protein synthesis and higher levels of ER stress, this pathway may reduce ER protein folding load in ER stressed-cells. Furthermore, eIF2α phosphorylation stimulates translation of selective mRNAs, including ATF4, that contain regulatory sequences, such as an inhibitory upstream open reading frame in 5´-untranslated regions. ATF4 is a transcription factor that has been known to control pro-survival genes in relation to protein folding, anti-oxidant response, and ER-stress-induced apoptosis (apoptosis occurs later during ER stress when the other arms of UPR fail to restore equilibrium). [bib_ref] An integrated stress response regulates amino acid metabolism and resistance to oxidative..., Harding [/bib_ref] Cleavage of ATF6 follows relatively quickly in response to ER stress. When ATF6 is released from GRP78, it moves to the Golgi apparatus where it is cleaved. This process causes the release of a functional cytosolic fragment of ATF6 (ATF6f). ATF6f translocates to the nucleus and induces gene transcriptions of enzymes that help folding, maturation, secretion, and ER-associated degradation (ERAD) of protein. ATF6 seems to modulate a relatively narrow range of genes participating in the ER protein quality control system, many of which are also influenced by the IRE1α/X-box-binding protein 1 (XBP1) pathway. This may be why mice lacking ATF6 are overtly normal with no apparent phenotype. Thus, the ATF6 pathway seems to fine-tune and enhance the functional capacity of UPR along with PERK and IRE1α, especially in chronic ER stress. [bib_ref] ATF6alpha optimizes long-term endoplasmic reticulum function to protect cells from chronic stress, Wu [/bib_ref] [bib_ref] Transcriptional induction of mammalian ER quality control proteins is mediated by single..., Yamamoto [/bib_ref] The IRE1α pathway is the most evolutionarily conserved one among 3 major branches of UPR. IRE1α has both serine/threonine protein kinase activity and site-specific endoribonuclease (RNase) activity. GRP78 dissociation from IRE1α or direct engagement of IRE1α with misfolded proteins promotes IRE1α dimerization and autophosphorylation, activating its RNase activity. IRE1α catalyzes the splicing of mRNA encoding XBP1 and generates a spliced variant (sXBP1) that functions as a transcription factor for genes associated with the structural and functional expansion of ER. Alone or with ATF6, sXBP1 induces chaperones, such as GRP78 and proteins, participating in ER biogenesis, lipid synthesis, ERAD, and protein secretion. [bib_ref] The unfolded protein response: controlling cell fate decisions under ER stress and..., Hetz [/bib_ref] Because the synthesis of XBP1 mRNA is up-regulated by ATF6 following ER stress, full activation of the IRE1α-XBP1 pathway may be delayed compared to the other pathways of UPR. [bib_ref] Functional and morphological impact of ER stress on mitochondria, Vannuvel [/bib_ref] However, transcriptional targets of XBP1 are not just limited to the restoration of the ER protein quality control system. The IRE1α/XBP1 pathway also participates in the induction of molecules associated with lipid metabolism, [bib_ref] Regulation of hepatic lipogenesis by the transcription factor XBP1, Lee [/bib_ref] immune and inflammatory responses, [bib_ref] TLR activation of the transcription factor XBP1 regulates innate immune responses in..., Martinon [/bib_ref] and cellular differentiation, [bib_ref] XBP1 controls diverse cell type-and condition-specific transcriptional regulatory networks, Acosta-Alvear [/bib_ref] so that it plays previously unrecognized roles in many critical cellular events. Consistent with these findings, mice lacking XBP1 display hypoplastic fetal livers leading to the death from anemia [bib_ref] An essential role in liver development for transcription factor XBP-1, Reimold [/bib_ref] and XBP1 is also known to be important in the terminal differentiation of B cells into highly secretory plasma cells. [bib_ref] Plasma cell differentiation requires the transcription factor XBP-1, Reimold [/bib_ref] Moreover, homozygous ire1α -/embryos show lethal defects in the differentiation of liver and B lymphocytes. [bib_ref] The unfolded protein response sensor IRE1alpha is required at 2 distinct steps..., Zhang [/bib_ref] Furthermore, in contrast to the specific endoribonuclease activity against XBP1, IRE1α also has non-specific RNase activity and may degrade ERmembrane-associated mRNAs to reduce the production of proteins, namely, regulated IRE1α-dependent decay (RIDD). [bib_ref] Decay of endoplasmic reticulum-localized mRNAs during the unfolded protein response, Hollien [/bib_ref] This cellular mechanism is well suited to complement other UPR pathways. Recently, increasing evidence has indicated that expression of a wide range of proteins, involved in diverse cellular processes beyond ER homeostasis, can be regulated through RIDD, explaining the critical involvements of UPR pathways in broad cellular events other than cellular protein homeostasis. [bib_ref] Getting RIDD of RNA: IRE1 in cell fate regulation, Maurel [/bib_ref] Cumulatively, these canonical aspects of UPR generally attempt to reduce ER stress by reducing the demand of protein folding, facilitating the protein degradation pathway, and increasing the expression of enzymes and ER chaperones that help in protein folding. If causative stress is very severe and the cell fails to resolve the protein folding defect, then these adaptive responses will initiate apoptosis. CCAAT/enhancer-binding protein homologous protein (CHOP), as a downstream effector of UPR, is known to mediate ER-stress-induced apoptosis. Thus, CHOP, along with GRP78, is widely used as a marker which indicates the presence of ER stress [bib_ref] Endoplasmic reticulum stress influences bronchial asthma pathogenesis by modulating nuclear factor κB..., Kim [/bib_ref] [bib_ref] Inhibition of endoplasmic reticulum stress alleviates lipopolysaccharide-induced lung inflammation through modulation of..., Kim [/bib_ref] and up-regulation of GRP78 is typically detectable earlier than CHOP .
## Non-canonical aspects of upr
More recently, there is accumulating evidence suggesting that ER stress and UPR pathways are closely cross-linked to diverse signaling outcomes having seemingly little to do with ER func- . Fungal exposure leads to the activation of endoplasmic reticulum (ER) stress and unfolded protein response (UPR) in the lung. Inhaled fungi possess a wide array of ligands that activate pattern recognition receptors (PRRs) expressed on structural cells (e.g. airway epithelial cells) and dendritic cells (DCs). Fungi can also produce large amounts of secreted enzymes, such as proteases, which disrupt tight junctions of airway epithelium. Initial recognition of fungi is followed by allergic sensitization and eosinophilic airway inflammation through close interactions between various facets of host immunity (not presented here). During this process, various cell types, including frontline cells (e.g. airway epithelial cells, DCs, and alveolar macrophages) and adaptive T and B cells, produce large amounts of cytokines/ chemokines as well as host defensive molecules. Increased protein folding demand in these cells results in ER stress and triggers UPR. UPR signaling is orchestrated by 3 ER-localized sensors, namely, PERK, IRE1, and ATF6. These adaptive responses together reduce protein folding demand, increase enzymes and chaperones involved in protein folding, and facilitate protein degradation pathway. However, when cells fail to resolve ER stress, UPR mediates ER stress-induced apoptosis.
tion as a protein folding factory. Indeed, our knowledge has expanded that there are miscellaneous non-canonical aspects of UPR providing mechanistic insight into the pathogenesis of various human inflammatory diseases. As comprehensively reviewed by Arensdorf and colleagues, [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] non-canonical UPR can arise from multiple points of canonical UPR through various mechanisms.
First of all, canonical UPR can result in translational regulation of specific proteins particularly through the PERK-eIF2α pathway. Since phosphorylation of eIF2α transiently suppresses almost 90% of cellular mRNA, [bib_ref] Extensive translatome remodeling during ER stress response in mammalian cells, Ventoso [/bib_ref] the PERK pathway can efficiently converge on a wide range of cellular processes, including inflammatory process. Specifically, NF-κB signaling, which is a well-known master transcription factor involved in pro-inflammatory cytokine production, leukocyte recruitment, or cell survival, [bib_ref] The nuclear factor NF-kappaB pathway in inflammation, Lawrence [/bib_ref] seems to be partly controlled by phosphorylation of eIF2α through suppressing translation of inhibitors of NF-κB (i.e. IκB). [bib_ref] Translational repression mediates activation of nuclear factor kappa B by phosphorylated translation..., Deng [/bib_ref] Moreover, this pathway also enables the translation of a number of transcription factors beyond ATF4. [bib_ref] Extensive translatome remodeling during ER stress response in mammalian cells, Ventoso [/bib_ref] [bib_ref] The zipper model of translational control: a small upstream ORF is the..., Yaman [/bib_ref] [bib_ref] Phosphorylation of eIF2 directs ATF5 translational control in response to diverse stress..., Zhou [/bib_ref] Secondly, the sensor proteins of canonical UPR, including IRE1α and PERK, can create stress-specific scaffolds on ER membrane. IRE1α and PERK have been demonstrated to become self-associated to form high-order oligomerization in cytosolic domains, followed by autophosphorylation through action of the kinase domain. [bib_ref] The unfolded protein response signals through high-order assembly of Ire1, Korennykh [/bib_ref] Furthermore, high-order assembly of UPR sensor proteins has been proposed to provide a specialized molecular microenvironment, which can be associated with low-affinity binding molecules with relatively high avidity. [bib_ref] Mammalian endoplasmic reticulum stress sensor IRE1 signals by dynamic clustering, Li [/bib_ref] Therefore, this allows an additional control principle in regulating UPR and other signaling pathways. The best example for protein interactions with phosphorylated and oligomerized IRE1α is tumor necrosis factor receptor-associated factor 2 (TRAF2), which is essential for the activation of several protein kinases. The IRE1α-TRAF2 complex recruits and activates c-JUN N terminal kinase (JNK), which induces the expression of genes involved in inflammation partly through phosphorylating the transcription factor activator protein 1. [bib_ref] Signal transduction by the JNK group of MAP kinases, Davis [/bib_ref] In addition, the IRE1α-TRAF2 complex is implicated in the recruitment of IκB kinase, which phosphorylates IκB, resulting in the degradation of IκB and the activation of NF-κB. [bib_ref] Autocrine tumor necrosis factor alpha links endoplasmic reticulum stress to the membrane..., Hu [/bib_ref] Consistent with these data, in the IRE1α -/mouse embryonic fibroblasts, decreased activity of JNK against ER stress [bib_ref] Coupling of stress in the ER to activation of JNK protein kinases..., Urano [/bib_ref] and impairment of ER stress-induced NF-κB activation and subsequent reduced expression of TNF-α 78 were observed. Regarding the PERK pathway, there are fewer known molecular interactions arising from stress-specific scaffolds than the IRE1α pathway. It has been demonstrated that PERK is required for the phosphorylation of nuclear factorerythroid 2-related factor-2 (Nrf-2), a key transcription factor implicated in cellular defense against oxidative insults, and cells harboring a targeted deletion of PERK exhibit failure to upregulate Nrf-2-target genes. [bib_ref] Nrf2 is a direct PERK substrate and effector of PERK-dependent cell survival, Cullinan [/bib_ref] Thirdly, many of the transcription factors in canonical UPR (e.g. ATF6, ATF4, XBP1) belong to the transcription factor family having basic leucine zipper (bZIP) domain and numerous molecular interactions between members of this family protein have been reported. [bib_ref] Deciphering B-ZIP transcription factor interactions in vitro and in vivo, Vinson [/bib_ref] Thus, these interactions among bZIP family members can significantly impact on gene expressions under ER stress. For instance, ATF6 is known to interact with CCAAT/enhancer-binding protein β, a member of the bZIP transcription factor family that participates in diverse cellular physiologies, including proliferation, differentiation, metabolism, and inflammation. [bib_ref] CCAAT/enhancer-binding proteins: structure, function and regulation, Ramji [/bib_ref] Furthermore, transcription factors in canonical UPR can affect global gene expression beyond their target-genes via sharing coregulatory molecules with constitutive transcription factors. [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] Through this mechanism, ATF6 disrupts the costimulatory interaction between cAMP-responsive element-binding protein (CREB), a member of the bZIP family, and CREB-regulated transcription coactivator 2 during ER stress associated with hepatic gluconeogenesis, inhibiting gluconeogenic program. [bib_ref] The CREB coactivator CRTC2 links hepatic ER stress and fasting gluconeogenesis, Wang [/bib_ref] Additionally, the expansion of UPR outcome can be achieved by regulating gene expression through RIDD of several other ER-localized mRNAs by the RNase activity of IRE1α, 68 by influencing the transcriptome and accompanying diverse physiologic processes through modulating the expression of multifunctional secondary UPR-regulated transcription factors, such as CHOP, [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] and by UPR-mediated regulation of mRNA through short single-stranded microRNAs, which promote the degradation of complementary mRNA. [bib_ref] Control of translation and mRNA degradation by miRNAs and siRNAs, Valencia-Sanchez [/bib_ref] Defining the non-canonical involvement of UPR pathways in diverse cellular events is an emerging area of investigation. All of these mechanisms may also explain biological consequences of ER stress and highlight the therapeutic potential of ER stress in a wide array of human pathologic conditions.
## Er stress and upr in the lung
Given that huge amounts of foreign materials are inhaled on a daily basis, recognition of these environmental cues should be accompanied by appropriate cellular response in the lung. Accordingly, close interactions between various cell types, including frontline immune cells and adaptive T and B cells, are essential. Importantly, these processes largely rely on cellular secretory function. Therefore, the functional integrity of ER is vital to maintain lung homeostasis against increasing demands of protein folding and secretion. [bib_ref] The UPR and lung disease, Osorio [/bib_ref] For examples, XBP-1 has been reported to be vital for ER/calcium store expansion and subsequent secretion of inflammatory cytokines in airway epithelium. [bib_ref] Airway epithelial inflammation-induced endoplasmic reticulum Ca2+ store expansion is mediated by X-box..., Martino [/bib_ref] Similarly, TLR signaling-induced activation of the IRE1α/ XBP1 pathway in macrophages is required for the production of certain pro-inflammatory cytokines, such as IL-6, IL-8, and TNF. [bib_ref] TLR activation of the transcription factor XBP1 regulates innate immune responses in..., Martinon [/bib_ref] Furthermore, the optimal functional competence of ER and UPR pathways are known to be essential in the development and differentiation of eosinophils, DCs, plasma cells, and subsets of T cells, thereby linking basic cell biology to broad im-munological outcomes in the lung as reviewed elsewhere. [bib_ref] Endoplasmic reticulum stress in immunity, Bettigole [/bib_ref] Previous research has also highlighted the involvement of ER stress and UPR in numerous pulmonary conditions associated with common environmental insults. For example, inhaled fine particulate matter (aerodynamic diameter <2.5 μm, PM2.5) activates ER stress, (especially the PERK/CHOP pathway) and causes ER stress-associated cellular apoptosis through a ROSdependent mechanism. [bib_ref] Airborne particulate matter selectively activates endoplasmic reticulum stress response in the lung..., Laing [/bib_ref] Additionally, the experimental exposure of cigarette smoke and aqueous extracts of cigarette smoke induces ER stress and PERK-mediated survival UPR activation in addition to direct oxidative effects in human lung epithelial cells [bib_ref] Cigarette smoke induces endoplasmic reticulum stress and the unfolded protein response in..., Jorgensen [/bib_ref] and mouse fibroblast, [bib_ref] Endoplasmic reticulum stress induced by aqueous extracts of cigarette smoke in 3T3..., Hengstermann [/bib_ref] respectively, which emphasizes the role of UPR in maintaining cellular redox homeostasis against oxidative stress. Moreover, XBP-1 mediates ER/Ca 2+ store expansion and potentiates IL-8 secretion during airway inflammation associated with Pseudomonas aeruginosa infection [bib_ref] Airway epithelial inflammation-induced endoplasmic reticulum Ca2+ store expansion is mediated by X-box..., Martino [/bib_ref] and IRE1α and PERK pathways are known to suppress viral protein synthesis under viral infection. [bib_ref] The impact of the unfolded protein response on human disease, Wang [/bib_ref]
## Role of er stress in allergic inflammation of safs
For years, there has been much progress in our knowledge on the implication of ER stress in inflammatory processes, which has been vastly reviewed elsewhere. [bib_ref] Regulation of the transcriptome by ER stress: non-canonical mechanisms and physiological consequences, Arensdorf [/bib_ref] [bib_ref] From endoplasmic-reticulum stress to the inflammatory response, Zhang [/bib_ref] As for the respiratory system, ER stress and UPR are closely interconnected with various cellular signaling networks, being involved in numerous inflammatory lung disorders including bronchial asthma. [bib_ref] Endoplasmic reticulum stress influences bronchial asthma pathogenesis by modulating nuclear factor κB..., Kim [/bib_ref] [bib_ref] The UPR and lung disease, Osorio [/bib_ref] [bib_ref] Endoplasmic reticulum stress in chronic obstructive lung diseases, Ribeiro [/bib_ref] [bib_ref] Endoplasmic reticulum stress enhances fibrotic remodeling in the lungs, Lawson [/bib_ref] [bib_ref] Inhibition of endoplasmic reticulum stress alleviates lipopolysaccharide-induced lung inflammation through modulation of..., Kim [/bib_ref] Disturbances in ER homeostasis can influence diverse aspects of allergic inflammatory process in the lung. First, many triggers of asthma have also been demonstrated as potent inducers of ER stress and UPR in the lung. [bib_ref] The UPR and lung disease, Osorio [/bib_ref] For instance, TLR4 activation by house dust mites (HDMs) in airway epithelium is known to be essential in allergic lung inflammation, [bib_ref] House dust mite allergen induces asthma via Toll-like receptor 4 triggering of..., Hammad [/bib_ref] and TLR4 signaling and the IRE1α/XBP1 arm of UPR coordinate immune response in the production of certain cytokines, such as IL-6 and TNF, in macrophages. [bib_ref] TLR activation of the transcription factor XBP1 regulates innate immune responses in..., Martinon [/bib_ref] Additionally, several proteins related to the structural and functional integrity of ER have been known to be implicated in the development of allergic lung inflammation. Reticulon-4 (RTN-4, also known as the Nogo family) is a member of the RTN family of proteins, which is largely restricted to ER and participates in shaping and structuring ER membranes. [bib_ref] A class of membrane proteins shaping the tubular endoplasmic reticulum, Voeltz [/bib_ref] Nogo-B, an isoform of Nogo, is predominantly expressed in the lung and expression of Nogo-B is markedly reduced in a murine model of allergic lung inflammation and human fatal asthmatic lungs. Interestingly, Nogo-knockout mice display exaggerated allergic lung inflammation and epithelial reconstitution of Nogo-B attenuates asthma-like phenotypes in these mice. [bib_ref] Epithelial reticulon 4B (Nogo-B) is an endogenous regulator of Th2-driven lung inflammation, Wright [/bib_ref] Allergen-induced mucin overproduction is one of the hallmarks of bronchial asthma. Given that airway mucins are large (~5,000 amino acid residues) glycoproteins and posttranslational modification of immature mucin is accomplished in ER, ER resident protein disulfide isomerase (PDI) anterior gradient homolog 2 (AGR2), which interacts with immature mucin, is likely to be involved in allergic lung inflammation. Consistent with this assumption, Agr -/mice display ER stress in airway epithelium and impaired mucin production during allergic lung inflammation. [bib_ref] AGR2 is induced in asthma and promotes allergeninduced mucin overproduction, Schroeder [/bib_ref] IRE1β, an isoform of IRE1, mainly expressed in epithelial cells of the gut and the lung, has been reported to be involved in allergen-induced goblet cell differentiation and airway epithelial mucin production partly through XBP-1-dependent transcription of AGR2. [bib_ref] The ER stress transducer IRE1β is required for airway epithelial mucin production, Martino [/bib_ref] Notably, a wellknown asthma-associated gene, orosomucoid-like 3 (ORM-DL3), belongs to ER membrane protein that is involved in membrane biogenesis. [bib_ref] Orm family proteins mediate sphingolipid homeostasis, Breslow [/bib_ref] Moreover, orm1Δ orm2Δ yeast that lacks the yeast members of the ORMDL family displays constitutive UPR and susceptibility to ER stress inducer. [bib_ref] Orm1 and Orm2 are conserved endoplasmic reticulum membrane proteins regulating lipid homeostasis..., Han [/bib_ref] A previous study also demonstrated that ORMDL3 is an allergen and a TH2 cytokine-inducible gene predominantly expressed in airway epithelium, and it induces expression of various mediators of allergic inflammation and UPR specifically the ATF6 pathway, highlighting a possible mechanistic link between the ER protein quality control system and bronchial asthma. [bib_ref] ORMDL3 is an inducible lung epithelial gene regulating metalloproteases, chemokines, OAS, and..., Miller [/bib_ref] Notably, increasing evidence has demonstrated that ER stress may be one of the key players in the development of severe allergic lung inflammation commonly refractory to conventional treatment, such as inhaled/systemic CS. [bib_ref] Endoplasmic reticulum stress and the related signaling networks in severe asthma, Kim [/bib_ref] One important study demonstrated the critical involvement of ER stress in bronchial asthma, especially neutrophilic asthma which commonly manifests the severe disease phenotype having CS resistance. [bib_ref] Endoplasmic reticulum stress influences bronchial asthma pathogenesis by modulating nuclear factor κB..., Kim [/bib_ref] In that study, representative ER stress markers, including GRP78 and CHOP, in the lung were increased in a murine model of neutrophilic asthma and asthmatic patients. Interestingly, inhibition of ER stress using a chemical chaperone, 4-phenylbutyric acid (4-PBA), resulted in the marked improvement of neutrophilic allergic lung inflammation and the reduction in protein levels of ER stress and UPR markers, including ATF6α, XBP-1, and p-eIF2α. However, dexamethasone treatment failed to decrease neither neutrophilic allergic lung inflammation nor elevation of ER stress or UPR proteins. In another study, allergeninduced airway fibrosis, a critical manifestation of severe asthma, was closely associated with ER resident protein 57 (ERp57), an ER-localized chaperone involved in glycoprotein folding and secretion. In that study, levels of ERp57 were predominantly increased in the epithelium of asthmatic patients and murine models of asthma, indicating the increased protein folding load. Notably, allergen-induced increases in collagen and smooth muscle actin, well-known fibrotic markers, were significantly decreased by specific ablation of epithelial ERp57. [bib_ref] Protein disulfide isomerase-endoplasmic reticulum resident protein 57 regulates allergen-induced airways inflammation, fibrosis,..., Hoffman [/bib_ref] Meanwhile, considering the pivotal role of ER stress in severe asthma, ER stress and UPR pathways may contribute to the pathogenesis of fungi-associated severe allergic lung inflammation, including SAFS . Although scarce information exists, interesting data from our group has shown the crucial involvement of these pathways in Aspergillus fumigatus-induced allergic inflammation. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] In this study, Aspergillus fumigatus ex-posure in mice resulted in characteristic features of fungi-induced allergic lung inflammation-including elevated pulmonary TH2-associated cytokines, such as IL-4, IL-5, and IL-13, total and Aspergillus fumigatus-specific IgE in serum, eosinophildominant allergic lung inflammation, and bronchial hyper-responsiveness. Interestingly, treatment with dexamethasone failed to improve Aspergillus fumigatus-induced allergic lung inflammation, suggesting that the murine model displays characteristic features of severe fungal allergic inflammation, such as SAFS. Protein levels of ER stress (GRP78 and CHOP) and UPR-related markers (p-IRE1α, p-eIF2α, XBP-1, and ATF-4) were also remarkably increased in the lung of Aspergillus fumigatus-exposed mice, and increases in GRP78 were observed in the lung tissues samples from patients with ABPA. Importantly, administration of 4-PBA markedly attenuated Aspergillus fumigatus-induced ER stress and improved characteristic features of Aspergillus fumigatus-induced allergic lung inflammation, while dexamethasone treatment did not. These findings suggest that ER stress may play a key role in the pathogenesis of fungal allergic lung inflammation, especially in the manifestation of severe phenotypes of the disease. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] Currently, it is not known exactly on the mechanism through which airway fungal exposure specifically induces eosinophilic allergic inflammation. However, one possible explanation would be related to NF-κB activation, a well-known master regulator of allergic inflammation. It has been demonstrated that close associations of ER stress and UPR with NF-κB signaling play a key role in ER stress-related inflammatory processes, specifically in lung inflammation. [bib_ref] Endoplasmic reticulum stress influences bronchial asthma pathogenesis by modulating nuclear factor κB..., Kim [/bib_ref] [bib_ref] Inhibition of endoplasmic reticulum stress alleviates lipopolysaccharide-induced lung inflammation through modulation of..., Kim [/bib_ref] Similarly, nuclear translocation of NF-κB p65 was remarkably increased in the lung tissues from the murine model of Aspergillus fumigatus-induced fungal allergic lung inflammation. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] In addition, respiratory exposure to Aspergillus fumigatus resulted in increases in pulmonary levels of type 2 cytokines, such as IL-4, IL-5, and IL-13, leading to eosinophilic lung inflammation. Importantly, treatment with an inhibitor of NF-κB into mice significantly reduced Aspergillus fumigatus-induced increases in the levels of type 2 cytokines and eosinophilic allergic inflammation. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] These findings suggest the crucial implication of NF-κB signaling in fungiinduced eosinophilic allergic lung inflammation.
In addition, it seems that several cross-talks between ER stress and other cellular inflammatory signaling platforms exist, all of which can further explain the molecular mechanism of severe allergic lung inflammation associated with fungi. These include phosphoinositide 3-kinase-δ (PI3K-δ), mitochondrial ROS (mtROS), and NLRP3 inflammasome.
## Cross-talks between er stress and cellular inflammatory signaling platforms leading to severe allergic inflammation in safs pi3k-δ
PI3Ks are lipid signaling kinases that control a variety of cru-cial cellular events. Class I PI3Ks are generally associated with cell membrane receptors, such as growth factor and cytokine receptors, and phosphorylate the 3´-position of inositol lipids to generate second messenger phosphatidylinositol-3, 4, 5-trisphosphate at the cell membrane, which serves as a cellular docking molecular platform for proteins possessing pleckstrinhomology-domain-containing proteins, such as AKT. [bib_ref] PI3K delta and PI3K gamma: partners in crime in inflammation in rheumatoid..., Rommel [/bib_ref] This process leads to downstream cascades of protein-protein interactions and phosphorylation, culminating in multiple biological consequences. Class I PI3Ks consist of four members (namely, PI3K-α, PI3K-β, PI3K-γ, and PI3K-δ), all of which are heterodimeric complexes having a catalytic p110 subunit (α, β, γ, or δ) in association with a regulatory subunit. While expressions of p110-α and p110-β isoforms are ubiquitous in most cell types, expressions of p110-γ and p110-δ isoforms are restricted to circulating leukocytes and both isoforms have been reported to play key roles in leukocyte signaling. [bib_ref] Signalling by PI3K isoforms: insights from gene-targeted mice, Vanhaesebroeck [/bib_ref] Particularly, PI3K-δ plays specific roles in a variety of immunologic processes involving antigen receptor signaling in T cells and B cells, mast cell degranulation, and the migration and activation of neutrophils and eosinophils. The crucial role of PI3Ks, especially PI3K-δ isoform in allergic lung inflammation has also been uncovered. [bib_ref] Involvement of PTEN in airway hyperresponsiveness and inflammation in bronchial asthma, Kwak [/bib_ref] [bib_ref] Inhibition of phosphoinositide 3-kinase delta attenuates allergic airway inflammation and hyperresponsiveness in..., Lee [/bib_ref] Blockade of PI3K-δ has also shown to attenuate CS-resistant severe inflammatory processes, including severe asthma and COPD. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] [bib_ref] Corticosteroid resistance in patients with asthma and chronic obstructive pulmonary disease, Barnes [/bib_ref] Therefore, inhibition of PI3K-δ is regarded as a promising target for developing a novel drug for chronic inflammatory airway disorders, especially for CS-resistant inflammation. PI3K pathways seem to be closely associated with ER stress and UPR. One report demonstrated that regulatory subunits of PI3Ks (p85) participate in the translocation and possibly stabilization of XBP-1, and that disruption of this interaction results in a severe defect in lowering ER stress. [bib_ref] The regulatory subunits of PI3K, p85alpha and p85beta, interact with XBP-1 and..., Park [/bib_ref] In addition, inhibition of PERK leads to decreased activation of PI3K/AKT signaling through regulating cellular localization of PTEN, a negative regulator of the PI3K pathway. [bib_ref] Feedback regulation on PTEN/AKT pathway by the ER stress kinase PERK mediated..., Zhang [/bib_ref] Furthermore, ER stress-induced Sirtuin 1 (SIRT1) expression, which is involved in ER stress-induced damage responses, has been shown to be regulated by the PI3K-AKT signaling pathway. [bib_ref] Endoplasmic reticulum (ER) stress induces sirtuin 1 (SIRT1) expression via the PI3K-Akt-GSK3β..., Koga [/bib_ref] Remarkably, we recently demonstrated the critical involvement of PI3K-δ in the regulation of ER stress in a murine model of SAFS . [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] In this study, the blockade of PI3K-δ ameliorated Aspergillus fumigatus-induced severe eosinophilic allergic lung inflammation that was refractory to dexamethasone. Furthermore, inhibition of PI3K-δ significantly reduced Aspergillus fumigatus-induced increases in ER stress and UPR markers, particularly in airway epithelial cells. Eventually, therapeutic effects of PI3K-δ blockade were closely linked to the attenuation of ER stress-associated NF-κB activation. Taken together, an important cross-talk between ER stress and PI3K-δ is likely to be implicated in the pathogenesis of SAFS. Oxidative stress has long been proposed as one of the essential features in chronic airway disorders. [bib_ref] Oxidants and the pathogenesis of lung diseases, Ciencewicki [/bib_ref] Exposure to various oxidants has been reported to cause several cardinal features of asthma, including allergic airway inflammation and airway hyper-responsiveness (AHR). [bib_ref] Systemic oxidative stress in asthma, COPD, and smokers, Rahman [/bib_ref] [bib_ref] A novel dithiol amide CB3 attenuates allergic airway disease through negative regulation..., Kim [/bib_ref] In the same manner, the functional incompetence of cellular anti-oxidant systems is also known to be closely implicated in the pathogenesis of bronchial asthma through inducing oxidative stress. ROS can impact on numerous aspects of the inflammatory process in the lung. First, ROS can activate a wide range of cellular signaling by themselves. ROS also interact with diverse biomolecules, such as lipids and proteins, thereby producing secondary mediators which possess a wide range of effects in the body. Moreover, ROS can directly cause protein modification and DNA damage. Furthermore, oxidative stress can be even more intensified through the recruitment of diverse inflammatory cells, another important endogenous source of ROS, to the lung. [bib_ref] Oxidants and the pathogenesis of lung diseases, Ciencewicki [/bib_ref] Given that quantities of ROS are vital for determining its specificity and function and that mitochondria are regarded as the most powerful source of intracellular ROS, mtROS should be tightly regulated in cells. Numerous factors can regulate mtROS in the processes of generation or elimination. [bib_ref] Physiological roles of mitochondrial reactive oxygen species, Sena [/bib_ref] Generation of mtROS can be controlled by various cellular stimuli, including increased cytosolic Ca 2+ concentration, activation of cellular signaling pathways involving immunoreceptors and cytokines, redox status of electron transport chain (ETC), and electrical gradient in inner mitochondrial membrane. [bib_ref] Physiological roles of mitochondrial reactive oxygen species, Sena [/bib_ref] [bib_ref] TCA cycle and mitochondrial membrane potential are necessary for diverse biological functions, Martínez-Reyes [/bib_ref] As for the removal of mtROS, several essential antioxidant systems, including glutathione peroxidases, peroxiredoxins, and catalase, can eliminate overproduced mtROS.
Furthermore, production of ROS and subsequent development of oxidative stress are known to be important in modulating the protein-folding capacity of ER. [bib_ref] Endoplasmic reticulum stress and oxidative stress: a vicious cycle or a double-edged..., Malhotra [/bib_ref] [bib_ref] Antioxidants reduce endoplasmic reticulum stress and improve protein secretion, Malhotra [/bib_ref] In particular, ER stress and mtROS can be interconnected, affecting both aspects of mtROS regulation in the cell. Previous reports have shown that increased leak of Ca 2+ from the ER lumen in response to ER stress or cellular oxidative stress can influence the generation of mtROS from ETC as a consequence of Ca 2+ accumulation in the mitochondria. [bib_ref] Physiological roles of mitochondrial reactive oxygen species, Sena [/bib_ref] [bib_ref] Endoplasmic reticulum stress induces calcium-dependent permeability transition, mitochondrial outer membrane permeabilization and..., Deniaud [/bib_ref] Then, increased generation of mtROS further impact on the Ca 2+ release channel in the ER membrane, [bib_ref] Nitric oxide-dependent modification of the sarcoplasmic reticulum Ca-ATPase: localization of cysteine target..., Viner [/bib_ref] thereby increasing Ca 2+ -release from ER. Furthermore, de-
## Fig. 2.
Cytoplasmic interactions involving endoplasmic reticulum (ER), mitochondria, and NLRP3 inflammasome may contribute to fungi-induced severe eosinophilic allergic inflammation in the lung. Initial fungal recognition activates the cell membrane-associated phosphoinositide 3-kinase delta (PI3K-δ) signaling pathway in various structural cells and immune cells. This process is followed by the downstream cascade of protein interactions and phosphorylation, leading to diverse biological consequences in these cells. Particularly, PI3K-δ modulates fungi-induced ER stress and UPR activation, especially in airway epithelial cells. Furthermore, mitochondrial generation of reactive oxygen species (mtROS) and NLRP3 inflammasome activation in response to respiratory fungal exposure are closely related to this process. The net result of these associations may play a key role in the pathogenesis of fungi-induced severe eosinophilic allergic inflammation in the lung.
creased mitochondrial functional integrities result in release of more antioxidant enzyme, such as glutathione in the mitochondrial matrix as well, perpetuating oxidative stress as a vicious cycle. [bib_ref] Physiological roles of mitochondrial reactive oxygen species, Sena [/bib_ref] [bib_ref] Endoplasmic reticulum stress and oxidative stress: a vicious cycle or a double-edged..., Malhotra [/bib_ref] Although the precise role of mtROS is not fully understood in allergic lung inflammation, a recent study showed that mtROS generation against common allergens is important in allergic inflammation, especially in airway structural cells. [bib_ref] NLRP3 inflammasome activation by mitochondrial ROS in bronchial epithelial cells is required..., Kim [/bib_ref] In that study, increased mtROS generation was observed in airway inflammatory cells and tracheal epithelial cells from ovalbumin (OVA)/lipopolysaccharide (LPS)-sensitized and OVA-challenged mice which showed neutrophil-dominant allergic inflammation. Interestingly, treatment with a potent mtROS scavenger, NecroX compound, 119 reduced allergen-induced mtROS generation in these cells and ameliorated various features of allergic lung inflammation through modulating NLRP3 inflammasome activation in airway epithelial cells. Furthermore, the therapeutic effect of mtROS scavenger was also demonstrated in another murine model of allergic lung inflammation induced by HDMs.
Importantly, mtROS may also be closely interconnected with ER stress in the pathogenesis of fungi-induced severe allergic lung inflammation. [bib_ref] Phosphoinositide 3-kinase-δ regulates fungus-induced allergic lung inflammation through endoplasmic reticulum stress, Lee [/bib_ref] In that study, increased production of mtROS was observed in airway inflammatory cells and tracheal epithelial cells from Aspergillus fumigatus-sensitized/challenged mice. Moreover, Aspergillus fumigatus-exposed mice showed decreased glutathione and increased glutathione disulfide levels in the lung compared to the control mice, which indicates the presence of oxidative stress. Treatment with a mtROS scavenger improved various features of Aspergillus fumigatus-induced severe eosinophilic allergic lung inflammation, and a mtROS scavenger also lowered Aspergillus fumigatus-induced elevations of ER stress markers in the lung. These results suggest that mtROS contribute to the modulation of ER stress in fungal allergic inflammation, and that this crosstalk may be involved in the development of SAFS .
## Nlrp3 inflammasome
NLRP3 inflammasome has been reported to be essential in anti-fungal immune response. [bib_ref] Syk kinase signalling couples to the Nlrp3 inflammasome for anti-fungal host defence, Gross [/bib_ref] However, the role of NLRP3 inflammasome in the pathogenesis of fungus-induced allergic lung inflammation remains poorly understood. Considering that ER stress can induce the release of diverse damage-associated molecular patterns (DAMPs) from mitochondria (e.g. mtROS, mitochondrial DNA [mtDNA], ATP, calcium), which are potent activators of NLRP3 inflammasome in the cytosol, [bib_ref] Endoplasmic reticulum stress activates the inflammasome via NLRP3-and Caspase-2-driven mitochondrial damage, Bronner [/bib_ref] interconnections between NLRP3 inflammasome and ER stress may play a role in the pathogenesis of fungi-induced allergic lung inflammation. Furthermore, because NLRP3 inflammasome has been reported to be associated with CS-resistant inflammation in the lung, 118,121 ER-NLRP3 inflammasome interactions are likely to be involved in the development of SAFS.
Interestingly, in our unpublished data, exposure to Aspergillus fumigatus resulted in NLRP3 inflammasome activation in the lung of mice, especially in bronchial epithelial cells. Expression of NLRP3 in the lung tissues from patients with ABPA was also increased. Furthermore, a selective inhibitor of NLRP3 inflammasome [bib_ref] A small-molecule inhibitor of the NLRP3 inflammasome for the treatment of inflammatory..., Coll [/bib_ref] reversed Aspergillus fumigatus-induced increases in IL-1β protein in the lung of mice, and treatment with IL-1βneutralizing antibody dramatically attenuated Aspergillus fumigatus-induced severe eosinophilic allergic lung inflammation. Notably, a mtROS scavenger also reversed the Aspergillus fumigatus-induced increases in IL-1β and attenuated characteristic features of Aspergillus fumigatus-induced severe inflammation. These results suggest that NLRP3 inflammasome in association with mtROS/ER stress may be one of the pivotal players in the development of SAFS .
# Conclusions
Previously unappreciated roles of fungi in bronchial asthma, especially in a more severe phenotype of the disease, have been increasingly demonstrated through numerous epidemiologic and translational studies. However, much knowledge on this field is unanswered so far. Recently, SAFS has been proposed as one of the fungal sensitization/allergy development-associated clinical disease entities to further investigate fungi-associated severe asthma phenotypes. Importantly, recent advances in our knowledge that intracellular organellar stress responses are closely implicated in the development of SAFS have broadened our understandings on this issue. ER stress and UPR are closely implicated in fungi-induced severe allergic lung inflammation, and amelioration of ER stress, especially in epithelial cell layer, may have the potential for treating the disease. More importantly, intracellular cross-talks between ER stress and other inflammatory signaling platforms may provide some clues for the vital role of fungi on severe allergic inflammation in humans. Further investigations on the interconnections between these stress responses and the other potential key mediators of CS resistance will expand our knowledge on SAFS in the future. Finally, efficacies of treatment strategies targeting ER-associated molecular networks in SAFS need to be thoroughly evaluated through well-designed clinical trials.
[table] Table: Diagnostic criteria of ABPA/ABPM and SAFS [/table]
|
Clinical Assessment and Management of Delirium in the Palliative Care Setting
Delirium is a neurocognitive syndrome arising from acute global brain dysfunction, and is prevalent in up to 42% of patients admitted to palliative care inpatient units. The symptoms of delirium and its associated communicative impediment invariably generate high levels of patient and family distress. Furthermore, delirium is associated with significant patient morbidity and increased mortality in many patient populations, especially palliative care where refractory delirium is common in the dying phase. As the clinical diagnosis of delirium is frequently missed by the healthcare team, the case for regular screening is arguably very compelling. Depending on its precipitating factors, a delirium episode is often reversible, especially in the earlier stages of a life-threatening illness. Until recently, antipsychotics have played a pivotal role in delirium management, but this role now requires critical reevaluation in light of recent research that failed to demonstrate their efficacy in mild-to moderate-severity delirium occurring in palliative care patients. Non-pharmacological strategies for the management of delirium play a fundamental role and should be optimized through the collective efforts of the whole interprofessional team. Refractory agitated delirium in the last days or weeks of life may require the use of pharmacological sedation to ameliorate the distress of patients, which is invariably juxtaposed with increasing distress of family members. Further evaluation of multicomponent strategies for delirium prevention and treatment in the palliative care patient population is urgently required.Key PointsElectronic supplementary material The online version of this article (
In inpatient palliative care settings, delirium prevalence increases from 13-42% on admission to 88% in the last weeks-hours of life. Delirium causes significant morbidity, including increased frequency of falls, increased cognitive and functional impairment, and significant patient and family psychological distress, and is associated with increased mortality.
Although antipsychotics are commonly used in the management of delirium in palliative care patients, recent research evidence in mild-to moderateseverity delirium suggests that antipsychotics are associated with both increased delirium symptoms and reduced patient survival.
While non-pharmacological delirium strategies should intuitively be an integral part of quality patient care, their role in the management of delirium in the palliative care context is unclear; outcome evidence in terms of their effectiveness in this patient population is still required.
# Introduction and aims
Delirium is a complex multifactorial syndrome resulting from global organic cerebral dysfunction. The prevalence of delirium is 18-35% in general medical inpatients and up to 50% in intensive care patients [bib_ref] Delirium in elderly people, Inouye [/bib_ref]. In a palliative setting, prevalence is reported as 13-42% on admission to inpatient palliative care units, increasing to 88% at the end of life (weeks-hours before death) [bib_ref] Delirium prevalence, incidence, and implications for screening in specialist palliative care inpatient..., Hosie [/bib_ref]. Advanced age and dementia are common predisposing risk factors across most healthcare settings, with a 56% incidence of delirium in patients with dementia [bib_ref] Delirium in elderly people, Inouye [/bib_ref]. Delirium has many negative consequences, including a strong association with higher morbidity and mortality [bib_ref] The frequency of missed delirium in patients referred to palliative care in..., De La [/bib_ref]. It is associated with an increased incidence of falls, longer hospital stays, and greater healthcare costs [bib_ref] Delirium in elderly people, Inouye [/bib_ref]. Delirium appears to worsen pre-existing dementia as well as increase the risk of 'de novo' dementia [bib_ref] Delirium in elderly people, Inouye [/bib_ref] [bib_ref] Association of delirium with cognitive decline in late life: a neuropathologic study..., Davis [/bib_ref]. Delirium impairs patient communication, thus challenging symptom assessment in palliative care patients.
The World Health Organization (WHO) defines palliative care as ''an approach that improves the quality of life of patients and their families facing the problem associated with life-threatening illness, through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual''. Palliative care is provided by specialists and generalists across various care settings to meet patient needs: specialist inpatient palliative care units (including those in acute care settings and stand-alone inpatient hospices), acute care with hospital consult teams, and community palliative care services. Palliative care is applicable anywhere in a person's illness trajectory, from diagnosis to end of life, and includes bereavement support (see . For the purposes of this paper, unless otherwise specified, we use 'end of life' to mean the presence of progressive life-limiting disease in a patient with a prognosis of weeks-months (as opposed to years), as proposed by Hui et al. [bib_ref] Concepts and definitions for ''actively dying,'' ''end of life,'' ''terminally ill,'' ''terminal..., Hui [/bib_ref] in a systematic review of palliative care definitions and concepts and 'dying phase' (viewed as synonymous with terminal phase or actively dying) as ''the hours or days preceding imminent death, during which time the patient's physiologic functions wane'' [bib_ref] Anticholinergic medications for managing noisy respirations in adult hospice patients, Kintzel [/bib_ref].
The aim of this review article is to provide a clinically oriented guide to the management of delirium in adult patients in palliative care settings. However, prompt and effective recognition of delirium and its etiologic presentation, together with patient and family-centered, ethical healthcare decision-making, and a preventive approach, are pre-requisites in an effective and comprehensive management plan, and will therefore be addressed.
## Literature search
The systematic literature search to inform this narrative review was conducted in Cochrane, Ovid MEDLINE, EMBASE, and SCOPUS databases from 2006 to 3 January 2017. Search terms included ''delirium'', ''confusion'', ''palliative'', ''prevention'', ''antipsychotic agents'', ''haloperidol'', ''alpha adrenergic agents'', ''cholinesterase inhibitors'', ''benzodiazepines'', ''therapeutics'', ''medications'', ''drugs'', ''pharmacology'', and ''treatment''. Results were limited to the English language. All original human clinical trials, systematic reviews, meta-analyses, and clinical practice guidelines were included. All case reports and editorials were excluded, as were papers that focused on alcohol withdrawal delirium (delirium tremens) and pediatric populations.
## Clinical features and standard diagnostic criteria for delirium
Although the global cerebral dysfunction associated with delirium is manifested as multiple symptoms and signs, collectively constituting a neurocognitive or neuropsychiatric disorder, the hallmark of delirium is impaired attention. The core clinical criteria for a diagnosis are codified in the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM), and the fifth edition (DSM-5)updated its itemized diagnostic criteria for delirium. This entailed changing the essential diagnostic feature to ''a disturbance in attention and awareness'', as compared to ''a disturbance of consciousness with reduced ability to focus, sustain or shift attention'' in the previous DSM version [fourth edition, text revision (DSM-IV-TR)]. Additionally, delirium is a change from baseline attention and awareness developing over a short period of time (hours-days) with fluctuations in severity over a 24-h time period. Other changes in cognition may also occur. Examples include disorientation, memory deficit, and disturbances in language, visuospatial ability, or perception. Additional clinical features of delirium (that are not part of the core DSM-5 criteria) include sleep-wake cycle disturbance, delusions, dysarthria, dysgraphia, emotional lability, and abnormal psychomotor activity (hypo-or hyperactivity). In countries other than the USA, the International Classification of Diseases (ICD) classification is frequently used. The ICD-10 criteria overlap with some of the DSM-5 features and also provide additional criteria for disturbances in psychomotor function, sleep-wake cycle, and emotional state. Sleep disturbance (with difficulty falling asleep and frequent awakening leading to sleepless nights) is one of the earliest prodromal symptoms of delirium [bib_ref] Progression of delirium in advanced illness: a multivariate model of caregiver and..., Kerr [/bib_ref]. Other prodromal features include irritability, anxiety, and restlessness [bib_ref] Progression of delirium in advanced illness: a multivariate model of caregiver and..., Kerr [/bib_ref]. Subsyndromal delirium (SSD) has also been described where a person has some but not all of the delirium features and so does not meet the full diagnostic criteria.
## Approach to clinical assessment of delirium
Delirium is a clinical diagnosis that is frequently overlooked or misdiagnosed by the healthcare team [bib_ref] The frequency of missed delirium in patients referred to palliative care in..., De La [/bib_ref] [bib_ref] Under-reporting of delirium in the NHS, Clegg [/bib_ref] [bib_ref] Are nurses recognizing delirium? A systematic review, Steis [/bib_ref]. This may be due to lack of recognition of delirium features, overlap with the clinical features of depression and dementia in particular, the occurrence of fluctuating symptom intensity and transient lucidity, and is compounded by inadequate delirium screening or an absence of regular screening [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref]. [fig_ref] Figure 2: Algorithm for the assessment and management of delirium in palliative care patients [/fig_ref] provides an algorithm for the assessment and management of delirium in palliative care.
## Delirium screening
The detection of delirium may be improved by the routine use of observational screening tools by nursing staff and cognitive screening tools or brief tests of attention, such as reciting the months of the years backwards. The choice of clinical tool may in part be determined by the level of staff training required, ease of administration and the potential level of patient, family, and staff burden [bib_ref] Practical assessment of delirium in palliative care, Leonard [/bib_ref]. It is also extremely useful to obtain a collateral history from a caregiver regarding a person's baseline cognitive function, and for caregivers to report any acute change in mental status to the healthcare team. The Single Question in Delirium (SQiD) asks the patient's family or friend: ''Do you feel that [patient's name] has been more confused lately?'' [bib_ref] Single Question in Delirium (SQiD): testing its efficacy against psychiatrist interview, the..., Sands [/bib_ref].
## Confusion assessment method (cam)
A commonly used screening instrument is the Confusion Assessment Method (CAM) [bib_ref] Clarifying confusion: the confusion assessment method. A new method for detection of..., Inouye [/bib_ref]. The full version of the CAM instrument has a total of nine items based on the DSM-III-R criteria. The CAM diagnostic algorithm consists of four items from the full CAM and has been validated in palliative care [bib_ref] Validation of the confusion assessment method in the palliative care setting, Ryan [/bib_ref]. In order to identify a CAM-positive delirium, the two essential features of ''acute onset and fluctuating course'' and ''inattention'' are required, in addition to ''disorganized thinking'' or ''altered level of consciousness''. The CAM is a copyrighted instrument which is available from the website for the Hospital Elder Life Program (HELP). Formal CAM training is essential to ensure reliability. The CAM was designed to be administered with a brief cognitive assessment tool, such as the Mini-Mental State Examination (MMSE) or Short Orientation Memory Concentration Test (SOMCT) as cognitive disturbance is also a diagnostic criterion for delirium [bib_ref] Mini-mental state'': a practical method for grading the cognitive state of patients..., Folstein [/bib_ref] [bib_ref] Validation of a short Orientation-Memory-Concentration Test of cognitive impairment, Katzman [/bib_ref]. The recently described 3D-CAM operationalizes the CAM using a 3-min assessment [bib_ref] 3D-CAM: derivation and validation of a 3-minute diagnostic interview for CAM-defined delirium:..., Marcantonio [/bib_ref]. [fig_ref] Table 1: Commonly used delirium screening tools[17,23,24] [/fig_ref] shows some commonly used delirium screening tools.
## Making a diagnosis of delirium
The ideal approach to making the diagnosis of delirium in palliative care is to conduct a clinical interview, obtain a collateral history, and assess the patient's cognition in a systematic manner [bib_ref] Delirium diagnosis, screening and management, Lawlor [/bib_ref] , which together allow the assessor to operationalize the DSM-5 delirium diagnostic criteria. These criteria consist of an attention and awareness disturbance; a temporal acuity of onset (hours-days); an additional cognitive deficit; a requirement that the cognitive disturbance is not better explained by an existing or evolving dementia or a comatose state; and, lastly, that there is evidence from the clinical assessment or investigations to indicate that the disturbance is due to a medical condition, a substance (intoxication or withdrawal), a toxin, or multiple factors. In clinical practice, the CAM is frequently used as a diagnostic tool. To a large degree, the CAM practically operationalizes the DSM-5 criteria for delirium, which is supported by their demonstrated significant agreement [bib_ref] A comparison of delirium diagnosis in elderly medical inpatients using the CAM,..., Adamis [/bib_ref]. Although this process is often straightforward and effectively administered, this is not always the case.
Challenges arise in the context of the palliative care setting where the burden of assessment and investigation has to be weighed against the need for thoroughness, especially in the dying phase. Furthermore, co-morbidity levels tend to increase with disease progression, as do levels of frailty, cachexia, fatigue and pain, which collectively add to the challenge of conducting a thorough diagnostic assessment of delirium. The phenomenological overlap between the diagnoses of dementia, delirium and depression warrants consideration and differentiation, especially the attentional deficit and acuity of onset that characterize delirium. The diagnosis of delirium superimposed on dementia is often very challenging and may only be exposed when therapeutic interventions to reverse delirium are only partly effective [bib_ref] Delirium diagnostic and classification challenges in palliative care: subsyndromal delirium, comorbid delirium-dementia,..., Leonard [/bib_ref]. Clinicians also need to recognize delirium psychomotor abnormalities in their diagnostic assessments. Examples of delirium assessment tools for diagnosis and severity ratings are shown in [fig_ref] Table 2: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, ICD-10 International Classification... [/fig_ref].
## Delirium subtype classification
Three clinical subtypes of delirium have been described according to the type of psychomotor activity: hyperactive, hypoactive, and mixed. In palliative care patients, the hypoactive and mixed subtypes are the most common. The hypoactive subtype is often missed [bib_ref] Hypoactive delirium: assessing the extent of the problem for inpatient specialist palliative..., Spiller [/bib_ref] as these patients become somnolent and withdrawn with impaired cognition but with no visible agitation, or is misdiagnosed as depression or fatigue especially as patients become frailer. However, patients with hypoactive delirium may also experience perceptual disturbances (e.g., hallucinations) and delusions in addition to experiencing considerable distress and having a poorer prognosis [bib_ref] Phenomenology of the subtypes of delirium: phenomenological differences between hyperactive and hypoactive..., Boettger [/bib_ref] [bib_ref] The delirium experience: delirium recall and delirium-related distress in hospitalized patients with..., Breitbart [/bib_ref] [bib_ref] A longitudinal study of motor subtypes in delirium: relationship with other phenomenology,..., Meagher [/bib_ref]. Diagnosing delirium in patients approaching the dying phase, especially the hypoactive subtype, is challenging given patients' reduced level of consciousness and reduced communication as their condition declines.
## Pathophysiology of delirium
The pathophysiology of delirium is complex and not fully understood. In view of both the multifactorial nature of delirium and multiple disease contexts in which delirium is likely to occur, it is perhaps not surprising that no single unifying theory exists that encompasses its diverse etiologic spectrum. An in-depth review of delirium pathophysiology is beyond the scope of this clinically oriented review; the neuropathogenesis of delirium has been comprehensively reviewed elsewhere [bib_ref] Neuropathogenesis of delirium: review of current etiologic theories and common pathways, Maldonado [/bib_ref]. Maldonado [bib_ref] Neuropathogenesis of delirium: review of current etiologic theories and common pathways, Maldonado [/bib_ref] has identified seven main theories that may, to varying degrees, be used to explain the neuropathogenesis of delirium; he suggests that these mechanistic hypotheses do not so much compete with each other as tend to complement and sometimes overlap each other. These theories relate to various cerebral clinicopathological entities, including neuroinflammation, neuronal aging, oxidative stress, various neurotransmitter alterations, neuroendocrine aberrations, and melatonin dysregulation, in addition to a breakdown in the integration of neuronal networks. More recently, Maldonado [bib_ref] Acute brain failure: pathophysiology, diagnosis, management, and sequelae of delirium, Maldonado [/bib_ref] has sought to integrate the theory of neurotransmitter dysregulation with the neuronal network dysconnectivity theory in a newly proposed ''systems integration failure hypothesis'' (SIFH). In the context of palliative care, in which many patients have a cancer diagnosis, the high prevalence of systemic inflammation is a possible contributor in the pathogenesis of delirium [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref]. Blood-brain barrier dysfunction in the context of systemic inflammation may result in cytokines and inflammatory mediators crossing into the central nervous system (CNS) and promoting neuroinflammation with consequent delirium [bib_ref] Neuropathogenesis of delirium: review of current etiologic theories and common pathways, Maldonado [/bib_ref] [bib_ref] The cholinergic system and inflammation: common pathways in delirium pathophysiology, Cerejeira [/bib_ref]. The fundamental therapeutic strategy for an episode of delirium entails treating its precipitating factors where possible and targeting its symptomatic distress, if necessary through pharmacological intervention [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref]. The mechanism of neurotransmitter imbalance (mostly cholinergic deficiency and/or relative excess of dopaminergic neurotransmission) as a final common pathway has to date largely underpinned the pharmacological approach of antidopaminergic antipsychotic use in the treatment of distressing delirium symptoms [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref] [bib_ref] Is there a final common neural pathway in delirium? Focus on acetylcholine..., Trzepacz [/bib_ref]. However, the extent of neurotransmitter disturbance in the pathogenesis of delirium is increasingly realized as being much broader and more complex than the dopaminergic-cholinergic imbalance. The most common additional neurotransmission disturbances include excess norepinephrine and/or glutamate; increased or decreased c-aminobutyric acid (GABA), N-methyl-D-aspartate (NMDA), and 5-hydroxytryptamine (5-HT; serotonin); and decreased melatonin [bib_ref] Neuropathogenesis of delirium: review of current etiologic theories and common pathways, Maldonado [/bib_ref]. These disturbances may to varying degrees provide other therapeutic targets for a pharmacologic approach to symptom management [bib_ref] Valproic acid for treatment of hyperactive or mixed delirium: rationale and literature..., Sher [/bib_ref] [bib_ref] Intravenous ketamine for rapid opioid dose reduction, reversal of opioid-induced neurotoxicity, and..., Winegarden [/bib_ref] [bib_ref] Preventive effects of ramelteon on delirium: a randomized placebo-controlled trial, Hatta [/bib_ref].
## Etiologic considerations
The etiology of delirium is often viewed as being multifactorial and this is invariably the case in the context of advanced disease in palliative care settings. Inouye and Charpentier's conceptual framework to describe the etiology of delirium in the elderly highlights the interrelationship between a patient's baseline vulnerability, as reflected by potential predisposing factors such as old age, and superimposed more acute precipitating factors such as infection [bib_ref] Precipitating factors for delirium in hospitalized elderly persons: predictive model and interrelationship..., Inouye [/bib_ref]. This model is readily applicable to delirium in palliative care, where baseline vulnerability relates to factors such as age, frailty, poor nutritional status, impaired functional status, chronic renal impairment, and pre-existing dementia. Precipitating factors for delirium include medications, especially psychoactive types, notably opioids and benzodiazepines; infection; metabolic factors such as organ dysfunction and electrolyte disturbance; hypoxia; and dehydration. Etiologic factors for patients with cancer are summarized in . A prospective study in a tertiarylevel palliative care unit identified a median of three (range 1-6) precipitants per episode of delirium in advanced cancer patients [bib_ref] Occurrence, causes, and outcome of delirium in patients with advanced cancer: a..., Lawlor [/bib_ref]. Similarly, frail geriatric patients have multiple precipitants for acute delirium [bib_ref] Predisposing and precipitating factors for delirium in a frail geriatric population, Laurila [/bib_ref]. Due to the multifactorial nature of delirium in the context of palliative care, the approach to both prevention and treatment accordingly warrant a multicomponent approach [bib_ref] Non-pharmacological interventions to prevent or treat delirium in older patients: clinical practice..., Abraha [/bib_ref].
## Prevention
A recent meta-analysis showed that multicomponent nonpharmacological interventions reduce the incidence of delirium by over 40% for older adult patients (without a terminal illness) in hospital or long-term residential care settings [bib_ref] Effectiveness of multicomponent nonpharmacological delirium interventions: a meta-analysis, Hshieh [/bib_ref]. The 2010 National Institute for Health and Clinical Excellence (NICE) guideline for the prevention of delirium (which excluded dying patients) made 13 recommendations for the prevention of delirium in at-risk adults, and also showed that this approach was cost effective on economic analysis [bib_ref] Synopsis of the National Institute for Health and Clinical Excellence guideline for..., O'mahony [/bib_ref]. A multicomponent intervention package can be applied by all members of the interprofessional team, with appropriate training and supportive healthcare system structures. However, at this time, the efficacy and applicability of these approaches for terminally ill patients is not clear [bib_ref] Effectiveness of multicomponent nonpharmacological delirium interventions: a meta-analysis, Hshieh [/bib_ref] [bib_ref] Delirium prevention in terminal cancer: assessment of a multicomponent intervention, Gagnon [/bib_ref] [bib_ref] In-facility delirium prevention programs as a patient safety strategy: a systematic review, Reston [/bib_ref].
## Non-pharmacological interventions to prevent delirium
In a seminal paper, Inouye et al. [bib_ref] A multicomponent intervention to prevent delirium in hospitalized older patients, Inouye [/bib_ref] reported on a multicomponent intervention for six delirium risk factors (cognitive impairment, sleep deprivation, immobility, visual and hearing impairment, and dehydration) in 852 hospitalized patients C70 years of age, which resulted in a significant reduction in both delirium incidence in the intervention group (9.9% vs. 15% in the usual care group) and duration of delirium episodes. This model of care was named HELP [bib_ref] The Hospital Elder Life Program: a model of care to prevent cognitive..., Inouye [/bib_ref]. The innovative HELP program for older patients now comprises multiple protocols, including protocols for orientation, therapeutic activities, fluid replacement, early mobilization, feeding assistance, vision, hearing, sleep enhancement, and review of psychoactive medications, and is led by trained volunteers and family members [bib_ref] NICE to HELP: operationalizing National Institute for Health and Clinical Excellence guidelines..., Yue [/bib_ref]. As part of fundamental clinical practice, clinicians should ensure that patients are hydrated, avoid unnecessary catheterization, optimize sleep hygiene, encourage mobilization, avoid sensory deprivation (with appropriate lighting and use of hearing and visual aids), and use verbal orientation and devices such as a visible clock and orientation board [bib_ref] Diagnosis, prevention, and management of delirium: summary of NICE guidance, Young [/bib_ref]. Another evidence-based strategy for delirium prevention includes comprehensive geriatric assessment in perioperative elderly patients [bib_ref] Reducing delirium after hip fracture: a randomized trial, Marcantonio [/bib_ref]. Seefor non-pharmacological strategies that can be implemented by the healthcare team and family for palliative care patients.
## Pharmacological interventions to prevent delirium
A recent Cochrane review [bib_ref] Interventions for preventing delirium in hospitalised non-ICU patients, Siddiqi [/bib_ref] found no clear evidence for a reduction in delirium incidence with cholinesterase inhibitors or perioperative antipsychotic medications from randomized controlled trials (RCTs) in hospitalized nonintensive care patients, except for one moderate-quality study comparing perioperative olanzapine with placebo in 400 elderly participants undergoing joint replacement surgery [bib_ref] Administration of olanzapine to prevent postoperative delirium in elderly joint-replacement patients: a..., Larsen [/bib_ref]. However, in this study, delirium duration and severity was greater in the olanzapine arm. A double-blind, placebo-controlled randomized trial of low-dose melatonin in 145 elderly internal medicine inpatients reported a reduced risk of delirium: 12% in melatonin treated group versus 31% in placebo group (p = 0.014) [bib_ref] Melatonin decreases delirium in elderly patients: a randomized, placebo-controlled trial, Al-Aama [/bib_ref]. A recent pilot study of melatonin in 30 advanced cancer inpatients reported a reduced incidence of delirium of 7% (1/14) in the melatonin arm versus 25% (4/16) in the placebo arm [bib_ref] Palliative Care Clinical Studies Collaborative. Randomised double blind placebo controlled phase II..., Agar [/bib_ref].
## Decision-making in the management approach
Although delirium complicates the last hours or days of life for most patients in palliative care settings [bib_ref] Delirium prevalence, incidence, and implications for screening in specialist palliative care inpatient..., Hosie [/bib_ref] , delirium is also characterized by the potential for reversal in certain circumstances, depending on the nature of the precipitant factors and the degree of baseline vulnerability. Antibiotics for infection and bisphosphonates for hypercalcemia are examples of treatable precipitants should such therapeutic interventions be consistent with the patient's goals of care. In palliative care, evidence suggests that approximately 50% of delirium episodes can be reversed [bib_ref] Occurrence, causes, and outcome of delirium in patients with advanced cancer: a..., Lawlor [/bib_ref] , especially those precipitated by medications, infection, and electrolyte abnormalities [bib_ref] Occurrence, causes, and outcome of delirium in patients with advanced cancer: a..., Lawlor [/bib_ref] [bib_ref] Reversibility of delirium in terminally ill patients and predicators of mortality, Leonard [/bib_ref]. However, reversible delirium and irreversible delirium often share a very similar clinical presentation; they may be indistinguishable until basic clinical (laboratory and sometimes radiological) investigations are completed. The apparent paradoxical connotations of delirium, in that it is ominously and almost inextricably linked with the terminal phase on the one hand and displays reversibility on the other (albeit perhaps earlier in the disease trajectory, but yet within the context of end of life, as we have defined it in this review) generates challenges for clinicians, patients, and their substitute decision makers (SDMs). These challenges and the patient's goals of care may combine to generate a decisional dilemma. The underpinnings of this dilemma relate to the risk of being unduly fatalistic (assuming irreversibility of delirium; failure to investigate clinically; and missing potentially reversible delirium precipitants) and using continuous deep (palliative) sedation inadvertently as a premature option. Death in this context becomes a selffulfilling prophecy. Alternatively, there is the risk of adopting an overly aggressive medicalized approach that could involve inappropriately burdensome and non-costeffective investigations and treatments in a truly non-reversible episode of delirium. Ultimately, the clinical outcome of delirium may therefore be strongly influenced by the management approach.
The ideal management decision-making process requires an individualized approach and clear, sensitive communication with family or other SDMs is imperative, as the patient will most often lack decisional capacity. The factors in the decisional mix will include clarification of the known clinical status; the trajectory of the illness and the suspected or known delirium precipitants; the functional status prior to the episode of delirium; the desire for further investigation, based on the patient's prior expressed goals of care; and the risk and burden of further investigation or treatment. Sometimes a timed trial of antibiotics may be necessary when the outcome of treating an infection (as a delirium precipitant) is difficult to predict. It should be remembered that patients and their families may be hugely appreciative of delirium reversal or part reversal, even for a short period so as to allow some meaningful communication between patient and family.
Regardless of whether clinical investigations are conducted to establish the nature of delirium precipitants or whether therapeutic interventions are pursued to achieve some degree of delirium reversal, the need for symptomatic treatment of delirium will require careful evaluation in the context of either end-of-life or terminal-phase care.
## Management of potentially reversible precipitants
Although approximately 50% of delirium episodes can be reversed in palliative care patients [bib_ref] Occurrence, causes, and outcome of delirium in patients with advanced cancer: a..., Lawlor [/bib_ref] , especially when the precipitants are medications, infection, or hypercalcemia [bib_ref] Occurrence, causes, and outcome of delirium in patients with advanced cancer: a..., Lawlor [/bib_ref] [bib_ref] Underlying pathologies and their associations with clinical features in terminal delirium of..., Morita [/bib_ref] , the decision to initiate treatment in an attempt to reverse an underlying cause(s) for delirium is dependent on a patient's estimated prognosis and goals of care. See Sect. 7 for a more in-depth discussion of this clinical challenge.
## Medication-induced delirium
Medications are an increasingly common precipitant of delirium [bib_ref] Drug-induced, dementia-associated and non-dementia, non-drug delirium hospitalizations in the United States, Lin [/bib_ref] as well as other neuropsychiatric adverse [bib_ref] Psychoactive medications and risk of delirium in hospitalized cancer patients, Gaudreau [/bib_ref] [bib_ref] Which medications to avoid in people at risk of delirium: a systematic..., Clegg [/bib_ref] [bib_ref] Effect of medications with anti-cholinergic properties on cognitive function, delirium, physical function..., Fox [/bib_ref] [bib_ref] Increasing anticholinergic burden and delirium in palliative care inpatients, Zimmerman [/bib_ref] [bib_ref] The impact of plasma cholinergic enzyme activity and other risk factors for..., Plaschke [/bib_ref]. All patients should have a careful assessment of their current and recent medication profile for potentially deliriogenic medications [bib_ref] DEL-FINE: a new tool for assessing the delirogenic properties of drugs of..., Bohmdorfer [/bib_ref] , with consideration of dose taper and/or discontinuation.
## Opioid-induced neurotoxicity
The majority of palliative care patients with end-stage disease will require treatment with opioids for pain and/or dyspnea. Opioid-induced neurotoxicity (OIN) is a syndrome of neuropsychiatric adverse effects that may occur with opioid therapy, and is exacerbated by a large dose, or rapid increase in opioid dose, and dehydration. The features of OIN are severe sedation, cognitive impairment, delirium, hallucinations, myoclonus, seizures, hyperalgesia, and allodynia. OIN is managed by opioid dose reduction if pain is well-controlled, or opioid rotation (also known as opioid 'switch') with an accompanying decrease in the equianalgesic dose of the new opioid by at least onethird because of incomplete cross tolerance [bib_ref] Opioid rotation in the management of refractory cancer pain, Indelicato [/bib_ref]. In addition, intravenous or subcutaneous hydration may be utilized.
## Dehydration
Clinically assisted hydration may be used in the management of patients with delirium, in particular when dehydration is thought to be a contributing factor. However, in a double-blind placebo-controlled randomized trial in 129 advanced cancer patients, subcutaneous hydration (hypodermoclysis) of 1000 mL of normal saline did not improve symptoms [bib_ref] Parenteral hydration in patients with advanced cancer: a multicenter, double-blind, placebo-controlled randomized..., Bruera [/bib_ref]. Memorial Delirium Assessment Scale (MDAS) scores worsened from baseline, with a non-significant trend for less deterioration in the hydration arm. Further research in this area of clinical practice is needed [bib_ref] Indications and practice of artificial hydration for terminally ill cancer patients, Nakajima [/bib_ref].
## Hypercalcemia
In patients with advanced cancer, corrected calcium levels above 3.0 mmol/L usually cause significant problems, including delirium. Management depends on the severity of symptoms, patient's wishes, and estimated life expectancy, and includes parenteral hydration with saline and treatment with bisphosphonate or calcitonin [bib_ref] Clinical practice. Hypercalcemia associated with cancer, Stewart [/bib_ref] [bib_ref] Delirium with severe symptom expression related to hypercalcemia in a patient with..., Delgado-Guay [/bib_ref]. The use of subcutaneous denosumab, a monoclonal antibody acting against the receptor activator of nuclear factor-jB (RANK) ligand, has recently been reported in 33 advanced cancer patients with bisphosphonate-refractory hypercalcemia [bib_ref] Denosumab for treatment of hypercalcemia of malignancy, Hu [/bib_ref]. In this study, denosumab was found to lower serum calcium levels in 21 of 33 (64%) patients within 10 days [bib_ref] Denosumab for treatment of hypercalcemia of malignancy, Hu [/bib_ref]. Due to an increased risk of hypocalcemia with denosumab, patients should be monitored post-treatment as they may require calcium and vitamin D supplementation [bib_ref] Hypercalcaemia and hypocalcaemia: finding the balance, Body [/bib_ref]. Results from larger studies with longer-term evaluation after denosumab treatment are needed.
## Brain tumor or metastasis
Although corticosteroids themselves have neuropsychiatric effects, including delirium [bib_ref] The neuropsychiatric complications of glucocorticoid use: steroid psychosis revisited, Dubovsky [/bib_ref] , they are often used to reduce peri-tumor edema, e.g., dexamethasone 4-8 mg daily in divided doses. In some patients with primary brain tumors, e.g., patients with progressive symptoms from intracranial glioblastoma multiforme, higher doses of corticosteroids may be required. There may also be a role for palliative radiotherapy depending on the patient's overall condition, burden of treatment, and goals of care.
## Pharmacological management of delirium in palliative care patients
Medications, in particular antipsychotics, have been utilized in the routine management of delirium in palliative care despite limited research evidence to support this practice [bib_ref] Treating an established episode of delirium in palliative care: expert opinion and..., Bush [/bib_ref]. [fig_ref] Table 4: Randomized trials for the pharmacological management of delirium in medical settings [/fig_ref] summarizes randomized trials for the management of delirium in non-intensive care populations. The Appendix in the Electronic Supplementary Material provides more in-depth information for medications that have been used in the management of delirium in palliative care. It should be noted that no medication is currently licensed for use in the management of delirium, so the use of medications for the indication of delirium is 'off-label'.
## Role of antipsychotics in delirium management
Despite limited RCT evidence [bib_ref] Antipsychotics in the treatment of delirium: a systematic review, Seitz [/bib_ref] , antipsychotics are commonly used in the management of delirium [bib_ref] Consensus and variations in opinions on delirium care: a survey of European..., Morandi [/bib_ref]. Back in 1993, the first edition of the Oxford Textbook of Palliative Medicine described haloperidol as ''the drug of choice in the treatment of delirium in the medically ill''. In 1999, the American Psychiatric Association's comprehensive practice guideline on delirium stated that ''antipsychotics have been the medication of choice in the treatment of delirium''. This practice was supported by a neurotransmitter hypothesis for the pathophysiology of delirium with a deficit of acetylcholine and an excess of dopamine for the pathophysiology of delirium, resulting in [bib_ref] Is there a final common neural pathway in delirium? Focus on acetylcholine..., Trzepacz [/bib_ref] , although it is now thought that many different neurotransmitter changes potentially lead to delirium [bib_ref] Neuropathogenesis of delirium: review of current etiologic theories and common pathways, Maldonado [/bib_ref]. Antipsychotics are divided into three classes:
1. First generation (previously known as 'typical'):
- Butyrophenones: haloperidol.
- Phenothiazines: e.g., levomepromazine (methotrimeprazine), chlorpromazine, prochlorperazine, promazine.
2. Second generation (previously known as 'atypical'):
- e.g., olanzapine, quetiapine, risperidone, ziprasidone, clozapine
## Third generation
- e.g., aripiprazole. [fig_ref] Table 5: Receptor affinities for selected antipsychotics [/fig_ref] shows the receptor affinities for antipsychotics commonly used for delirium.
In elderly patients and patients with dementia, the use of first-, second-, and third-generation antipsychotics is associated with an increased risk of death, stroke, and transient ischemic attack [bib_ref] Risk of death in elderly users of conventional vs. atypical antipsychotic medications, Wang [/bib_ref] [bib_ref] Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of..., Schneider [/bib_ref]. (In the meta-analysis of randomized placebo-controlled trials with atypical antipsychotics by Schneider et al. [bib_ref] Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of..., Schneider [/bib_ref] , the duration of included studies ranged from 6 to 26 weeks.) Thus, in patients with co-morbid dementia, the use of antipsychotics should be avoided. If deemed clinically necessary, consider quetiapine if the oral route is possible, as a retrospective study of over 90,000 patients with dementia found the number needed to harm (NNH) = 50 with quetiapine compared with NNH = 26 with haloperidol [bib_ref] Antipsychotics, other psychotropics and the risk of death in patients with dementia:..., Maust [/bib_ref]. For delirious patients who also have Parkinson's disease or dementia with Lewy bodies (DLB), second-generation antipsychotics (e.g., quetiapine) are preferred because first-generation antipsychotics increase the risk of disease exacerbation and extrapyramidal side effects (EPS) in such patients [bib_ref] Neuroleptic sensitivity in patients with senile dementia of Lewy body type, Mckeith [/bib_ref]. In these patient groups, alternatives to antipsychotics should also be considered, such as a benzodiazepine for agitation.
Haloperidol has been the 'practice standard' antipsychotic for delirium management in palliative care for many years, in part due to familiarity with its use in clinical practice, the versatility of multiple routes of administration, especially the subcutaneous route, and the absence of placebo-controlled trial evidence prior to a recent study by Agar et al. [bib_ref] Efficacy of oral risperidone, haloperidol, or placebo for symptoms of delirium among..., Agar [/bib_ref] (see below). The use of haloperidol may also be derived from emergency psychiatric services where haloperidol combined with a benzodiazepine or promethazine is used in the management of psychosis-induced aggression for rapid tranquilization [bib_ref] Haloperidol plus promethazine for psychosis-induced aggression, Huf [/bib_ref]. Haloperidol is predominantly metabolized by the liver; its metabolites include reduced haloperidol, an active metabolite, as well as a potentially neurotoxic pyridinium metabolite [bib_ref] Pharmacokinetics of haloperidol, Kudo [/bib_ref]. [fig_ref] Figure 4: Metabolism of haloperidol[100][101][102] [/fig_ref] shows the metabolism of haloperidol. Small observational studies in cancer patients have compared haloperidol, risperidone, and olanzapine with aripiprazole (n = 21, case-matched) [bib_ref] Haloperidol, risperidone, olanzapine and aripiprazole in the management of delirium: a comparison..., Boettger [/bib_ref] or with quetiapine (n = 27) [bib_ref] Novel therapeutic strategies for delirium in patients with cancer: a preliminary study, Tanimukai [/bib_ref] with improvement in MDAS scores, but more EPS reported with haloperidol. In some countries, levomepromazine (methotrimeprazine) or chlorpromazine are used for symptomatic relief of agitated delirium due to their sedative effect [bib_ref] Evidence for the use of levomepromazine for symptom control in the palliative..., Dietz [/bib_ref] [bib_ref] Evidence-based treatment of delirium in patients with cancer, Breitbart [/bib_ref] [bib_ref] Frequency and outcome of neuroleptic rotation in the management of delirium in..., Shin [/bib_ref].
In 2012, a Cochrane review for drug therapy for delirium in terminally ill adult patients concluded that ''there remains insufficient evidence to draw conclusions about the role of drug therapy'' [bib_ref] Drug therapy for delirium in terminally ill adult patients, Candy [/bib_ref]. The NICE-commissioned Clinical Guideline 103 on delirium recommended a limited role for antipsychotics: ''If a person with delirium is distressed or considered a risk to themselves or others … consider giving short-term haloperidol or olanzapine'' (recommendation 1.6.4). It should be noted that this comprehensive NICE guideline excluded ''people receiving end-of-life care'', defined as ''last few days of their life''. Increasingly, authors have described potential concerns and harms with antipsychotics [bib_ref] First do no harm… Terminal restlessness or drug-induced delirium, White [/bib_ref] [bib_ref] Doing damage in delirium: the hazards of antipsychotic treatment in elderly persons, Inouye [/bib_ref] , as well as confirming a lack of evidence in recent systematic reviews and meta-analysis for the use of haloperidol and other antipsychotics in the use of delirium treatment or prevention in hospitalized adults [bib_ref] Efficacy and safety of haloperidol for inhospital delirium prevention and treatment: a..., Schrijver [/bib_ref] [bib_ref] Antipsychotic medication for prevention and treatment of delirium in hospitalized adults: a..., Neufeld [/bib_ref].
Challenging this historical dogma of antipsychotic use for delirium management in palliative care, Agar et al. [bib_ref] Efficacy of oral risperidone, haloperidol, or placebo for symptoms of delirium among..., Agar [/bib_ref] recently published the results of an Australian multi-site, double-blind, parallel-arm, dose-titrated RCT of oral risperidone, haloperidol, or placebo with rescue subcutaneous midazolam for delirium management in the first reported adequately powered RCT in a palliative care population. Adult inpatients receiving hospice or palliative care with confirmed delirium (n = 247) received oral study medications over a 72-h period. (There was also an option to continue the blinded study medication for an additional 48 h if a partial response occurred or to allow a dose taper with symptom resolution.) Study participants had mild to moderate delirium (median baseline MDAS scores ranged from 13.7 to 15.1), and were predominantly male (65.6%) with a mean age of 74.9 years. The majority had a cancer diagnosis (88.3%) and baseline Australia-modified Karnofsky Performance Status (AKPS) scores ranged from 30% to 50%. Participants were randomly assigned risperidone, haloperidol, or placebo as oral solutions. The antipsychotic dosing schedule was the same for both haloperidol and risperidone, and adjusted for age. For participants B65 years, a 0.5 mg loading dose was administered along with a first dose of 0.5 mg, followed by 0.5 mg every 12 h. If the sum of Nursing Delirium Screening Scale (Nu-DESC) scores (items 2, 3, and 4-for ''inappropriate behavior, inappropriate communication and illusions or hallucinations'': range 0-6) was C1 at the most recent assessment, the dose could be titrated by 0.25 mg on day 1, and then by 0.5 mg to a maximum dose of 4 mg/day. For participants [65 years, the loading, first, and maximum doses were 50% of the aforementioned doses. The placebo solution was similarly titrated using matching volumes of solution. Doses were reduced if adverse effects, delirium resolution (defined as MDAS score \7 for 48 h), or resolution of symptoms (sum of Nu-DESC scores\1 for 48 h) occurred. Subcutaneous 'rescue' midazolam (not age adjusted) was given in a dose of 2.5 mg every 2 h as needed. All study participants also received treatment of their reversible delirium participants where clinically indicated and non-pharmacological management approaches, consisting of hydration, vision and hearing aids, family presence, and reorientation, ''as appropriate''. The two main study comparisons were placebo versus risperidone and placebo versus haloperidol. In the intention-to-treat analysis, participants in the risperidone and haloperidol arms had higher delirium symptom scores (p = 0.02 and p = 0.009, respectively) than the placebo arm at study end. In addition, significantly less rescue midazolam was used in the placebo arm. There were statistically significant greater mean extrapyramidal symptoms (as assessed by the Extrapyramidal Symptom Rating Scale) in both active arms than in the placebo arm. Median survival was 26 days in the placebo arm, 17 days in the risperidone arm, and 16 days in the haloperidol arm. In a post hoc analysis of this RCT, the authors reported that participants receiving an antipsychotic were approximately 1.5 times more likely to die. The results of this RCT confirm the importance of maximizing non-pharmacological strategies and reversing delirium precipitants, if appropriate, in patients with delirium of mild to moderate severity. However, further research is required in the management of palliative care patients with severe delirium, or with reduced performance status and increased frailty.
## Adverse effects of antipsychotics
Due to their variation in receptor affinities, antipsychotics vary in their propensity for extrapyramidal, sedative, anticholinergic, and hypotensive side effects (see [fig_ref] Table 6: Profile of adverse effects for selected antipsychotics at therapeutic doses [/fig_ref]. Cardiac adverse effects include prolongation of the ratecorrected QT (QTc) interval and arrhythmias. As only short-term use of antipsychotics may be used in the management of delirium, the longer-term adverse effects of antipsychotics, such as endocrine and metabolic adverse effects, are not discussed here.
## Extrapyramidal side effects
EPS comprise dystonia, akathisia, parkinsonism, and tardive dyskinesia. While antipsychotic effects are associated with striatal dopamine D 2 receptor occupancy of 65-70%, there is a significantly increased risk of EPS when the receptor occupancy is in the range of 80% or higher [bib_ref] Treatments for schizophrenia: a critical review of pharmacology and mechanisms of action..., Miyamoto [/bib_ref].
Other risk factors for EPS include high or rapidly increasing antipsychotic dose, especially during the first days of treatment, increasing age, and dementia. There is a reduced risk of EPS with second-generation antipsychotics due to 5-HT 2 antagonism and D 2 partial agonism (quetiapine has the lowest risk). However, EPS can occur at higher doses, especially risperidone [6 mg/day [bib_ref] Evidence-based treatment of delirium in patients with cancer, Breitbart [/bib_ref].
There is a reduced tendency to develop antipsychotic-induced parkinsonism with antipsychotics that have anticholinergic activity [bib_ref] Neuroleptic-induced movement disorders: an overview, Sachdev [/bib_ref] , such as levomepromazine (methotrimeprazine) and quetiapine, as compared with other antipsychotics in their respective classes with less anticholinergic activity.
## Seizures
Antipsychotics may lower the seizure threshold, and should be used with caution in patients with a seizure history. The risk approximates to the degree of antipsychotic sedation and is dose dependent. Haloperidol is the lowest-risk antipsychotic, and there is a lower likelihood with secondgeneration antipsychotics. The highest risk of precipitating seizures is with sedating antipsychotics, e.g., levomepromazine (methotrimeprazine) and chlorpromazine.
## Neuroleptic malignant syndrome
Neuroleptic (antipsychotic) malignant syndrome (NMS) is precipitated by dopamine antagonists. It is an infrequent, idiosyncratic, and potentially fatal syndrome that occurs in \1% of patients prescribed an antipsychotic. The clinical features of severe rigidity, hyperthermia, altered mental status, and autonomic dysfunction develop over daysweeks. Hyporeflexia also occurs. Management includes stopping the causative drug and general supportive care. Benzodiazepines may reduce muscle rigidity. Bromocriptine has been used in severe cases, and intravenous dantrolene in acute medical settings [bib_ref] Management of serotonin syndrome and neuroleptic malignant syndrome, Katus [/bib_ref]. Antipsychotics should not be prescribed for patients who have recovered from NMS as they have a 30-50% risk of recurrence.
## Corrected qt interval prolongation
An increased risk of sudden cardiac death can occur in users of both first-generation and second-generation antipsychotics [bib_ref] Atypical antipsychotic drugs and the risk of sudden cardiac death, Ray [/bib_ref]. Some antipsychotics have a higher propensity for prolongation of the QTc interval, e.g., thioridazine (drug withdrawn from Canada and other countries) and ziprasidone [bib_ref] QTc prolongation by psychotropic drugs and the risk of torsade de pointes, Wenzel-Seifert [/bib_ref]. With QTc prolongation, especially [500 ms, patients are at risk of developing 'torsades de pointes' (polymorphic ventricular tachycardia). In addition to antipsychotic dose, risk for drug-induced QTc prolongation may be increased with the presence of other risk factors such as electrolyte abnormality (e.g., hypokalemia, hypomagnesemia, hypocalcemia), drug interaction with the concurrent use of another QT-prolonging medication, congenital long QT syndrome, bradycardia, female sex, and age [65 years [bib_ref] Drug-induced prolongation of the QT interval, Roden [/bib_ref] [bib_ref] High prevalence of corrected QT interval prolongation in acutely ill patients is..., Pickham [/bib_ref] (a list of QT-prolonging drugs and drug interactions can be found online. An electrocardiogram should be considered in patients with a cardiac history and prolonged dosing of antipsychotic, but in a very distressed delirious palliative care patient the need for antipsychotic administration for symptomatic management may outweigh the potential risk of QTc interval prolongation.
## Role of benzodiazepines in delirium management
Palliative care practitioners have been cautious in prescribing benzodiazepines in their patients due to concerns for causing and/or exacerbating delirium. A systematic review examining the association between multiple medication classes and delirium concluded that benzodiazepines may be associated with an increased risk of developing delirium [odds ratio (OR) 3, 95% confidence interval (CI) 1.3-6.8] using matched analysis data from one moderatequality case-control study of 1341 mixed surgical group patients [bib_ref] Which medications to avoid in people at risk of delirium: a systematic..., Clegg [/bib_ref]. In particular, longer-acting benzodiazepines and higher doses of benzodiazepines within a 24-h period increased the risk of delirium. Benzodiazepines are used as first-line agents in the management of alcohol or sedativehypnotic withdrawal. Benzodiazepines may be utilized in agitated patients with Parkinson's disease or DLB because of the risk of EPS with antipsychotics (see ESM Appendix).
In 1996, Breitbart et al. [bib_ref] A double-blind trial of haloperidol, chlorpromazine, and lorazepam in the treatment of..., Breitbart [/bib_ref] conducted a seminal study in 30 hospitalized AIDS patients with delirium. There was no placebo control group in this randomized, double-blind comparison trial of haloperidol, chlorpromazine, and lorazepam. The lorazepam arm was stopped early as all six patients who received lorazepam either refused to take the drug or had the treatment discontinued due to the development of adverse effects (including over-sedation, disinhibition, ataxia, and increased confusion). In a cohort study of 261 hospitalized cancer patients, patients who received a benzodiazepine dose (assessed as cumulative daily equivalent dose of oral lorazepam) of [2 mg/day had a two times increased risk of developing delirium (as assessed by the Nu-DESC) compared with patients exposed to lower doses [bib_ref] Psychoactive medications and risk of delirium in hospitalized cancer patients, Gaudreau [/bib_ref].
Recently, Ferraz Gonçalves and colleagues [bib_ref] Comparison of haloperidol alone and in combination with midazolam for the treatment..., Gonçalves [/bib_ref] reported that a protocol of combined haloperidol and midazolam by the intramuscular or subcutaneous route was significantly more effective than haloperidol alone for agitated palliative care inpatients, with ''cognitive impairment as assessed by the Short Confusion Assessment Method''. It should be noted that the use of nonpharmacological approaches in the inpatient palliative care unit was not reported as part of their protocol. Further studies are required to better define the role of benzodiazepines in the management of severe agitated delirium.
## Other medications used for treatment of delirium
Other medications that have been examined for a potential role in delirium treatment include methylphenidate, modafinil, valproic acid, gabapentin, ondansetron, and melatonin [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref] [bib_ref] Evidence-based treatment of delirium in patients with cancer, Breitbart [/bib_ref] [bib_ref] Adjunctive valproic acid in management-refractory hyperactive delirium: a case series and rationale, Sher [/bib_ref]. They are not currently recommended for routine clinical practice due to limited evidence. A recent systematic review found a lack of efficacy of acetylcholinesterase inhibitors for treating or preventing delirium in older adults, in addition to increased mortality in a prematurely terminated study with rivastigmine [bib_ref] Acetylcholinesterase inhibitors for delirium in older adults, Tampi [/bib_ref] [bib_ref] Effect of rivastigmine as an adjunct to usual care with haloperidol on..., Van Eijk [/bib_ref]. Dexmedetomidine, an a 2 -receptor agonist with analgesic and opioid-sparing properties used for sedation in intensive care settings by intravenous infusion, may potentially have a role in the management of delirium in palliative care patients [bib_ref] Dexmedetomidine: a review of its use for sedation in the intensive care..., Keating [/bib_ref] [bib_ref] Review article: dexmedetomidine: does it have potential in palliative medicine?, Prommer [/bib_ref].
## Delirium clinical practice guidelines
Few published current delirium clinical guidelines are aimed for patients at the end of life [bib_ref] Quality of clinical practice guidelines in delirium: a systematic appraisal, Bush [/bib_ref]. Delirium guidelines are often limited in their ability to make strong formal recommendations due to the relative lack of high-quality research at this time, resulting in some recommendations for certain guidelines being derived from expert consensus-based statements [bib_ref] Clinical practice guidelines for delirium management: potential application in palliative care, Bush [/bib_ref]. However, by identifying the knowledge gaps, guidelines may assist in the prioritization of research activities. Another challenge for guidelines in general is that their content needs to be regularly updated to ensure that current research knowledge is reflected and optimally translated into clinical practice. Appropriate implementation strategies and sustainability plans for guidelines are crucial, as guidelines alone may not improve clinical outcomes [bib_ref] Do guidelines improve the process and outcomes of care in delirium?, Young [/bib_ref].
## Quality improvement in delirium care
Although not specific to palliative care, the following nonexhaustive list of resources offers useful quality improvement recommendations that can be implemented by the interprofessional team to actively improve delirium care. NICE quality standard 63 recommends five key priorities to improve delirium care. Similarly, the Australian Commission on Safety and Quality in Health Care has recently published a Delirium Clinical Care Standard following on from its initiative entitled ''A Better Way to Care: Safe and High-Quality Care for Patients with Cognitive Impairment (Dementia and Delirium) in Hospital''. The Canadian Patient Safety Institute has developed a Safer Healthcare Now!: Prevention and Management of Delirium: ''Getting Started Kit'' intervention for the prevention and management of delirium.
## Palliative sedation
The development of delirium in dying patients is a poor prognostic sign [bib_ref] Agitation and delirium at the end of life: ''we couldn't manage him, Breitbart [/bib_ref]. The management of refractory agitated delirium (and/or other distressing and refractory symptoms) at the end of life may require the judicious use of proportionate sedative medication to reduce patient distress [bib_ref] Palliative sedation in dying patients: ''we turn to it when everything else..., Lo [/bib_ref] [bib_ref] End-of-life delirium: issues regarding recognition, optimal management and the role of sedation..., Bush [/bib_ref]. By facilitating a more 'peaceful' death with appropriate sedation, family distress will also be relieved.
## Definition of palliative sedation
Palliative sedation (PS), or sedation in the terminal phase, has been defined as ''the intentional administration of sedative drugs in dosages and combinations required to reduce the consciousness of a terminal patient as much as necessary to adequately relieve 1 or more refractory symptoms'' [bib_ref] Palliative sedation: a review of the research literature, Claessens [/bib_ref]. Continuous PS should only be considered when death is anticipated within 2 weeks or less [bib_ref] Canadian Society of Palliative Care Physicians Taskforce. Framework for continuous palliative sedation..., Dean [/bib_ref]. Appropriately titrated PS is an ethically and legally accepted intervention, and does not shorten life [bib_ref] Palliative sedation in end-of-life care and survival: a systematic review, Maltoni [/bib_ref]. All discussions with the patient and/or family or SDM should be documented and include the indications for initiating PS. During PS, the level of sedation should be regularly monitored using a clinical assessment tool, e.g., Richmond Agitation-Sedation Scale (RASS) [bib_ref] The richmond agitation-sedation scale, Sessler [/bib_ref] or the palliative version, RASS-PAL [bib_ref] The Richmond Agitation-Sedation Scale modified for palliative care inpatients (RASS-PAL): a pilot..., Bush [/bib_ref] , in addition to the patient's level of comfort or discomfort. Other symptom management medications, e.g., opioids for pain, should be continued. Patients, family members, and the healthcare team may require emotional support, information provision, and have any concerns with PS addressed.
## Medications used for palliative sedation
Various medications are used for PS. Midazolam is the drug of first choice due its rapid onset of action, but the selection of medication will depend on local drug availability, care location, and the indication for PS. Other medications that are used for PS include levomepromazine (methotrimeprazine), chlorpromazine, lorazepam, and phenobarbital [bib_ref] Palliative sedation in end-of-life care and survival: a systematic review, Maltoni [/bib_ref] [bib_ref] ESMO Clinical Practice Guidelines for the management of refractory symptoms at the..., Cherny [/bib_ref] (see [fig_ref] Table 7: Medications commonly used for palliative sedation in the management of refractory agitated... [/fig_ref] and the ESM Appendix). Propofol, an ultra-fast-acting general anesthetic agent best suited to use in a monitored setting, has been used as a continuous intravenous infusion when standard treatments have failed [bib_ref] When nothing helps: propofol as sedative and antiemetic in palliative cancer care, Lundström [/bib_ref]. Dexmedetomidine has been used as continuous subcutaneous infusion in the management of refractory delirium in a palliative care unit [bib_ref] A case report of dexmedetomidine used to treat intractable pain and delirium..., Hilliard [/bib_ref]. It should be noted that the use of medications for PS is based on case series and expert opinion, and that evidence regarding efficacy of PS for symptom control remains insufficient at this time [bib_ref] ESMO Clinical Practice Guidelines for the management of refractory symptoms at the..., Cherny [/bib_ref] [bib_ref] Palliative pharmacological sedation for terminally ill adults, Beller [/bib_ref]. Further research is needed on the efficacy and harms of different pharmacologic interventions for PS, and determining appropriate dosing and titration strategies.
## Education and support
Delirium causes considerable distress for patients and their families, as well as professional caregivers [bib_ref] The delirium experience: delirium recall and delirium-related distress in hospitalized patients with..., Breitbart [/bib_ref] [bib_ref] Impact of delirium and recall on the level of distress in patients..., Bruera [/bib_ref]. Patients, even those with hypoactive delirium, may experience perceptual disturbances, delusions, disorientation, and feel threatened and anxious [bib_ref] Delirium in patients with cancer: assessment, impact, mechanisms and management, Lawlor [/bib_ref]. Namba et al. [bib_ref] Terminal delirium: families' experience, Namba [/bib_ref] reported that 70% of 20 bereaved family members expressed distress at observing delirium in their relatives in the last 2 weeks of life, and identified a need for supportive information. Cohen et al. [bib_ref] Delirium in advanced cancer leading to distress in patients and family caregivers, Cohen [/bib_ref] interviewed 37 family caregivers of adult patients with advanced cancer who had recently recovered from an episode of delirium. Family caregivers ''who had expected 'confusion' found the experience less distressing'' [bib_ref] Delirium in advanced cancer leading to distress in patients and family caregivers, Cohen [/bib_ref]. Family members need support from the healthcare team; this includes advice on how to respond to a delirious patient, non-pharmacological strategies, as well as information about delirium [bib_ref] The experiences of caregivers of patients with delirium, and their role in..., Finucane [/bib_ref]. Psychoeducational interventions that have been reported as beneficial include verbal information and a delirium information leaflet for family members [bib_ref] Delirium in advanced cancer: a psychoeducational intervention for family caregivers, Gagnon [/bib_ref] [bib_ref] Usefulness of the leaflet-based intervention for family members of terminally ill cancer..., Otani [/bib_ref] and 'debriefing' patients who have recovered from an episode of delirium [bib_ref] The delirium experience: what is the effect on patients, relatives and staff..., Partridge [/bib_ref]. Delirium is a diverse and complex 'endpoint' clinical syndrome resulting from potentially a myriad of interconnected pathophysiological mechanisms. Differing treatment strategies are likely required for differing delirium precipitating factors, pathophysiological variables/mechanisms, subtypes and phenomenological differences, and in early palliative care patients as opposed to patients at the 'end of life' or in the 'dying phase'. As delirium management in the context of palliative care focuses more on targeting symptoms rather than simple receptor targeting, further research is needed as to the benefit and efficacy of pharmacotherapy and multicomponent strategies for reducing morbidity and improving quality of life in this patient population. In the meantime, non-pharmacological strategies should be optimized and antipsychotics used judiciously until more high-quality evidence exists to guide practice in delirium management in palliative care across care settings and the illness trajectory.
Acknowledgements The authors would like to thank Research Administrative Assistant Monisha Kabir for her invaluable assistance in the preparation of the manuscript.
## Compliance with ethical standards
Conflicts of interest Authors Shirley H. Bush, Sallyanne Tierney, and Peter G. Lawlor have no conflicts of interest to disclose.
Funding No sources of funding were used in the preparation of this narrative review.
Open Access This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, distribution, and reproduction in any medium, provided 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.
[fig] Figure 2: Algorithm for the assessment and management of delirium in palliative care patients. AP antipsychotic, BDZ benzodiazepine, CAM Confusion Assessment Method, DSM-5 Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, EOL end of life, EPS extrapyramidal side effects, ICD-10 International Classification of Diseases, 10th revision [/fig]
[fig] Figure 4: Metabolism of haloperidol[100][101][102] (note large interindividual variations in haloperidol pharmacokinetics). CPHP 4-(4-chlorophenyl)-4-hydroxypiperidine, CYP cytochrome P450, EPS extrapyramidal side effects [/fig]
[table] Table 1: Commonly used delirium screening tools[17,23,24] [/table]
[table] Table 2: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, ICD-10 International Classification of Diseases, 10th revision [/table]
[table] Table 4: Randomized trials for the pharmacological management of delirium in medical settings [/table]
[table] Table 5: Receptor affinities for selected antipsychotics (derived from Procyshyn et al.[92] Howard et al.[93], and Lal et al.[94]) [/table]
[table] Table 6: Profile of adverse effects for selected antipsychotics at therapeutic doses (derived from Procyshyn et al.[92]) [/table]
[table] Table 7: Medications commonly used for palliative sedation in the management of refractory agitated delirium within the last 2 weeks of life[140,[147][148][149][150] Administration: SC or IV:Loading dose (2.5-5 mg), then start continuous infusion at lowest effective dose Titrate by 0.5-1 mg/h increments q30 min according to clinical response and depth of sedation required Up to maximum 6 mg/h as continuous infusion (CSCI/CIVI) LorazepamLorazepam less amenable to rapid titration up or down than midazolam because of its slower pharmacokineticsAdministration: SC or IV; intermittent bolus: Starting dose: 0.05 mg/kg q2-4 h when administered by intermittent bolus Chlorpromazine Administration: IV, deep IM, PR (PO) Starting dose 12.5 mg q4-12 h (slow IV/IM), or 3-5 mg/h CIVI, or 25-100 mg q4-12 h PR (Doses of 25-375 mg/24 h reported [139]) Phenobarbital (phenobarbitone) Administration: SC or IV (IM, PR, PO) Loading bolus: up to 200 mg Then regular dosing: up to 800 mg/24 h (Doses of up to 2400 mg/24 h and 3400 mg/24 h reported [140, 141]) CIVI continuous intravenous infusion, CSCI continuous subcutaneous infusion, IM intramuscular, IV intravenous, PO by mouth, PR per rectum, qx h every x h, qx min every x min, SC subcutaneous [/table]
|
Peripheral Protein Quality Control as a Novel Drug Target for CFTR Stabilizer
Conformationally defective cystic fibrosis transmembrane conductance regulator (CFTR) including rescued F508-CFTR is rapidly eliminated from the plasma membrane (PM) even in the presence of a CFTR corrector and potentiator, limiting the therapeutic effort of the combination therapy. CFTR elimination from the PM is determined by the conformation-dependent ubiquitination as a part of the peripheral quality control (PQC) mechanism. Recently, the molecular machineries responsible for the CFTR PQC mechanism which includes molecular chaperones and ubiquitination enzymes have been revealed. This review summarizes the molecular mechanism of the CFTR PQC and discusses the possibility that the peripheral ubiquitination mechanism becomes a novel drug target to develop the CFTR stabilizer as a novel class of CFTR modulator.
# Introduction
Cystic fibrosis (CF) is one of the most lethal autosomal-recessive diseases caused by mutation in CFTR [bib_ref] CFTR modulators: shedding light on precision medicine for cystic fibrosis, Lopes-Pacheco [/bib_ref]. CFTR mutations are classified as I-VII according to their properties (I-protein synthesis defect, II-maturation defect, III-gating defect, IV-conductance defect, V-reduced quantity, VI-reduced PM stability, VII-no mRNA transcription). The most prevalent CF causing mutation, F508, was classically categorized as class II mutation. However, rescued F508 (r F508)-CFTR by corrector (e.g., VX-809/lumacaftor) or low temperature culture shows class III and VI phenotypes [bib_ref] Altered chloride ion channel kinetics associated with the delta F508 cystic fibrosis..., Dalemans [/bib_ref] [bib_ref] From CFTR biology toward combinatorial pharmacotherapy: expanded classification of cystic fibrosis mutations, Veit [/bib_ref]. Although drug targets of the class II or III mutations are well studied, that of the class VI mutation are not because the mechanism of CFTR PM stability regulation are still veiled by numerous undefined molecules involved in CFTR PQC system. In this review, we summarize accumulated findings regarding the CFTR PQC from the molecular and environmental aspects and also discuss the potential of recently identified PQC machineries including endocytic adaptors and ubiquitination enzymes as targets for CFTR stabilizer which anchors the functional channel at the PM and reduces the degradation .
## Cftr instability at the pm
Nascent wild-type (WT) CFTR is N-glycosylated at the endoplasmic reticulum (ER) during translation and folded by the aid of chaperones such as calnexin (CNX), HSP70 and HSP90 [bib_ref] CFTR and chaperones: processing and degradation, Amaral [/bib_ref] [bib_ref] Folding of CFTR is predominantly cotranslational, Kleizen [/bib_ref] [bib_ref] Role of calnexin in the ER quality control and productive folding of..., Okiyoneda [/bib_ref] [bib_ref] N-glycans are direct determinants of CFTR folding and stability in secretory and..., Glozman [/bib_ref] [bib_ref] Mechanisms of CFTR folding at the endoplasmic reticulum, Kim [/bib_ref]. Properly folded CFTR is then sorted to the Golgi apparatus and processed to complex glycosylation while misfolded CFTR is retained in the ER and consequently degraded by ER-associated degradation (ERAD). The CFTR ERAD is associated with several ER QC processes such as chaperones binding, ubiquitination and retro-translocation from the ER to cytosol [bib_ref] Sequential quality-control checkpoints triage misfolded cystic fibrosis transmembrane conductance regulator, Younger [/bib_ref] [bib_ref] The role of the UPS in cystic fibrosis, Turnbull [/bib_ref] [bib_ref] Gp78 cooperates with RMA1 in endoplasmic reticulum-associated degradation of CFTRDeltaF508, Morito [/bib_ref] [bib_ref] The endoplasmic reticulum-associated Hsp40 DNAJB12 and Hsc70 cooperate to facilitate RMA1 E3-dependent..., Grove [/bib_ref] [bib_ref] TorsinA participates in endoplasmic reticulum-associated degradation, Nery [/bib_ref] [bib_ref] RNF185 is a novel E3 ligase of endoplasmic reticulum-associated degradation (ERAD) that..., El Khouri [/bib_ref] [bib_ref] Endoplasmic reticulum protein quality control is determined by cooperative interactions between Hsp/c70..., Matsumura [/bib_ref] [bib_ref] VAMP-associated Proteins (VAP) as receptors that couple cystic fibrosis transmembrane conductance regulator..., Ernst [/bib_ref] [bib_ref] Non-native conformers of cystic fibrosis transmembrane conductance regulator NBD1 are recognized by..., Gong [/bib_ref] [bib_ref] Trafficking and function of the cystic fibrosis transmembrane conductance regulator: a complex..., Mcclure [/bib_ref] [bib_ref] Dissection of the Role of VIMP in endoplasmic reticulum-associated degradation of CFTR..., Hou [/bib_ref]. Properly folded and matured CFTR is trafficked to the PM to function as an ATP-regulated ion-channel [bib_ref] Cystic fibrosis transmembrane conductance regulator has an altered structure when its maturation..., Chen [/bib_ref].
The CFTR is internalized by clathrin-mediated endocytosis (CME) and recycled back to the PM. Conformationally defective CFTR produced by genetic mutations (e.g., F508, T70) and/or environmental stresses (e.g., heat) selectively undergoes ubiquitination at the PM by PQC machineries. The ubiquitinated CFTR is rapidly internalized and delivered to lysosome for degradation [bib_ref] Misfolding diverts CFTR from recycling to degradation: quality control at early endosomes, Sharma [/bib_ref] [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref].
Internalized CFTR could be de-ubiquitinated at endosomes by deubiquitinase (DUB) and recycled back to the PM depending on the conformational states. The class VI mutations render the CFTR unstable at the PM. Additionally, the class I, II and some class III CFTR mutants also show PM instability [bib_ref] The delta F508 mutation decreases the stability of cystic fibrosis transmembrane conductance..., Lukacs [/bib_ref] [bib_ref] C-terminal truncations destabilize the cystic fibrosis transmembrane conductance regulator without impairing its..., Haardt [/bib_ref] [bib_ref] A mutation in the cystic fibrosis transmembrane conductance regulator generates a novel..., Silvis [/bib_ref] [bib_ref] CFTR potentiators partially restore channel function to A561E-CFTR, a cystic fibrosis mutant..., Wang [/bib_ref] [bib_ref] From CFTR biology toward combinatorial pharmacotherapy: expanded classification of cystic fibrosis mutations, Veit [/bib_ref]. N-glycosylation, especially the core-glycosylation, determines the CFTR PM stability likely by affecting the CFTR conformational stability [bib_ref] N-glycans are direct determinants of CFTR folding and stability in secretory and..., Glozman [/bib_ref] [bib_ref] Modulation of endocytic trafficking and apical stability of CFTR in primary human..., Cholon [/bib_ref]. Protein translation kinetics is also a significant factor that modulates proper cotranslational folding. Knock down (KD) of ribosomal protein L12 (RPL12) increases F508-CFTR PM expression and stability [bib_ref] Ribosomal stalk protein silencing partially corrects the F508-CFTR functional expression defect, Veit [/bib_ref]. RPL12 KD might affect protein translation kinetics associated with co-translational protein folding efficiency [bib_ref] Synonymous codons direct cotranslational folding toward different protein conformations, Buhr [/bib_ref] and thereby improve CFTR thermodynamic stability which also determines the CFTR PM stability [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref] [bib_ref] Correction of both NBD1 energetics and domain interface is required to restore..., Rabeh [/bib_ref]. Thus, correcting the CFTR structural defects at the ER could improve the PM stability.
## Environmental stresses affecting the cftr pm stability
## Infection and inflammation
The CFTR loss of function induces airway surface liquid (ASL) dysregulation which impairs clearance of infected bacteria and/or fungi, and increases the concentration of other soluble signal mediators such as cytokines, chemokines and growth factors. Pseudomonas aeruginosa (PA) is one of the most common bacteria found in CF respiratory tissue and responsible for lung injury in CF [bib_ref] Early infection and progression of cystic fibrosis lung disease, Koch [/bib_ref] [bib_ref] Cystic fibrosis lung environment and Pseudomonas aeruginosa infection, Bhagirath [/bib_ref]. PA destabilizes PM CFTR by inhibiting endocytic recycling [bib_ref] Pseudomonas aeruginosa inhibits endocytic recycling of CFTR in polarized human airway epithelial..., Swiatecka-Urban [/bib_ref]. PA secretes CFTR inhibitory factor (Cif) that stabilizes complex formation of ubiquitin (Ub) specific peptidase 10 (USP10) and GTPase activating protein (SH3 domain) binding protein 1 (G3BP1) and inhibits CFTR-USP10 interaction. Cif inhibits internalized CFTR sorting to recycling pathway by suppressing USP10 dependent CFTR de-ubiquitination at endosome, resulting in the lysosomal degradation of WT-CFTR [bib_ref] A Pseudomonas aeruginosa toxin that hijacks the host ubiquitin proteolytic system, Bomberger [/bib_ref]. PA also activates transforming growth factor β1 (TGF-β1) signaling that is an important modifier of lung disease severity in CF [bib_ref] Plasma TGF-β 1 in pediatric cystic fibrosis: potential biomarker of lung disease..., Harris [/bib_ref]. TGF-β1 inhibits functional PM expression of WT-CFTR and F508-CFTR by reducing mRNA level [bib_ref] Tgf-β1 inhibits Cftr biogenesis and prevents functional rescue of F508-Cftr in primary..., Snodgrass [/bib_ref] [bib_ref] Tgf-beta downregulation of distinct chloride channels in cystic fibrosis-affected epithelia, Sun [/bib_ref] although its role in the PQC remains unknown.
## Heavy metals
More than 10 ppb of arsenic induces the WT-CFTR ubiquitination and lysosomal degradation via c-Cbl in CF bronchial epithelial (CFBE) cells [bib_ref] Arsenic promotes ubiquitinylation and lysosomal degradation of cystic fibrosis transmembrane conductance regulator..., Bomberger [/bib_ref]. Importantly, the phenotype of arsenic toxicity overlaps with CF patient [bib_ref] Arsenic promotes ubiquitinylation and lysosomal degradation of cystic fibrosis transmembrane conductance regulator..., Bomberger [/bib_ref] [bib_ref] Elevated sweat chloride levels due to arsenic toxicity, Mazumdar [/bib_ref]. Cadmium (Cd) is a major component of cigarette smoke (CS), and its inhalation is associated with decreased pulmonary function and chronic obstructive pulmonary disease. Cd reduces CFTR PM level, but it remains unknown if it reduces the PM stability [bib_ref] Cadmium regulates the expression of the CFTR chloride channel in human airway..., Rennolds [/bib_ref].
## Cigarette smoke
Cigarette smoke is a major risk factor of chronic obstructive pulmonary disease and interferes with CFTR functionality. Ten minutes of CS exposure transiently suppresses CFTR function, induces internalization and decreases ASL height in human bronchial epithelial (HBE) cells [bib_ref] Cigarette smoke exposure induces CFTR internalization and insolubility, leading to airway surface..., Clunes [/bib_ref]. CS promotes CFTR internalization in BHK cells and results in increased insolubility of CFTR and colocalization with vimentin, a filament protein associated with aggresome Ca 2+ dependently. This observation suggesting that CS induces PM CFTR destabilization by stimulating internalization and aggregation in addition to suppressing CFTR functionality [bib_ref] Cigarette smoke exposure induces CFTR internalization and insolubility, leading to airway surface..., Clunes [/bib_ref] [bib_ref] Cigarette smoke-induced Ca2 + release leads to cystic fibrosis transmembrane conductance regulator..., Rasmussen [/bib_ref].
## Molecular machineries determining the cftr pm stability
## Endocytosis adaptors and tethering factors
Endocytosis is the critical step of elimination of PM CFTR as a part of PQC and is regulated by several molecules. WT-CFTR is internalized slowly by CME while misfolded r F508-CFTR endocytosis is accelerated [bib_ref] Misfolding diverts CFTR from recycling to degradation: quality control at early endosomes, Sharma [/bib_ref] [bib_ref] The short apical membrane half-life of rescued {Delta}F508-cystic fibrosis transmembrane conductance regulator..., Swiatecka-Urban [/bib_ref] [bib_ref] Enhanced cell-surface stability of rescued DeltaF508 cystic fibrosis transmembrane conductance regulator (CFTR)..., Varga [/bib_ref] [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. KD of CME adaptor AP-2 µ2 subunit or disabled 2 (DAB2) stabilizes r F508-CFTR at the PM by inhibiting endocytosis [bib_ref] Dab2 is a key regulator of endocytosis and post-endocytic trafficking of the..., Fu [/bib_ref].
CFTR has a postsynaptic density 95, disks large, zonula occludens-1 (PDZ) binding motif at C-terminus and binds with Na + /H + exchanger regulatory factor (NHERF1) PDZ domain. NHERF1 tethers CFTR with Ezrin and works as a scaffold protein that supports CFTR efficient channel activation and apical PM localization [bib_ref] Na + /H + exchanger regulatory factor 1 overexpressiondependent increase of cytoskeleton..., Favia [/bib_ref] [bib_ref] Stabilizing rescued surface-localized δf508 CFTR by potentiation of its interaction with Na(..., Arora [/bib_ref] [bib_ref] A molecular switch in the scaffold NHERF1 enables misfolded CFTR to evade..., Loureiro [/bib_ref]. NHERF1 also binds to misfolded F508-CFTR and increases the PM stability by inhibiting carboxy terminus of HSP70-interacting protein (CHIP) Ub ligase interaction [bib_ref] A molecular switch in the scaffold NHERF1 enables misfolded CFTR to evade..., Loureiro [/bib_ref]. An exchange protein directly activated by cAMP1 (EPAC1) selective activating cAMP analog 007-AM promotes WT-CFTR and NHERF1 interaction and increases CFTR PM stability in CFBE cells by suppressing endocytosis [bib_ref] EPAC1 activation by cAMP stabilizes CFTR at the membrane by promoting its..., Lobo [/bib_ref]. EPAC1 activation can rescue F508-CFTR PM expression, and its effect is further improved with VX-809 combination [bib_ref] EPAC1 activation by cAMP stabilizes CFTR at the membrane by promoting its..., Lobo [/bib_ref].
The CFTR-associated ligand (CAL) negatively regulates F508-CFTR PM abundance through its PDZ domain. CAL inhibition enhances the functional stability of F508-CFTR at the apical PM, implying an attractive therapeutic target for CFTR PM stabilizer [bib_ref] A stabilizing influence: CAL PDZ inhibition extends the half-life of F508-CFTR, Cushing [/bib_ref]. However, CAL also interacts with syntaxin 6 (STX6) and Golgilocalized E3-ligase membrane associated RING-CH type finger 2 (MARCH2) and regulates WT-CFTR PM expression [bib_ref] Ubiquitination and degradation of CFTR by the E3 ubiquitin ligase MARCH2 through..., Cheng [/bib_ref].
Filamin-A (FLN-A) is a membrane tethered actin adaptor protein and interacts with CFTR N-terminus region. S13F mutation of CFTR compromises FLN-A binding and consequently destabilizes the PM CFTR [bib_ref] Direct interaction with filamins modulates the stability and plasma membrane expression of..., Thelin [/bib_ref]. FLN-A binds with both WT and r F508-CFTR at similar level, however, its contribution to the CFTR PQC remains unclear.
## Protein kinases
The CFTR PM stability is regulated by phosphorylation. CFTR is predominantly phosphorylated at the R domain and also at nucleotide binding domain 1 (NBD1) and C-terminus residues by protein kinase A (PKA), protein kinase C (PKC), casein kinase II (CK2) and AMP-activated protein kinase (AMPK) for the channel function [bib_ref] Phosphorylation of protein kinase C sites in NBD1 and the R domain..., Chappe [/bib_ref] [bib_ref] Mechanistic insight into control of CFTR by AMPK, Kongsuphol [/bib_ref] [bib_ref] Contribution of casein kinase 2 and spleen tyrosine kinase to CFTR trafficking..., Luz [/bib_ref]. CK2 is predicted to regulate CFTR PM stability by phosphorylation at Thr-1471 where NHERF1 could interact [bib_ref] Detection of phospho-sites generated by protein kinase CK2 in CFTR: mechanistic aspects..., Venerando [/bib_ref]. Lemur tyrosine kinase 2 (LMTK2) phosphorylates CFTR at Ser-737 [bib_ref] Peptide microarray analysis of substrate specificity of the transmembrane Ser/Thr kinase KPI-2..., Wang [/bib_ref] and its KD or mutation at CFTR Ser-737 suppresses the endocytosis and increases CFTR PM density and stability [bib_ref] LMTK2-mediated phosphorylation regulates CFTR endocytosis in human airway epithelial cells, Luz [/bib_ref]. However, LMKT2 KD only modestly improves the PM function of r F508-CFTR [bib_ref] LMTK2-mediated phosphorylation regulates CFTR endocytosis in human airway epithelial cells, Luz [/bib_ref]. Spleen tyrosine kinase (SYK) phosphorylates CFTR at Tyr-512 and decreases CFTR PM levels possibly by triggering endocytosis [bib_ref] Antagonistic regulation of cystic fibrosis transmembrane conductance regulator cell surface expression by..., Mendes [/bib_ref]. Mixed-lineage kinase 3 (MLK3) pathway regulates not only F508-CFTR ERQC, but also the PQC by regulating the CFTR proteostasis [bib_ref] Unravelling druggable signalling networks that control F508del-CFTR proteostasis, Hegde [/bib_ref]. Inhibition of MLK3 pathway could regulate F508-CFTR folding/degradation switch by impairing interaction with PQC machinery such as HSP70/HSP90 Organizing Protein (HOP) [bib_ref] Unravelling druggable signalling networks that control F508del-CFTR proteostasis, Hegde [/bib_ref].
## Chaperones
Molecular chaperones selectively interact with and stabilize unfolded or partially folded protein to acquire a functionally active conformation. Nascent CFTR interacts with a panel of chaperones and co-chaperones including HSC70, HSP70, HSP90, and CNX at the ER [bib_ref] The common variant of cystic fibrosis transmembrane conductance regulator is recognized by..., Yang [/bib_ref] [bib_ref] Participation of the endoplasmic reticulum chaperone calnexin (p88, IP90) in the biogenesis..., Pind [/bib_ref] [bib_ref] Perturbation of Hsp90 interaction with nascent CFTR prevents its maturation and accelerates..., Loo [/bib_ref] [bib_ref] Delta F508 CFTR pool in the endoplasmic reticulum is increased by calnexin..., Okiyoneda [/bib_ref]. Even at the post-ER compartments, conformationally defective CFTR such as unfolded r F508-CFTR is recognized by chaperone/cochaperone complex [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. HSC70/HSP90 complex selectively interacts with unfolded r F508-CFTR at the post-Golgi and this interaction is crucial for the unfolding dependent ubiquitination [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. KD of HSC70/HSP90 complex (HSP90, HSC70, HOP, AHA1, DNAJB2, DNAJA1, BAG1) increases the r F508-CFTR PM stability in HeLa cells [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. The HSC70/HSP90 complex is also essential for maintaining kinetic and thermodynamic stability of r F508-CFTR at the PM by reshaping the CFTR conformation during energetic destabilization [bib_ref] Chaperones rescue the energetic landscape of mutant CFTR at single molecule and..., Bagdany [/bib_ref]. This chaperone activity also maintains r F508-CFTR channel function at the PM [bib_ref] Chaperones rescue the energetic landscape of mutant CFTR at single molecule and..., Bagdany [/bib_ref]. Thus, modulating the chaperone activity would be a viable target for attenuating the ubiquitination and for stabilizing the CFTR function at the PM.
## Ubiquitination enzymes
Ubiquitination determines CFTR elimination not only at the ER, but also from the PM. Ubiquitination is mediated by a sequential action of E1, E2, and E3 enzymes and this modification could be removed by DUB. Specifically, E3 Ub ligase has been proposed to determine the substrate specificity. CHIP is the first identified E3 ligase responsible for the CFTR PQC [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. Consistent with the action at the ER [bib_ref] The Hsc70 co-chaperone CHIP targets immature CFTR for proteasomal degradation, Meacham [/bib_ref] , CHIP selectively interacts with unfolded F508-CFTR at the post-Golgi through the HSC70/HSP90 chaperones. CHIP KD reduces the ubiquitination of unfolded F508-CFTR, resulting in the decelerated endocytosis and lysosomal delivery in HeLa cells [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. CHIP KD also stabilizes r F508-CFTR at the PM of polarized CFBE cells.
E3 ligase c-Cbl may play a role in the CFTR peripheral QC, but its contribution could be modest since its KD slightly increases r F508-CFTR PM stability in CFBE cells. c-Cbl also binds with WT-CFTR and decreases the PM stability without affecting the ubiquitination, suggesting that c-Cbl could regulate constitutive PM turnover of folded CFTR by inducing endocytosis through its C-terminus adaptor function [bib_ref] c-Cbl facilitates endocytosis and lysosomal degradation of cystic fibrosis transmembrane conductance regulator..., Ye [/bib_ref].
Nedd4-2 is a member of homologous to the E6-AP carboxyl terminus (HECT) E3 which may regulate the CFTR PM expression. Nedd4-2 KD reduces F508-CFTR ubiquitination at the ER, and increases the PM expression and function in CF pancreatic adenocarcinoma cell 1 (CFPAC1) and IB3-1 cells [bib_ref] Rescue of DeltaF508-CFTR by the SGK1/Nedd4-2 signaling pathway, Caohuy [/bib_ref]. Nedd4-2 binds both WT-and F508-CFTR while its role in the WT-CFTR ubiquitination remains controversial [bib_ref] Nedd4-2 does not regulate wt-CFTR in human airway epithelial cells, Koeppen [/bib_ref]. However, Nedd4-2 KD does not stabilize the PM r F508-CFTR in CFBE cells, implying its marginal contribution to the CFTR PQC [bib_ref] Nedd4-2 does not regulate wt-CFTR in human airway epithelial cells, Koeppen [/bib_ref]. Nedd4-2 is unlikely a viable CF drug target because its knock out (KO) induces CF-like lung phenotype by excessive function of epithelial Na + Channel (ENaC) [bib_ref] Deletion of the ubiquitin ligase Nedd4L in lung epithelia causes cystic fibrosislike..., Kimura [/bib_ref] [bib_ref] Nedd4-2 and the regulation of epithelial sodium transport, Rotin [/bib_ref].
A number of DUBs regulate the CFTR turnover. USP10, a DUB localized at early endosomes, interacts with WT-CFTR and reduces the CFTR poly-ubiquination in CFBE cells. The USP10mediated deubiquitination enhances the endocytic recycling of WT-CFTR [bib_ref] The deubiquitinating enzyme USP10 regulates the post-endocytic sorting of cystic fibrosis transmembrane..., Bomberger [/bib_ref]. The role of USP10 in the PM stability of conformationally defective CFTR such as r F508-CFTR remains unclear.
Recently, we have discovered RING finger and FYVE like domain containing E3 Ub protein ligase (RFFL) as a novel component of the CFTR PQC machineries by a comprehensive siRNA screen in CFBE cells. RFFL selectively recognizes unfolded r F508-CFTR through the disordered regions. RFFL promotes K63-linked polyubiquitination of the unfolded CFTR in post-Golgi, resulting in accelerated endocytosis and lysosomal degradation. Importantly, RFFL directly interacts with conformationally defective CFTR such as r F508-CFTR, but not with folded WT-CFTR at the PM and endosomes. Moreover, the RFFL-mediated ubiquitination is conformation dependent as it selectively ubiquitinates thermally unfolded NBD1. RFFL KD enhances the functional PM expression of r F508-CFTR in the presence of VX-809, and this effect is further improved by inhibiting the HSC70-dependent ubiquitination machinery. Thus, RFFL plays an important role in the chaperone-independent CFTR PQC mechanism in HBE cell models.
## Cftr modulators affecting the cftr pm stability pharmacological chaperones and chemical chaperones
Pharmacological chaperones affect the CFTR PM stability by direct stabilization. CFTR corrector VX-809 is the first food and drug administration (FDA) approved CFTR corrector in combination with VX-770/ivacaftor (known as Orkambi). VX-809 selectively improves the processing of misfolded CFTR by stabilizing NBD1-membrane spanning domain (MSD) interface but not other misfolded proteins such as human ether-à-go-go-related gene (hERG) mutants [bib_ref] Correction of the F508del-CFTR protein processing defect in vitro by the investigational..., Van Goor [/bib_ref] [bib_ref] Mechanism-based corrector combination restores F508-CFTR folding and function, Okiyoneda [/bib_ref] [bib_ref] Increased efficacy of VX-809 in different cellular systems results from an early..., Farinha [/bib_ref]. VX-809 repairs not only the CFTR folding defect at the ER but also the CFTR PM instability. VX-809 washout prolongs F508-CFTR functional sustainability [bib_ref] Correction of the F508del-CFTR protein processing defect in vitro by the investigational..., Van Goor [/bib_ref] , suggesting that improvement of the CFTR folding at the ER could increase the thermal stability and proper co-and/or post-translational modifications that renders CFTR more energetic robust conformations even at the PM. VX-809 also promotes F508-CFTR and NHERF1 interaction, that may increase the PM stability [bib_ref] Stabilizing rescued surface-localized δf508 CFTR by potentiation of its interaction with Na(..., Arora [/bib_ref]. C3 (CFcor-325/VRT-325) and C4 (Corr-4a) also extend r F508-CFTR PM stability in CFBE cells probably by directing binding [bib_ref] Correctors promote maturation of cystic fibrosis transmembrane conductance regulator (CFTR)-processing mutants by..., Wang [/bib_ref] [bib_ref] Enhanced cell-surface stability of rescued DeltaF508 cystic fibrosis transmembrane conductance regulator (CFTR)..., Varga [/bib_ref] although their effect could be not specific to the conformationally defective CFTR [bib_ref] Correction of the F508del-CFTR protein processing defect in vitro by the investigational..., Van Goor [/bib_ref]. Chemical chaperones such as glycerol also increases the r F508-CFTR PM stability probably by non-specifically improving the conformational stability [bib_ref] Mechanism-based corrector combination restores F508-CFTR folding and function, Okiyoneda [/bib_ref].
## Cftr potentiators
The first FDA approved CFTR potentiator VX-770 improves the gating defect of some CFTR mutants. However, chronic VX-770 treatment destabilizes the PM r F508-CFTR in CFBE and F508 homozygous CF patient HBE (CF-HBE) cells [bib_ref] Potentiator ivacaftor abrogates pharmacological correction of F508 CFTR in cystic fibrosis, Cholon [/bib_ref] [bib_ref] Some gating potentiators, including VX-770, diminish F508-CFTR functional expression, Veit [/bib_ref]. Importantly, chronic VX-770 treatment diminishes the VX-809 therapeutic efficacy by stimulating the elimination of PM r F508-CFTR [bib_ref] Potentiator ivacaftor abrogates pharmacological correction of F508 CFTR in cystic fibrosis, Cholon [/bib_ref] [bib_ref] Some gating potentiators, including VX-770, diminish F508-CFTR functional expression, Veit [/bib_ref]. In addition to VX-770, several CFTR potentiators including P1 (VRT-532) and P2 (PG-01) also decrease the r F508-CFTR PM stability [bib_ref] Some gating potentiators, including VX-770, diminish F508-CFTR functional expression, Veit [/bib_ref]. VX-770 and other potentiators could destabilize a variety of CFTR rare mutants referred to as CFTR2 mutants including E92K and L1077P at the PM . Thus, several CFTR potentiators may decrease the thermal stability of metastable mutant CFTR at the PM by inducing conformational change that positively affects for channel gating but negatively affects stability. High-throughput screening has identified several novel CFTR potentiators such as class A analog 4 (A04) and class P analog 12 (P12) that could not destabilize the PM r F508-CFTR [bib_ref] Potentiators of defective F508-CFTR gating that do not interfere with corrector action, Phuan [/bib_ref].
## Proteostasis regulating drugs
Proteostasis regulating drugs that affect array of proteins regulating CFTR folding and QC also affect the CFTR PM stability. Histone deacetylase (HDAC) inhibitor suberoylanilide hydroxamic acid (SAHA) alters expression of a subset of CF-interacting gene products (e.g., chaperones and DAB2) and sustains PM expression of F508-CFTR in CFBE cells [bib_ref] Reduced histone deacetylase 7 activity restores function to misfolded CFTR in cystic..., Hutt [/bib_ref]. Tissue transglutaminase (TGM2) inhibitor cystamine also stabilizes F508-CFTR at the PM of airway epithelial cells by restoring BECN1 interactome which is sequestrated by CFTR dysfunction [bib_ref] Targeting autophagy as a novel strategy for facilitating the therapeutic action of..., Luciani [/bib_ref] [bib_ref] Targeting the intracellular environment in cystic fibrosis: restoring autophagy as a novel..., Villella [/bib_ref]. MLK3 pathway inhibitor oxozeaenol has been reported to be effective in correcting the F508-CFTR proteostasis defect in the primary HBE cells [bib_ref] Use of kinase inhibitors to correct F508-CFTR function, Trzcinska-Daneluti [/bib_ref]. Oxozeaenol could stabilize F508-CFTR at the PM as MLK3 KD reduces mature F508-CFTR elimination by PQC [bib_ref] Unravelling druggable signalling networks that control F508del-CFTR proteostasis, Hegde [/bib_ref].
## Cavosonstat and cal inhibitor
HSP70/HSP90 Organizing Protein is adaptor protein which coordinates HSP70 and HSP90 function in protein folding and regulates CFTR maturation and PM stability [bib_ref] Hop: more than an Hsp70/Hsp90 adaptor protein, Odunuga [/bib_ref] [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. HOP S-nitrosylation by S-nitrosoglutathione (GSNO) induces HOP degradation and increases F508-CFTR PM expression. Levels of S-nitrosothiols such as GSNO are low in CF airway [bib_ref] Decreased levels of nitrosothiols in the lower airways of patients with cystic..., Grasemann [/bib_ref] and S-nitrosothiol decreases the internalization rate of r F508-CFTR in HBE cells [bib_ref] Decreased levels of nitrosothiols in the lower airways of patients with cystic..., Grasemann [/bib_ref] [bib_ref] Airway nitric oxide levels in cystic fibrosis patients are related to a..., Grasemann [/bib_ref] [bib_ref] S-Nitrosothiols increases cystic fibrosis transmembrane regulator expression and maturation in the cell..., Zaman [/bib_ref]. Cavosonstat (N91115) is an orally bioavailable inhibitor of GSNO reductase and restores GSNO levels [bib_ref] Pharmacokinetics and safety of cavosonstat (N91115) in healthy and cystic fibrosis adults..., Donaldson [/bib_ref]. Cavosonstat is the first CFTR stabilizer in phase II trials, but it was not beneficial for improvement of lung function in combination with ivacaftor.
CFTR-associated ligand binds CFTR via a PDZ interaction domain and targets CFTR for lysosomal degradation. CAL inhibition increases the PM stability of F508-CFTR [bib_ref] A stabilizing influence: CAL PDZ inhibition extends the half-life of F508-CFTR, Cushing [/bib_ref] and cell penetrating CAL inhibiting peptide is established [bib_ref] Intracellular delivery of peptidyl ligands by reversible cyclization: discovery of a PDZ..., Qian [/bib_ref]. CAL inhibitor has been developed as a cell surface CFTR stabilizer in pre-clinical level while its therapeutic efficacy and conformational selectivity remain unclear.
## Ub ligase inhibitors
RING finger protein 5 (RNF5/RMA1) is an ER associated E3 Ub ligase that regulates early stage CFTR proteostasis at the ER [bib_ref] Sequential quality-control checkpoints triage misfolded cystic fibrosis transmembrane conductance regulator, Younger [/bib_ref]. A RNF5 inhibitor Inh-2 identified by homology modeling and virtual ligand screening causes significant rescue of F508-CFTR in immortalized and primary HBE cells from CF patients [bib_ref] Pharmacological inhibition of the ubiquitin ligase RNF5 rescues F508del-CFTR in cystic fibrosis..., Sondo [/bib_ref]. Intriguingly, Inh-2 modestly increases mature F508-CFTR half-life and this stabilization effect is further improved by VX-809. While the contribution of RNF5 in the CFTR peripheral QC remains unclear, RNF5 inhibitor may be useful to overcome the CFTR instability.
Currently, CHIP and RFFL are the only Ub ligases responsible for the CFTR peripheral QC [bib_ref] Peripheral protein quality control removes unfolded CFTR from the plasma membrane, Okiyoneda [/bib_ref]. Thus, inhibiting their activity could selectively reduce the ubiquitination and elimination of unfolded CFTR from the PM, improving the limited efficacy of CF combination therapy. CHIP binds and regulates a number of substrates via chaperones [bib_ref] The co-chaperone CHIP regulates protein triage decisions mediated by heat-shock proteins, Connell [/bib_ref]. Moreover, inhibiting the CHIP activity induces deleterious effect as the CHIP KO mice result in the abnormal phenotypes including ataxia and pre-mature death 1 . In contrast, RFFL could bind and regulate a limited number of substrates because of its nature of direct binding to the CFTR through the disordered regions. More importantly, inhibiting the RFFL activity seems to have no venomousness since the RFFL KO mice exhibit no abnormal phenotype [bib_ref] CARP2 deficiency does not alter induction of NF-kappaB by TNFalpha, Ahmed [/bib_ref]. Therefore, counteracting RFFL activity may provide a preferable therapeutic approach as a CFTR stabilizer that is a class of drugs that extends the PM resident time of CFTR class VI mutants. Although future studies 1 http://www.informatics.jax.org/marker/phenotypes/MGI:1891731 are needed to validate the impact on F508-CFTR in CF-HBE cells, developing agents selectively inhibiting RFFL-mediated CFTR ubiquitination may help improve the efficacy of CF pharmacological therapy.
# Conclusion and perspective
Beside the progresses of CF pharmacological therapy, stabilizing the cell surface CFTR remains challenging and is necessary to improve the limited therapeutic efficacy. Recent studies have revealed some of the CFTR PQC mechanism eliminating the functional but conformationally defective CFTR from the PM. Understanding the CFTR PQC mechanism help the development of the CFTR stabilizer, a novel class of CFTR modulator necessary to establish the robust CF pharmacological therapy.
# Author contributions
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.
# Funding
This work was supported by JSPS KAKENHI grant numbers JP25893275, 15H05643, and 15H01192.
September 2018 | Volume 9 | Article 1100
Frontiers in Pharmacology | www.frontiersin.org
Conflict of Interest Statement:TO has a patent pending in Japan for methodology to identify inhibitors of RFFL-mediated CFTR ubiquitination (2017-047626).The remaining author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.Copyright © 2018 Fukuda and Okiyoneda. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
|
Fluorographene based Ultrasensitive Ammonia Sensor
Single molecule detection using graphene can be brought by tuning the interactions via specific dopants. Electrostatic interaction between the most electronegative element fluorine (F) and hydrogen (H) is one of the strong interactions in hydrogen bonding, and here we report the selective binding of ammonia/ammonium with F in fluorographene (FG) resulting to a change in the impedance of the system. Very low limit of detection value of ~0.44 pM with linearity over wide range of concentrations (1 pM-0.1 μM) is achieved using the FG based impedance sensor, andthisscreen printed FG sensor works in both ionized (ammonium) and un-ionized ammonia sensing platforms. The interaction energies of FG and NH 3 /NH 4 + are evaluated using density functional theory calculations and the interactions are mapped. Here FGs with two different amounts of fluorinecontents −~5 atomic% (C 39 H 16 F 2 ) and ~24 atomic% (C 39 H 16 F 12 ) -are theoretically and experimentally studied for selective, high sensitive and ultralow level detection of ammonia. Fast responding, high sensitive, large area patternable FG based sensor platform demonstrated here can open new avenues for the development of point-of-care devices and clinical sensors.
Several articles have been reported in the past discussing the development of various gas sensors, and ammonia is one of the gases taken seriously due to its natural and industrial origin [bib_ref] Ammonia Sensors and Their Applications -A Review, Timmer [/bib_ref] [bib_ref] Mass Ammonia Inhalation Burns-Experience in the Management of 12 Patients, Leung [/bib_ref] [bib_ref] Exhaust Gas Sensors for Automotive Emission Control, Riegel [/bib_ref] [bib_ref] The Electronic Nose Applied to Dairy Products: A Review, Ampuero [/bib_ref]. But low level detection of ammonia (< 2 ppb for environmental monitoring and < 50 ppb for breath analysis) is still lacking a state of the art detection method, and its development is highly desirable for the development of clinical ammonia sensors 1 . Low dosages of ammonia can cause problems to human respiratory, skin, eyes etc., and its prolonged exposure can leads to pulmonary oedema 2 . Since ammonia is also excreted from human body in the form of urea, breathe ammonia measurement (measuring the quantity in the exhaled gas) can be a diagnostic tool for monitoring the function of kidney or ulcers caused by Helicobacter pyloribacterial stomach infection. The amount of exhaled air available will be minimal and hence high sensitive ammonia breathe sensors are highly demanding, but lacking at present. Apart from ammonia gas sensors, analysis of ammonia (ionized/un-ionized) level in blood is also of interest in medicine. Blood ammonia monitoring has of interests in sports medicines (0.1 to 10 ppm) and pediatrics (> 100 μ mol/L) too [bib_ref] Exhaust Gas Sensors for Automotive Emission Control, Riegel [/bib_ref] [bib_ref] The Electronic Nose Applied to Dairy Products: A Review, Ampuero [/bib_ref].
Presently various sensing platforms are existing for ammonia: metal oxide based sensors [bib_ref] Gas-Sensing Characteristics of Modified MoO 3 Thin Films using Ti-Overlayers for NH..., Imawan [/bib_ref] [bib_ref] Tin Oxide Gas Sensors: An Analytical Comparison of Gas-Sensitive abd Non-Gas-Sensitive Thin..., Huebner [/bib_ref] (by conductance change due to the chemisorption of gas molecule), catalytic ammonia sensors (by catalytic activity of certain metals towards ammonia gas and the ammonia concentration dependency in the charge carriers) 9,10 , conducting polymer based ammonia sensors (by a two-fold oxidation-reduction mechanism) [bib_ref] DC-Resistometric Urea Sensitivity Device Utilizing a Conducting Polymer Film for the Gas-Phase..., Palmqvist [/bib_ref] [bib_ref] Regeneration of an ElectropolymerizedPolypyrrole Layer for the AmperometricDetetion of Ammonia, Heiduschka [/bib_ref] , optical and spectrometric ammonia detection (cause in the coloration up on ammonia exposure as a measure) [bib_ref] Inexpensive Near-Infrared Diode-Laser-Based Detection System for Ammonia, Feher [/bib_ref] , gas permeable membranes based selective detection techniques etc. are the most frequently used techniques [bib_ref] Hybrid Microfabricated Device for Field Measurement of Atmospheric Sulfur Dioxide, Ohira [/bib_ref] [bib_ref] Wet Effluent Denuder Coupled Liquid/Ion Chromatography Systems, Simon [/bib_ref]. But these methods lack the synergy of selective, sensitive, cost effective, and fast detection platform. A clinical ammonia sensor demands the above mentioned features along with a limit of detection value (LOD) of ~50 ppb within a response time of a few minutes. Existing selective optical methods are inadequate for the development of a portable (point of care (POCs)) ammonia sensors where limited sample needs to be detected using economically viable routes [bib_ref] Ammonia Sensors and Their Applications -A Review, Timmer [/bib_ref].
Graphene based biosensors are receiving tremendous scientific attention due to the engineering possibilities of graphene by bringing specificity and sensitivity in the device via doping and defects. Further, robustness and availabilities of different detections mechanisms of the graphene based devices (electronic transducers (field effect transistor), electrochemical (conductometry, potentiometric, amperometric, and impedance) sensors etc.) attract the development of graphene based sensors [bib_ref] Engineered 2D Nanomaterials-Protein Interfaces for Efficient Sensors, Tadi [/bib_ref] [bib_ref] Differential Response of Doped/Defective Graphene and Dopamine to Electric Fields: A Density..., Ortiz-Medina [/bib_ref]. Electrostatic interaction can bring specificity in the binding of molecules to graphene and hence graphene based sensors are capable of detecting individual gas molecules [bib_ref] Detection of Individual Gas Molecules Adsorbed on Graphene, Schedin [/bib_ref]. Charge transfer kinetics between graphene and adsorbed molecules can be tuned by doping, and hence the response time of a graphene based sensor can also be engineered. An external molecule can change the carrier density of graphene surface and itmayalso act as chemically doped entity on graphene driven by the electrostatic interactions among them [bib_ref] Self-Activated Transparent All-Graphene Gas Sensor with Endurance to Humidity and Mechanical Bending, Kim [/bib_ref]. Heteroatom doping on graphene can enhance the catalytic activity of the neighborhood of the dopant, and this can enhance the adsorption of molecules and affect the conductivity of graphene backbone. Moreover, it is also important to study the interaction between adsorbed molecules and graphene to understand the charge transfer mechanism, and tailoring the response by defects/dopants 20 .
Tang et al. have reported Density Functional Theory (DFT) calculations based graphene oxide (GO) -ammonia interactions [bib_ref] Adsorption and Dissociation of Ammonia on Graphene Oxides: A First-Principle Study, Tang [/bib_ref]. Adsorption of ammonia on GO is stronger than graphene due to the presence of active functional groups such as hydroxyl and epoxy -where they form hydrogen bonding with ammonia leading to a charge transfer process. But the poor stability of GO in various solutions [bib_ref] Deoxygenation of Exfoliated Graphite Oxide Under Alkaline Conditions: A Green Route to..., Fan [/bib_ref] , low thermal stability, instability of GO under long exposure of ammonia (possible reduction), and difficulties in controlling/benchmarking the extend of oxidation (C/O ratio) are still bottlenecks in the development of GO based gas sensors. Moreover, GO is an electrical insulator and also less electrochemically active. Recently, Ghosh et al. reported a reduced GO (RGO) based conductometric ammonia sensor with the LOD of 200ppm and also mentioned its some extend of selectivity towards ammonia [bib_ref] Chemically Reduced Graphene Oxide for Ammonia Detection at Room Temperature, Ghosh [/bib_ref]. The detection mechanism of RGO sensor is same as that in GO with an enhanced electrical conduction platform due to the reduction of some functional groups. Hence RGO cannot be a recommendable sensing platform for ammonia sensing. But, hydrogen bonding possibilities of ammonia with other thermally and chemically stable functional groups of graphene may be a viable option for graphene based ammonia sensor. Hydrogen bonding between the most electronegative element fluorine (F) and ammonia will be an ideal option, and hence fluorographene with optimum fluorine content (to optimize the electrical conductivity) can selectively detect ammonia with very high sensitivity.
Relatively high thermal stability of fluorographene (300-400 °C)will help the sustainability of FG based sensors. In the recent past, one of the authors has extensively studied the properties of FG derived from fluorinated graphite polymer ((CF 0.25 ) x ) [bib_ref] Power limiting in fluorinated graphene oxide: An insight into the nonlinear optical..., Chantharasupawong [/bib_ref] [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref] [bib_ref] Improved Heterogeneous Electron Transfer Kinetics of Fluorinated Graphene Derivatives, Boopathi [/bib_ref] [bib_ref] Fluorinated Graphene Oxide; A New Multimodal Material for Biological Applications, Romero-Aburto [/bib_ref]. It was found that fluorination alters the physical, chemical, electrical and electrochemical properties of graphene/GO. The high polarity of C-F bond even modifies the dielectric permittivity of graphene and it also induce sparamagnetism to graphene/GO backbone [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref]. The charge transfer studies on fluorinated graphene systems indicate that the density of states near the Fermi level can also be altered by the fluorine doping [bib_ref] Improved Heterogeneous Electron Transfer Kinetics of Fluorinated Graphene Derivatives, Boopathi [/bib_ref]. Hence it is established that FG is stable even after the exposure to various alkaline/acid environments and also after elevated heating (100 °C).
In the present study, we report the interactions of FG and ammonia (unionized and ionized (ammonium)) using DFT calculations. The interactions between graphene and ammonia are also explored to compare the strength of FG-ammonia interactions. The FGs in the present study are derived from fluorinated graphene oxide (FGO, with 2-types of FGOs − 5 at.% fluorine and 24~ at.% fluorine) and the role of residual functional groups in the interactions is not studied. Later, using Fourier Transform-Infrared (FT-IR) studies, we found that the interaction between fluorine and ammonia is prominent than other hydrogen bonding interactions via residual groups. FG based screen printed electrode is developed and its efficacy in ammonia sensing using impedance spectroscopy is demonstrated. Large area patternability of chemically derived FG is shown using a soft-lithography process, indicating the possibilities of FG based different sensor geometries.
# Results and discussions
Interaction Energy Calculations. FG has a carbon honeycomb lattice like in graphene with covalent fluorine (F) attached to carbon via sp 3 hybridization. FG is an atomically layered material as it is evident from the electron microscope images given in the supporting information, [fig_ref] Figure 1: Optimized graphene/FG and ammonia/ammonium structures [/fig_ref] ,B. The bond lengths were calculated from fully optimized structures of graphene and FG (~5at%) and the values are given in the supporting information, . It is evident from the Table S1 that the edge bond lengths of C-C and C= C are less than the internal bond lengths of the same, and it is due to the strains at the edges, as reported by other researchers [bib_ref] Tunable Assembly of sp3 Cross-Linked 3D Graphene Monoliths: A First-Principles Prediction, Jiang [/bib_ref]. lists the basis set superposition error (BSSE) corrected interaction energies of graphene and FG complexes with ammonia and ammonium ion (NH 4 + ). The DFT studies reveal that FG is having strong interactions (− 14.15 kcal/ mol)with ammonia compared to graphene (− 1.29 kcal/mol) due to the strong hydrogen bonding between C-F in FG and N-H in ammonia. The optimized structures of complexes are shown in [fig_ref] Figure 1: Optimized graphene/FG and ammonia/ammonium structures [/fig_ref]. From the optimized structures, it is found that the N-H groups of ammonia are interacting with fluorine atoms present in the FG. The simulated closest distance of approach between FG and ammonia is about 2.90 Å whereas in case of graphene, the distance is found to be 3.10 Å. Moreover, an increase in the C-F bond length of FG from 1.416 Å to 1.426 Å is also observed after the complex formation with ammonia, and it is later verified using Fourier Transform Infrared spectroscopy (FT-IR) studies. These results indicate the presence of weak H-bonding (result of a dipolar electrostatic interaction) involved between the N-H group of ammonia and fluorine of FG. In the case of ammonium ion, even though the interaction energy of FG-NH 4 + (− 53.72 kcal/mol) is higher than the G-NH 4 + , the difference is less than that in the case of ammonia. This might be due to the enhanced electrostatic interaction between positively charged of ammonium ion and graphene (π electrons).
The optimized structures and electrostatic interactions between ammonia/ionized ammonia (NH 4 + ) and graphene/FG were calculated from HOMO and LUMO energies of the corresponding complexes. The calculated energy gaps are shown in [fig_ref] Table 2: HOMO-LUMO gaps for various complexes [/fig_ref]. It is evident from the [fig_ref] Table 2: HOMO-LUMO gaps for various complexes [/fig_ref] that the energy gaps between NH 3 /NH 4 + and FG is less than that of NH 3 /NH 4 + and graphene. This can be due to the strong interactions of NH 3 or NH 4 + with FG, making the energy differences between HOMO and LUMO smaller. Further, the electrostatic potential maps of the interactions are shown in [fig_ref] Figure 2: Electrostatic potential plots for [/fig_ref]. The positive charged 1s orbital of hydrogen on NH 3 /NH 4 + has strong electrostatic interaction with negatively charged 2p orbital of fluorine on FG, leading to a hydrogen bonding (electrostatic) interaction. The interaction maps show considerable orbital overlapping of NH 3 and NH 4 + with FG, while little interaction between NH 3 and graphene is observed as it is reported by other researchers [bib_ref] Soft-lithographic processed soluble micropatterns of reduced graphene oxide for wafer-scale thin film..., Jia [/bib_ref]. But NH 4 + shows a better interaction with graphene due to its positive charge.
## Ft-ir and micro raman analysis. the theoretical prediction of interaction between fg and ammonia
is experimentally validated using FT-IR and micro Raman spectroscopy. Chemically derived (explained in the method section) FG powder (low resolution transmission electron microscope (TEM) image and high angle annular dark field TEM images are shown in supporting information [fig_ref] Figure 1: Optimized graphene/FG and ammonia/ammonium structures [/fig_ref] , indicate the crystallinity and two dimensional sheet like morphology of FG) having 5 at. % F content (calculated using X-ray photoelectron spectroscopy (XPS), the C1s spectrum of FG is given in supporting information [fig_ref] Figure 1: Optimized graphene/FG and ammonia/ammonium structures [/fig_ref] is used for the investigation. The FT-IR spectrum of FG is also simulated and is shown in(y-axis of theoretical curve is absorbance), and it exactly matches with that of FG (experimental) [bib_ref] Power limiting in fluorinated graphene oxide: An insight into the nonlinear optical..., Chantharasupawong [/bib_ref] [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref]. The presence of covalent C-F linkages in FG is further confirmed from the F 19 Nuclear Magnetic Resonance (NMR) spectrum, as shown in the supporting information, [fig_ref] Figure 1: Optimized graphene/FG and ammonia/ammonium structures [/fig_ref]. It has to be noted that the absorption peak at ~3250 cm −1 in theoretical FG is due to the = CH stretching of the end terminated hydrogen while in the case of FG experimental spectrum it is due to the -OH stretching from residual functional groups. In, FG-0s depicts the experimental FTIR spectrum
## Complex
## Homo (ev) lumo (ev)
Band-gap (eV) (y-axis is % transmittance unlike the theoretical spectrum) of FG without ammonia exposure. The covalent C-F peak at ~1200 cm −1 is clear in both theoretical and experimental spectra. The FG has been exposed to ammonia gas (pressure ~1 psi) for different exposure times, and the corresponding FT-IR spectra are shown in. The increase in exposure time increases the intensity of C-F vibrations along with a continuous red shift in the vibration frequency. This increase in the intensity of C-F vibration along with the red shift is the characteristic feature of normal hydrogen bonding [bib_ref] Red-, Blue-, or No-Shift in Hydrogen Bonds: A Unified Explanation, Joseph [/bib_ref]. It is also noticed that vibration frequency bands corresponding to residual -OH stretching or other oxygen functionalities (residual) of FG are not changed after the exposure of ammonia. This indicates that ammonia has a preferential binding with fluorine than with other functional groups. The micro-Raman spectra of graphene and FG before and after exposing to ammonia are shown in. All the spectra show typical graphitic Raman signatures corresponding to G and D peaks at 1600 and 1350 cm −1 , respectively. The G band of graphene originates from the vibration of in-plane sp 2 carbon. The D band is attributed to the defects in graphene. There are no considerable changes in the positions and intensity ratios of D and G bands before and after the exposure of ammonia indicating that the graphene structure is not disturbed after the ammonia exposure (structure distortion or reduction of oxygen functionalities via ammonia exposure is well reported the literature) [bib_ref] Deoxygenation of Exfoliated Graphite Oxide Under Alkaline Conditions: A Green Route to..., Fan [/bib_ref]. Impedance Sensor. The above mentioned selective interaction of ammonia/ammonium with fluorine in FG via the electrostatic hydrogen bonding is utilized for the development of an ammonia sensor, as explained in the following section. Electrochemical impedance spectroscopy (EIS) was used to study the binding of ammonia on FG modified screen printed electrodes [fig_ref] Figure 2: Electrostatic potential plots for [/fig_ref]. Theillustrates the Nyquist plots, graph between the real (Z′ ) and imaginary (− Z′ ′ ) parts of impedance, using the screen printed electrode (SPE, shown in, and supporting information -where working electrode is modified with graphene samples as explained in the method section) obtained after the gradual increase in concentration of ammonia (ionized ammonia) from 1 pM to 0.1 μ M (i-vii).
[formula] 1 FG-NH 3 − 0.08023 − 0.20139 0.121 2 Graphene-NH 3 − 0.07987 − 0.2022 0.122 3 FG-NH 4 + − 0.18358 − 0.30316 0.119 4 Graphene-NH 4 + − 0.16822 − 0.30455 0.136 [/formula]
It is evident from thethat the charge transfer resistance (R ct ) increases with increase in ammonium ion concentration. The increase in R ct is attributed to the binding of ammonium to the FG via hydrogen bonding. In order to confirm that the observed change in impedance is due to surface modification of SPE and not due to superimposed effects, ratio of charge transfer resistance for the desired concentration (R ct (C i )) and charge transfer resistance of the blank FG electrode (R ct (C o )) is plotted against the logarithm of ammonia concentration.can be linearly fitted tothe following equation; R ct (C i )/R ct (C o ) = 1.723 + 0.055 log C NH3 . This shows that the sensor works in a linear concentration of ammonia in the range of 1 pM to 0.1 μ M with a correlation coefficient of 0.998. Sensitivity of the sensor can be deduced from slope of this curve, resulting to a value of ~0.055 M −1 . The limit of detection (LOD) of the sensor has been calculated according to Long and Wineforder method 32 :
[formula] = . * . . − LOD [(3 3 S D ) a]/b(1) [/formula]
where S.D. is the standard deviation of the blank, a is the linear coefficient, and b is the angular coefficient (sensitivity). The calculated LOD is found to be ~0.44pM. Relative change in EIS data is more reliable for sensing applications than absolute impedance. Previous discussion indicates that a graph between the change in R ct (Δ R ct ) values and the logarithm of ammonia concentrationsreveals a linear detection range for ammonia concentrations in the range of 1 pM to 0.1 μ M with. A curve between logarithm of frequency and imaginary part of impedance (− Z") has also been plotted (supplementary information,.shows that − Z" increases continuously with increase in ammonia concentration.
Further, EIS measurements are also conducted by direct purging of ammonia in to the phosphate buffer solution (PBS) (10 mM, pH 7.0) containing the mixture of 5 mM[Fe(CN) 6 ] 4− (ferrocyanide) and 5 mM[Fe(CN) 6 ] 3− (ferricyanide). The experimental set up is shown in supporting information. The impedance spectra (supporting informationshow a continuous increase in impedance with ammonia bubbling time. This also indicates that the interaction between FG and ammonia and the fast response in the sensing (within a few seconds).
In order to check the efficacy of FG based ammonia sensor for direct ammonia detection, ammonia gas has been directly exposed to the FG modified electrodes (< 1 psi pressure, 7 mm 2 area) and the EIS was conducted immediately after the exposure. The EIS spectra are recorded for different times of ammonia exposure as mentioned in the, and the corresponding Nyquist plots are shown. The variation in R ct with exposure time seems to be linearly increased with the exposure of ammonia. The values of R ct are linearly increased up to 30 s and then the response becomes lessen.
Further the impedance response of the bare (SPE), graphene oxide (GO), graphene and FGO coated electrodes with increase in concentration of ammonia (ionized ammonia) is studied. Sensitivity calibration curves (concentrations of NH 4 + vsΔ R ct ) over a range of ammonia concentration from 1 pM to 0.1 μ M for the various modified electrodes are shown in. From the plot, it is clear that the GO coated SPE has the lowest sensitivity towards ammonia (sensitivity 61 Ω M −1 ), while the sensitivity has been slightly increased to 104 Ω M −1 after the reduction of GO. FG showed the highest sensitivity (293 Ω M −1 ). However, the sensitivity of the FGO (125 Ω M −1 ) is less than that of FG, but higher than that of GO. It has been discussed in our previous reports that GO and FGO have similar structure, morphology and C/O ratio 27 . The only difference between them is the presence of fluorine (here ~5 at.%) in FGO. Hence this enhanced sensitivity of FG/FGO is due to the presence of fluorine in the honeycomb lattice.
Reversibility (whether the binding of analyte is reversible or not) of an electrochemical sensor is one of important features of a chemical sensor. In the present case, reversibility of the electrode has been checked by performing the impedance studies after direct exposing to NH 3 gas (as discussed in the previous section) followed by a simple washing of electrode with deionized water (running water). The decrease in the impedance (R ct ) after washing is evident from the, indicating that the electrodes are reversible.
In order to study the effect of higher amount of fluorine doped graphene (C 39 H 16 F 12 ) for ammonia sensing, DFT calculations were conducted on ~24 atomic % fluorine containing graphene (HFG) (the optimized structure is shown in supporting information, . The BSSE corrected interaction energy for HFG-NH 3 is found to be − 3.926210211 kcal/mol. This indicates that HFG-NH 3 interactions are weaker than that of FG-NH 3 . The C-F bond lengths before and after NH 3 stabilization are calculated as 1.378 Å and 1.380 Å respectively. These values are in agreement with the recent calculations of bond lengths of fluorgraphene [bib_ref] Diffusion of Fluorine on and between Graphene Layers, Sadeghi [/bib_ref]. Unlike in the case of FG, there is negligible change in C-F bond length of HFG after ammonia stabilization. Moreover, the closest distance between NH 3 and HFG is found to be 3.76 Å, which is larger than that in FG (2.9 Å). This indicates that the binding of HFG with NH 3 is rather weak. The poor interaction between HFG and NH 3 is further confirmed experimentally via EIS studies (supporting information, . Here the change in impedance with NH 3 concentration is found to be minimal. The increase in fluorination will increases the defect levels and also decreases the electrical conductivity [bib_ref] Power limiting in fluorinated graphene oxide: An insight into the nonlinear optical..., Chantharasupawong [/bib_ref] [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref] [bib_ref] Improved Heterogeneous Electron Transfer Kinetics of Fluorinated Graphene Derivatives, Boopathi [/bib_ref] [bib_ref] Fluorinated Graphene Oxide; A New Multimodal Material for Biological Applications, Romero-Aburto [/bib_ref]. Hence this study indicates that the amount of F in FG needs to be optimized for the development of a practical ammonia sensor. This optimization is important in applied aspects of other doped graphene too, where the enhanced defect levels and interactions can adversely affect the properties. The added advantage of chemically derived FG is the ability to make micro-electrodes and patternable devices using simple techniques such as soft-lithography [bib_ref] Electrochemistry and Soft Lithography: A Route to 3-D Microstructures, Jackman [/bib_ref]. Large area electrode patterns constructed using FG are shown in supporting information . Since soft-lithography (polydimethylsiloxane (PDMS) stamps) based imprinting (solvent assisted micro-molding technique, solvent used is dimethyl formamide (DMF)) can be made on substrates like plastics/glasses/cellulosic papers etc., sensors with flexibilities and visible light transmitting can be developed using these chemically derived atomic layers. A Raman mapping conducted on such an FG pattern is also shown in supporting showing the uniformity of patterns. This indicates that the FG based ammonia sensors can bring features such as sensitivity, selectivity, fast response and reversibility along with other novel aspects of modern POCs and biosensors such as transparency and mechanical flexibility.
# Conclusions
This study reveals the possibility of making a commercial ammonia sensor (both gaseous and ionized) by doping graphene with appropriate amount of fluorine -bringing all the key features (sensitivity, selectivity, fast response and reversibility) required for a practical chemisensor. The interactions between ammonia (both ionized and un-ionized) and FG were studied using DFT calculations, and the results were compared with that of graphene-ammonia/ammonium ion systems. An augmented electrostatic interaction is observed in FG-ammonia/ammonium ion (− 14.15 kcal/mol and − 53.72 kcal/mol respectively) systemsthan that of graphene -ammonia/ammonium ion (− 1.29 kcal/mol and − 37.26 kcal/mol), and it is established through the hydrogen bonding interactions via fluorine and hydrogen. FG modified screen printed electrodes were studied for ammonia sensing in both solution and gas phases, and a very low limit of detection of 4.4 × 10 −13 M (0.44 pM) with linearity over a wide range of concentrations (1 pM-0.1 μ M) is achieved using this FG based impedance sensing platform. This study also points out the importance of optimization of dopant levels in graphene for its best sensing performance. Further, the possibility of large scale microelectrode patterning ability using these chemically derived FG with the aid of soft-lithography is demonstrated and it reveals the promises of FG towards the development of flexible and transparent ammonia sensors and POCs. Recent research indicates that such a flexible impedance sensing device can be a futuristic in-vivo non-invasive diagnostic platform for static and dynamic continuous monitoring modes.
# Methods
## Computational methods. thestructures of fgs and graphene, and the interactions were studied via
Density Functional Theory calculations using Gaussian 09 soft-ware package. All the structures were subjected to full geometry optimizations without any constraints at the M05-2X/6-31G*level. This method was chosen as it appears to be better compared to the more popular B3LYP alternatives when modeling non-bonding interaction [bib_ref] Assessment of the Performance of the M05-2X and M06-2X Exchange-Correlation Functionals for..., Hohenstein [/bib_ref].
All stationary points were characterized as minima after verifying the presence of all real frequencies. Single point energy calculations were carried out at the M05-2X/6-311+ + G** level. The interaction energy (IE) was calculated using equation [bib_ref] Mass Ammonia Inhalation Burns-Experience in the Management of 12 Patients, Leung [/bib_ref] as the difference between the total energy of the complex (E G-X or E FG-X ) and the sum of the total energy of the parent 2D-nano material (E graphene (E G ) or E FG ) and the binding molecule (E X , X = NH 3 or NH 4
[formula] + ). = + − IE E ( E E )(2) [/formula]
Gor(FG) X GorFG X The interaction energy was corrected for basis set superposition error (BSSE) using the counter poise (CP) correction scheme (which is within 1 kcal/mol).
Frontier molecular orbitals of the complexes were studied using the energy level gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) respectively. Materials Synthesis and Characterization. Fluorinated graphite polymer (cat. No. 42537) was purchased from Alfa Aesar. Graphite polymer (particle size < 20 microns), ammonium chloride and KMnO 4 were procured from Sigma-Aldrich. All chemicals were analytical reagent grade and used without further purification. Aqueous solutions were prepared using Millipore water received from Milli-Q system (Millipore Inc.).
The preparation of GO and FGO was adopted from the previous report [bib_ref] Power limiting in fluorinated graphene oxide: An insight into the nonlinear optical..., Chantharasupawong [/bib_ref] [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref] [bib_ref] Improved Heterogeneous Electron Transfer Kinetics of Fluorinated Graphene Derivatives, Boopathi [/bib_ref] [bib_ref] Fluorinated Graphene Oxide; A New Multimodal Material for Biological Applications, Romero-Aburto [/bib_ref]. A mixture of concentrated sulfuric acid and phosphoric acid in 9:1 ratio (360 mL: 40 mL) was added to graphite (3.0 g, 1 wt equiv.). It was followed by gradual flake by flake addition of KMnO 4 (18.0 g, 6 wt equiv.). The reactants were then heated to 90 °C and stirred for 12 h. The reactants were brought it down to room temperature and poured onto ice with 30% H 2 O 2 (3 mL) and stirred for 2 h. The obtained solid suspension was washed several times successively with excess of deionized water, 30%HCl, and ethanol. Finally, it was coagulated with 200 mL of ether and filtered over a PTFE membrane with a 0.22 μ m pore size. Fluorinated GO (FGO) was also prepared in similar method by taking Fluorinated graphite polymer instead of graphite as the starting material [bib_ref] Power limiting in fluorinated graphene oxide: An insight into the nonlinear optical..., Chantharasupawong [/bib_ref] [bib_ref] Fluorinated Graphene Oxide for Enhanced S and X-band Microwave Absorption, Sudeep [/bib_ref]. Two different types of FGOs (having different fluorine content) were resulted from the phase separation of the material while H 2 O 2 was added to the mixture. The reduction of functional groups in FGO results to the formation of FG and it was conducted using electrochemical reduction. The synthesized GO or FGO (4 mg) was well dispersed in 2 ml of deionized water by sonication for 2 hours. Then 5 μ L of the solution was drop casted on a conductive electrode surface (e.g. SPE, carbon) and dried under ambient conditions to get GO or FGO coated SPE. GO or FGO was electrochemically reduced by CV scanning from 0.0 to − 1.5 V in N 2 purged 0.1 M pH 5.0 PBS (K 2 HPO 4 /KH 2 PO 4 ) for 20 cycles, and then rinsed with water and dried at room temperature. For direct sensing of ammonia gas, GO or FGO was thermally reduced by placing at 90 °C for 3 h in vacuum oven.
Electrochemical impedance studies (EIS) were conducted using a Biologic potentiostat, model SP-300. Low-cost, screen-printed electrodes (schematic of the device is shown in supporting information, [fig_ref] Figure 2: Electrostatic potential plots for [/fig_ref] were procured from Zensor (Taiwan). These consist of a 3 mm diameter working electrode and an arc-shaped auxiliary electrode (both made of graphitic carbon powder) and a Ag/AgCl pellet reference electrode all on a 50 × 13 mm plastic substrate. The EIS measurements were also conducted in the 3-electrode system consists of glassy carbon electrode as working electrode Ag/AgCl (Sat. KCl) as reference electrode and spiral platinum wire as counter electrode. EIS measurements were carried out at half-wave peak potential of the redox mixture in PBS solution (10 mM, pH 7.0) containing a mixture of 5 mM [Fe(CN) 6 ] 4− and 5 mM[Fe(CN) 6 ] 3 over the frequency range 10 5 -0.01 Hz with 5 mV as the alternating current amplitude. Using the redox probe (5 mM [Fe(CN) 6 ] 3−/4− ), change in charge transfer resistance (R ct ) at electrode/electrolyte interface has been investigated in electrochemicalimpedance.
FTIR analysis was conducted in transmittance mode on a Bruker (model: Alpha) spectrometer. The micro-Raman spectra of GO and FGO samples were studied using LabRamXploRA Raman spectroscope (excitement wavelength 632 nm).
Supporting Information. TEM images of FG, XPS and NMR data of FG, schematics of screen printed electrode and ammonia sensing set up are provided in the supporting file. Theoretically evaluated bond lengths observed in free graphene and FG is also provided in the supporting file. SEM image and Raman mapping of large area printed FG microelectrodes is provided. Direct ammonia sensing data is also provided. Ammonia sensing results (theoretical and experimental) with high fluorine containing graphene is also provided. This material is available free of charge via the Internet at http://pubs.acs.org.
[fig] Figure 1: Optimized graphene/FG and ammonia/ammonium structures: (A) graphene-ammonia (C 39 H 16 -NH 3 ), (B) FG -ammonia (C 39 H 16 F 2 -NH 3 ), (C) graphene-ammonium ( [/fig]
[fig] Figure 2: Electrostatic potential plots for (A) graphene-ammonia, (B) FG-ammonia, (C) grapheneammonium, and (D) FG-ammonium. [/fig]
[fig] Figure 4: (A) Nyquist plots of FG coated electrode on SPE sensor for varying ammonium ion (NH 4 + ) concentrations (i. blank, ii. 1 pM, iii. 10 pM, iv. 100 pM, v. 1 nM, vi. 10 nM, and vii. 0.1 μ M), (inset) the photograph of an FG coated SPE sensor, (B) normalized charge transfer resistance for various ammonium ion concentrations, (C) Nyquist plots showing increased impedance with increase in direct NH 3 exposing time, (D) R ct values with varying direct NH 3 exposures. [/fig]
[fig] Figure 3: (A) FT-IR spectra of pristine FG (theoretical (y-axis is absorbance) and experimental (FG 0S, y-axis transmittance)) and after exposure of ammonia gas (at a pressure ~1psi). (B) Raman spectra (632 nm excitation).Scientific RepoRts | 6:25221 | DOI: 10.1038/srep25221 4.4 × 10 −13 M (0.44 pM) limit of detection. The linear relationship could be characterized using the linear equation: Δ R ct (kΩ) = 3.832 + 0.293 log C NH3 (kΩ). This FG modified SPE exhibits a sensitivity of 0.293 kΩ M −1 with a correlation coefficient of 0.997 and standard deviation of 0.018 kΩ [/fig]
[fig] Figure 5: (A) Sensitivity of various electrodes towards ammonium ion ( [/fig]
[table] Table 2: HOMO-LUMO gaps for various complexes. [/table]
|
Daughters and Mothers Against Breast Cancer (DAMES): Main outcomes of a randomized controlled trial of weight loss in overweight mothers with breast cancer and their overweight daughters
BACKGROUND: Few studies to date have used the cancer diagnosis as a teachable moment to promote healthy behavior changes in survivors of cancer and their family members. Given the role of obesity in the primary and tertiary prevention of breast cancer, the authors explored the feasibility of a mother-daughter weight loss intervention. METHODS: A randomized controlled trial of a mailed weight loss intervention was undertaken among 68 mother-daughter dyads (n 5 136), each comprised of a survivor of breast cancer (AJCC stage 0-III) and her adult biological daughter. All women had body mass indices 25 kg/m 2 and underwent in-person assessments at baseline, 6 months, and 12 months, with accelerometry and exercise capacity performed on a subset of individuals. All women received a personalized workbook and 6 newsletters over a 1-year period that promoted weight loss; exercise; and a nutrientrich, low-energy density diet. A total of 25 dyads received individually tailored instruction (INDIVIDUAL), 25 dyads received teamtailored instruction (TEAM), and 18 dyads received standardized brochures (CONTROL). RESULTS: The trial met its accrual target, experienced 90% retention, and caused no serious adverse events. Significant differences in baseline to 12-month changes were observed between INDIVIDUAL versus CONTROL mothers for body mass index, weight, and waist circumference (WC); significant differences also were observed in the WC of corresponding daughters (P <.05). Significant differences were found between INDIVID-UAL versus CONTROL and TEAM versus CONTROL dyads for WC (P 5.0002 and .018, respectively), minutes per week of physical activity (P 5.031 and .036, respectively), and exercise capacity (P 5.047 for both). CONCLUSIONS: Significant improvements in lifestyle behaviors and health outcomes are possible with tailored print interventions directed toward survivors of cancer and their family members. For greater impact, more research is needed to expand this work beyond the mother-daughter dyad. Cancer , who helped with this project; all were compensated for their work. We thank Nike for donating shoe chips. We also are grateful to our study participants and the following oncology care providers: Drs.
# Introduction
The link between obesity and the risk of postmenopausal breast cancer is well known, and similar associations are acknowledged for cancers of the endometrium, kidney, esophagus, colorectum, and pancreas. [bib_ref] The role of obesity in cancer survival and recurrence, Demark-Wahnefried [/bib_ref] Moreover, as improvements in early detection and treatment are made against the backdrop of increasing rates of obesity, more survivors of cancer are now obese and at an increased risk of prevalent comorbidities such as cardiovascular disease, diabetes, and second malignancies (and possibly progressive cancer). [bib_ref] Nutrition and physical activity guidelines for cancer survivors, Rock [/bib_ref] Therefore, overweight and its attendant risks are fast replacing cachexia as the most prevalent nutritional problem among patients with cancer. [bib_ref] Nutrition and physical activity guidelines for cancer survivors, Rock [/bib_ref] Organizations such as the American Society of Clinical Oncology are currently developing toolkits that capitalize on the teachable moment of cancer to promote weight control. [bib_ref] Riding the crest of the teachable moment: promoting long-term health after the..., Demark-Wahnefried [/bib_ref] However, the impact of a cancer diagnosis is not confined to the patient alone. In defining the term "cancer survivor," the National Coalition for Cancer Survivorship includes family, friends, and caregivers,because the impact of cancer is far-reaching. Observational studies have suggested that the mother-daughter relationship may be particularly affected by a cancer diagnosis, especially breast cancer. 5, [bib_ref] Mothers and daughters: intertwining relationships and the lived experience of breast cancer, Wiggs [/bib_ref] Sinicrope et al 7 have suggested that the motherdaughter relationship can be leveraged specifically to deliver messages regarding cancer prevention and control.
Only a few health promotion interventions to date have capitalized on the mother-daughter bond to promote healthier behaviors, such as contraception, [bib_ref] Rural mothers' experiences and perceptions of their role in pregnancy prevention for..., Noone [/bib_ref] exercise, [bib_ref] Daughters and mothers exercising together (DAMET): a 12-week pilot project designed to..., Ransdell [/bib_ref] [bib_ref] Daughters and mothers exercising together: effects of home-and community-based programs, Ransdell [/bib_ref] osteoporosis prevention, [bib_ref] A mother-based intervention trial for osteoporosis prevention in children, Winzenberg [/bib_ref] and weight control. [bib_ref] Effects of an obesity prevention program on the eating behavior of African..., Stolley [/bib_ref] Overall, these interventions have been successful, although all have occurred among mother-daughter dyads in which the daughter was a child or adolescent. To our knowledge, to date there are no published reports of mother-daughter interventions among adult dyads, studies that navigate the complex relationship that has been characterized as "the closest and most profound psychological and emotional intergenerational bond," yet one that is acknowledged as difficult. [bib_ref] Mothers and daughters: intertwining relationships and the lived experience of breast cancer, Wiggs [/bib_ref] [bib_ref] Factors associated with breast cancer prevention communication between mothers and daughters, Sinicrope [/bib_ref] The DAMES (Daughters And MothErS Against Breast Cancer) trial endeavored to capitalize on the mother-daughter bond and the teachable moment created by a cancer diagnosis 3 to promote weight loss in overweight or obese women recently diagnosed with breast cancer and their overweight or obese daughters. If feasible and promising, the DAMES trial could offer an expedient way to promote both primary and tertiary prevention: tertiary prevention given that obesity is a poor prognostic indicator for cancer-related and overall survival for the patient diagnosed with breast cancer and primary prevention for her daughter who is at increased risk by virtue of family history and weight status. [bib_ref] The role of obesity in cancer survival and recurrence, Demark-Wahnefried [/bib_ref] Specific aims of this National Cancer Institute-sponsored, 3-armed, randomized controlled trial (RCT) were to explore the feasibility of a mother-daughter weight loss intervention and evaluate whether an individual approach in which mothers and daughters work in parallel to achieve diet and exercise goals or a team-based approach in which mother-daughter dyads work as a team to achieve these goals yielded greater reductions in body mass index (BMI) from baseline to 12-month follow-up. We hypothesized that a mother-daughter intervention would be feasible and that compared with an individual approach, the teambased approach would yield superior results. Support for the team-based approach is provided by literature regarding weight control and exercise in healthy populations in which having a "buddy" results in increased communication, reinforcement, and support, [bib_ref] Twelve month adherence of adults who joined a fitness program with a..., Wallace [/bib_ref] and in cancer populations in which dyadic-based interventions have been shown to promote joint problem-solving and reciprocal coping, leading to superior self-efficacy and quality of life, as well as higher rates of adherence and retention. [bib_ref] Quality of life of couples dealing with cancer: dyadic and individual adjustment..., Kim [/bib_ref] [bib_ref] Individual and dyadic relations between spiritual well-being and quality of life among..., Kim [/bib_ref] [bib_ref] Project Genesis: assessing the efficacy of problem-solving therapy for distressed adult cancer..., Nezu [/bib_ref]
# Materials and methods
## Overview
The DAMES trial was a 2-center, single-blind, parallel group RCT in which a total of 68 dyads (each comprised of an overweight or obese postmenopausal survivor of breast cancer and her overweight or obese adult daughter) underwent baseline assessment and were then randomly assigned to 1 of 3 study conditions: 1) a tailored diet and exercise intervention that was delivered in parallel and individually to mothers and daughters (25 dyads) (INDIVIDUAL); 2) a tailored diet and exercise intervention that emphasized the mother-daughter bond in a team-based approach (25 dyads) (TEAM); or 3) an attention control arm that received standardized diet and exercise materials (18 dyads) (CON-TROL). Each of the interventions delivered 7 installments (1 workbook followed by 6 newsletters) of mailed materials over a 1-year period. Dyads were reassessed at 6 months and 12 months of follow-up [fig_ref] Figure 1: CONSORT [/fig_ref] years of diagnosis with no evidence of progressive disease or second primary tumors and who had a biological daughter who was aged 21 years were eligible. Dyad daughters had to have no previous diagnoses of cancer, with the exception of nonmelanoma skin cancer. Both mothers and daughters had to meet the following criteria: 1) a BMI of 25 kg/m 2 to 39.9 kg/m 2 ; 2) no preexisting medical condition(s) that would preclude adherence to an unsupervised exercise program (eg, untreated stage 3 hypertension, severe orthopedic conditions or being scheduled for a hip or knee replacement, paralysis, endstage renal disease, dementia, unstable angina, history of a recent myocardial infarction, or congestive heart failure or pulmonary conditions requiring hospitalization or oxygen use within 6 months) or to a diet high in fruits and vegetables (ie, taking pharmacologic doses of warfarin); 3) ability to speak and write English and the completion of at least the sixth grade and thereby the ability to comprehend the intervention materials; 4) community dwelling in the United States, Puerto Rico, or Guam
## Accrual/enrollment
Dyad members were recruited through Web sites, flyers, community presentations, and listserves; however, the primary route for recruitment was through the cancer registries of Duke University and UT-MDACC. Letters of invitation with accompanying screening surveys (see criteria above) were mailed to stage-eligible and age-eligible cases, and a preaddressed, postage-paid envelope was provided for return. Women returning the completed screener received an incentive of 10 postage stamps. Permission and information needed to contact daughters was obtained from survivors of breast cancer who were deemed eligible, and a similar invitation and survey was then posted to daughters. Once an eligible dyad was identified, an in-person baseline assessment was scheduled at Duke University or UT-MDACC (if participants resided within a 60-mile radius and agreed to report) or through a home visit with a visiting nurse for those who were unable or unwilling to travel. Signed informed consent was obtained before baseline data were collected.
## Measures
The following objective and self-reported measures were taken at baseline and repeated at 6 months and 12 months of follow-up unless otherwise indicated.
## Clinical assessments
Participants' heights and weights were measured while they were wearing light clothing and no shoes.Weights were assessed on calibrated and zeroed scales. Heights (taken only at baseline) were performed at maximal upright stance on the inhale with the use of a Frankfort plane. BMI was calculated in kg/m 2 and served as an outcome of primary interest. Waist circumference was assessed at the level of the iliac crest at exhale using a nonstretch tape measure.Systolic and diastolic blood pressure were measured after participants had been seated for at least 5 minutes and on the arm ipsilateral to the affected breast. Participants with blood pressures above 179 mm Hg systolic or 109 mm Hg diastolic were placed "on hold" until written clearance was obtained by their physician to participate in the unsupervised exercise intervention.
A symptom-limited cardiopulmonary exercise test (CPET) to assess exercise capacity (VO 2peak ) was performed on a convenience sample of 47 participants who lived within a 60-mile radius of Duke University or UT-MDACC and reported to these institutions for their assessments. The CPET was performed on an electronically braked cycle ergometer with breath-by-breath gas analysis (MedGraphics; St. Paul, Minn) (TrueOne; Parvo Medics, Sandy, Utah) with continuous 12-lead electrocardiographic monitoring according to CPET guidelines for clinical populations. [bib_ref] Cardiorespiratory exercise testing in clinical oncology research: systematic review and practice recommendations, Jones [/bib_ref] Oxyhemoglobin saturation was monitored continuously using pulse oximetry, and blood pressure was measured noninvasively by manual auscultatory sphygmomanometry every 2 minutes. Data were recorded as the highest 30-second average within the last 2 minutes of the CPET.
## Self-reported measures
Two-part telephone interviews were administered at all 3 timepoints and included the following components: 1) 2 random 24-hour dietary recalls were conducted by trained interviewers using the interactive Nutrition Data System-Revised software (NCC Food and Nutrient Database System Version 2006, Minneapolis, Minn) and data regarding total energy intake; energy from solid fats, added sugars, and alcohol; servings of fruits and vegetables, legumes, total and whole grains, dairy, and meat; and intakes of sodium and saturated fat were used to derive a Healthy Eating Index score as defined by Guenther et al 21 and using methods described by Miller et al [bib_ref] Development and evaluation of a method for calculating the Healthy Eating Index-2005..., Miller [/bib_ref] ; 2) physical activity using the Leisure-Time Exercise Questionnaire of Godin et al 17 ; 3) health-related quality of life using the Short Form-36 instrument 23 ; 4) 2 items on self-efficacy: How confident or sure are you that you could walk or do another type of exercise for at least 30 minutes on 5 or more days of the week? How confident or sure are you that you could regularly limit the number of calories you eat or drink? (examples of portion control, substitution of low-energy for high-energy density foods, and behavioral strategies for consuming fewer calories were provided [anchors: very sure (1), sure (2), somewhat sure (3), unsure (4), and very sure (5)]); 5) social support related to healthful behavior change with respect to diet and exercise, using validated scales developed by Sallis et al 25 ; and 6) self-reported weight. Sociodemographic factors, such as age, race, marital and smoking status, educational attainment, income adequacy, and distance (in miles) between the residences of the mother and daughter were collected only at baseline. Response cards and food portion visuals were mailed to participants before the scheduled telephone call to enhance the quality of the data collected and to expedite the survey. At 6-month and 12month follow-up, all participants were asked whether they experienced any hospitalizations or changes in health; details for each event were gathered along with information to discern whether the event was or was not attributable to the intervention. Objective data on physical activity were captured for all participants accrued at the Duke University site via accelerometry (Actigraph, Fort Walton Beach, Fla). Programmed actigraphs were mailed to subjects along with instructions for the collection of 1 weeks' worth of data. Procedures analogous to those described by Sloane et al [bib_ref] Comparing the 7-day physical activity recall with a triaxial accelerometer for measuring..., Sloane [/bib_ref] were used to excerpt data that were then correlated with measures of self-report.
## Randomization
After complete baseline data were collected for both members of each dyad, dyads were randomly assigned by an off-site statistician to 1 of 3 arms (INDIVIDUAL, TEAM, or CONTROL) within 2 strata defined based on the race of the mother (white/nonwhite).
## Interventions
## Common elements
All groups received a welcome letter informing them that they would receive 6 additional informational mailings on a bimonthly basis over the next year and 6 brief surveys. All participants were informed that they would receive an incentive of $5 for each survey completed and returned within a 3-week window. This letter also was accompanied by a workbook that was personalized with the participant's name. Participants were encouraged to call a tollfree study telephone number if they had questions or problems or needed to report an adverse event.
In addition to equal contact, interventions also shared equal content with all materials written at a sixthgrade reading level. Intervention materials reinforced goals proposed by the American Cancer Society 27 and the US dietary guidelines.Materials promoted portion control and diets high in nutrients and low in energy as well as 150 minutes per week of aerobic exercise and twiceweekly strength training. However, interventions differed with respect to tailoring.
## Individual arm
Mothers and daughters assigned to this arm each received individually tailored print materials. For example, the initial workbook was not only personalized with the participant's name, but the initial pages also delineated individual weight goals and the kilocalorie levels required to achieve desired rates of weight loss using the Mifflin-St. Jeor equation (kcal/day 5 2161 1 10(wt) 1 6.25 (ht) 25(age)). [bib_ref] Comparison of predictive equations for resting metabolic rate in healthy nonobese and..., Frankenfield [/bib_ref] In addition, the 3 major foods contributing the highest percentage of kilocalories to each participant's diet were identified from the dietary recalls performed at baseline, and individuals were either directed to lower-calorie substitutes or provided with guidance on portion control. Introductory pages also included tailored feedback on how current intakes of saturated fat and fruits and vegetables as well as physical activity compared with the national guidelines. In keeping with social cognitive theory, which provided the primary behavioral framework for the interventions,participants were encouraged to keep records of their food intake and physical activity (self-monitoring), as well as to problem-solve on overcoming perceived barriers to healthy behaviors and to set incremental and achievable goals.As stated, participants were surveyed bimonthly on their progress and plans, as well as their perceived barriers and readiness to pursue lifestyle changes and responses were used to inform tailored messages. [bib_ref] Stages of change and decisional balance for 12 problem behaviors, Prochaska [/bib_ref] The 6 subsequent newsletters provided tailored messages regarding progress toward goals, along with appropriate reinforcement (if progress was charted) or encouragement (if progress was stalled) and feedback on portion control and overcoming barriers. Newsletter messages also were framed on readiness to pursue lifestyle changes, and thus elements of the transtheoretical model of behavior change were drawn on to engage participants with the level of information best able to motivate them. [bib_ref] Stages of change and decisional balance for 12 problem behaviors, Prochaska [/bib_ref] Mothers and daughters assigned to this arm also received supplies and equipment to assist them with selfmonitoring, such as logbooks and reference manuals (T-Factor 2000; WW Norton and Company, New York, NY) or Web sites (eg, mypyramid.org) to tally their intakes of kilocalories and saturated fat. They also received portion control tableware (Portion Doctor; Portion Health Products, St. Augustine Beach, Fla); iPods with prerecorded selections to set walking pace (Apple, Cupertino, Calif); and shoe chips (Nike Inc, Beaverton, Ore) to monitor steps taken, minutes of physical activity, and kilocalories burned. [fig_ref] Figure 2: Illustration of intervention materials used in the DAMES [/fig_ref] shows an illustration of the study materials.
## Team arm
Mothers and daughters assigned to the team-based intervention received information and supplies identical to those in the individual arm, but also received information on their other team member. Here, concepts of interdependence theory (ie, structuring goals to guide motherdaughter interactions to ultimately achieve outcomes) and the theory of communal coping (ie, cooperative problemsolving to deal with individual and common stressors) were drawn on to leverage the mother-daughter bond by encouraging effective communication between partners that would enhance their sense of confidence in planning, coordinating, and carrying forth strategies to increase mutual benefit. [bib_ref] Coping as a communal process, Lyons [/bib_ref] As an example, if a dyad member was charting success at meeting exercise goals, their next newsletter would provide positive reinforcement and also encourage them to share (in a helpful way) what had worked for them with their partner. Likewise, if a dyad member was experiencing a setback, they were provided with suggestions to get back on track and their partner was encouraged to provide them with helpful support.
## Control arm
Mothers and daughters assigned to this arm received a copy of the National Cancer Institute brochure Facing Forward (NIH Publication No. 10-2424) and the American Institute for Cancer Research publication Facts on Weight Management and Cancer, which were included in their binder personalized with their name. Subsequent brochures were mailed on a bimonthly basis and included American Institute for Cancer Research brochures (New American Plate, A Healthy Weight for Life, Getting Active-Staying Active, and Moving Toward a Plant-based Diet) and pamphlets from the American Heart Association (Managing Your Weight and Cholesterol, Blood Pressure and Weight Tracker) and the American College of Sports Medicine (Fit Over 40). These brochures were accompanied by a cover letter that encouraged the participant to read the brochure and then place it in their binder for easy reference. Bimonthly surveys assessed the perceived helpfulness of each brochure.
# Statistical analysis
DAMES was a feasibility trial and therefore an emphasis was placed on the accrual, delivery, and acceptance of the intervention, attrition, and adverse events. Preestablished benchmarks were the achievement of: 1) targeted accrual within a 9-month period; 2) "good" adherence as noted by the completion of at least 4 of 6 of the written surveys; 3) a retention rate of at least 80%; and 4) an absence of serious adverse events that were directly attributable to the intervention. General linear models were used to test for arm differences in adherence as measured by the number of completed surveys. Because other feasibility endpoints were dichotomous, their arm differences were tested using chi-square tests.
Because weight loss was the goal of each intervention, changes in BMI from baseline to 12 months were also of interest. The sample size for this feasibility study was derived using assumed differences between the control and each experimental arm of 0.48 (effect sizes in the "medium-to-large range" as defined by Cohen 33 ) using a 2-sided alpha of .05 and a power of 0.80. The effect of the interventions by arm on other measures of adiposity (eg, body weight and waist circumference), health-related outcomes (eg, blood pressure and cardiorespiratory fitness), and quality of life were explored in a similar fashion. To estimate the arm effects on change in outcomes from baseline to 12 months, repeated-measures models were used to account for the covariance between the mothers and daughters. In addition, general linear models were used to estimate these arm effects within mothers and daughters separately. Although testing largely compared both experimental arms with the control group, for BMI (a key outcome) the 2 experimental groups also were compared with one another. Because this was a feasibility study, a 2sided alpha level of .05 was used for all tests without attempting to control for multiplicities. All analyses were performed using SAS statistical software (version 9.3; SAS Institute Inc, Cary, NC).
# Results
## Feasibility endpoints
## Accrual
The DAMES trial achieved its accrual target and also met racial/ethnic benchmarks within the established time frame. However, the 3% accrual rate was much lower than the 10% rate that was initially forecast and therefore the study required a patient pool that was triple the original estimate. Approximately 38% of respondents completed the screening questionnaire, and among responders the leading reasons for nonparticipation in the trial were the absence of a biological daughter and overall disinterest (a frequent written-in comment suggested that respondents would be interested in participating if they could partner with someone other than their daughter). Women diagnosed with breast cancer (mothers) who either refused, failed to respond, or were ineligible to participate in the trial did not differ from those who were enrolled in terms of race, age, or stage of disease (all P > .05). Characteristics of the overall study sample are reported in [fig_ref] TABLE 1: Study Sample Characteristics a [/fig_ref] and did not differ by arm assignment with respect to tumor features, demographics, study site, smoking status, or BMI. The study sample was diverse in terms of race/ethnicity and socioeconomic status. The majority of the mothers had stage I or II breast cancer, and most women had a BMI of 30 kg/m 2 (obese).
Adherence and retention [fig_ref] TABLE 2: Adherence [/fig_ref] shows data regarding adherence as ascertained by completion of the intermittent surveys. The benchmark of completion of at least 4 of the 6 surveys (67%) was only achieved in the control group, which had significantly higher completion rates compared with either of the intervention arms (P 5 .0019). Response rates were also found to be significantly higher among mothers compared with daughters across all 3 study arms (P 5 .0342). Retention rates varied from 84% to 100% across arms (90% overall); no significant differences were observed between arms or for any sociodemographic or disease-related characteristics.
## Adverse events
A total of 95 health-related events were reported among the 136 women over the 1-year study period, 9 of which were serious. No differences were noted between study arms and only 2 events (1 report in the CONTROL arm and other in the INDIVIDUAL arm) were attributable to the intervention (nonserious knee and/or hip soreness with exercise). [fig_ref] TABLE 3: Mean Values [/fig_ref] documents precision estimates for various study outcomes by arm across time. Significant reductions in BMI were observed in mothers assigned to the INDIVID-UAL arm compared with those in the CONTROL arm, although such differences were not observed among daughters and only a trend was seen in dyads overall. Similar findings were observed in the INDIVIDUAL arm for weight loss as measured as a continuous variable. It is interesting to note that 39.1% of mothers as well as 39.1% of daughters in the INDIVIDUAL arm lost a clinically significant amount of body weight ( 5% of their baseline weight), whereas the respective rates among mothers and daughters in the TEAM arm and CON-TROL arm were 21.7% and 33.3% and 27.8% and 35.3%, respectively; no significant between-arm differences were observed. Although analyses were performed on actual weights, a significant correlation (q 5 0.940; P < .0001) was observed between actual and self-reported weight (data not shown). No significant differences were Original Article observed between the INDIVIDUAL and TEAM arms for BMI. Waist circumference proved to be a more sensitive measure of adiposity and dyads assigned to both the TEAM and INDIVIDUAL arms experienced significant reductions compared with those assigned to the CON-TROL arm. Significant differences also were observed among mothers and daughters in the CONTROL versus INDIVIDUAL arms. Likewise, VO 2peak change scores were found to be significantly improved in both of the Compared with controls, dyads in both the TEAM and INDIVIDUAL arms experienced significantly greater increases in self-reported minutes of moderate to vigorous physical activity (supported by trends of total metabolic equivalent task [MET] hours per week of physical activity captured via self-report and accelerometry, both of which were found to be significantly correlated [Spearman q, 0.300; P 5 .009]). No between-arm differences were noted with regard to energy intake or diet quality. Similarly, no between-arm differences were observed with regard to potential mediators of the intervention (eg, social support or self-efficacy). Significant correlations were found between adherence (as measured by survey completion) and change in BMI in both the CONTROL and INDIVIDUAL arms, but not for the TEAM arm; Pearson correlation coefficients were q of 0.417 (P 5 .015), q of 0.294 (P 5 .048), and q of 20.038 (P 5 .802), respectively.
## Effects
# Discussion
To the best of our knowledge, the DAMES trial is the first lifestyle intervention that attempted to capitalize on the teachable moment of a cancer diagnosis together with leveraging the mother-daughter bond to promote intergenerational lifestyle change in mothers with breast cancer and their biological daughters. Similar to previous studies conducted in healthy mother-daughter dyads that were aimed at healthful dietary changes and increased exercise, the overall retention rate for the DAMES trial was 90% and therefore was comparable to the 78% to 100% range established by these trials, [bib_ref] Daughters and mothers exercising together: effects of home-and community-based programs, Ransdell [/bib_ref] [bib_ref] Effects of an obesity prevention program on the eating behavior of African..., Stolley [/bib_ref] while also reporting no serious adverse events attributable to the intervention.
Although the DAMES trial was undertaken to assess feasibility and was not fully powered to test for differences in outcomes, several statistically significant improvements in physical activity and fitness and reductions in adiposity (as measured by BMI, body weight, and waist circumference) were observed with both minimal interventions of personally and iteratively tailored print materials. Moreover, the magnitude of effects also appears clinically significant for many of these outcomes. For example, increases in the Healthy Eating Index of 2 points as noted in the mothers and daughters on the INDIVIDUAL arm are considered slightly above a "small effect," whereas improvements of roughly 5 points, as noted among TEAM mothers and daughters, are considered significant. [bib_ref] A new multivariate measurement error model with zero-inflated dietary data, and its..., Zhang [/bib_ref] The doubling of the number of minutes of physical activity within the intervention arms and the finding that these increases equate to an annual weight loss of 0.97 kg to 1.59 kg also bodes for clinical significance. However, the large variation in physical activity, particularly among daughters on the TEAM arm, is cause for caution and may be influenced somewhat by 3 participants who became marathon runners over the course of the study year (these women were obvious outliers, but because their data were valid we included them in our analysis). Finally, the weight loss literature indicates that reductions of 5% body weight are found to improve several health parameters, such as serum glucose, lipids, and blood pressure. [bib_ref] Benefits of modest weight loss in improving cardiovascular risk factors in overweight..., Wing [/bib_ref] Thus, the finding that these minimal interventions resulted in a weight loss of 5% in a substantial percentage of participants (ie, 21.7%-39.1%) is also clinically significant. Perhaps the reason that we were unable to detect differences between the INDIVIDUAL and TEAM tailored interventions was due to a lack of power, especially among mothers, in whom the percentage achieving 5% weight loss was most discrepant and did not align with our original hypothesis (ie, that the teambased approach would produce effect sizes of the greatest magnitude).
The benefits of family-based approaches for weight loss may depend partly on the nature of the relationship between family members and the age/developmental stage of the target. Supporting this idea, a review conducted by McLean et al [bib_ref] Family involvement in weight control, weight maintenance and weight-loss interventions: a systematic..., Mclean [/bib_ref] found that programs that treated overweight couples together resulted in greater weight loss for both partners, but that programs that treated overweight children and their overweight parents only yielded positive results for the children (to the best of our knowledge, however, to date this has only been explored in dependent children). Given the complex nature of family relationships, more research is needed to understand how best to involve family members in diet and exercise interventions. Indeed, family involvement can either entail enlisting a family member as a supporter of the survivor's behavior change or actively involving and treating the family member along with the survivor. [bib_ref] Family involvement in weight control, weight maintenance and weight-loss interventions: a systematic..., Mclean [/bib_ref] [bib_ref] The treatment and prevention of obesity: a systematic review of the literature, Glenny [/bib_ref] The latter approach is consistent with family systemsand interpersonal theories [bib_ref] Understanding health behavior change among couples: an interdependence and communal coping approach, Lewis [/bib_ref] and emphasizes relationship factors such as closeness, communication, and the quality of the relationship in the behavior change process. [bib_ref] Understanding health behavior change among couples: an interdependence and communal coping approach, Lewis [/bib_ref] Although this main outcomes analysis did not examine relationship characteristics as potential moderators of effect, studies suggest that we may also need to carefully consider family member characteristics when deciding whether a communal or social support approach is more appropriate. For example, Brownell et al [bib_ref] Treatment of obese children with and without their mothers: changes in weight..., Brownell [/bib_ref] found that obese adolescents who were treated alone lost significantly more weight than obese adolescents who attended a weight loss program with their mothers (who were not obese). However, in an intervention in which overweight adults were encouraged to invite up to 3 partners of their choosing to join them in a weight loss program, Gorin et al 41 found that those with at least 1 successful partner lost significantly more weight at 6 months, 12 months, and 18 months. Together, these studies suggest that more work is needed to determine whether survivors of cancer and their family members can benefit from communal approaches. Moreover, data from the current study also suggest that the impact of communal approaches may vary based on outcome. For example, compared with the INDIVIDUAL-based approach, the TEAM-based approach appeared to generate nearly double the increase in diet quality, but exhibited fewer improvements in weight loss. Although the cause of this variable effect is unknown, one potential explanation could be that the TEAM-based intervention tied into mother-daughter communications that have been traditionally practiced, such as recipe sharing and food preparation.Weaker than expected effects within the TEAM-based approach for other outcomes may have been due to the lack of intensive skill training in areas of active listening, requesting assistance, and providing optimal support. [bib_ref] A systematic review and meta-analysis of psychosocial interventions for couples coping with..., Badr [/bib_ref] Perhaps the more important lesson learned from the current study lies not in the comparative effect of the interventions but rather in the study population (ie, mothers and daughters). As indicated in the literature, the mother-daughter relationship is complex, and although it can be positive, it also is subject to strain.The added pressure of a breast cancer diagnosis and issues surrounding weight control can further stress the relationship, 6 making participation in a mother-daughter weight loss intervention perhaps less productive than if survivors of cancer were provided free choice in selecting a teammate for the partner-based intervention, as in the aforementioned study by Gorin et al. [bib_ref] Involving support partners in obesity treatment, Gorin [/bib_ref] Through implementation of the current study, we found that focusing recruitment solely on mother-daughter dyads served as a considerable barrier because many survivors of breast cancer did not have biological daughters or did not want to participate in a weight loss intervention with them. Likewise, similar barriers were observed among daughters. Our team has conducted several mailed print interventions and the response rate to the DAMES trial was considerably lower; our response rate was 3%, compared with the response rate of 42% experienced in FRESH START, which recruited 543 survivors of breast and prostate cancer over a similar time period. [bib_ref] Main outcomes of the FRESH START trial: a sequentially tailored, diet and..., Demark-Wahnefried [/bib_ref] In addition, the absence of relative improvements in perceived social support or satisfaction with the mother-daughter relationship, especially within the TEAM arm, after working toward a common goal over the 1-year intervention period are of interest, although it could be posited that high baseline levels may have reduced the ability to detect change. Nevertheless, the observation that adherence was significantly greater in the CONTROL arm compared with the INTERVEN-TION arm, and that adherence was only related to BMI in the CONTROL and INDIVIDUAL arms but not among members of the TEAM arm, suggests that other factors may be responsible (eg, the intervention spurring membership in weight loss and fitness programs).
The lack of changes in social support or self-efficacy as well as the lack of an association between these measures and changes in objective outcomes are disconnects, particularly because the intervention was grounded in social cognitive theory,interdependence theory,and the theory of communal coping. [bib_ref] Coping as a communal process, Lyons [/bib_ref] However, similar disconnects have been found in fully powered trials either for global behavior change 46 or for behavior change within specific domains. [bib_ref] Change in self-efficacy partially mediates the effects of the FRESH START intervention..., Mosher [/bib_ref] Therefore, the finding that we observed changes in BMI but did not observe changes in these intermediate constructs is hardly a rare phenomenon and also could be due to ceiling effects or a lack of statistical power. Indeed, the lack of statistical power was a primary limitation of the current study, but one that was balanced by numerous strengths, including a strong RCT design; objective measures; excellent retention; and, even if enrollment was low, the sample accrued was diverse and not biased in terms of race, age, or stage of disease.
Therefore, the data from the current study, plus the resulting precision estimates that demonstrate the benefits of a minimal intervention comprised of 1 workbook and a series of 6 iteratively tailored newsletters, are compelling and call for future interventions that are directed toward survivors of cancer and selected partners. Given the evidence of a greater impact with partner-based interventions that emanate from other studies, there is a need to test a similar intervention in other patient-partner dyads; to test the added value of intensive skill training in active listening and other supportive techniques; and to experiment with other means of dissemination, such as Web-based platforms to achieve broader reaching impact.
# Funding support
Funding for this research was provided by the following National Institutes of Health grants: R21 CA122143, P30 CA16672, and
[fig] Figure 1: CONSORT (Consolidated Standards Of Reporting Trials) diagram for the current study is shown. BMI indicates body mass index. [/fig]
[fig] Figure 2: Illustration of intervention materials used in the DAMES (Daughters and MothErS Against Breast Cancer) trial is shown. [/fig]
[table] TABLE 1: Study Sample Characteristics a [/table]
[table] TABLE 2: Adherence (Completion Rates) to Periodic Surveys Used to Inform Newsletters [/table]
[table] TABLE 3: Mean Values (SD) for Study Outcomes Across Study Arms at Baseline, 6 Months, and 12 Months [/table]
|
Clot or Not?
# Introduction
Cardiac masses are rare, with a prevalence of 0.15% in echocardiographic series. [bib_ref] MR imaging of cardiac tumors and masses: a review of methods and..., Motwani [/bib_ref] The differential diagnosis for cardiac masses includes primary cardiac tumors, cardiac metastases, vegetations, thrombi, fluid-filled lesions, and artifacts. [bib_ref] MR imaging of cardiac tumors and masses: a review of methods and..., Motwani [/bib_ref] These entities have widely different implications for management and prognosis, highlighting the importance of accurately identifying the etiology of a cardiac mass. Echocardiography is a key tool for the identification and diagnosis of cardiac masses, relying on the appearance and location of the mass. Multimodality imaging permits the integration of echocardiography and other noninvasive imaging modalities to improve the characterization and differentiation of cardiac masses. This integrated approach allows tissue characterization without the use of a cardiac biopsy and its associated risks.
We report a case of multiple right ventricular (RV) cardiac masses characterized using an integrated imaging approach.
## Case presentation
An 85-year-old cachectic man, an ex-smoker, presented to the emergency department with a 3-month history of progressive exertional dyspnea. His medical history was notable for laryngeal cancer resected in 2010 and recent diagnosis of invasive urothelial carcinoma without prostatic invasion, status post resection with curative intent 3 months before presentation.
His blood pressure on arrival was 121/64 mm Hg, with a heart rate of 77 beats/min and mild hypoxemia. Laboratory values revealed stable high-sensitivity troponin T of 20 ng/L (upper limit of normal, 15 ng/L) and prostate-specific antigen of 1.2 mg/L (upper limit of normal, 4.0 mg/L).
On point-of-care ultrasound, the emergency physician identified RV masses, raising the differential diagnosis of tumors versus thrombi. Computed tomographic angiography described extensive bilateral lobar and segmental arterial filling defects compatible with pulmonary emboli, accompanied by masslike filling defects within the RV cavity and features of right heart strain. Low-dose thrombolysis was considered but withheld because of the nonthrombotic echocardiographic appearance of the masses and the patient's hemodynamic stability.
Formal transthoracic echocardiography revealed moderate RV enlargement with global hypokinesis and multiple large masses in all segments of the right ventricle (inflow, body, and outflow). Masses were adherent to the lateral free wall, the septal aspect of the tricuspid annulus, the papillary muscles, and the RV apex , Videos 1 and 2).
The masses were multilobulated and had similar echogenicity to the myocardium without independent mobility. No right atrial or inferior vena cava masses were observed [fig_ref] Figure 3: Plethoric inferior vena cava [/fig_ref].
There was secondary mild tricuspid valve inflow obstruction (mean transtricuspid gradient, 2 mm Hg at 80 beats/min) [fig_ref] Figure 4: Mild RV inflow obstruction [/fig_ref] and moderate tricuspid regurgitation (Video 3), with an estimated pulmonary artery systolic pressure of 59 mm Hg. There was a small circumferential pericardial effusion without echocardiographic features of tamponade , Video 1).
The constellation of these echocardiographic findings suggested tumor rather than multiple thrombi; however, superimposed thrombi could not be excluded.
To further characterize the masses, cardiac magnetic resonance imaging (MRI) was performed. Cardiac MRI confirmed the location and burden of the RV masses , Videos 4 and 5).
Tissue characteristics were consistent with tumor without superimposed thrombi: T1 isointense [fig_ref] Figure 7: A [/fig_ref] , T2 hyperintense [fig_ref] Figure 7: A [/fig_ref] , first-pass perfusion positive [fig_ref] Figure 7: A [/fig_ref] , and late gadolinium enhancement (LGE) heterogeneously positive.
The patient died of his disease within 1 month of diagnosis of cardiac involvement.
# Discussion
Echocardiography is the first-line imaging modality for cardiac masses because it is a portable, widely available, noncontrast, and dynamic technology. It can also be used by nonechocardiographers to enhance their clinical assessments. [bib_ref] Focused cardiac ultrasound: recommendations from the American Society of Echocardiography, Spencer [/bib_ref] In this case, in view of the point-of-care ultrasound and computed tomographic results, a presumptive diagnosis of multiple pulmonary emboli was made, and the emergency department physician considered low-dose thrombolysis 3 on the basis of the ''clot'' burden and RV strain. Appropriately, formal echocardiography was performed given the uncertainty of the etiology of the masses. The imaging findings on echocardiography, particularly the heterogeneity of the RV masses, their distribution, and motion synchronous with myocardium, suspected invasion of the RV lateral wall, as well as the thickened pericardium, suggested tumor rather than thrombi. Therefore, thrombolysis was deferred on the basis of the echocardiographic findings.
Considering our patient's history of recent urothelial carcinoma, cardiac metastases were strongly suspected. We proceeded to Apical four-chamber view demonstrating a heterogeneous mass adjacent to the lateral RV free wall, possibly invading the myocardium (star). There were separate heterogeneous, multilobed masses on the papillary muscles (asterisk). Moderate RV dilation was noted. There was a small pericardial effusion.
## Figure 2
Off-axis view demonstrating the masses on the papillary muscles in the RV body (asterisk) and a separate mass originating in the RV apex extending into the right ventricular outflow tract (solid circle). The masses were not independently mobile. RV inflow and outflow steady-state free precession MRI sequence demonstrating the RV masses isointense to the myocardium originating from the base of the papillary muscles (asterisk), right ventricular outflow tract (solid circle), RV lateral wall (star), and tricuspid valve annulus (diamond).
## Figure 6
Four-chamber steady-state free precession MRI sequence demonstrating the RV masses isointense to the myocardium originating from the base of the papillary muscles (asterisk) and RV lateral wall (star) and tricuspid valve annulus (diamond).
perform cardiac MRI to further characterize the masses, with the goal of establishing a noninvasive tissue diagnosis. MRI permits differentiation between thrombus and tumor. The appearance of thrombi depends on the acuity of the thrombi. Acute thrombi appear hyperintense on T1 and T2 sequences because of the presence of oxygenated hemoglobin. Subacute thrombi appear hyperintense on T1 imaging and hypointense on T2 imaging because of the paramagnetic effects of methemoglobin and the loss of water content. Over time, the methemoglobin is replaced by fibrous tissue, and the thrombus continues to lose water content, leading to the hypointense appearance of chronic thrombi on T1-and T2weighted imaging. In contrast, most tumors are isointense on T1 and hyperintense on T2 sequences. Being avascular, acute and chronic thrombi do not perfuse on first-pass perfusion imaging. Conversely, tumors, being vascular structures, will perfuse on first-pass perfusion imaging. Tumors may heterogeneously enhance on late gadolinium sequences, but thrombi do not usually enhance with LGE. If a fibrous rim develops around the chronic thrombus, it can peripherally enhance on LGE. [bib_ref] Value of CMR for the differential diagnosis of cardiac masses, Pazos-L Opez [/bib_ref] Therefore, the tissue characteristics of our patient (T1 isointense, T2 hyperintense, first-pass perfusion positive, LGE heterogeneously positive) were consistent with tumor.
Cardiac computed tomography (CT) and cardiac positron emission tomography are two other imaging modalities that can also be considered to characterize cardiac masses. Cardiac CT is recognized for its high spatial resolution. An advantage of CT is the ability to identify areas of calcification. Calcified tissue is not susceptible to radiofrequency pulses and will therefore remain hypointense on all sequences on MRI. Conversely, areas of calcification are easily identified on CT. [bib_ref] Cardiac masses on cardiac CT: a review, Kassop [/bib_ref] The diagnostic advantage of positron emission tomography is twofold; first, it can distinguish metabolically active tissues from non-metabolically active tissue (most benign masses), and second, it can identify recurrence of the primary tumor and distant metastatic foci. [bib_ref] Utility of FDG PETCT in assessment of cardiac masses, Shrikanthan [/bib_ref] In our case, positron emission tomography may have offered additional insight by confirming the masses were metabolically active and identifying recurrence at the primary site or other metastatic lesions. [bib_ref] MR imaging of cardiac tumors and masses: a review of methods and..., Motwani [/bib_ref] Cardiac involvement in urothelial carcinoma is rare. The prevalence in autopsy studies is 3.9%.Only a small number of cases have been identified before autopsy, with a total of 15 cases having been reported. Cardiac metastases from urothelial carcinoma most frequently involve the right ventricle (eight of 15 cases), but they have also been reported in the left ventricle, pericardium, and right atrium. [bib_ref] Metastatic urothelial carcinoma to pericardia manifested by dyspnea from cardiac tamponade during..., Hattori [/bib_ref] [bib_ref] Metastatic cardiac tumor from urothelial carcinoma detected by transthoracic echocardiography: a case..., Nakashima [/bib_ref] Identification of cardiac metastases in urothelial carcinoma has significant prognostic implications. Hattori et al. [bib_ref] Metastatic urothelial carcinoma to pericardia manifested by dyspnea from cardiac tamponade during..., Hattori [/bib_ref] described a poor prognosis associated with cardiac urothelial metastasis, with most patients dying shortly after their diagnosis or their first clinical encounter.
In the context of the patient's clinical history of urothelial carcinoma, the masses likely represent cardiac urothelial metastasis. Biopsy was considered to confirm the diagnosis but deferred given the age and poor prognosis of the patient, as well as a persistent thrombocytopenia.
# Conclusion
Multimodality imaging is useful in differentiating thrombi from cardiac tumors. Differentiating intracardiac thrombus from tumor in patients with an active malignancy has significant implications for management and prognosis. In this case, echocardiography significantly altered the early management of this patient in deferring thrombolysis and suggesting the etiology of the masses as tumors. Cardiac MRI allowed the confirmation of the RV masses as cardiac tumors, likely urothelial metastases, conferring a poor prognosis.
## Supplementary data
Supplementary data related to this article can be found at https://doi. org/10.1016/j.case.2017.09.010.
[fig] Figure 3: Plethoric inferior vena cava (25 mm) without luminal masses. [/fig]
[fig] Figure 4: Mild RV inflow obstruction (mean transtricuspid gradient, 2 mm Hg at 80 beats/min). [/fig]
[fig] Figure 7: A) T1-weighted sequence demonstrating the masses are isointense to the myocardium. (B) T2-weighted sequence demonstrating heterogeneous hyperintensity of the RV masses. (C) Positive first-pass perfusion following administration of gadodiamide (0.1 mmol/kg). [/fig]
|
Q Fever Endocarditis in Romania: The First Cases Confirmed by Direct Sequencing
Infective endocarditis (IE) is a serious, life-threatening disease with highly variable clinical signs, making its diagnostic a real challenge. A diagnosis is readily made if blood cultures are positive, but in 2.5 to 31% of all infective endocarditis cases, routine blood cultures are negative. In such situations, alternative diagnostic approaches are necessary. Coxiella burnetii and Bartonella spp. are the etiological agents of blood culture-negative endocarditis (BCNE) most frequently identified by serology. The purpose of this study is to investigate the usefulness of molecular assays, as complementary methods to the conventional serologic methods for the rapid confirmatory diagnostic of Q fever endocarditis in patients with BCNE. Currently, detection of C. burnetii by culture or an antiphase I IgG antibody titers >800 represents a major Duke criterion for defining OPEN ACCESS the diagnosis of endocarditis due to Bartonella spp. We used indirect immunofluorescence assays for the detection of IgG titers for C. burnetii, B. henselae and B. quintana in 57 serum samples from patients with clinical suspicion of IE. Thirty three samples originated from BCNE patients, whereas 24 were tested before obtaining the blood cultures results, which finally were positive. The results of serologic testing showed that nine out of 33 BCNE cases exhibited antiphase I C. burnetii IgG antibody titer >800, whereas none has IgG for B. henselae or B. quintana. Subsequently, we used nested-PCR assay for the amplification of C. burnetii DNA in the nine positive serum samples, and we obtained positive PCR results for all analyzed cases. Afterwards we used the DNA sequencing of amplicons for the repetitive element associated to htpAB gene to confirm the results of nested-PCR. The results of sequencing allowed us to confirm that C. burnetii is the causative microorganism responsible for BCNE. In conclusion, the nested PCR amplification followed by direct sequencing is a reliable and accurate method when applied to serum samples, and it may be used as an additional test to the serological methods for the confirmatory diagnosis of BCNE cases determined by C. burnetii.
IE, while a titers of >800 for IgG antibodies to either B. henselae or B. quintana is used for
# Introduction
Infective endocarditis is a serious, life-threatening disease with highly variable clinical signs that are making the condition a diagnostic challenge. A diagnosis is readily made if blood cultures are positive, but in 2.5 to 31% of all infective endocarditis cases, routine blood cultures are negative [bib_ref] Simplified serological diagnosis of endocarditis due to Coxiella burnetii and Bartonella, Rolain [/bib_ref] [bib_ref] Endocarditis due to rare and fastidious bacteria, Brouqui [/bib_ref] [bib_ref] Diagnostic methods. Current best practices and guidelines for identification of difficult-to-culture pathogens..., Houpikian [/bib_ref]. This variation in incidence could be explained by several factors, including: (i) differences in the diagnostic criteria used; (ii) specific epidemiological factors, as for fastidious zoonotic agents; (iii) variations in the early use of antibiotics before the blood sampling; (iv) differences in sampling strategies; or (v) involvement of unknown pathogens [bib_ref] Comprehensive diagnostic strategy for blood culture-negative endocarditis, a prospective study of 819..., Fournier [/bib_ref] [bib_ref] Contribution of systematic serological testing in diagnosis of infective endocarditis, Raoult [/bib_ref].
Blood culture negative endocarditis (BCNE) was recognized by Osler at the beginning of last century [bib_ref] Chronic infectious endocarditis, Osler [/bib_ref]. Recently, many publications in European countries have demonstrated a significant involvement of Coxiella burnetti, Bartonella henselae, and B. quintana in patients with BCNE [bib_ref] Infective endocarditis due to Bartonella spp. and Coxiella burnetii experience at a..., Siciliano [/bib_ref] [bib_ref] Blood culture-negative endocarditis in a reference center, etiologic diagnosis of 348 cases, Houpikian [/bib_ref]. C. burnetii is one of the most encountered fastidious agents in BCNE. Q fever is characterized by its clinical polymorphism and the presentation of the disease is variable, with both acute and chronic manifestations [bib_ref] Coxiella burnetii infection of aortic aneurysms or vascular grafts, report of 30..., Botelho-Nevers [/bib_ref] [bib_ref] Cardiac valves in patients with Q fever endocarditis, microbiological, molecular, and histologic..., Lepidi [/bib_ref]. Following acute infection, 1 to 5% of patients progress to chronic infection, which can develop after months to several years after acute Q fever infection, the longest interval being 20 years after infection [bib_ref] From acute Q fever to endocarditis, serological follow-up strategy, Landais [/bib_ref] [bib_ref] The value of follow-up after acute Q fever infection, Healy [/bib_ref].
Endocarditis is the main form of chronic Q fever (78% of all chronic Q fever cases) [bib_ref] Natural history and pathophysiology of Q fever, Raoult [/bib_ref]. The most exposed persons are patients with preexistent valvular disease or vascular defects, especially aortic aneurysm and aortic stents and prostheses, immunocompromised patients, and pregnant women [bib_ref] Chronic Q fever-related dual-pathogen endocarditis, case series of three patients, Kampschreur [/bib_ref] [bib_ref] Chronic Q fever. Ninety-two cases from France, including 27 cases without endocarditis, Brouqui [/bib_ref] [bib_ref] Risks factors and prevention of Q fever endocarditis, Fenollar [/bib_ref]. The estimated risk of transformation from acute infection to Q fever endocarditis in patients with preexisting valvulopathy is approximately 40% [bib_ref] Risks factors and prevention of Q fever endocarditis, Fenollar [/bib_ref]. Because symptoms of Q fever endocarditis are protean and not specific, diagnosis is often delayed, only after significant valvular damage has occurred, resulting in an increasing mortality rate. Some authors proposed that all patients with acute Q fever be investigated by a transthoracic echocardiography [bib_ref] Endocarditis after acute Q fever in patients with previously undiagnosed valvulopathies, Fenollar [/bib_ref].
The diagnosis of Q fever endocarditis requires both clinical endocarditis and isolation or serologic evidence of C burnetii. Because Q fever endocarditis is a chronic illness, a single serum specimen is sufficient for diagnosis. A phase I IgG titers of 800 or greater is one of the major modified Duke criteria [bib_ref] Q fever, epidemiology, diagnosis, and treatment, Hartzell [/bib_ref]. Previously studies showed that PCR with serum samples may be helpful in establishing an early diagnosis of chronic Q fever [bib_ref] Molecular genetic methods for the diagnosis of fastidious microorganisms, Fenollar [/bib_ref].
Currently there is no data concerning the incidence of Q fever endocarditis cases among Romanian population. The first Q fever cases in Romania were registered in 1947 in Constanta County [bib_ref] Q fever epidemiology in Roumania, Cracea [/bib_ref]. The most recent data about Q fever in Romania were represented by urban sporadic cases reported in the period 1981-1987 [bib_ref] Dogaru, D. Q fever urban cases in Romania, Crăcea [/bib_ref].
# Results and discussion
## Case definitions
Patients were considered to have definite blood culture-negative endocarditis if the results of standard blood cultures were negative, and if clinical and echocardiographic findings met the Duke criteria for infective endocarditis. All serum samples tested by serological methods for detection of IgG to C. burnetii, B. quintana and B. henselae originated from patients with clinical suspicion of BCNE. According to the modified Duke criteria a single positive blood culture for Coxiella burnetii or antiphase I IgG antibody titers >800 represents a major criterion for definite infective endocarditis [bib_ref] Proposed modifications to the duke criteria for the diagnosis of infective endocarditis, Li [/bib_ref]. The results of serological testing showed that 9 out of 33 serum samples exhibited antiphase I C. burnetii IgG antibody titers >800, while none of samples has IgG for B. henselae or B. quintana. The modified Duke criteria used for definite infective endocarditis (IE) diagnosis of analyzed patients with clinical suspicions of BCNE are presented in [fig_ref] Table 1: The modified Duke criteria used for defining infective endocarditis [/fig_ref]. Diagnosis of IE is definite if there are 2 major criteria or 1 major and 3 minor criteria or 5 minor criteria [bib_ref] Guidelines on the prevention, diagnosis, and treatment of infective endocarditis (new version..., Habib [/bib_ref] [bib_ref] Clinical presentation, etiology, and outcome of infective endocarditis in the 21st century..., Murdoch [/bib_ref]. Eight out of nine investigated cases fulfilled 2 major criteria (antiphase I C. burnetii IgG antibody titer >800 and when there is a vegetation or a new valvular regurgitation) for defining IE, while the remaining case fulfilled one major criterion (antiphase I C. burnetii IgG antibody titers >800) and one minor criterion (fever 38 °C), being classified as a possible case.
It is well-known that endocarditis is the most common presentation of chronic Q fever. Initially thought to be a rare disorder, later it has been estimated to account for up to 5% of all endocarditis cases worldwide [bib_ref] Molecular detection of Coxiella burnetii in the sera of patients with Q..., Fenollar [/bib_ref]. It occurs almost exclusively in patients who have pre-existing valvular disease or who are immunocompromised. Unlike typical cases of endocarditis, the clinical presentation of endocarditis from chronic Q fever is often nonspecific and lacks many of the typical features of subacute, bacterial endocarditis, such as usual clinical and echocardiographic features common to typical cases of endocarditis [bib_ref] Q fever endocarditis, A case report and review of the literature, Deyell [/bib_ref]. Thus, the diagnosis is often significantly delayed or even missed, resulting in significant morbidity and mortality. Despite increasing awareness, recent studies show a mean delay of seven months from symptom onset to diagnosis [bib_ref] Changing clinical presentation of Q fever endocarditis, Houpikian [/bib_ref] [bib_ref] Q fever endocarditis, A surgical view and a word of caution, Mesana [/bib_ref]. Without prompt recognition and adequate antimicrobial therapy, the course of Q fever endocarditis is severe and potentially fatal [bib_ref] Q fever endocarditis, A surgical view and a word of caution, Mesana [/bib_ref].
The diagnosis of Q fever endocarditis is hampered by the inability to culture C. burnetii using routine media. As a strict obligate intracellular bacterium, it can only be cultured in living cell lines, or embryonated chicken eggs, but the cultures cannot be easily performed in most laboratories, and the technique is restricted to biosafety level 3 laboratories. Thus, the diagnosis of chronic Q fever, therefore, relies on serological testing, being characterized by increased titres against the phase I antigen.
## Pcr and sequencing results
Detection of C. burnetii DNA by PCR is an important diagnostic method that could be used on different types of clinical specimens (blood, serum, infected heart valves) [bib_ref] Highly sensitive real-time PCR for specific detection and quantification of Coxiella burnetii, Klee [/bib_ref]. During the last years, several PCR based diagnostic assays have been developed to detect C. burnetii DNA in cell cultures and clinical samples. These assays used conventional PCR [bib_ref] Q fever, epidemiology, diagnosis, and treatment, Jhartzell [/bib_ref] , nested PCR [bib_ref] Plasmid based differentiation and detection of Coxiella burnetii in clinical samples, Willems [/bib_ref] [bib_ref] Detection of Coxiella burnetii specific DNA in blood samples from Japanese patients..., Kato [/bib_ref] [bib_ref] Clinical evaluation of a new PCR assay for detection of Coxiella burnetii..., Zhang [/bib_ref] [bib_ref] Direct identification of Coxiella burnetii plasmids in human sera by nested PCR, Zhang [/bib_ref] or real-time PCR conditions with LightCycler [bib_ref] Comparison of PCR and serology assays for early diagnosis of acute Q..., Fournier [/bib_ref] [bib_ref] Molecular detection of Coxiella burnetii in the sera of patients with Q..., Fenollar [/bib_ref] , SYBR Green [bib_ref] Measurement of the antibiotic susceptibility of Coxiella burnetii using real time PCR, Boulos [/bib_ref] or TaqMan chemistry. The target sequences of the assays originated from single chromosomal genes like com1, on plasmids (QpH1, QpRS) or within the transposase gene of insertion element IS1111 that is present in 20 copies in the genome of the C. burnetii Nine Mile RSA493 strain [bib_ref] Complete genome sequence of the Q-fever pathogen Coxiella burnetii, Seshadri [/bib_ref]. Due to the multicopy number of the IS1111 element, the corresponding PCR is very sensitive.
The results of PCR assays performed on DNA from serum samples positive for antiphase I C. burnetii IgG showed that nested-PCR assay permitted the amplification of C. burnetii DNA. Thus, nested-PCR led to obtaining the amplification products of the repetitive element IS1111a associated to htpAB transposase gene in the second round of amplification for all analyzed DNAs .
## Figure 1.
Gel electrophoresis of amplification products from the second round of repetitive element associated to htpAB gene in analyzed DNA samples of Q fever endocarditis cases. Line 1-Gene Ruler 100 bp Plus DNA Ladder (Fermentas); 2-5 and 7-8-DNA samples from Q fever endocarditis cases, 6-negative serum for C. burnetii; 9-negative control (pure water).
## Experimental section
## Patients
In this study we analyzed the role of C. burnetii as causative agent of blood culture negative infective endocarditis among patients with clinical diagnostic of infective endocarditis. The blood-cultures were performed for 102 patients hospitalized in three Institutes for Cardiovascular Diseases from Bucharest, Timisoara and Targu-Mures, and in Clinical Hospital for Infectious Diseases from Cluj. For each patient, a standardized questionnaire was filled by the physician in charge and logged into a database. The information filled in the questionnaire were consisted of: known preexisting valvular defect, type of valve involved (native/bioprosthetic/mechanical valve, and its position: aortic, mitral, tricuspid, pulmonary); previous antibiotic therapy; clinical symptoms; and laboratory results. Patients were considered to have possible or definite endocarditis according to the modified Duke criteria.
## Indirect immunofluorescence assays
57 serum samples were tested by indirect immunofluorescence assay (IFA) for detection of IgG antibodies to C. burnetii, B. henselae and B. quintana. From these samples 33 originated from patients with BCNE, and the rest of samples were tested before obtaining the blood cultures results, which finally were positive. We used IFA kits (Vircell, Spain) for detection of IgG antiphase I and antiphase II C. burnetii, and for IgG to B. quintana and IgG to B. henseale. The presence of IgG titers >800 to C. burnetii or B. quintana or B. henselae were considered positive for endocarditis diagnosis. Furthermore, we analyzed the positive serum samples for C. burnetii with antiphase I IgG antibody titers >800 using molecular methods for the confirmation of serological results.
# Molecular methods
## Dna extraction
The serum samples from nine patients with antiphase I C. burnetii IgG antibody titer >800, were used for DNA extraction. Total genomic DNA was extracted from 200 microliters of serum using the QIAamp blood kit (Qiagen, Hilden, Germany) according to the manufacturer instructions. DNA was resuspended in fifty microliters of elution buffer. Genomic DNAs were stored at 4 °C until their use as templates in PCR assays and subsequently at −20 °C. DNA samples have been handled carefully to avoid the risk of cross-contamination. DNA extraction, mix preparation, and PCR were performed in different rooms to prevent PCR carryover contamination. No positive control was used to prevent lateral contamination (i.e., contamination caused by PCR products amplified in other tubes in the same assay). DNA extracted from serum specimens of blood donors was used every 4 specimens as a negative control.
## Pcr assay
We used a nested-PCR assay for detection of the repetitive element IS1111 associated to htpAB transposase gene (GenBank accession number M80806). This repetitive element is present in multiple copies in the genome of C. burnetii strains (e.g., there are 20 copies of this element in the C. burnetii Nine Mile I genome), which increase the detection sensitivity of this pathogen in serum samples [bib_ref] Molecular genetic methods for the diagnosis of fastidious microorganisms, Fenollar [/bib_ref]. In the first round of amplification we used IS111F1 and IS111R1 primers, which were designed to amplify a 485-bp fragment of the repetitive element IS1111, while the second round of amplification was performed using the IS111F2 and IS111R2 primers, which amplify an internal 260-bp fragment from the same target [bib_ref] Molecular detection of Coxiella burnetii in the sera of patients with Q..., Fenollar [/bib_ref]. The sequence of specific primers used in nested-PCR reactions, and the molecular size of the amplicons are presented in [fig_ref] Table 2: The primer sequences used in nested-PCR assay for the repetitive element IS1111a... [/fig_ref]. In the nested-PCR assay, each gene fragment amplified separately on 2700 Applied Biosystems instrument using necessary components provided by Promega. The components used in each type of PCR reaction are described in [fig_ref] Table 3: The components used in nested-PCR reactions [/fig_ref]. The parameters for the amplification cycles used in each PCR experiment are presented in [fig_ref] Table 4: The conditions used for the amplification of repetitive element IS1111 of htpAB... [/fig_ref]. PCR products from nested-PCR assay were separated in a 1.5% agarose gel for 1 h at 100 V, stained with ethidium bromide and detected by UV transillumination.
## Sequencing of pcr products and sequence analysis
The sequencing of the amplicons from nested-PCR has been used to confirm the PCR results. The amplicons from the second round of nested-PCR were sequenced in both directions using the BigDye V3.1 kit as described by the manufacturer. Sequencing products have been resolved using an ABI 3100 automated Avant Genetic Analyzer (Applied Biosystems). Sequence analysis was performed with BioEdit program, which permitted obtaining of the consensus sequences that were compared with similar sequences from BLAST. The sequences obtained showed a sequence similarity of 100% with that of the GenBank prototype strain sequence. Thus, sequencing of the amplicons from the second round of PCR reaction has permitted to confirm that amplification products belong to C. burnetii. These two molecular tests were used together for the first time to investigate the BCNE cases with C. burnetii in Romania.
# Conclusions
We propose that all patients with clinical suspicion of IE be tested serologically for evidence of infection with other agents such as C. burnetii in parallel with performing blood cultures. This is the first report in this country for using the molecular methods to confirm Q fever endocarditis cases on the serum samples from eight confirmed cases and from one possible case of Q fever endocarditis tested by nested-PCR, based on repetitive element IS1111a of the transposase gene. This assay exhibited high sensitivity and specificity and led us to obtain specific amplification products, which were subsequently confirmed by direct sequencing to belong to C. burnetii, confirming previous studies showing that this nested-PCR assay presents a specificity of 100% and a sensitivity of one C. burnetii DNA copy [bib_ref] Molecular detection of Coxiella burnetii in the sera of patients with Q..., Fenollar [/bib_ref]. In conclusion, our results have demonstrated that nested-PCR amplification, followed by direct sequencing, is a reliable and accurate method when applied to serum samples and can be used as a supplementary diagnosis tool for BCNE cases.
[table] Table 1: The modified Duke criteria used for defining infective endocarditis (IE) diagnosis applied to the nine patients. [/table]
[table] Table 2: The primer sequences used in nested-PCR assay for the repetitive element IS1111a of htpAB transposase. [/table]
[table] Table 3: The components used in nested-PCR reactions. [/table]
[table] Table 4: The conditions used for the amplification of repetitive element IS1111 of htpAB transposase gene. [/table]
|
Microbial Diversity and Sulfur Cycling in an Early Earth Analogue: From Ancient Novelty to Modern Commonality
Life emerged and diversified in the absence of molecular oxygen. The prevailing anoxia and unique sulfur chemistry in the Paleo-, Meso-, and Neoarchean and early Proterozoic eras may have supported microbial communities that differ from those currently thriving on the earth's surface. Zodletone spring in southwestern Oklahoma represents a unique habitat where spatial sampling could substitute for geological eras namely, from the anoxic, surficial light-exposed sediments simulating a preoxygenated earth to overlaid water column where air exposure simulates oxygen intrusion during the Neoproterozoic era. We document a remarkably diverse microbial community in the anoxic spring sediments, with 340/516 (65.89%) of genomes recovered in a metagenomic survey belonging to 200 bacterial and archaeal families that were either previously undescribed or that exhibit an extremely rare distribution on the current earth. Such diversity is underpinned by the widespread occurrence of sulfite, thiosulfate, tetrathionate, and sulfur reduction and the paucity of sulfate reduction machineries in these taxa. Hence, these processes greatly expand lineages mediating reductive sulfur-cycling processes in the tree of life. An analysis of the overlaying oxygenated water community demonstrated the development of a significantly less diverse community dominated by well-characterized lineages and a prevalence of oxidative sulfur-cycling processes. Such a transition from ancient novelty to modern commonality underscores the profound impact of the great oxygenation event on the earth's surficial anoxic community. It also suggests that novel and rare lineages encountered in current anaerobic habitats could represent taxa that once thrived in an anoxic earth but have failed to adapt to earth's progressive oxygenation.IMPORTANCE Life on earth evolved in an anoxic setting; however, the identity and fate of microorganisms that thrived in a preoxygenated earth are poorly understood. In Zodletone spring, the prevailing geochemical conditions are remarkably similar to conditions prevailing in surficial earth prior to oxygen buildup in the atmosphere. We identify hundreds of previously unknown microbial lineages in the spring and demonstrate that these lineages possess the metabolic machinery to mediate a wide range of reductive sulfur processes, with the capacity to respire sulfite, thiosulfate, sulfur, and tetrathionate, rather than sulfate, which is a reflection of the differences in sulfur-cycling chemistry in ancient versus modern times. Collectively, such patterns strongly suggest that microbial diversity and sulfur-cycling processes in a preoxygenated earth were drastically different from the currently observed patterns and that the Great Oxygenation Event has precipitated the near extinction of a wide range of oxygen-sensitive lineages and significantly altered the microbial reductive sulfur-cycling community on earth.
S ulfur is one of the most abundant elements on earth, exhibiting a wide range of oxidation states . Microorganisms have evolved a plethora of genes and pathways for exploiting sulfur-redox reactions for energy generation. Reductive processes employ sulfur oxyanions or elemental sulfur as terminal electron acceptors in anaerobic respiratory schemes linked to heterotrophic or autotrophic growth. Oxidative processes, on the other hand, employ sulfides or elemental sulfur as electron donors, powering chemolithotrophic and photosynthetic growth.
Thermodynamic considerations limit reductive sulfur processes to habitats where oxygen is limited. This habitat restriction is reflected in the global distribution of microorganisms that reduce sulfate (SO , sulfite (SO , thiosulfate (S 2 O 3 22 ), tetrathionate (S 4 O , and elemental sulfur (S 0 ) (henceforth collectively referred to as SRM) in permanently and seasonally anoxic and hypoxic habitats in marine [bib_ref] The life sulfuric: microbial ecology of sulfur cycling in marine sediments, Wasmund [/bib_ref] [bib_ref] Sánchez-Andrea I. 2021. The bacterial sulfur cycle in expanding dysoxic and euxinic..., Vliet [/bib_ref] , freshwater (4), terrestrial [bib_ref] High unique diversity of sulfate-reducing prokaryotes characterized in a depth gradient in..., Schmalenberger [/bib_ref] , and subsurface (6) ecosystems. Sulfate is highly abundant on the current earth. Hence, sulfate reduction dominates reductive processes in the global sulfur cycle, although the reduction and disproportionation of the intermediate sulfur species, e.g., sulfur [bib_ref] Elemental sulfur reduction in the deep-sea vent thermophile, Thermovibrio ammonificans, Jelen [/bib_ref] , sulfite [bib_ref] Elemental sulfur reduction in the deep-sea vent thermophile, Thermovibrio ammonificans, Jelen [/bib_ref] , thiosulfate, and tetrathionate [bib_ref] Cryptic roles of tetrathionate in the sulfur cycle of marine sediments: microbial..., Mandal [/bib_ref] , could be significant in localized settings.
The history of earth's sulfur cycle is a prime example of a geological-biological feedback loop, where the evolution of biological processes is driven by, and dramatically impacts, the earth's biogeochemistry. The earth's surface was completely anoxic during the first two billion years of its history, and the availability and speciation of various sulfur species differed greatly from their current values. Sulfate levels were significantly lower than to current values in oceanic water , with estimates of ,200 mM to 1mM from the Archean up to the Paleoproterozoic (2.3 gigayears ago [Gya]) eras [bib_ref] The Archean sulfur cycle and the early history of atmospheric oxygen, Canfield [/bib_ref] [bib_ref] Sulfidic anion concentrations on early earth for surficial origins-of-life chemistry, Ranjan [/bib_ref] [bib_ref] The evolution of the sulfur-cycle, Canfield [/bib_ref] [bib_ref] Calibration of sulfate levels in the Archean ocean, Habicht [/bib_ref]. On the other hand, intermediate sulfur species appear to have played an important role in shaping the ancient sulfur cycle [bib_ref] Sulfur isotope fractionation during bacterial reduction and disproportionation of thiosulfate and sulfite, Habicht [/bib_ref]. Modeling suggests that mM levels of SO were attained in the Archean anoxic shallow surficial aquifers as a result of the dissolution of the volcanic SO 2 prevailing in aquatic habitats [bib_ref] Sulfidic anion concentrations on early earth for surficial origins-of-life chemistry, Ranjan [/bib_ref]. Isotopic studies have demonstrated the importance of elemental sulfur, sulfite, and thiosulfate reduction in the Archean era [bib_ref] Sulfur isotope fractionation during bacterial reduction and disproportionation of thiosulfate and sulfite, Habicht [/bib_ref] [bib_ref] Early archaean microorganisms preferred elemental sulfur, not sulfate, Philippot [/bib_ref].
The evolution of life (3.8 to 4.0 Gya) in the early Archean era and the subsequent evolution of major bacterial and archaeal clades in the late Archean and early Proterozoic eras [bib_ref] The timetree of Prokaryotes: new insights into their evolution and speciation, Marin [/bib_ref] occurred within this background of anoxia and characteristic sulfur chemistry. As such, it has been speculated that organisms using intermediate forms of sulfur were likely more common than sulfate-reducing organisms [bib_ref] Sulfur isotope fractionation during bacterial reduction and disproportionation of thiosulfate and sulfite, Habicht [/bib_ref]. However, while isotopic fractionation, modeling, and microscopic studies could provide clues on prevailing sulfur speciation patterns and prevalent biological processes, the identity of microorganisms mediating such processes is unknown. This knowledge gap is due mostly to constrains on the preservation of nucleic acids and other biological macromolecules, with the oldest successful DNA-sequenced sample being only 1.2 million years old [bib_ref] Million-year-old DNA sheds light on the genomic history of mammoths, Van Der Valk [/bib_ref].
Investigation of the microbial community in modern ecosystems with conditions resembling those prevailing in the ancient earth could provide important clues to the nature and identity of microorganisms that thrived under conditions prevailing prior to earth's oxygenation. In Zodletone spring, a surficial anoxic spring in southwestern Oklahoma, anoxic, surficial, light-exposed conditions are maintained in the sediments by the constant emergence of sulfide-saturated water at the spring source from anoxic underground water formations in the Anadarko Basin, along with gaseous hydrocarbons, which occur in seeps in the general vicinity. These surficial anoxic conditions also support a sulfur chemistry characterized by high levels of sulfide, sulfite, sulfur (soluble polysulfide), thiosulfate, and a low level of sulfate, as reported previously [bib_ref] Insights into chemotaxonomic composition and carbon cycling of phototrophic communities in an..., Buhring [/bib_ref] [bib_ref] Bacterial diversity and sulfur cycling in a mesophilic sulfide-rich spring, Elshahed [/bib_ref] [bib_ref] Metatranscriptomic analysis of a high-sulfide aquatic spring reveals insights into sulfur cycling..., Spain [/bib_ref]. Microbial diversity using 16S rRNA amplicon surveys have reported a higher level of phylogenetic diversity in the anoxic spring sediments and the affiliation of a fraction of the spring community with previously recognized sulfur-metabolizing lineages, as well as the high proportion of phylogenetically novel taxa in the spring anoxic sediments [bib_ref] Bacterial diversity and sulfur cycling in a mesophilic sulfide-rich spring, Elshahed [/bib_ref] [bib_ref] Novel high-rank phylogenetic lineages within a sulfur spring (Zodletone Spring, Oklahoma), revealed..., Youssef [/bib_ref] [bib_ref] Fine-scale bacterial beta diversity within a complex ecosystem, Youssef [/bib_ref]. As such, the prevailing conditions at the spring source are reminiscent of Novel Diversity in an Early Earth Anoxic Analogue mBio ancient metabolic capacities prevailing on the earth's surface in the late Archean/early Proterozoic eras as noted previously [bib_ref] Insights into chemotaxonomic composition and carbon cycling of phototrophic communities in an..., Buhring [/bib_ref]. Furthermore, the sediments at the source of the spring are overlaid by an airexposed water column, and prior microsensor measurements and detailed geochemical analysis [bib_ref] Insights into chemotaxonomic composition and carbon cycling of phototrophic communities in an..., Buhring [/bib_ref] [bib_ref] Metatranscriptomic analysis of a high-sulfide aquatic spring reveals insights into sulfur cycling..., Spain [/bib_ref] demonstrated that oxygen intrusion leads to a vertical oxygen gradient (from oxic in the top 1 mm, to hypoxic in the middle, to anoxic in deeper layers overlaying the sediments) (see [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref] in the supplemental material). As such, contrasting communities between the anoxic sediments and the oxygen-exposed water column could provide a glimpse into how oxygen evolution has altered such communities. Here, we combined metagenomic, metatranscriptomic, and amplicon-based approaches to fully characterize the microbial community in Zodletone spring. Our results provide a glimpse of the community mediating the ancient sulfur cycle, significantly expand the overall microbial diversity by the description of a wide range of novel lineages, and greatly increase the number of lineages documented to mediate reductive sulfur processes in the microbial tree of life.
# Results
Novel phylogenetic diversity in Zodletone sediments. Metagenomic sequencing of the spring sediments yielded 281 Gbp, of which 79.54% assembled into 12-Gbp contigs, with 6.8-Gbp contigs longer than 1 Kbp. A total of 1,848 genomes were binned, but only 683 passed quality-control criteria, and 516 remained after dereplication (see [fig_ref] TABLE S1 ,: XLSX file, 0 [/fig_ref] in the supplemental material). These metagenome-assembled genomes (MAGs) represented 64 phyla or candidate phyla (53 bacterial and 11 archaeal), 127 classes, 198 orders, and 300 families [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref]. A diversity assessment utilizing small subunit ribosomal protein S3 from assembled contigs (n = 2,079), as well as a complementary 16S rRNA Illumina sequencing effort (n = 309,074 amplicons), identified a higher number of taxa (82 phyla and 1,679 species in the ribosomal protein S3 data set, and 69 phyla and 1,050 species in 16S rRNA data set) (see [fig_ref] FIG 2: Novelty, rarity, and phylum-level makeup in Zodletone sediment and water communities [/fig_ref] in the supplemental material). Nevertheless, the overall community composition profiles generated from all three approaches were broadly similar [fig_ref] FIG 2: Novelty, rarity, and phylum-level makeup in Zodletone sediment and water communities [/fig_ref] , suggesting that the MAG list largely reflects the sediment microbial community.
An assessment of the novelty and degree of uniqueness of sediment MAGs identified a remarkably high number of previously undescribed lineages (1 phylum, 14 classes, 43 orders, and 97 families) as well as lineages exhibiting rare global distribution (LRD) pattern (11 phyla, 24 classes, 45 orders, and 113 families) in the spring [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref]. We define LRD lineages as those represented by 5 genomes or less in the Genome Taxonomy Database release 95 (GTDB r95). At the family level, 132 (25.58%) and 208 (40.03%) genomes clustered into 97 novel and 113 LRD families, respectively, bringing the proportion of genomes belonging to novel or LRD families in Zodletone sediments to 65.89%. The high level of novelty in the sediment MAGs is reflected in an average relative evolutionary divergence (RED) value of 0.76, which is a value that is slightly lower than the median RED value for the designation of a novel family (0.77) [bib_ref] A complete domain-to-species taxonomy for Bacteria and Archaea, Parks [/bib_ref].
The Chloroflexota (n = 69), Planctomycetota (n = 47), Bacteroidota (n = 43), Desulfobacterota (n = 43), Spirochaetota (n = 28 genomes), Patescibacteria (n = 20 genomes), and the archaeal phylum Nanoarchaeota (n = 21) were the most abundant phyla in Zodletone spring sediments, albeit representing only 52.52% of the total number of recovered genomes (see Text S1 in the supplemental material; [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref] ; see in the supplemental material). An extreme paucity of genomes belonging to the Proteobacteria (6 genomes) and Firmicutes (12 genomes), which are widely distributed and abundant taxa in current biomes, and the absence of oxygen-generating Cyanobacteria (0 genomes) were observed [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref]. Therefore, in addition to expanding the number of novel lineages (classes, orders, and families) and greatly enriching available genomes in rare, poorly represented taxa, our results highlight the uniqueness and distinction of the microbial community thriving in Zodletone spring sediments, compared with those of present earth environments studied so far. Oxygen intrusion reduces the proportion of novel and rare lineages in Zodletone spring. Metagenomic sequencing of the oxygen-exposed overlaying water column community yielded 323 Gbp, of which 80.07% assembled into 3.6-Gbp contigs, with 3.1-Gbp contigs of .1 Kbp. A total of 883 genomes were binned, with only 114 remaining after dereplication. Of these genomes, 62 belonged to families shared with the sediment community, while 52 were water specific. Genomes recovered from the water column belonged to a significantly lower number of phyla (n = 27), classes (n = 37), orders (n = 52), and families (n = 79) than those from the euxinic sediments [fig_ref] TABLE S1 ,: XLSX file, 0 [/fig_ref]. The community exhibited a much lower level of novelty and rarity at the phylum, class, order, and family levels than those of the sediment community [fig_ref] FIG 2: Novelty, rarity, and phylum-level makeup in Zodletone sediment and water communities [/fig_ref]. Water-specific genomes (n = 52) belonged mostly to well-characterized microbial lineages, e.g., families Rhodobacteraceae and Rhodospirillaceae in Alphaproteobacteria; families Thiomicrospiraceae, Halothiobacillaceae, Acidithiobacillaceae, Burkholderiaceae, Chromatiaceae, and Methylothermaceae in Gammaproteobacteria; families Sulfurimonadaceae and Sulfurovaceae in phylum Campylobacterota; and well described families in the phyla Bacteroidota and Desulfobacterota (Text S1; [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref] ; [fig_ref] TABLE S1 ,: XLSX file, 0 [/fig_ref]. Collectively, this information demonstrates a pattern where the intrusion of oxygen is negatively correlated with the presence of previously undescribed and LRD lineages, which are prevalent in the sediment.
Reductive sulfur processes dominate Zodletone spring sediment communities. A total of 149 genomes (28.9% of all genomes), belonging to 32 phyla, 51 classes, 69 orders, and 97 families were involved in at least 1 reductive sulfur processes [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] ; see [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] in the supplemental material). By comparison, only 21 sediment genomes (4.06% of all genomes) encoded at least 1 sulfur oxidation pathway [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] ; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. The reductive sulfur community in the spring exhibited two unique traits as follows: first, a majority of genomes encoding such capacities belonged to novel (47 genomes) or LRD (66 genomes) lineages [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , and second, sulfite, polysulfide, thiosulfate, and tetrathionate reduction capacities appear to be more prevalent than sulfate-reduction capacities in the sediment genomes.
Sulfate reduction capacity was encoded in only 18 sediment genomes [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] but exhibited a unique community composition, when compared with wellstudied marine and terrestrial habitats [bib_ref] The life sulfuric: microbial ecology of sulfur cycling in marine sediments, Wasmund [/bib_ref] [bib_ref] Sulphate reduction and sulphur cycling in lake sediments: a review, Holmer [/bib_ref] [bib_ref] Metagenomes and metatranscriptomes shed new light on the microbial-mediated sulfur cycle in..., Vavourakis [/bib_ref]. Sulfate reduction capacities were observed in mostly previously undescribed or LRD lineages within the Zixibacteria, Acidobacteriota (members of family UBA6911, equivalent to Acidobacteria group 18), Myxococcota, Bacteroidota, Planctomycetota, and candidate phylum OLB16 (1 genome),
[formula] FIG 1 Legend (Continued) [/formula]
the tree represent (from innermost to outermost)the following: cultured status at the order level (cultured versus uncultured), abundance in GTDB based on the number of available genomes (abundant with more than 5 genomes, rare with 5 genomes or less, and novel with no genomes in GTDB), percentage database enrichment (calculated as number of genomes belonging to a certain order binned in the current study as a percentage of the number of genomes belonging to the same order in GTDB), energy conservation capabilities depicted by colored circles (salmon, aerobic respiration; orange, Fe 31 respiration; yellow, nitrate/nitrite reduction; dark green, reductive sulfur processes; lime green, nitrogen oxidation; cyan, oxidative sulfurprocesses; pink, respiratory hydrogen oxidation; and purple, photosynthesis), and the number of MAGs belonging to each order binned from the sediment (blue bars) and the water (orange bars). For orders with 20 or more genomes, the family-level delineation is shown in . These orders are Anaerolineales , Bacteroidales , Sedimentisphaerales , Spirochaetales , Syntrophales , and Woesearchaeales . [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. Sulfite (but not sulfate) reduction via the DsrAB1DsrC1DsrKMJOP system was identified in only 8 genomes belonging to 7 families within the phyla Planctomycetes, Chloroflexota, Spirochaetota, and Desulfobacterota [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. On the other hand, the sulfite reduction capacity within Zodletone spring sediment solely via the Asr/Hdr system was rampant, being encountered in 104 genomes belonging to 28 phyla, 43 (8 novel and 9 LRD) classes, 56 (18 novel and 12 LRD) orders, and 72 (31 novel and 25 LRD) families [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S6b; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] , with a gene organization of the asr locus adjacent to the hdr locus in the majority of genomes . Asr-encoding genomes in the sediment included members of previously undescribed and LRD lineages within the Chloroflexota, Desulfobacterota, Planctomycetota, and Bacteroidota. The capacity was also rampant in the yet-uncultured bacterial phyla, of which many have a fairly limited global distribution (e.g., the candidate phyla CSSED10-310, FCPU426, RBG-13-66-14, SM23-31, SZUA-182, UBP14, Aureabacteria, Sumerlaeota, and Krumholzibacteriota). Zodletone dissimilatory sulfite reductase and the anaerobic sulfite reductase sequences clustered with reference sequences from the same phylum, generally showing no evidence of LGT.
Sulfur (polysulfide) reduction capacities were observed in 20 Zodletone sediment genomes that encoded psrABC genes [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , S5c, and S6c; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. In addition, representatives of the cytoplasmic sulfurhydrogenase I (HydABCD system) and/or II (ShyABCD system) were identified in 119 Zodletone sediment genomes [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. However, the direct involvement of these enzymes in an ETS-associated respiration is not yet clear (Text S1).
Sediment genomes also encoded thiosulfate disproportionation and reduction capacities. The quinone-dependent membrane-bound molybdopterin-containing thiosulfate reductase PhsABC was encoded in 11 genomes belonging to 6 phyla ( . Within these genomes, only two (a Chloroflexota family UBA6092 genome and a Desulfatiglandales family HGW15 genome) also encoded a dissimilatory sulfite reductase (the Asr system) akin to the Gammaproteobacteria thiosulfate-disproportionating pure culture members [bib_ref] The alternative electron acceptor tetrathionate supports B12-dependent anaerobic growth of Salmonella enterica..., Price-Carter [/bib_ref] , where the final products of thiosulfate disproportionation are expected to be only hydrogen sulfide [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S5d; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. On the other hand, 5 of the 11 phsABC-encoding Zodletone genomes also encoded the sulfite dehydrogenase SoeABC system, akin to Desulfobacterota and Firmicutes pure culture members, where the final products of thiosulfate disproportionation are expected to be both hydrogen sulfide and sulfate (28, 29) [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S5d; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref]. In addition to the phsABC system, 14 Zodletone genomes belonging to 6 phyla encoded a rhodanase-like enzyme (EC 2.8.1.1 or EC 2.8.1.3) for thiosulfate disproportionation, as well as enzymes for both sulfite oxidation (by means of reversal of sulfate reduction via Sat1AprAB or the sulfite dehydrogenase SoeABC), and sulfite reduction (via the dissimilatory sulfite reductases Dsr or Asr), where the final products of thiosulfate disproportionation are expected to be both hydrogen sulfide and sulfate (30-34) [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S5d; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref].
Tetrathionate reduction capacities were identified in 105 sediment genomes. Seventythree Zodletone sediment genomes from 14 phyla encoded the octaheme tetrathionate reductase (OTR) enzyme [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S6e). In addition to Otr, 68 Zodletone genomes from 14 phyla encoded the Ttr enzyme system [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] and S5e). As shown previously in Salmonella enterica serovar Typhimurium [bib_ref] The alternative electron acceptor tetrathionate supports B12-dependent anaerobic growth of Salmonella enterica..., Price-Carter [/bib_ref] , in the presence of means for thiosulfate disproportionation/reduction and sulfite reduction, the thiosulfate produced as a result of tetrathionate reduction could be further reduced to sulfide. Out of the 105 sediment genomes encoding the Otr, and/or Ttr enzymes, only 12 genomes also encoded thiosulfate and sulfite reduction enzymes.
Within lineages mediating reductive sulfur processes in Zodletone sediments (n = 98), a wide range of substrates supporting sulfidogenesis were identified ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. They included hexoses (26% to 87% of sulfidogenic lineages); pentoses (30% to 41% of sulfidogenic lineages); amino acids and peptides (39% of lineages); short-chain fatty acids, e.g., lactate, propionate, butyrate, and acetate (22% to 73% of lineages); long-chain fatty acids (29% of lineages); aromatic hydrocarbons (3% of lineages); and short-chain alkanes (6% of lineages). Autotrophic capacities with hydrogen as the electron donor were identified in 28% of sulfidogenic lineages.
Transcriptomic analysis. Transcriptional expression of genes involved in S species reduction/disproportionation was analyzed in the spring sediments [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. All S species reduction/disproportionation genes discussed above were identified and mapped to 51 distinct phyla. Total transcription levels of the Asr system were 4 times higher than those of the Dsr system, which is consistent with the higher number of Zodletone sediment genomes encoding the Asr system than that of the Dsr system. Asr system genes were mapped to 11 phyla; while DSR genes mapped to 4 phyla [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. Sulfate reduction genes (Sat, AprAB, and QmoABC) were also transcribed with major contributions from 4 phyla. Transcription of the thiosulfate disproportionating rhodanese-like (EC 2.8.1.1 or EC 2.8.1.3), thiosulfate reductase phsABC, tetrathionate reduction genes ttrABC, octaheme tetrathionate reductase otr, psrABC for polysulfide reduction, and cytoplasmic sulfurhydrogenases I and II (hyd/shy systems) were also identified [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] ; Text S1, for detailed contributions of taxa).
Oxidative sulfur processes dominate the Zodletone water community. Reductive sulfur processes were extremely sparse in the water community [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] ; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; Text S1). In contrast, oxidative sulfur processes dominated the water commu- Novel Diversity in an Early Earth Anoxic Analogue mBio nity, with pathways encoding sulfide, sulfur, thiosulfate, tetrathionate, and/or sulfite oxidation to sulfate present in 59/114 (51.8%) of water genomes, belonging to 13 phyla, 16 classes, 25 orders, and 43 families, respectively. The oxidative sulfur community in the water belonged to mostly well-characterized lineages [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , with only 8 and 10 genomes involved in oxidative sulfur processes belonging to previously undescribed and LDR families, respectively. A complete SOX system, putatively mediating oxidation of a wide range of reduced sulfur-species to sulfate, was encoded in genomes belonging to well-characterized families within Proteobacteria (11 genomes in Acidithiobacillaceae, Burkholderiaceae, Halothiobacillaceae, Rhodobacteraceae, and Thiomicrospiraceae) and Campylobacterota (3 genomes in the family Sulfurimonadaceae) [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. The capacity for sulfide oxidation to sulfur (sulfide dehydrogenase and/or the sulfide:quinone oxidoreductase Sqr) was encoded in 39 water genomes (Text S1; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. Only 2 of the above 39 genomes (a Proteobacteria genome and a Nitrospirota genome) encoded the capacity to further oxidize the sulfur/ polysulfide to sulfite via the reversal of the Dsr system. The capacity for sulfite oxidation to sulfate via the reversal of AprAB1QmoABC system, the sulfite dehydrogenase (quinone) SoeABC, or the sulfite dehydrogenase (cytochrome) SorAB system was encoded in 1, 22, and 3 genomes, respectively (Text S1; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. Finally, for thiosulfate oxidation, eight water genomes (Proteobacteria and Flavobacteriaceae) encoded thiosulfate to tetrathionate oxidation capacities via either the thiosulfate dehydrogenase tsdA (EC 1.8.2.2) or doxAD (EC 1.8.5.2). Two of these 8 genomes also encoded tetrathionate hydrolase (tetH) [bib_ref] Identification of a gene encoding a tetrathionate hydrolase in Acidithiobacillus ferrooxidans, Kanao [/bib_ref] that is known to cleave tetrathionate to thiosulfate, sulfur, and sulfate (Text S1; [fig_ref] Table S2 ;: Zod Metabat [/fig_ref] ; [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. Simultaneous identification of the SOX system and both forms of sulfide dehydrogenase (fccAB and Sqr) imply that these two genomes encode the capacity for complete thiosulfate oxidation to sulfate.
# Discussion
The microbial community in Zodletone spring sediments exhibited a high level of phylogenetic diversity, novelty, and rarity [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref]. Conversely, representatives of lineages that predominate in most present earth environments, e.g., Proteobacteria, Firmicutes, and Cyanobacteria, were absent or extremely sparse within the sediments. The community in the spring sediments was also characterized by a high proportion of SRM and the prevalence of lineages mediating the reduction of sulfur cycle intermediates (sulfite, thiosulfate, tetrathionate, and elemental sulfur) over sulfate reducers. Many of the organisms mediating reductive sulfur-cycling processes belonged to novel and LRD lineages [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref] , hence expanding the range of SRM within the tree of life .
What drives the assembly, propagation, and maintenance of such a diverse, novel, and distinct community in the spring sediments? The high level of diversity, novelty, and rarity within Zodletone spring sediment SRM community could be attributed to two main factors. First, a wide range of sulfur cycle intermediates are available in concentrations much higher than sulfate, in contrast to sulfate predominance in current ecosystems [bib_ref] The life sulfuric: microbial ecology of sulfur cycling in marine sediments, Wasmund [/bib_ref]. Such a pattern selects for a more diverse community of SRM in the spring than that of predominantly sulfate-driven marine and freshwater ecosystems [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref]. Second, additional factors usually constraining SRM growth in several habitats, such as diel or seasonal intrusion of oxygen, Fe and NO 3 [bib_ref] Competition for electron donors among nitrate reducers, ferric iron reducers, sulfate reducers,..., Achtnich [/bib_ref] [bib_ref] Controls on volatile fatty acid concentrations in marine sediments, Glombitza [/bib_ref] , recalcitrance of available substrates [bib_ref] Microbial metabolisms in a 2.5-km-deep ecosystem created by hydraulic fracturing in shales, Daly [/bib_ref] [bib_ref] Slow microbial life in the seabed, Jørgensen [/bib_ref] , temperature [bib_ref] Bacterial sulfate reduction above 100°C in deep-sea hydrothermal vent sediments, Jørgensen [/bib_ref] [bib_ref] Psychrophilic properties of sulfatereducing bacteria in Arctic marine sediments, Scholze [/bib_ref] , pH [bib_ref] Metagenomes and metatranscriptomes shed new light on the microbial-mediated sulfur cycle in..., Vavourakis [/bib_ref] [bib_ref] Sulfate reduction at low pH to remediate acid mine drainage, Sánchez-Andrea [/bib_ref] [bib_ref] The microbial sulfur cycle at extremely haloalkaline conditions of soda lakes, Sorokin [/bib_ref] , salinity [bib_ref] Sulfate-reducing bacteria and their activities in cyanobacterial mats of Solar Lake, Teske [/bib_ref] , and pressure extremes [bib_ref] Microbial metabolisms in a 2.5-km-deep ecosystem created by hydraulic fracturing in shales, Daly [/bib_ref] [bib_ref] Active sulfur cycling in the terrestrial deep subsurface, Bell [/bib_ref] , or combinations thereof, are absent in the spring. Therefore, while the reductive global sulfur cycle appears to be dominated by a few sulfate-reducing lineages within Desulfobacterota, and to a lesser extent Firmicutes, as well as Thermodesulfobacteria and Archaeoglobus in high-temperature habitats, the SRM community in Zodletone is extremely diverse, encompassing a wide range of previously undescribed and LRD lineages [fig_ref] FIG 4: Sulfur cycle in Zodletone spring [/fig_ref].
Sulfate-reducing organisms are the most prevalent component of the reductive sulfur cycle in most marine and aquatic ecosystems. Aspects of the ecology (2), physiology [bib_ref] A post-genomic view of the ecophysiology, catabolism and biotechnological relevance of sulphate-reducing..., Rabus [/bib_ref] , and biochemistry (48-50) of dissimilatory sulfate reduction have been investigated extensively [bib_ref] Dissimilatory sulfate-and sulfurreducing prokaryotes, Rabus [/bib_ref]. While not the most prevalent process, the sulfate-reducing community in Zodletone spring sediment exhibited a unique composition, with members of Zixibacteria, Acidobacteriota, Myxococcota, Bacteroidota, Planctomycetota, and candidate phylum OLB16 constituting the major players, as well as rare and novel lineages within Desulfobacterota (Desulfatiglandales and order C00003060), with scarce representation of canonical Desulfobacterota sulfate reducers (1 genome). While the identification of the dissimilatory sulfate-reducing machinery in some of these lineages (e.g., Zixibacteria, Acidobacteriota, and Planctomycetota) has been shown before [bib_ref] Expanded diversity of microbial groups that shape the dissimilatory sulfur cycle, Anantharaman [/bib_ref] [bib_ref] Peatland Acidobacteria with a dissimilatory sulfur metabolism, Hausmann [/bib_ref] [bib_ref] Novel taxa of Acidobacteriota involved in seafloor sulfur cycling, Flieder [/bib_ref] , these members rarely appear to be the dominant players in a single ecosystem.
Compared with sulfate reduction, the ecology and diversity of microbial dissimilatory sulfite reduction has not been studied extensively. The biochemistry of the process has been examined in sulfate reducers, when grown on sulfite (51), as well as in a few other dedicated sulfite reducers, such as members of Desulfitobacterium (55), Salmonella (56), Shewanella (57), and Wolinella (58). A recent study suggested the importance and the ancient nature of sulfite reduction in an extreme thermophilic environment in a limited diversity biofilm [bib_ref] Phylogenomic analysis of novel Diaforarchaea is consistent with sulfite but not sulfate..., Colman [/bib_ref]. We document a plethora of microorganisms within the phyla Planctomycetes, Chloroflexota, Spirochaetota, and Desulfobacterota encoding the dissimilatory sulfite reductase DSR, as well as 72 additional families (31 novel and 25 LRD) encoding the anaerobic sulfite reductase. These organisms expand the known sulfite reduction capacity within the domain Bacteria . Furthermore, the novelty or rarity of some of these families is a reflection of the dearth of current habitats that could support this mode of metabolism, once predominant on ancient earth.
The bulk of knowledge on thiosulfate reduction and or disproportionation comes from studies in pure cultures, e.g., members of Desulfobulbaceae (e.g., Desulfocapsa) (28, 60) and the genera Desulfovibrio and Desulfomonile (61-63) within Desulfobacterota, the gammaproteobacterium Pantoea agglomerans [bib_ref] Sulfur disproportionation by the facultative anaerobe Pantoea agglomerans SP1 as a mechanism..., Obraztsova [/bib_ref] , and members of Thermodesulfobacteria (65) and Firmicutes [bib_ref] Thiosulfate disproportionation by Desulfotomaculum thermobenzoicum, Jackson [/bib_ref]. Radioisotope tracing of different sulfur atoms showed a significant contribution of thiosulfate disproportionation to the sulfur cycle in marine [bib_ref] A thiosulfate shunt in the sulfur cycle of marine sediments, Jørgensen [/bib_ref] , as well as freshwater, sediments [bib_ref] The sulfur cycle of freshwater sediments: role of thiosulfate, Jørgensen [/bib_ref]. However, the lack of a marker gene for the process hinders ecological culture-independent studies. Similar to sulfite, the high levels and constant generation of thiosulfate in Zodletone sediments sustain a highly diverse thiosulfate-reducing (thiosulfate reductase plus a sulfite reduction complex) or thiosulfate-disproportionating (thiosulfate reductase plus both sulfite reduction and sulfite oxidation systems) community with major contributions from novel or rare families in Acidobacteriota, Chloroflexota, Desulfobacterota, KSB1, Myxococcota, and Spirochaetota. Finally, the extremely high levels of zero valent sulfur, available as soluble polysulfide, result in enriching the community with a plethora of polysulfide-reducing organisms.
As described above, this study infers that the microbial communities thriving under ancient conditions of anoxia and a high proportion of sulfur cycle intermediates were extremely diverse. In comparison, communities in the oxygen-exposed water column were markedly less diverse. What drives this drastic shift in diversity and community structure? We argue that oxygen introduction into the system is responsible for such a shift, as evident by the shift to oxidative sulfur processes in the water samples. The prevailing conditions in the water column are hence more akin to microbial communities that would thrive in sulfur-rich yet air-exposed habitat on the current earth.
The comparison presented here between both communities could demonstrate putatively how ancient metabolic pathways and lineages mediating them have been curtailed due to oxygen evolution and predominance in the current surficial earth. The evolution of oxygenic photosynthesis has led to the steady and inexorable accumulation of O 2 in Earth's atmosphere (the great oxidation event [GOE]), with the rise of atmospheric O 2 to 1% to 5% of current levels between 2.4 and 2.1 billion years (Gyr) ago, and its accumulation to values comparable to modern values 500 to 600 million years ago (Mya) [bib_ref] The rise of oxygen in Earth's early ocean and atmosphere, Lyons [/bib_ref]. Due to the expected sensitivity and lack of adaptive mechanisms to cope with atmospheric oxygen in multiple strict anaerobes, as well as the chemical instability of multiple S species in an oxygenated atmosphere, the GOE exerted a profound negative impact on anaerobic surficial life forms (the oxygen catastrophe) leading to the first and arguably most profound extinction event in earth's history [bib_ref] The rise of oxygen in Earth's early ocean and atmosphere, Lyons [/bib_ref]. In addition to suppressing anaerobiosis in atmospherically exposed habitats, the GOE also led to a significant change in the S cycle, from one based on atmospheric inputs to one dependent on oxidative weathering leading to the release of a huge amount of sulfate derived from the oxidation of pyrite and the dissolution of sulfate minerals (69), hitherto a minor by-product of Archean abiotic and biotic reactions [bib_ref] Sulfur isotope fractionation during bacterial reduction and disproportionation of thiosulfate and sulfite, Habicht [/bib_ref] [bib_ref] Some applications of thermochemical data to problems of ore deposits; part 2,..., Holland [/bib_ref]. Therefore, it appears that the loss of niches associated with geological transformations could be one of the possible explanations for high extinction rates for microorganisms on earth, as well as the constant identification of rare, novel taxa within anaerobic settings. It is notable that phylogenetically novel branches with extremely rare distribution on earth (defined as phyla with 5 genomes or less in GTDB) have been identified consistently in anaerobic habitats.
In summary, by examining microbial diversity in Zodletone spring, we greatly expand the overall diversity within the tree of life via the discovery and characterization of a wide range of novel lineages and significantly enrich the representation of a wide range of LRD lineages. We also describe a unique sulfur-cycling community in the spring that is largely dependent on sulfite, thiosulfate, sulfur, and tetrathionate, rather than sulfate, as an electron acceptor. Given the remarkable similarity to conditions prevailing prior to the GOE, we consider the spring an invaluable portal with which to investigate the community thriving on the earth's surface during these eras and posit that GOE precipitated the near extinction of a wide range of phylogenetically distinct oxygen-sensitive lineages and drastically altered the reductive sulfur-cycling community from sulfite, sulfur, and thiosulfate reducers to predominantly sulfate reduction in the current earth.
# Materials and methods
Site description and geochemistry. Zodletone spring is located in the Anadarko Basin of western Oklahoma (N34.99562°W98.68895°). The spring arises from underground, where water is pumped out slowly along with sediments. Sediments settled at the source of the spring, a boxed square of 1 m 2 [fig_ref] FIG 1: Phylogenomics of the 516 bacterial [/fig_ref] , are overlaid with water that collects and settles in a concrete pool erected in the early 1900s. The settled water is 50 cm deep above the sediments and is exposed to atmospheric air. Water and sediments originating from the spring source are highly reduced due to the high dissolved sulfide levels (8 to 10 mM) in the spring sediments. Microsensor measurements show a completely anoxic (oxygen levels of ,0.1 mM) and highly reduced source sediments. Oxygen levels slowly increase in the overlaid water column from 2 to 4 mM in the 2 mm above the source to complete oxygen exposure at the top of the water column [bib_ref] Insights into chemotaxonomic composition and carbon cycling of phototrophic communities in an..., Buhring [/bib_ref]. The spring geochemistry has been monitored regularly during the last 2 decades [bib_ref] Insights into chemotaxonomic composition and carbon cycling of phototrophic communities in an..., Buhring [/bib_ref] [bib_ref] Bacterial diversity and sulfur cycling in a mesophilic sulfide-rich spring, Elshahed [/bib_ref] [bib_ref] Barite deposition mediated by photootrophic sulfide-oxidizing bacteria, Senko [/bib_ref] and is remarkably stable. The spring is characterized by low levels of sulfate (50 to 94 mM), with higher levels of sulfite (0.21 mM), elemental sulfur (0.1 mM), and thiosulfate (0.52 mM) [bib_ref] Metatranscriptomic analysis of a high-sulfide aquatic spring reveals insights into sulfur cycling..., Spain [/bib_ref] [bib_ref] Barite deposition mediated by photootrophic sulfide-oxidizing bacteria, Senko [/bib_ref].
Sampling. Samples were collected from the source sediments and standing overlaid water in sterile containers and kept on ice until they were transported to the lab (;2-h drive), where they were processed immediately. For metatranscriptomics, samples were collected at three different time points, namely, morning (9:15 a.m.), afternoon (2:30 p.m.), and evening (5:30 p.m.) in June 2019. They were stored on dry ice and then transferred to the lab where they were stored at 280°C until being processed for RNA extraction within a week.
Nucleic acid extraction. DNA was extracted directly from 0.5 g of source sediments. For water samples, water was filtered on 0.2-mm sterile filters. DNA was directly extracted from filters (20 filters, 10 L of water samples). Extraction was conducted using the DNeasy PowerSoil kit (Qiagen, Valencia, CA, USA). RNA was extracted from 0.5-g sediment samples using RNeasy PowerSoil total RNA kit (Qiagen) according to the manufacturer;s instructions.
16S rRNA gene amplification, sequencing, and analysis. Triplicate DNA extractions were performed for both sediment and water samples from the Zodletone spring. To characterize the microbial diversity based on 16S rRNA gene sequences, we used the Quick-16S next-generation sequencing (NGS) library prep kit (Zymo Research, Irvine CA), following the manufacturer's protocol. For amplification of the V4 hypervariable region, we used a mix of modified versions of primers 515F-806R (72), tailored to provide better coverage for several underrepresented microbial lineages. They included 515FY (59-GTGYCAGCMGCCGCGGTAA) (73), 515F-Cren (59-GTGKCAGCMGCCGCGGTAA, for Crenarchaeota) (74), 515F-Nano (59-GTGGCAGYCGCCRCGGKAA, for Nanoarchaeota) [bib_ref] Comparative analysis of microbial diversity across temperature gradients in hot springs from..., Podar [/bib_ref] , and 515F-TM7 (59-GTGCCAGCMGCCGCGGTCA for TM7/Saccharibacteria) (75) as forward mix and 805RB (59-GGACTACNVGGGTWTCTAAT) (76) and 805R-Nano (59-GGAMTACHGGGGTCTCTAAT, for Nanoarchaeota) (74) as reverse mix. Purified barcoded amplicon libraries were sequenced on a MiSeq instrument (Illumina Inc., San Diego, CA) using a v2 500-cycle kit, according to the manufacturer's protocol. Demultiplexed forward Novel Diversity in an Early Earth Anoxic Analogue mBio and reverse reads were imported as paired fastq files into QIIME2 v. 2020.8 [bib_ref] Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2, Bolyen [/bib_ref] for analysis. The DADA2 plugin was used to trim, denoise, pair, purge chimeras, and select amplicon sequence variants (ASVs), using the command "qiime dada2 denoise-paired." Between 44,000 and 194,000 nonchimeric sequences were obtained for the individual samples. The ASVs were classified taxonomically in QIIME2 using a trained classifier built based on the Silva-138-99 rRNA sequence database. The ASVs were assigned to 1,643 taxonomic categories corresponding to taxonomic level 7 (species and above) and to 932 genera (level 6). Alpha rarefaction curves indicated a saturation of observed sequence features (ASVs) at a sequencing depth of 70,000 to 80,000 sequences. Metagenome sequencing, assembly, and binning. Metagenomic sequencing was conducted using the services of a commercial provider (Novogene, Beijing, China) using two lanes of the Illumina HiSeq 2500 system for each of the water and sediment samples. Transcriptomic sequencing using Illumina HiSeq 2500 2 Â 150-bp paired-end technology was conducted using the services of a commercial provider (Novogene Corporation). Metagenomic reads were assessed for quality using FastQC followed by quality filtering and trimming using Trimmomatic v.0.38 [bib_ref] Trimmomatic: a flexible trimmer for Illumina sequence data, Bolger [/bib_ref]. High-quality reads were assembled into contigs using MegaHit (v.1.1.3) with minimum kmer of 27, maximum kmer of 127, kmer step of 10, and minimum contig length of 1,000 bp. Bowtie2 was used to calculate sequencing coverage of each contig by mapping the raw reads back to the contigs. Assembled contigs were searched for ribosomal protein S3 (rpS3) sequences using a custom hidden Markov model (HMM) built from Uniprot reference sequences assigned to the KEGG orthologies (Kos) K02982 and K02984 (corresponding to the bacterial, and archaeal RPS3, respectively) using hmmbuild (HMMER 3.1b2). rpS3 sequences were clustered at 99% identity (ID) using CD-HIT as suggested previously for a putative species cutoff for rpS3 data [bib_ref] Mediterranean grassland soil C-N compound turnover is dependent on rainfall and depth,..., Diamond [/bib_ref]. Taxonomic affiliations of rpS3 groups were identified using Diamond BLAST against the GTDB r95 database [bib_ref] A complete domain-to-species taxonomy for Bacteria and Archaea, Parks [/bib_ref].
Contigs from the sediment and water assemblies were binned into draft genomes using both Metabat [bib_ref] MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction..., Kang [/bib_ref] and MaxBin2 [bib_ref] MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic..., Wu [/bib_ref]. DasTool was used to select the highest quality bins from each metagenome assembly [bib_ref] Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy, Sieber [/bib_ref]. CheckM was used for the estimation of genome completeness, strain heterogeneity, and contamination [bib_ref] CheckM: assessing the quality of microbial genomes recovered from isolates, single cells,..., Parks [/bib_ref]. Genomic bins showing contamination levels higher than 10% were further refined based on the taxonomic affiliations of the binned contigs, as well as the GC content, tetranucleotide frequency, and coverage levels using RefineM [bib_ref] Recovery of nearly 8,000 metagenomeassembled genomes substantially expands the tree of life, Parks [/bib_ref]. Low-quality bins (.10% contamination) were cleaned by removal of the identified outlier contigs, and the percentage completeness and contamination were again rechecked using CheckM.
Genome classification, annotation, and metabolic analysis. Taxonomic classifications followed the Genome Taxonomy Database (GTDB) release r95 [bib_ref] A complete domain-to-species taxonomy for Bacteria and Archaea, Parks [/bib_ref] and were carried out using the classify_workflow in GTDB-Tk (v.1.1.0) [bib_ref] GTDB-Tk: a toolkit to classify genomes with the, Chaumeil [/bib_ref]. Phylogenomic analysis utilized the concatenated alignment of a set of 120 single-copy bacterial genes and 122 single-copy archaeal genes (24) generated by the GTDB-Tk. A maximum-likelihood phylogenomic tree was constructed in FastTree using the default parameters [bib_ref] FastTree 2-approximately maximum-likelihood trees for large alignments, Price [/bib_ref].
Annotation and metabolic analysis. Protein-coding genes were predicted using Prodigal [bib_ref] Prodigal: prokaryotic gene recognition and translation initiation site identification, Hyatt [/bib_ref]. GhostKOALA [bib_ref] BlastKOALA and GhostKOALA: KEGG tools for functional characterization of genome and metagenome..., Kanehisa [/bib_ref] was used for the functional annotation of every predicted open reading frame in every genomic bin and to assign protein-coding genes to KEGG orthologies (KOs).
Analysis of sulfur-cycling genes. To identify taxa mediating key sulfur-transformation processes in the spring sediments, we mapped the distribution of key sulfur-cycling genes in all genomes and deduced capacities in individual genomes by documenting the occurrence of entire pathways (as explained below in detail). This information was confirmed subsequently by phylogenetic analysis and examining contiguous gene organization in processes requiring a multisubunit and/or multigene. Furthermore, expression data were used from three time points to identify the fraction of the community that is metabolically actively involved in the process. An analysis of Sulfur (S) cycling capabilities was conducted on individual genomic bins by building and scanning hidden Markov model (HMM) profiles as explained below. To build the sulfur gene HMM profiles, Uniprot reference sequences for all genes with an assigned KO number were downloaded and aligned using Clustal Omega [bib_ref] Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega, Sievers [/bib_ref] , and the alignment was used to build an HMM profile using hmmbuild (HMMER 3.1b2) [bib_ref] Challenges in homology search: HMMER3 and convergent evolution of coiled-coil regions, Mistry [/bib_ref]. For genes not assigned a KO number (e.g., otr, tsdA, and tetH), a representative protein was compared against the KEGG database using BLASTP, and significant hits (those with E values of ,e-80) were downloaded and used to build HMM profiles as explained above. The custom-built HMM profiles were then used to scan the analyzed genomes for significant hits using hmmscan (HMMER 3.1b2) (90) with the option -T 100 to limit the results to only those profiles with an alignment score of at least 100. Further confirmation was achieved through phylogenetic assessment and tree building procedures, in which potential candidates identified by hmmscan were aligned to the reference sequences used to build the custom HMM profiles using Clustal Omega (89), followed by maximum likelihood phylogenetic tree construction using FastTree [bib_ref] FastTree 2-approximately maximum-likelihood trees for large alignments, Price [/bib_ref]. Only candidates clustering with reference sequences were deemed true hits and were assigned to the corresponding KO. Details on the genes examined for evidence of sulfate, sulfite, polysulfide, tetrathionate, and thiosulfate reduction; thiosulfate disproportionation; and various sulfur oxidation capacities are provided in the supplemental material (Text S1).
Phylogenetic analysis and operon organization of S cycling genes. The phylogenetic affiliation of the S cycling proteins AsrB, Otr, PhsC, PsrC, and DsrAB was examined by aligning the Zodletone genome predicted protein sequences to Uniprot reference sequences using MAFFT [bib_ref] MAFFT: a novel method for rapid multiple sequence alignment based on fast..., Katoh [/bib_ref]. The DsrA and DsrB alignments were concatenated in MEGA X [bib_ref] MEGA X: Molecular Evolutionary Genetics Analysis across computing platforms, Kumar [/bib_ref]. All alignments were used to construct maximum likelihood phylogenetic trees in RAxML [bib_ref] RAxML-NG: a fast, scalable and user-friendly tool for maximum likelihood phylogenetic inference, Kozlov [/bib_ref]. The R package genoPlotR (94) was used to produce gene maps for the DSR and ASR loci in Zodletone genomes using the Prodigal predicted gene starts, ends, and strand direction.
Transcription of sulfur-cycling genes. A total of 21.4 million, 27.9 million, and 22.5 million 150-bp paired-end reads were obtained from the morning, afternoon, and evening transcriptome sequencing (RNA-seq) libraries. Reads were pseudoaligned to all Prodigal-predicted genes from all genomes using Kallisto with default settings [bib_ref] Near-optimal probabilistic RNA-seq quantification, Bray [/bib_ref]. The calculated transcripts per million (TPM) were used to obtain total transcription levels for genes identified from genomic analysis as involved in S cycling in the spring.
Additional metabolic analysis. For all other non-sulfur-related functional predictions, combined GhostKOALA outputs of all genomes belonging to a certain order (for orders with 5 genomes or less; n = 206) or family (for orders with more than 5 genomes; n = 85) were checked for the presence of groups of KOs constituting metabolic pathways (https://github.com/nohayoussef/Zodletone_Metagenomics). The list of these 291 lineages is shown in . The presence of at least 80% of KOs assigned to a certain pathway in at least one genome belonging to a certain order/family was used as an indication of the presence of that pathway in that order/family. Such criteria were used for the prediction of autotrophic capabilities, as well as catabolic heterotrophic degradation capabilities of sugars, amino acids, long-chain fatty acids, short-chain fatty acids, anaerobic benzoate degradation, anaerobic short-chain alkane degradation, aerobic respiration, nitrate reduction, nitrification, and chlorophyll biosynthesis. Glycolytic and fermentation capabilities were predicted by feeding the GhostKOALA output to KeggDecoder [bib_ref] Potential for primary productivity in a globally-distributed bacterial phototroph, Graham [/bib_ref]. Proteases, peptidases, and protease inhibitors were identified using BLASTP against the MEROPS database (97) [bib_ref] FeGenie: a comprehensive tool for the identification of iron genes and iron..., Garber [/bib_ref] was used to predict the presence of iron reduction and iron oxidation genes in individual bins.
Data availability. The whole-genome shotgun project was submitted to GenBank under BioProject identifier (ID) PRJNA690107 and BioSample IDs SAMN17269717 (for the sediment metagenome) and SAMN17269718 (for the water metagenome). The individual assembled MAGs have been deposited at DDBJ/ENA/GenBank under accession numbers JAFFZZ000000000 to JAFGPI000000000. The versions described in this paper are the first versions, JAFFZZ010000000 to JAFGPI010000000. Metagenomic raw reads for the sediment and the water are available under SRA accession numbers SRX9813571 and SRX9813572. RNA-seq reads generated in this study are available under SRA accession numbers SRX9810743, SRX9810744, and SRX9810745 for the morning, afternoon, and evening samples, respectively.
# Supplemental material
Supplemental material is available online only. TEXT S1, DOCX file, 0.1 MB.
[fig] FIG 2: Novelty, rarity, and phylum-level makeup in Zodletone sediment and water communities. Genomes belonging to novel (orange), and LRD (blue) lineages are shown as a percentage of total binned genomes in the sediment (A) and the water (B) communities. The sum of novel and LRD genome percentages is shown in gray.Novel Diversity in an Early Earth Anoxic Analogue mBio as well as rare and novel lineages within Desulfobacterota(Fig. 4, Fig. S4; seeFig. S5ain the supplemental material; [/fig]
[fig] FIG 4: Sulfur cycle in Zodletone spring. (A) Diagram of sulfur transformations predicted to take place in the spring. Different sulfur species are shown in black boxes. Reduction reactions are depicted by purple arrows, oxidation reactions are depicted by red arrows [/fig]
[fig] FIG S1 ,: PDF file, 0.1 MB. FIG S2, PDF file, 0.04 MB. FIG S3, PDF file, 0.2 MB. FIG S4, PDF file, 1.8 MB. FIG S5, PDF file, 2.3 MB. FIG S6, PDF file, 2 MB. [/fig]
[table] Table S2 ;: Zod Metabat.279FIG 3 Family-level delineation for orders with 20 or more genomes. The maximum likelihood trees were constructed in FastTree (86) based on the concatenated alignments of 120 and 122 single-copy genes obtained from GTDB-TK(85). Bootstrap support values are shown as bubbles for nodes with .70% support. Families are color coded. To the right of the trees, tracks are shown for cultured status at the family level (cultured versus uncultured) and abundance in GTDB based on the number of available genomes (abundant with more than 5 genomes, rare with 5 genomes or less, and novel with no genomes in GTDB). [/table]
[table] TABLE S1 ,: XLSX file, 0.1 MB.TABLE S2, XLSX file, 0.1 MB. TABLE S3, XLSX file, 0.1 MB.ACKNOWLEDGMENTSThis work was supported by the National Science Foundation grants 2016423 (to N.H.Y. and M.S.E.) and 2016371 to M.P. [/table]
|
Aligning Method with Theory: A Comparison of Two Approaches to Modeling the Social Determinants of Health
There is increasing interest in the study of the social determinants of maternal and child health. While there has been growth in the theory and empirical evidence about social determinants, less attention has been paid to the kind of modeling that should be used to understand the impact of social exposures on well-being. We analyzed data from the nationwide 2006 Canadian Maternity Experiences Survey to compare the pervasive disease-specific model to a model that captures the generalized health impact (GHI) of social exposures, namely low socioeconomic position. The GHI model uses a composite of adverse conditions that stem from low socioeconomic position: adverse birth outcomes, postpartum depression, severe abuse, stressful life events, and hospitalization during pregnancy. Adjusted prevalence ratios and 95% confidence intervals from disease-specific models for low income (\20,000/year) compared to high income (C80,000/year) ranged from a low of 1.43 (1.09-1.85) for adverse birth outcomes to a high of 5.69 (3.59-8.84) for stressful life events. Estimates from the GHI model for experiencing three to five conditions yielded a prevalence ratio of 18.72 (9.29-35.77) and a total population attributable fraction of 78%. While disease-specific models are important for uncovering etiological factors for specific conditions, models that capture GHIs might be an attractive alternative when the focus of interest is on measuring and understanding the myriad consequences of adverse social determinants of health.
# Introduction
Socioeconomic disparities in health have been well documented since the birth of epidemiology and public health in the mid nineteenth century. The ''eras'' of epidemiology reflect the evolution of the disease-specific model, from the pursuit of a single cause (i.e., the germ theory) of infectious diseases, to the consideration of myriad causes (i.e., the web of causation) of specific conditions with the rise of chronic diseases as the major public health threat [bib_ref] Choosing a future for epidemiology: I. Eras and paradigms, Susser [/bib_ref]. And while much of the focus of epidemiology is on etiologic contributors to single specific health conditions or behaviors (e.g., cardiovascular diseases, obesity, adequacy of prenatal care), social epidemiology has taken a specific interest in the myriad health consequences of social exposures (e.g., low income or poverty, ethnic density of neighborhoods, immigration, discrimination) [bib_ref] Socioeconomic status and perinatal outcomes in a setting with universal access to..., Joseph [/bib_ref] [bib_ref] Socioconomic differences in perinatal health and disease, Mortensen [/bib_ref] [bib_ref] The effects of racial density and income incongruity on pregnancy outcomes, Pickett [/bib_ref] [bib_ref] Ethnic density effects on maternal and infant health in the Millennium Cohort..., Pickett [/bib_ref] [bib_ref] Inequality: an underacknowledged source of mental illness and distress, Pickett [/bib_ref].
Theoretical perspectives from the social sciences such as medical sociology suggest that the effects of social exposures such as socioeconomic position (SEP) are fundamental causes of poor health and do not cause a single health outcome but, rather, have a generalized impact on well-being [bib_ref] Cumulative impact of sustained economic hardship on physical, cognitive, psychological, and social..., Lynch [/bib_ref] [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref]. Evidence supporting this perspective includes the persistence and concentration of morbidity and mortality among the poor compared to wealthier populations across the centuries despite major shifts in the major causes of death during this time from infectious diseases to chronic conditions [bib_ref] Health and the life course: Why safety nets matter, Bartley [/bib_ref] [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref]. The resurgence of interest in social conditions within epidemiology in recent decades [bib_ref] The Black report and beyond: What are the issues?, Macintyre [/bib_ref] may have deepened our understanding of the relationships between social position and health but did so while remaining attached to the disease-specific analytic approach pervasive within epidemiology. The diseasespecific approach of identifying etiologic factors-both biological and social-is appropriate when a particular condition or problem is of interest, for example, to uncover the primary determinants of smoking during pregnancy to design effective interventions. But when the primary focus is on a social exposure (e.g., poverty, discrimination) and its impact on overall health or well-being, the diseasespecific model is theoretically incompatible with this research agenda. The disease-specific modeling of fundamental social conditions may mis-or underestimate their impact on health [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref].
In practice, the disease-specific model treats all potential causes (predictors) as if they had the same ontological status. However, SEP differs in several aspects from other more proximate determinants commonly studied in epidemiology, such as tobacco smoking, diet, environmental contaminants or genes. First, SEP may lead directly and indirectly to ill health through several complex pathways that change and evolve over time with the history of a given society [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref]. Social organization is thus seen as a ''distal'', ''upstream'', or ''fundamental'' force that put individuals at ''risk of risks'' [bib_ref] Health and the life course: Why safety nets matter, Bartley [/bib_ref] [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref] [bib_ref] Socioeconomic position, Lynch [/bib_ref]. Second, as noted earlier, the effects of social organization on health are nonspecific [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] The contribution of the social environment to host resistance: The fourth Wade..., Cassel [/bib_ref]. Third, the negative effects of low social position accumulate longitudinally (i.e., the longer the exposure the greater the risk) and cluster cross-sectionally (i.e., individuals at the bottom of the social scale are more likely to experience multiple adverse outcomes) [bib_ref] Health and the life course: Why safety nets matter, Bartley [/bib_ref]. These considerations suggest that the disease-specific approach of studying only one disease manifestation of social determinants prevents us from capturing the full impact of social causes on health [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref]. And while there are times when researchers with an exclusive interest on multifactoral determinants of a single health condition should rely on the disease-specific model, the growing interest in social disparities with a primary focus on social determinants of well-being would benefit from the use of alternative models that facilitate evaluation of a generalized effect of social factors on outcomes [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref].
We draw from the sociological literature which distinguishes two types of models, etiologic or disease-specific models and consequences models or models that measure the generalized impact of a particular social exposure. Models that examine the consequences or generalized impact of social exposures have as their point of departure the social cause, not the disease [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref] [bib_ref] The social determinants of health: Coming of age, Braveman [/bib_ref] [bib_ref] Reducing racial and social-class inequalities in health: The need for a new..., Syme [/bib_ref]. While the conceptual or theoretical underpinnings of the two models are quite different, the operational distinctions are quite simple and focus on the outcome variable [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref]. While the disease-specific model uses a single health outcome, models of the consequences of social exposures use a composite outcome. Use of composite outcomes is not new to the field of health [bib_ref] Composite endpoints for clinical trials: Current perspectives, Kleist [/bib_ref] [bib_ref] Latent variable regression for multiple discrete outcomes, Bandeen-Roche [/bib_ref]. In the case of the study of the consequences of social exposures, the composite outcome must be grounded in sound theory; those conditions included in the dependent variable should have solid evidence supporting its relation to the social exposure of interest. Despite its promise, models of the generalized health impact (GHI) of social exposures have been applied by social scientists in studies on mental health [bib_ref] Social structure, stress and mental health: Competing conceptual and analytic models, Aneshensel [/bib_ref] [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref] [bib_ref] The use of multiple outcomes in stress research: A case study of..., Horwitz [/bib_ref] but have not been widely applied by epidemiologists.
Our objective was to apply the (GHI) model to the study of maternal and newborn well-being using a national survey of Canadian childbearing women and to compare it with the traditional disease-specific approach of examining one outcome at a time. Reproductive health may be suitable for this endeavour since social disparities have been amply demonstrated for multiple pregnancy related outcomes.
# Methods
## Study population
The Maternity Experiences Survey (MES) is a populationbased survey conducted by Statistics Canada during 2006-2007 on behalf of the Public Health Agency of Canada. The MES target population consisted of biological mothers who were age 15 and older at the time of their babies' singleton live birth in Canada and lived with their infants at the time of the survey. A stratified simple random sample was selected without replacement, using recent births drawn from the Census 2006 sampling frame. The sample was stratified on province or territory in which the mother resided at the time of the census and on maternal age (\20 years, C20 years). Among 8,542 women selected from the frame, 8,244 were estimated to be eligible cases based on the target criteria. The questionnaire was successfully completed by 6,421 women (77.9% response rate). After applying the survey weights, which were adjusted for non-response, these women represented approximately 76,500 Canadian women. The data were collected in a 45 min computer-assisted telephone interview by professional female interviewers in English, French and 13 non-official languages. Paper versions were used when telephone interviews were not feasible. Information on postal code of the respondent was used to link the data to the 2006 Canadian census to characterize residential neighborhood. Further details of the survey design and methods have been reported elsewhere [bib_ref] The Canadian maternity experiences survey: Design and methods, Dzakpasu [/bib_ref] [bib_ref] Factors associated with perceived stress and stressful life events in pregnant women:..., Kingston [/bib_ref].
## Outcome measures
To compare the disease-specific and the social consequences models, we chose a priori several conditions that were strongly associated with high levels of deprivation and low SEP [bib_ref] Socioeconomic status and perinatal outcomes in a setting with universal access to..., Joseph [/bib_ref] [bib_ref] Intimate partner violence as a risk factor for postpartum depression among Canadian..., Beydoun [/bib_ref] [bib_ref] How much does low socioeconomic status increase the risk of prenatal and..., Goyal [/bib_ref] [bib_ref] Testing a sociomedical model for preterm delivery. Paediatric and Perinatal, Misra [/bib_ref] [bib_ref] The relationship of prenatal care and pregnancy complications to birthweight in Winnipeg..., Mustard [/bib_ref] [bib_ref] The epidemiology of hospitalized postpartum depression in New York state, Savitz [/bib_ref] [bib_ref] Violence against pregnant women: Prevalence, patterns, risk factors, theories, and directions for..., Taillieu [/bib_ref].
a. Adverse birth outcomes was a composite measure defined by the presence of low birthweight (\2,500 g), preterm birth (\completed 37 weeks) or small for gestational age (birthweight below the 10th percentile of a Canadian population-based sex-and gestational age-specific reference) [bib_ref] A new and improved populationbased Canadian reference for birth weight for gestational..., Kramer [/bib_ref]. These measures were constructed based on maternal reports of gestational age, infant sex and birthweight. The resulting rates of singleton preterm birth and small for gestational age in the survey were consistent with national surveillance data based on birth certificates. b. Postpartum Depression was assessed using the Edinburgh Post-Natal Depression Scale, a ten item screening tool to identify postpartum depression at the time of its administration [bib_ref] Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale, Cox [/bib_ref]. A score of 13 or more out of a maximum possible of 30 was used to indicate the presence of postpartum depression. Validation studies have showed that the scale can detect depression in postpartum women with 86% sensitivity and 78% specificity [bib_ref] Comparative analysis of the performance of the Postpartum Depression Screening Scale with..., Beck [/bib_ref] [bib_ref] The validation of the Edinburgh Post-natal Depression Scale on a community sample, Murray [/bib_ref]. c. Serious abuse was defined as the combination of threats and physical or sexual abuse suffered right before, during or after pregnancy. The MES contained a section to assess abuse and violence. The questions were adapted from the Violence Against Women Surveyto capture abuse and violence during the childbearing year. Threats or potential hurting acts were defined by the occurrence of direct threats of physical harm, throwing objects at them and being pushed, grabbed or shoved in a way that could have hurt respondents. Physical or sexual abuse included at least one episode of slapping, kicking, hitting with a weapon, beating, choking, use of a gun or knife, and being forced into unwanted sexual activity. Thus, serious abuse involves at least two episodes of violence of different kinds. d. Hospitalization during pregnancy was defined as an affirmative answer to the question of whether respondents stayed at a hospital overnight before labour and the birth. e. Frequent stressful life events were considered as such when respondents identified three or more stressful events (out of 13) in the 12-month period before the baby's birth [bib_ref] Factors associated with perceived stress and stressful life events in pregnant women:..., Kingston [/bib_ref].
In the disease-specific models, each of the previous conditions is considered as a separate outcome. In contrast, to be consistent with the theory underpinning the GHI models we created a compound outcome variable by counting the number of conditions experienced by respondents [bib_ref] Research in mental health: Social etiology versus social consequences, Aneshensel [/bib_ref] [bib_ref] Composite endpoints for clinical trials: Current perspectives, Kleist [/bib_ref] [bib_ref] The use of multiple outcomes in stress research: A case study of..., Horwitz [/bib_ref] , and categorizing them into 0, 1, 2 and 3-5 conditions. Exposures a. Total household income, before taxes and deductions, of all household members from all sources in the 12-month period preceding the interview. The unexposed group was considered to be composed of households with annual incomes of 80,000 dollars and higher and exposed groups were those in the income brackets of \20,000, 20,000-49,999, 50,000-79,999 dollars, and unknown income. b. Neighborhood deprivation was a contextual variable assessed by the proportion of households whose income in 2005 was below the Statistics Canada Low income after-tax cut-off (LICO-AT). The LICO-AT identifies households spending 20 percentage points more of their after-tax income than the average family in the region on food, shelter and clothing, thus leaving less income available for other expenses such as health, education, transportation and recreation. The LICO-AT cut-offs are differentiated by size of family and area of residence. Proportions of LICO-AT were computed at the dissemination area level, which is the smallest standard geographic area for which all census data are disseminated, with a population of 400-700 persons. c. Immigrant status was categorized into Canadian-born, recent immigrants (\10 years of stay in Canada), and long term immigrants (C10 years of stay).
Variables for confounder control were maternal age, parity, and place of residence. For the latter, place of residence was categorized into urban areas (Census Metropolitan Areas [CMA] and Census Agglomerations [CA], with an urban core of at least 10,000 inhabitants) and rural areas (non CMA/CA zones), following the Statistics Canada Standard Geographical Classification.. Levels of these variables are specified in [fig_ref] Table 1: Characteristics of the Maternity Experiences Survey respondents, 2006-2007, [/fig_ref].
# Analytic methods
Survey weights were used to account for the unequal probabilities of selection of respondents and thus obtain unbiased point estimates representative of the Canadian population. Special procedures for the analysis of survey data (SURVEYFREQ and SURVEYLOGISTIC) (SAS version 9.2, SAS Institute Inc., Cary, NC) were used to obtain weighted proportions and Odds Ratios (OR) with 95% confidence intervals (95% CI) using the Taylor Series method of variance estimation.
For modeling the five disease-specific conditions, the logistic model was used to compute ORs for each condition separately. To model the generalized impact of social exposures, and to take into account the multiple categories of our compound outcome, we used the multinomial model to obtain ORs for the occurrence of one, two, and three to five conditions, relative to none. The general regression equation of the multinomial model with a single predictor is given by Log(p j /pJ) = a j ? b j x, where the response levels are j = 1, …, J -1 and the baseline response category is J.
To avoid overestimating relative risks, odds Ratios were converted to Prevalence Ratios (PR) using a simple formula that provides a good approximation to estimates of the relative risk when direct estimation is not feasible [bib_ref] What's the relative risk? A method of correcting the odds ratio in..., Zhang [/bib_ref]. In the formula PR i = (P i /P 0 ) = OR i /[(1 -P 0 ) ? (P 0 9 OR i )], P i is the weighted proportion of cases in the exposure level i, P 0 is the weighted proportion of cases among the non-exposed and OR i is the OR for exposure level i. This conversion further allowed comparing the disease specific and social consequence models in terms of population attributable fractions (PAF), based on the weighted proportion of cases at each exposure level and on the adjusted PRs, as expressed in the formula PAF = (Pd i *((PR i -1)/PR i )), where Pd i is the weighted proportion of cases in stratum i and PR i is the adjusted PR in stratum i [bib_ref] Use and misuse of population attributable fractions, Rockhill [/bib_ref]. PAFs facilitated the comparison of the etiologic fraction attributed to low SEP between the diseasespecific and GHI models. Missing data were very low for most variables and therefore were not considered in the analyses, with the exception of household income, for which we created a category labelled ''Unknown'' to prevent a significant drop in the sample size.
The study was approved by the St. Michael's Hospital Research Ethics Board and by the Research Data Centre Access Granting Committee of Statistics Canada.
# Results
Among the 6,421 respondents in the MES, 406 women (6.3%) were excluded due to missing or invalid responses in at least one of the outcomes or covariates. The final sample for analyses was 6,015 (weighted N = 71,400).
About four out of 10 women had at least one adverse condition [fig_ref] Table 1: Characteristics of the Maternity Experiences Survey respondents, 2006-2007, [/fig_ref]. Among affected women, 70% had only one and seven percent three to five outcomes. The proportion of women experiencing at least one outcome decreased with advanced age, higher household income and lower neighborhood deprivation. This pattern was more evident as the number of concomitant conditions increased. Among affected women, the number of concomitant conditions increased while the proportion of immigrants decreased.
Regression models for each single condition showed moderate to strong associations between low household income and each outcome, with PR ranging from 1.4 to 5.7, after adjustment [fig_ref] Table 2: Prevalence ratios of reporting a single condition or compound health problems among... [/fig_ref]. In comparison, the multinomial model showed two types of gradients; the well-known gradient by which the lower the income the poorer the outcomes, and a new gradient by which, within each exposed income group, the PR increase with the number of conditions, reaching a prevalence ratio of 17 for women in the lowest income households having 3-5 conditions. The prevalence of 3-5 conditions was 10.36% among women living in households making \20,000 dollars (626/6,039 from [fig_ref] Table 1: Characteristics of the Maternity Experiences Survey respondents, 2006-2007, [/fig_ref] versus 0.54% among those whose household income was 80,000 or above (125/23,042 from [fig_ref] Table 1: Characteristics of the Maternity Experiences Survey respondents, 2006-2007, [/fig_ref].
In the fully adjusted model, neighborhood deprivation was not consistently associated with the outcomes.
In contrast to the findings for the socioeconomic exposures, logistic models show that being an immigrant was associated with higher risk of postpartum depression but lower risk of abuse, hospitalizations and stressful life events, particularly among recent immigrants. Multinomial models show a trend towards lower risk of concomitant adverse outcomes, particularly among recent immigrants.
Another approach to compare the magnitude of effects is to examine the Population Attributable Fraction (PAF) for the social exposures. Here we show PAFs for income. Low income PAFs ranged from 15 to 55% for single conditions based on the single outcome model [fig_ref] Table 3: Population attributable fractions [/fig_ref]. In comparison, PAFs in the composite condition model were substantially larger at 51 and 78% for 2 and 3-5 conditions, respectively.
# Discussion
To our knowledge, this is the first study to apply the GHI approach to model pregnancy related outcomes. In a representative sample of Canadian childbearing women, we found that, compared to the disease-specific model, a GHI model detects stronger effects of social position on pregnancy related outcomes. Both approaches showed the well documented gradient of decreasing risk with increasing household income. In addition, we were able to demonstrate a strong dose-response relationship using the GHI model. The stronger gradient is explained, in part, by having fewer individuals with conditions related to low SEP in the 'condition free' category of the GHI model which is not true for the disease specific models.
Immigrants also exhibited linear trends according to the number of adverse outcomes, but in contrasting directions. Even after controlling for household and neighborhood income, immigrants were less likely to experience multiple conditions, particularly recent immigrants. This observation is consistent with the ''healthy migrant effect'' and also suggestive of its loss with increasing time spent in the new country [bib_ref] Glossary: Migration and health, Urquia [/bib_ref].
Strengths of our study are the use of a high-quality nationally representative survey and the simultaneous consideration of a wide array of adverse outcomes. Such approaches are particularly relevant for research focussing on social determinants for example in the study of social disparities. Our findings are strengthened by the existence of a dose-response relationship between lower income and increasing number of adverse outcomes. Although we focused our attention on household income, which cannot capture the full complexity of social location [bib_ref] Socioeconomic position, Lynch [/bib_ref] , we also considered additional indicators, such as immigration, which was also associated with the compound outcome in a dose-response fashion. Several weaknesses exist. First, as data are self-reported, recall bias is always a possibility. However, our variables focused on the 2-year period preceding the interview and bias resulting from inaccurate recall is likely to be small. Second, the main exposure was total household income since disposable income after application of redistributive policies was not available in the survey, which would be more accurate as a measure of the material resources actually available to the households. The use of pre-tax income is likely to bias the estimates toward the null, yet we found strong and consistent associations. It is unlikely that reverse causation could explain our findings as prospective studies suggest that the cross-sectional associations between income and health chiefly reflect the influence of income on health rather than the opposite [bib_ref] Cumulative impact of sustained economic hardship on physical, cognitive, psychological, and social..., Lynch [/bib_ref] [bib_ref] Income and mortality: The shape of the association and confounding New Zealand..., Blakely [/bib_ref]. Unfortunately, we did not have life course data to fully explore the problem of reverse causation. In the case of immigrant status, it is a fixed attribute that cannot be affected by the outcomes. Third, our list of outcomes and exposures is not exhaustive and we were constrained by what we could include by the survey. With regard to outcomes, we chose a priori a limited number of known consequences of low individual and neighborhood SEP for comparison purposes but the use of a different set of outcomes may result in different effect estimates. We anticipate that, if low SEP is associated with each single outcome, the use of the GHI model would reveal similar patterns. Moreover, the GHI approach is well suited for documenting the impact of social determinants on well being but may be less appropriate for revealing the mechanisms or pathways by which social factors result in adverse health given the outcome is a composite of many variables, some of which may have unique pathways resulting from deprivation. We did not control for correlates of income, such as maternal education and marital status, because of potential overcontrol or colinearity issues. Adjustment for smoking and substance use was discarded since these are conceptualized as mediators of the relationship of interest. Finally, while we used logistic and multinomial regression, alternative approaches, such as structural equation modeling might have been employed to explore the same research questions. We anticipate that if such approaches were used, a similar set of findings would result.
Despite these limitations, our application of the GHI model provides further evidence supporting the hypothesis that the negative consequences of social position cluster among the socially disadvantaged. While socioeconomic gradients constitute one of the most robust findings in social and perinatal epidemiology, our study reveals a less known gradient towards the simultaneous occurrence of multiple adverse outcomes associated with increasing disadvantage. Income inequalities accounted for 51 and 78% of the excess risk of having 2 and 3-5 conditions, respectively, supporting their role as a fundamental cause [bib_ref] Social conditions as fundamental causes of disease, Link [/bib_ref]. While we provided evidence supporting the existence of a clustering of multiple adverse outcomes with increasing deprivation, we did not have appropriate longitudinal data to test the related hypothesis that, at a given level of lower social position, the occurrence of concomitant adverse outcomes would be higher among those who have been exposed longer or repeatedly to low SEP [bib_ref] Cumulative impact of sustained economic hardship on physical, cognitive, psychological, and social..., Lynch [/bib_ref] [bib_ref] Health and the life course: Why safety nets matter, Bartley [/bib_ref]. Such approaches might be useful for examining specific questions around social exposures such as discrimination [bib_ref] Racial discrimination and the blackwhite gap in adverse pregnancy outcomes: A review, Giurgescu [/bib_ref] or issues of deprivation such as the weathering hypothesis [bib_ref] Understanding and eliminating racial inequalities in women's health in the United States:..., Geronimus [/bib_ref] [bib_ref] Black/white differences in the relationship of maternal age to birthweight: A population-based..., Geronimus [/bib_ref]. Further research in this area will benefit from adopting a life-course perspective [bib_ref] Health disparities beginning in childhood: A life-course perspective, Braveman [/bib_ref] and analysing longitudinal datasets.
Our findings have important implications for research and practice. From an analytic perspective, the diseasespecific model underestimates the negative impact of low SEP on health. In particular, it overlooks the fact that socially disadvantaged individuals are also affected by related conditions other than the one under investigation. Thus, the broad non-specific effects of SEP on health domains may be better captured by the GHI model. From a policy perspective, the understanding that low SEP is not only independently associated with various adverse outcomes but also with their simultaneous occurrence suggests that greater health gains may be achieved if investments focus on reducing the social inequities behind the health disparities rather than on tackling proximate risk factors that may hopefully prevent one but not all negative consequences of low social position [bib_ref] Estimation of health benefits from a local living wage ordinance, Bhatia [/bib_ref] [bib_ref] Impact of income maintenance on low birth weight: Evidence from the gary..., Keherer [/bib_ref]. It is important to clarify that the GHI model does not intend to replace disease-specific research, which is the model of choice when the interest is to unveil the mechanisms and pathways for specific and well-defined health outcomes. However, when the goal is to weigh the non-specific sequels of SEP on a general domain of health, such as mental health or reproductive health, the GHI model has clear advantages.
[table] Table 1: Characteristics of the Maternity Experiences Survey respondents, 2006-2007, (weighted N = 71,395) by number of composite health conditions Measured as proportion of households in a dissemination area living at or below the Statistics Canada Low Income Cutoff (LICO) [/table]
[table] Table 2: Prevalence ratios of reporting a single condition or compound health problems among Canadian women in the maternity experiences survey ( [/table]
[table] Table 3: Population attributable fractions (PAF) for single conditions or number of concomitant conditions among women in the Canadian Maternity Experiences Survey, 2006-2007 PAF for single conditions PAF for concomitant conditions [/table]
|
Comparative acute efficacy and tolerability of OROS and immediate release formulations of methylphenidate in the treatment of adults with attention-deficit/hyperactivity disorder
Background:The main aim of this study was to compare the safety and efficacy of IR MPH administered three times daily to those of once daily OROS-MPH.Methods: Subjects were outpatient adults satisfying full diagnostic criteria for DSM-IV ADHD between 19 and 60 years of age. Data from two independently conducted 6-week placebo controlled, randomized clinical trials of IR-MPH (tid) and of OROS-MPH were pooled to create three study groups: Placebo (N = 116), IR-MPH (tid) (N = 102) and OROS-MPH (N = 67).Results: Eight-five percent (N = 99) of placebo treated subjects, 77% (N = 79) of the IR-MPH (tid) treated subjects, and 82% (N = 55) of the OROS-MPH treated subjects completed the 6-week trial. Total daily doses at endpoint were 80.9 ± 31.9 mg, 74.8 ± 26.2 mg, and 95.4 ± 26.3 mg in the OROS-MPH, IR-MPH (tid), and placebo groups, respectively. At endpoint, 66% (N = 44) of subjects receiving OROS-MPH and 70% (N = 71) of subjects receiving IR-MPH (tid) were considered responders compared with 31% (N = 36) on placebo.Conclusion:Comparison of data from two similarly designed, large, randomized, placebocontrolled, trials, showed that equipotent daily doses of once daily OROS-MPH had similar efficacy to that of TID administered IR MPH.Trial Registration:The trial of OROS-MPH was registered at clinicaltrials.gov, number NCT00181571.
# Background
Attention deficit hyperactivity disorder (ADHD) is a persistent disorder associated with high levels of morbidity and disability across the lifecycle. Although long concep-tualized as a pediatric disorder, it is now estimated to affect between 3%-5% of adults in this country [bib_ref] The prevalence and effects of adult attention deficit/hyperactivity disorder on work performance..., Kessler [/bib_ref] [bib_ref] The prevalence and correlates of adult ADHD in the United States: Results..., Kessler [/bib_ref]. The extant literature documents several similarities between adult and pediatric ADHD in terms of psychiatric comor-bidity and neuropsychological deficits, as well as a neural and genetic underpinning that supports the syndromatic continuity of ADHD across the lifecycle. Recent work has also documented the severe functional impairments associated with ADHD including educational and occupational under attainment, driving accidents, addictive behaviors and a wide range of interpersonal deficits adversely impacting all aspects of life [bib_ref] Gender effects of attention deficit hyperactivity disorder in adults, revisited, Biederman [/bib_ref] supporting the need for the identification of safe and effective treatments for this disorder in adults.
While methylphenidate (MPH) remains one of the leading pharmacological treatments of pediatric ADHD, there is a limited literature on its safety and efficacy in the treatment of adults with ADHD. In contrast to equivocal results observed in early studies that used daily doses of approximately 0.5 mg/kg of immediate release methylphenidate (IR MPH) [bib_ref] Diagnosis and treatment of minimal brain dysfunction in adults: a preliminary report, Wood [/bib_ref] [bib_ref] Methylphenidate effects on symptoms of attention deficit disorder in adults, Mattes [/bib_ref] [bib_ref] A controlled study of methylphenidate in the treatment of attention deficit disorder,..., Wender [/bib_ref] [bib_ref] Attention deficit disorders in adults, Gualtieri [/bib_ref] [bib_ref] The efficacy of 2 different dosages of methylphenidate in treating adults with..., Bouffard [/bib_ref] , clearer patterns of response were documented in studies using doses of approximately 1 mg kg of IR-MPH [bib_ref] A double-blind, crossover comparison of methylphenidate and placebo in adults with childhood-onset..., Spencer [/bib_ref] [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref]. These findings suggest that MPH is effective in the treatment of adults with ADHD when used in weight adjusted doses equipotent to those used in pediatrics.
Results from a recent, large randomized, six week, placebocontrolled clinical trial documented that osmotic release methylphenidate (OROS-MPH) at daily average doses of 1 mg/kg/day was also highly effective and well tolerated in the treatment of adults with ADHD [bib_ref] A randomized, placebo-controlled trial of OROS-methylphenidate in adults with attention-deficit/hyperactivity disorder, Biederman [/bib_ref]. Since Spencer et al [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref] administered MPH three times daily and the long acting formulation of OROS-MPH was designed to provide day-long coverage, these results suggest that a key component of efficacy observed in these studies may also be day long pharmacological coverage. Although OROS-MPH was designed to mimic the pharmacokinetic profile of IR MPH administered three times daily, is unclear if these two formulations of MPH are equally tolerated and effective since there have been no head to head comparisons conducted.
In the absence of randomized head to head comparisons, existing studies that used OROS and IR formulations of MPH could be compared if they employed similar methodology. We have conducted two six week, randomized, placebo-controlled studies of OROS-MPH and IR MPH that used nearly identical dosing and assessment methodology. The main aim of this study was to compare the safety and efficacy of equipotent doses of IR MPH administered TID to those of once daily OROS-MPH. To this end we used data from two similarly designed, large, randomized, placebo-controlled, six week trials of IR and OROS-MPH in adults with DSM-IV ADHD [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref] [bib_ref] A randomized, placebo-controlled trial of OROS-methylphenidate in adults with attention-deficit/hyperactivity disorder, Biederman [/bib_ref]. We hypothesized that once daily OROS-MPH would show similar efficacy and tolerability to that of IR-MPH administered three times a day.
# Methods
## Subjects
Subjects in both studies were outpatient adults with ADHD between 19 and 60 years of age. To be included, subjects had to satisfy full diagnostic criteria for DSM-IV ADHD based on clinical assessment and confirmed by structured diagnostic interview. There was no overlap in participation between the study samples. We excluded potential subjects if they had clinically significant chronic medical conditions, abnormal baseline laboratory values, I.Q. <80, delirium, dementia, or amnestic disorders, other clinically unstable psychiatric conditions (i.e., bipolar disorder, psychosis, suicidality), drug or alcohol abuse or dependence within the six months preceding the study, or previous adequate trial of methylphenidate. We also excluded pregnant or nursing females. The human research committee of the institutional review board (IRB) approved these studies and all subjects completed a written informed consent.
## Procedures
Randomized Trial of IR-MPH (tid) [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref] This was a double-blind, randomized, 6 week, placebocontrolled, parallel design study of MPH in the treatment of adult ADHD. Patients were randomized to MPH or placebo at a ratio of 2.5:1. Weekly supplies of MPH or placebo were dispensed by the pharmacy in identically appearing 5 and10 mg capsules. Study physicians prescribed medication under double blind conditions in three times per day dosing (7:30 am, noon and 5 pm). Study medication was titrated (forced titration) up to 0.5 mg/kg/day by weekone, 0.75 mg/kg/day by week two and 1.0 mg/kg/day by week three, in TID dosing, unless adverse effects emerged. The dose could have been increased to a maximum of 1.3 mg/kg by weeks 5 and 6 if efficacy was partial and treatment was well tolerated. Other psychoactive medications were not permitted during the protocol.
Randomized Trial of OROS-MPH [bib_ref] A randomized, placebo-controlled trial of OROS-methylphenidate in adults with attention-deficit/hyperactivity disorder, Biederman [/bib_ref] This was a double blind, randomized, 6-week, placebocontrolled, parallel design study of OROS-MPH. Patients were randomized to OROS-MPH or placebo at a ratio of 1:1. Medication was titrated to optimal response (a maximum daily dose of 1.3 mg/kg; initial dose of 36 mg). During titration to optimal response, dose was increased by 36 mg/day for only those subjects who failed to attain an a priori definition of improvement defined by a clinical global impressions scale-improvement (CGI-Improvement) score of 1 or 2 and a reduction in the Adult ADHD Investigator Symptom Rating Scale (AISRS) score larger than 30%) and who did not experience adverse effects. All doses of OROS-MPH and placebo were delivered in identically appearing tablets.
## Assessment
To assess inclusion and exclusion criteria, all subjects underwent a comprehensive clinical assessment which included a psychiatric evaluation by a board certified psychiatrist, structured diagnostic interview, medical history, vital signs, and laboratory assessments (liver function tests, complete blood count, and electrocardiogram). The structured diagnostic interview used was the Structured Clinical Interview for DSM-IV (SCID), supplemented for childhood disorders by modules (DSM-IV ADHD and conduct disorder) from the Kiddie SADSE (Epidemiologic Version).. This interview was selected because it diagnoses both lifetime and current month psychopathology and has been used extensively in clinical and research settings [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref] [bib_ref] Patterns of psychiatric comorbidity, cognition, and psychosocial functioning in adults with attention..., Biederman [/bib_ref].
To have been given a full diagnosis of adult ADHD, the subject must have: a)met full DSMIVR criteria (at least 6 of 9 symptoms) for inattentive and/or hyperactive/impulsive subtypesby the age of seven as well as within the past month (i.e. ADHD-IA, ADHD-HI and ADHD-C subjects were enrolled); b)described a chronic course of ADHD symptomatology from childhood to adulthood and c) endorsed a moderate or severe level of impairment attributed to the ADHD symptoms.
Overall severity and change in severity of ADHD was assessed with the Clinical Global Impression Scale (CGI). The CGI includes Global Severity (1 = not ill, to 7 = extremely ill) and Global Improvement (1 = very much improved, to 7 = very much worse) Scales. The Adult ADHD Investigator System report Scale (AISRS) [bib_ref] Diagnostic approaches to adult attentiondeficit/hyperactivity disorder, Spencer [/bib_ref] was used to assess each of the 18 individual criteria symptoms of ADHD in DSMIV on a severity grid (0 = not present; 3 = severe; overall minimum score = 0; maximum score = 54). To assess symptoms of depression and anxiety, we used the 17-item Hamilton Depression Scale (HAM-D, minimum = 0; max = 52) [bib_ref] A rating scale for depression, Hamilton [/bib_ref] and the Hamilton Anxiety Scale (HAM-A, minimum = 0; max = 56) [bib_ref] The assessment of anxiety states by rating, Hamilton [/bib_ref]. A global measure of psychosocial functioning (global assessment of functioning (GAF) scale) was rated according to guidelines in DSM-IV. Adverse events were elicited by spontaneous reports through open-ended questions at each visit. Weight and vital signs were obtained at each visit and an EKG was performed at baseline and endpoint. Both raters and subjects were blind to treatment assignment.
# Statistical analysis
Analyses were intention to treat (ITT) with the exception that subjects must have been assessed on drug or placebo for at least one week (92% [bib_ref] A large, double-blind, randomized clinical trial of methylphenidate in the treatment of..., Spencer [/bib_ref] and 95% [bib_ref] A randomized, placebo-controlled trial of OROS-methylphenidate in adults with attention-deficit/hyperactivity disorder, Biederman [/bib_ref] of randomized returned for at least 1 assessment). A mixedeffects model repeated measures approach was used to account for missing data in our longitudinal assessments of safety (i.e. weight and vital signs) and efficacy. Models assessing symptom improvement of the primary measure of outcome (AISRS score) were adjusted for any demographic or clinical differences at baseline between the groups and baseline AISRS score. Omnibus and pair wise comparisons were made with post estimation Wald tests such that χ 2 statistics are reported for continuous data.
Continuous and categorical data were tested with ANOVA and Pearson's χ 2 , respectively for non-longitudinal data (i.e. demographics at baseline, prevalence of adverse effects or response at endpoint, etc). Statistical significance was determined at p < 0.05. For simplicity of exposition, placebo subjects from both studies were pooled into a single placebo group. Thus, three groups were compared: Placebo (N = 116), IR-MPH (tid) (N = 102) and OROS-MPH (N = 67).
# Results
Clinical and demographic characteristics are presented in [fig_ref] Table 1: Demographic and Clinical Characteristics at Baseline [/fig_ref]. Although there were small but statistically significant differences in age between the groups (statistically significant for OROS-MPH versus placebo), each of the three groups were in their mid thirties, on average. There were no statistically significant differences in gender, ADHD age at onset, number of symptoms, or clinical impression of severity at baseline.
At baseline, there were small statistically significant but not clinically meaningful differences in global assessment of functioning and ratings of anxiety that were slightly worse in the OROS-MPH than in the IR MPH groups [fig_ref] Table 1: Demographic and Clinical Characteristics at Baseline [/fig_ref]. Also, there was a small but significant difference in the AISRS score between the three groups (F(2.282) = 4.1, p = 0.02) that was accounted for by a difference between the OROS-MPH (30.1 ± 5.9) and the placebo (32.1 ± 7.9) subjects .
There were no differences in dose at endpoint between IR-MPH (tid) and OROS-MPH (0.97 ± 0.21 mg/kg versus 0.99 ± 0.32 mg/kg; p = 0.09) but both were statistically significantly lower than placebo (1.15 ± 0.21 mg/kg; p = 0.001). Total daily doses at endpoint were 80.9 ± 31.9 mg, 74.8 ± 26.2 mg, and 95.4 ± 26.3 mg in the OROS-MPH, IR-MPH (tid), and placebo groups, respectively. Adjusting for baseline AISRS value, there were statistically significant treatment effects compared to placebo for both the IR-MPH (tid) (Wald χ 2 (6) = 40.9, p < 0.0001) and the OROS-MPH (Wald χ 2 (6) = 23.7, p = 0.0006) groups. There was not a statistically (Wald χ 2 (6) = 7.8, p = 0.3) or clinically significant difference between the IR-MPH (tid) and OROS-MPH treated subjects .
The rate of improvement according to the psychiatrist rated clinical global impression (CGI-I) for ADHD was statistically significantly higher for both IR-MPH (tid) [bib_ref] The prevalence and correlates of adult ADHD in the United States: Results..., Kessler [/bib_ref] (1) = 27.7, p < 0.001) and OROS-MPH (χ 2 (1) = 21.2, p < 0.001) groups compared with placebo , and there were no statistically (χ 2 (1) = 0.008, p = 0.9) or clinically significant differences between the two formulations of MPH. Forty percent (N = 46), 80% (N = 80), and 69% (N = 46) of the placebo, IR-MPH (tid) and OROS-MPH groups, respectively attained a 30% reduction of baseline Clinical Ratings of Improvement Clinical Ratings of Improvement. Clinical Ratings of ADHD Symptoms Clinical Ratings of ADHD Symptoms.
AISRS scores at endpoint. Both the IR-MPH (tid) and the OROS-MPH groups were statistically significantly more likely to have a 30% reduction in symptoms than placebo (p < 0.001) and were not different from one another (p = 0.1).
At endpoint, 66% (N = 44) of subjects receiving OROS-MPH and 70% (N = 71) of subjects receiving IR-MPH (tid) were considered responders compared with 31% (N = 36) on placebo (χ 2 (2) = 38.1, p < 0.001), using our a priori definition of response of much or very much improved on the CGI-I plus more than a 30% reduction in symptoms on the AISRS. Both active medication groups were statistically significantly more likely to demonstrate this level of improvement compared with placebo (p < 0.001) but not when compared to one another (p = 0.6).
The rate of adverse effects reported over the study period is presented in [fig_ref] Table 2: Adverse Effects [/fig_ref]. Both the IR-MPH (tid) and the OROS-MPH treated subjects were more likely to report dry mouth, decreased appetite, sleep difficulties and moodiness than were subjects treated with placebo [fig_ref] Table 2: Adverse Effects [/fig_ref]. There was a statistically significant greater weight loss in the IR-MPH (tid) (-2.1 ± 2.4 kg, p < 0.001) and the OROS-MPH groups (-2.8 ± 1.9 kg, p < 0.001) than in the placebo group(0.02 ± 1.7 kg); differences in weight loss between OROS-MPH and IR-MPH (tid) were small but statistically significant (p = 0.03). In addition subjects treated with OROS-MPH were more likely than subjects treated with IR-MPH (tid) to report having dry mouth and decreased appetite [fig_ref] Table 2: Adverse Effects [/fig_ref]. Complaints of GI difficulties were statistically significantly elevated in the OROS-MPH subjects only.
There were no serious adverse events reported. Specific adverse effects leading to drop out in the placebo group were irritability (N = 1), fatigue (N = 1), increased pulse/ racing heart (N = 1) and elevated blood pressure (N = 2). Adverse effects leading to dropout in the IR-MPH subjects were jitteriness (N = 2), irritability (N = 2), depression, (N = 1), anxiety (N = 1), over-focus (N = 1), headache (N = 1), insomnia (N = 1), and elevated blood pressure (N = 5).
Adverse effects leading to dropout in the OROS-MPH groups were jitteriness (N = 1), irritability (N = 3), depression (N = 1), anxiety (N = 1), increased pulse/racing heart (N = 2) and elevated blood pressure (N = 1).
Changes from baseline to endpoint on cardiac measures are presented in [fig_ref] Table 3: Cardiac Parameters and Baseline and Endpoint [/fig_ref]. As expected, there were small but statistically significant differences in diastolic blood pressure and pulse between both active treatment groups and placebo subjects. There were no differences between the IR-MPH (tid) and the OROS-MPH groups with the exception of a smaller increase in pulse in OROS-MPH group that, although statistically significant (p = 0.049), would be of limited clinical significance. Outliers analysis of cardiovascular data revealed that significantly more subjects treated with OROS-MPH and IR-MPH (tid) attained clinically significant elevated heart rate (>100 bpm: 6 (9%) and 3 (3%) vs 1 (1%) respectively, χ 2 (2) = 8.5, p = 0.02; maximum value observed was 114 bpm). Other cardiovascular outliers were not significantly more likely to occur in subjects treated with active medication than with placebo: systolic blood pressure (>140 mmHg: 5 (8%) and 14 (14%) vs. 6 (5%), χ 2 (2) = 5.5, p = 0.06; maximum value observed was 156 mmHg), diastolic blood pressure (>90 mmHg: 2 (3%) and 8 (8%) vs. 3 (3%), χ 2 (1) = 4.1, p = 0.1; maximum value observed was 102), or QTC interval (>460 msecs: 1 (2%) and 1 (2%) vs. 2 (2%), χ 2 (1) = 0.1, p = 0.9; maximum value observed was 488 msecs). For none of these clinical outliers was the difference between IR-MPH (tid) and OROS-MPH statistically significant.
# Discussion
Comparison of data from two similarly designed, large, randomized, placebo-controlled, 6-week trials of IR and OROS-MPH in adults with DSM-IV ADHD, showed that once daily OROS-MPH had similar efficacy and tolerability to that of equipotent daily doses of IR-MPH administered three times per day. These results confirmed the study hypothesis that once daily OROS-MPH is as effective as three times daily IR MPH at similar doses. The find- ing that the same daily dose of once daily OROS-MPH was as effective as that of IR-MPH (tid) is consistent with the pharmacokinetic profile of OROS-MPH that was formulated to deliver three doses of IR MPH across the day with once daily administration.
The magnitude of response observed in this study is consistent with results reported in over 250 controlled trials of stimulants in pediatric ADHD using similar weight corrected daily doses of 1 mg/kg/day. Moreover, the response rate of 70% observed for both formulations of MPH in adults with ADHD at daily doses of 1 mg/kg are much higher than the average 52% response rate reported in early studies of MPH using average daily doses of 0.5 mg/ kg [bib_ref] Diagnosis and treatment of minimal brain dysfunction in adults: a preliminary report, Wood [/bib_ref] [bib_ref] Methylphenidate effects on symptoms of attention deficit disorder in adults, Mattes [/bib_ref] [bib_ref] A controlled study of methylphenidate in the treatment of attention deficit disorder,..., Wender [/bib_ref] [bib_ref] Attention deficit disorders in adults, Gualtieri [/bib_ref] [bib_ref] The efficacy of 2 different dosages of methylphenidate in treating adults with..., Bouffard [/bib_ref]. Taken together, these results support the hypothesis that daily doses of 1 mg/kg/day are needed to attain a robust response to MPH in the treatment of adults with ADHD.
Despite the relatively high daily doses of MPH used, both formulations of MPH were well tolerated as manifested in the high completion rate (~ 80% for both formulations) and absence of serious adverse events. While subjects treated with OROS-MPH formulation more commonly reported dry mouth, decreased appetite, and GI com-plaints than subjects treated with IR-MPH (tid), only 14% of each group dropped from the study due to adverse effects. Treatment with both formulations of MPH was associated with similarly small but statistically significant effects in diastolic blood pressure and pulse relative to placebo. Although abnormal values in cardiovascular parameters (maximum values observed were 114 bpm for heart rate, 156 mmHg for systolic blood pressure, 102 mmHg for diastolic blood pressure, and 488 msecs for QTc interval) were not associated with medical complications in this study, adults with ADHD should be monitored for changes in blood pressure and weight loss when receiving treatment with MPH.
The documentation that once daily OROS-MPH is as effective as IR MPH administered three times daily has important clinical implications. Because adults with ADHD tend to be forgetful, once daily administration of OROS-MPH may lead to better compliance. Also, since the delivery of MPH via OROS is gradual and the MPH cannot be as easily extracted from the OROS tablet this formulation has a lower abuse potential than IR-MPH. This is so because the abuse potential of MPH is thought to be due to the rapid onset of blockade of the presynaptic dopamine transporter (DAT) in the brain [bib_ref] Variables that affect the clinical use and abuse of methylphenidate in the..., Volkow [/bib_ref]. Thus, the more gradual rise of plasma MPH concentration with the OROS formulation of MPH could lead to a slower onset of blockade of the presynaptic DAT and therefore a lower risk for detection of euphoria [bib_ref] Variables that affect the clinical use and abuse of methylphenidate in the..., Volkow [/bib_ref]. Spencer et al [bib_ref] A PET Study Examining Pharmacokinetics, Detection and Likeability, and Dopamine Transporter Receptor..., Spencer [/bib_ref] compared the relationship between peripheral and central pharmacokinetic properties of IR and OROS-MPH and their impact on abuse liability potential using C-11 altropane and positron emission tomography (PET) and found further support of this hypothesis.
The results of this study should be viewed in light of methodological limitations. Our criteria for ADHD were somewhat stricter than that DSM-IV. We required full childhood onset, whereas DSM-IV only requires some significant symptoms in childhood. The diagnosis of ADHD and assessment of ADHD symptoms also relied on selfreport. Although the usefulness of self reports of ADHD symptoms are limited in pediatric samples, there is evidence that self reports of adults with ADHD correspond very well to corroborating histories provided by parents and spouses [bib_ref] Use of self-ratings in the assessment of symptoms of attention deficit hyperactivity..., Murphy [/bib_ref]. Also, assessment of adverse effects relied on spontaneous reports and such open-ended questioning of adverse events limits the precision of characterization that may be found in structured rating scales. While the current study is reassuring, strong inferences about tolerability and safety that require a much larger sample size to fully assess the occurrences of rare adverse events. Future work that can directly assess any improvements in treatment compliance associated with once-daily treatments for ADHD are also needed.
# Conclusion
Despite these limitations, comparison of data from two similarly designed, large, randomized, placebo-controlled trials showed that OROS-MPH has similar efficacy and tolerability as that of IR-MPH administered three times per day. These results indicate that MPH is highly effective for the treatment of adults with ADHD when delivered in appropriate doses and dosed across the day. More work is needed to evaluate whether these short-term benefits extend to the long-term.
[table] Table 1: Demographic and Clinical Characteristics at Baseline [/table]
[table] Table 2: Adverse Effects [/table]
[table] Table 3: Cardiac Parameters and Baseline and Endpoint [/table]
|
Proteomic approaches for characterizing renal cell carcinoma
Renal cell carcinoma is among the top 15 most commonly diagnosed cancers worldwide, comprising multiple subhistologies with distinct genomic, proteomic, and clinicopathological features. Proteomic methodologies enable the detection and quantitation of protein profiles associated with the disease state and have been explored to delineate the dysregulated cellular processes associated with renal cell carcinoma. In this review we highlight the reports that employed proteomic technologies to characterize tissue, blood, and urine samples obtained from renal cell carcinoma patients. We describe the proteomic approaches utilized and relate the results of studies in the larger context of renal cell carcinoma biology. Moreover, we discuss some unmet clinical needs and how emerging proteomic approaches can seek to address them. There has been significant progress to characterize the molecular features of renal cell carcinoma; however, despite the large-scale studies that have characterized the genomic and transcriptomic profiles, curative treatments are still elusive. Proteomics facilitates a direct evaluation of the functional modules that drive pathobiology, and the resulting protein profiles would have applications in diagnostics, patient stratification, and identification of novel therapeutic interventions.
# Background
Renal cell carcinoma (RCC) is among the top 15 most commonly diagnosed cancers in men and women, with an estimated 73,000 newly diagnosed cases in the United States and 403,000 newly diagnosed cases worldwide. RCC is a heterogeneous disease comprised of three major histopathological subtypes: clear cell renal cell carcinoma (ccRCC), papillary renal cell carcinoma (pRCC), and chromophobe renal cell carcinoma (chRCC); in addition to more rare and benign subtypes that include collecting duct RCC, papillary adenoma, hybrid oncocytic chromophobe, multilocular cystic clear cell carcinoma and oncocytomas. The predominant histology of RCC is ccRCC accounting for ~ 75% of cases, followed by pRCC, which is further divided into two distinct subtypes accounting for 15% of cases, and chRCC which accounts for ~ 5% of cases. To understand the underlying molecular alterations that drive RCC oncogenesis, The Cancer Genome Atlas (TCGA) has performed extensive genomic, epigenomic and transcriptomic profiling of ccRCC, pRCC, and chRCC. Unique to ccRCC was the dysregulation of chromosome 3p and associated genes, including the ubiquitous loss of the hypoxic signaling regulator VHL and the genes BAP1, PBRM1, SETD2, which have been shown to follow loss of VHL and are linked to aggressive disease. Type I pRCC was observed to have a high frequency of chromosome 7, 16, and 17 gain and genomic aberrations involving MET, while Type 2 pRCC showed a strong association with CDKN2A dysregulation. Genomic alterations in chRCC were quite distinct relative to ccRCC, pRCC, and even other cancer types, observing substantial chromosome copynumber loss, with the majority of tumors displaying loss of chromosomes 1, 2, 6, 10, 13, and 17. Subsequent pan-renal studies have attempted to delineate the unique and shared features of each histology, revealing a high degree of molecular heterogeneity across RCC tumors that could be used to identify up-regulated cellular pathways, immune-related signatures, and patient survival with respect to each histology. Moreover, this molecular information, when combined with histopathological examination, has also provided insight into the nephron cell types from which RCC originates, revealing ccRCC and pRCC arise from proximal tubule epithelial, while chRCC is associated with the distal tubule epithelium. Overall, the genomic and transcriptomic profiling of RCC has provided unique insight into the molecular basis of renal oncogenesis that is complementary to histopathological examination.
Genomic profiling techniques, such as next generation sequencing (NGS), can identify the altered DNA sequences resulting from somatic mutations that are associated with carcinogenesis, as well as serve as a potential screening tool for early diagnosis. However, sequencing DNA and cataloging the altered basepair profiles provides little information related to the functional consequences of a mutation, and some have questioned the utility genomic-based approaches for identifying actionable targets for therapeutic intervention. Similarly, although transcriptomic analyses offer a high-throughput methodology for assessing gene expression via the measurement of protein-coding mRNA transcript abundance, transcriptomic profiling cannot fully elucidate the functional modules that regulate cellular processes. Mounting evidence has indicated that paired mRNA transcript and protein expression is not robustly correlated in normal tissues or in tumors, highlighting the rationale for prioritizing protein measurements for an accurate representation of biological systems. Furthermore, mRNA transcript abundance offers minimal insight into posttranslational modifications (PTMs) of proteins, such as phosphorylation, glycosylation, or ubiquitination, which serve as additional layers of gene regulation by impacting protein function, stability, protein-protein interactions, and cellular location. Finally, considering that the majority of drug targets are proteins and protein-based analysis is the most common technique utilized in the clinical setting, delineating target proteins of interest have direct translational applications. Seeking to address the shortcomings of genomic and transcriptomic profiling, proteomic technologies offer a comprehensive methodology for determining differential global protein and PTM abundance, facilitating a direct analysis of the functionally active molecules and cellular mechanisms dysregulated in the disease state.
There have been a myriad of studies that have utilized proteomic technologies to explore the protein profiles of tissues and biological fluids in an effort to identify the differentially expressed proteins associated with RCC . In this review, we summarize the details of these studies, focusing our discussion on the samples analyzed, experimental design, proteomic techniques employed, and the results reported in the context of RCC characterization at the protein level. In addition, our review encompasses several emerging approaches and future directions that can be explored to provide further insight into altered protein profiles associated with RCC.
## Tissue profiling
Analysis of tumor tissues offers the most direct method to identify dysregulated protein expression or protein PTM profiles resulting from aberrant genomic events in RCC. Including additional examination of benign or normal adjacent tissues (NATs) facilitates comparative proteomic profiling to identify differentially expressed proteins, with the end goal of delineating the aberrant cellular processes associated with RCC or elucidating potential disease protein biomarkers. As a result of these features, the majority of the studies discussed in this review have examined tumor tissues from RCC patients; with these studies representative of various proteomic technologies, RCC histologies, and disease severity. ccRCC tissue profiling ccRCC accounts for the majority of RCC cases and as a result studies focused on profiling ccRCC are far more numerous than studies examining other RCC histologies. Potentially confounding the summarization of proteomic experiments in this review is the classification of ccRCC as RCC in several studies, resulting from RCC initially being divided into two major histologies prior to 2004 and ccRCC being viewed as conventional RCC. For the purposes of this review, we will consider studies describing the characterization of kidney tumors as RCC to be ccRCC unless additional information related to the genomic background, pathological classification, or anatomical location would suggest a different histology.
Two-dimensional electrophoresis (2-DE) is an early proteomic technology, utilizing a dual separation approach based on a protein's isoelectric point and molecular weight on a polyacrylamide gel. The individual spots visualized by this methodology are then excised and subjected to mass spectrometry analysis to identify and annotate the protein spots, with differences in protein spot intensity between experimental conditions inferring differential protein abundance and identification of dysregulated proteins. Multiple studies have leveraged this technique to identify differential expressed protein profiles between ccRCC and NAT samples, followed by immunoblotting or immunohistochemical validation in independent samples.
Perroud et al., identified 31 proteins that were differentially expressed between ccRCC tumor tissues and NATs derived from four patients, validating the overexpression of HSP27 (HSPB1) and PKM2 via immunoblotting. Pathway analysis indicated most of the differential Established proteomic approaches used for investigation of RCC biological samples. Comparative two-dimension electrophoresis (left) assesses protein abundance differences based on spot intensity (Quantitation), with subsequent mass spectrometry analysis identifying the proteins from the excised spot (Identification). Label-free quantitation (middle) entails mass spectrometry analysis of individual samples, with peptides identified at the MS2 level (Identification), and peptide abundance based on peak intensity determined at the MS1 level (Quantitation). Protein abundance is inferred from peptide abundance measurements. In some label-free quantitation-based experiments, spectral counting is employed, wherein protein abundance is inferred by the number of mass spectrometry spectra generated for each peptide derived from the precursor protein. Isobaric labeling (right) methods involve the labeling of peptides derived from individual samples with mass tags that include reporter ions and mixing of samples prior to mass spectrometry analysis. Peptide Identification and Quantitation information is obtained at the MS2 level in the same spectra. Protein abundance is inferred from peptide abundance measurementsexpressed proteins were annotated as metabolic-related, prompting their investigation of the metabolite profiles in pooled urine samples, which revealed increased abundance of glycolysis by-product sorbitol. Raimondo et al.also showed metabolic profiles are altered in ccRCC, and found immune response to be up-regulated in tumor tissues. Validation focused on the proteins that were upregulated, with the authors verifying overexpression of ANXA2, LGALS1, PPIA, and FABP7. ANXA2 overexpression was primarily associated with the plasma membrane, and prompted several follow-up studies by this same group focused on delineating differences in protein abundance of plasma membrane domains between tumors and NATs. In addition to identifying several other proteins of interest, including VDAC1, BSG, and THY1, these studies described a strategy of annotating the cell surface proteome of renal cancer cells that may have applications in identifying novel therapeutic targets. In another study, comparison of nine cases of ccRCC (paired tumor tissue and NAT) revealed the differential expression of 34 proteins, including the protein NDRG1, which was found to be elevated in ccRCC tissues, with subsequent validation experiments via IHC revealing the protein's nuclear localization (versus membrane localization) was associated with a favorable prognosis, and functional assays revealing a potential tumor suppressor role of NDRG1 in ccRCC. There are several inherent disadvantages of 2-DE profiling, including issues related to reproducibility, solubilization and visualization of hydrophobic proteins, low-throughput, and a narrow dynamic range of protein expression detection. To circumvent these drawbacks, Multidimensional Protein Identification Technology (MudPIT) was developed, which relies on proteolytic digestion of proteins into peptide products and subsequent fractionation prior to mass spectrometry analysis. This "shotgun proteomic" approach facilitates several strategies for quantitation, including label-free quantitation (LFQ), and isobaric reporter tag labelling of peptides. Reports leveraging LFQ to profile ccRCC have mirrored earlier 2-DE studies in terms of the altered cellular pathways resulting from ccRCC pathobiology, albeit identifying far more differentially expressed proteins between tumors and NATs. Atrih et al.used LFQ to examine the disparate protein profiles between ccRCC tissues and NATs, revealing almost 600 proteins to be differentially expressed. Using IHC, the authors validated the expression pattern of two proteins, CORO1A and ADFP, which were shown to have increased in abundance in ccRCC samples. Interestingly, the authors found that CORO1A was not overexpressed in renal cancer cells, but instead in the infiltrating lymphocytes localized in the tumor microenvironment, whereas ADFP overexpression was associated with tumor cells. This latter result highlights one drawback of bulk tissues analyses that many proteomic studies employ, specifically the loss of spatial information related to protein expression, as well as the mixing of different cell populations during sample homogenization.
Isobaric labelling, including technologies such as isobaric tags for relative and absolute quantitation (iTRAQ) and tandem-mass-tag (TMT), offer a strategy to reduce the time needed for data acquisition via sample multiplexing. Following proteolytic digestion, samples are labelled with a mass tag reporter ion that when subjected to fragmentation in the mass spectrometer, enables the measurement of peptide (and subsequently, protein) abundance across multiple samples. An early report applied this approach, identifying a total 324 differentially expressed proteins between a tumor and NAT sample, with 99 robustly detected in replicate analyses. Included in their iTRAQ experiment were two "control" samples, a transitional cell carcinoma (TCC) case and a kidney tissue sample from a patient with end stage glomerulonephritis, facilitating a comparison of other renal disorders and ccRCC. Although the authors examined their dataset in relation to several previous publications and validated several proteins, including VIM, SERP-ING1, NNMT, and LDHA via IHC or immunoblotting, they did not report the expression pattern of these proteins in their "control" samples, thus missing the opportunity to identify ccRCC-specific protein candidates. In another study that incorporated sample fractionation for deeper profiling, a larger cohort of patients were examined using the iTRAQ approach, wherein a pooled reference sample was included to link multiple iTRAQ sample plexes. In this proteomic-based discovery phase, the authors identified 55 proteins differentially expressed between ccRCC and NAT samples. From this list of proteins, the authors prioritized candidates that had gene ontology (GO) annotation as "secretory" and could serve as potential serum biomarkers, validating five proteins-ENO1, HSPE1, HSPB1, AHNAK, and LDHA via immunoblotting and/or IHC tissue microarray (TMA) profiling. Mirroring their proteomic results, the proteins ENO1, HSPB1, LDHA, and AHNAK were elevated in ccRCC, while HSPE1 was down-regulated. Of note, the authors did not verify any of these candidates in serum samples, missing the opportunity to link tissue protein profiles to circulating serum protein profiles.
Moving beyond comparisons of ccRCC tumor and NATs tissues only, several studies have sought to delineate protein expression patterns associated with disease severity, including stratifying patient samples based on tumor grade, tumor stage or comparisons of primary and metastatic lesions. Perroud et al.leveraged a LFQ approach to characterize the proteome of fifty kidney tissues comprising equal number of normal tissues and Fuhrman grades 1-4 of ccRCC tissues. Prior to their evaluation of archival formalin-fixed paraffin embedded (FFPE) tissues, the authors first assessed whether the FFPE process introduced any bias in the proteomic data, finding only a few proteins displayed varied abundance between sample matched frozen tissues. Following this preliminary result, the authors then examined the 50 FFPE tissues, wherein they identified 105 proteins that showed significant differential abundance across the four grades and normal tissues. The protein NPM1, which has been shown to impact nucleolar morphology, showed increasing expression from normal tissue and G1 tumors to G2, G3, and G4 tumors. Considering the Fuhrman grading system is based on the nucleolus morphology and sizing, the increasing expression of NPM1 in higher grade tissues served as a robust, positive control. Pathway analysis of the differentially expressed proteins revealed higher expression proteins involved in glycolysis (PGK1, ALDOA-C, ENO1) xenobiotic metabolism (ALDH4A1, ALDH1A1, ALDH9A1), and down-regulation of proteins associated with apoptosis (AIFM1) were associated with higher tumor grade and are representative of the cellular mechanisms involved in ccRCC progression. Interestingly, the authors found that proteins clusters could differentiate G1/2 tumors and G3/4 tumors, mirroring the disparate prognosis of patients and morphological differences observed between low grade ccRCC versus high grade ccRCC. Another study examining proteome profiles associated with tumor stage in ccRCC using a modified 2-DE approach (2D-DIGE), with the resulting protein profiles visualized using a principle component analysis (PCA) showing a clear separation between nonneoplastic kidney tissues and pT1, pT2, and pT3 staged tumors. Similar to increasing tumor grade, increasing ccRCC tumor stage associated with alterations in metabolism, specifically up-regulation of glycolysis with a parallel decrease in TCA cycle protein expression. The authors highlighted and validated the differential expression of several select proteins, PHB, PRDX3, and S100A9, using IHC and immunoblotting. While PRDX3 downregulation was associated with increasing tumor stage, S100A9 and PHB increased in abundance and were also associated with tumor grade. An interesting feature was detected in the immunoblot results for PHB, wherein the expected 30 kD band showed minimal difference across the four experimental conditions, while a slightly larger band at 40 kD was discriminatory and may suggest a post-translation modification on PHB increased in higher stage disease. To investigate the protein alterations associated with metastatic disease, Laird et al.used an immunofluorescence antibody-based approach to profile a TMA composed of 138 ccRCC tissues, 14 papillary tissues, 103 renal vein tumor thrombus (VTT) tissues, and 69 metastatic tissues. Evaluating a panel of antibodies comprised of MKI67, TP53, VEGFR1 (FLT1), VEGFD, SNAIL (SNAI1), and SLUG (SNAI2) paired with a technology called Automated Quantitative Analysis (AQUA) that allows for an unbiased assessment of protein expression, the authors showed all the proteins, except for VEGFD, were significantly elevated in metastatic disease relative to VTT and primary lesions. Although VEGFD did not show an association with disease severity, both VEGFD and VEGFR1 expression were prognostic in primary tumors, with elevated expression associated with reduced cancer specific survival. Less clear was any discriminatory value for these markers between ccRCC and papillary primary tumors, which would be relevant in the context of ccRCC displaying elevated angiogenic signaling resulting from a pseudo-hypoxia phenotype due to VHL inactivation. Using an iTRAQ-based approach, Masui et al.characterized six cases of ccRCC tissues and patient-matched NATs, and six unmatched metastatic tissues, identifying 29 proteins that were differentially expressed between primary and metastatic lesions. Three proteins, LGALS1, PFN1, and YWHAZ, were selected for validation via immunoblot in the same patient cohort and IHC using TMAs derived from an independent cohort, which showed increased expression in primary ccRCC relative to NATs, and increased expression in metastatic lesions relative to primary tumors. An additional assessment was done to determine the prognostic value of these three candidate proteins which showed PFN1 was associated with a poor prognosis. In a follow-up study, the same team performed a deeper investigation of this proteomic dataset to identify the cellular pathways associated with metastasis, revealing the metabolic pathways glycolysis, pyruvate metabolism, and the TCA cycle to be highly dysregulated. They validated the proteomic results using a PCR array of metabolic genes and found a high degree of concordance for select pathways between the two datasets. Overall, these studies reveal that many of the cellular processes that are dysregulated in early stage ccRCC tumors, in particular metabolism-related pathways, are maintained and amplified during disease progression.
With the advent of high-throughput technologies for genomic, transcriptomic, and proteomic characterization, several studies have begun to incorporate multiple levels of molecular information to facilitate "multi-omic" profiling of tissue samples. Integration of multiple omic data types allows for researchers to begin to link genomic alterations to observed phenotypes, as well as identify aberrant regulatory mechanisms that may not be detected with a single omic dataset. Leveraging a proteotranscriptomic approach, Neely et al.performed a series of comparative transcriptomic and proteomic experiments to identify altered gene profiles associated with the molecular phenotype of ccRCC. The authors identified 342 proteins that were differentially expressed between ccRCC and NATs samples, with subsequent pathway analysis results further supporting ccRCC to be defined as disease with significant metabolic dysfunction. A subset of the samples were previously characterized at the transcript level, and integration of both data types revealed ~ 70% of the transcripts and proteins showed a positive correlation, ~ 94% of the differentially expressed mRNA-protein pairs showed a positive correlation, although the authors noted this correlation was not linear. In this same report, the authors also identified several candidate proteins associated with aggressive ccRCC, including CFL1, PFN1, NNMT, and ALDOA that were found to be elevated in Stage 4 disease. Interestingly, in this particular dataset, CFL1, PFN1, and ALDOA were found to be decreased in early stages relative to NATs. In another study, FFPE ccRCC tissues and patient matched NATs that were previously characterized at the transcript level, were characterized at the protein level, wherein the authors found disparate cellular pathways over-represented at each biological domain. At the protein level, metabolic pathways were robustly shown to be altered, whereas transcriptomics indicated immune response pathways were more significantly impacted. The antigen presentation pathway was observed to be dysregulated in both datasets, with CD74 showing significant differential expression at both the mRNA and protein levels. Specific to the proteomics results was the observed down-regulation of SIRT3, which was not detected at the transcriptomic level, and highlights the added benefit of employing complementary technologies for molecular characterization. Recently, our lab led the Clinical Proteomics Tumor Analysis Consortium (CPTAC) effort to interrogate ccRCC tumors using a proteogenomic approach, comprehensively profiling and integrating genomic, epigenomic, transcriptomic, proteomic, and phosphoproteomic measurements. In addition to delineating novel features of ccRCC at the genomic level, including the identification of a subset of tumors displaying a high degree of chromosome instability and chromosomal translocation involving 3p as a prominent event, we also showed there are four major subtypes of ccRCC defined by disparate tumor microenvironment compositions and proteomic signatures. These four subtypes-CD8+ Inflamed, CD8− Inflamed, VEGF Immune Desert, and Metabolic Immune Desert-were not only predicted to have distinct responses to immune checkpoint and anti-VEGF therapies, but also predicted patient survival.
In an effort to expand therapeutic options available to ccRCC patients in the clinical setting, we leveraged our phosphoproteomic results to prioritize the identification of phospho-substrates of kinases with current FDAapproved inhibitors, revealing signaling associated with MAPK/ERK, PI3K/AKT/mTOR, and G2/M cell cycle stalling to be elevated in ccRCC tumors and potential druggable targets. Global proteomic results revealed the up-regulation of immune response pathways and hypoxic signaling, as well as the previously described alterations in cellular metabolic pathways. Interestingly, we showed that the down-regulation of oxidative phosphorylation was only detectable at the protein level, and not the mRNA level, which not only highlights the added benefit of multi-omic profiling to gain unique insight into ccRCC biology, but also cautions against using mRNA levels as a surrogate for protein expression. The continual observation of altered metabolic profiles in ccRCC prompted Zhang et al.to examine alterations lysine succinylation between ccRCC and NATs, which has been previously linked to metabolism regulation and is an abundant PTM on mitochondrial proteins. Employing a TMT quantitation strategy and antibody-based enrichment of succinylated peptides, the authors integrated global and succinlyome measurements, finding many of the abundances changes of succinylated peptides were the result of global protein changes. Together, these results show the utility of multi-omic profiling using complementary technologies to deepen our understanding of ccRCC, and the unique insight that can be gained when integrating measurements at various levels of gene expression and regulation.
IHC serves as an orthogonal methodology to validate protein expression patterns and enables researchers to determine the spatial distribution of a protein across different areas of the tumor and surrounding tumor microenvironment. Although IHC multiplexing technologies are emerging, the number of proteins that can be simultaneously profiled is still relatively low. MALDI Mass Spectrometry Imaging (MALDI-MSI) is an established technique that greatly expands the types of biomolecules that can be directly measured, including proteins, peptides, lipids, glycans, and small molecules, and links these molecular profiles to specific anatomical locations. Oppenheimer et al.used MSI to assess the tumor margin of ccRCC resected tissues from 75 patients and determined the distribution of protein abundance across four distinct regions: tumor, tumormargin, NAT margin, and NAT. Their results robustly showed the up-regulation of S100A10, S100A11, MIF, and TMSB10 in tumors, while mitochondrial proteins such as COX6C, COX5B, COX7C, UQCRG, and CYCS were down-regulated, mirroring previous studies examining lysed bulk tissue samples. Interestingly, the authors found many of the down-regulated proteins were also dysregulated in the tumor margins of the resected tissue, suggesting that molecular alterations precede changes in cell morphology, a feature that has been well established in other tumor types. Utilizing MALDI-MSI, another study examined the spatial distribution of protein and lipid profiles across ccRCC tumor and NAT samples to delineate differential abundance of these biomolecules. Proteins significantly up-regulated in tumors included TMSB10, TMSB4X, HBB, HBD, HBG1, and FABP7, while COX5B, COX5A, FABP1 were downregulated. This particular cohort of patient samples included information related to disease recurrence, facilitating the identification of protein profiles that could distinguish non-recurrent and recurrent ccRCC. Only two proteins-DEFA1 (elevated in tumors) and LYZ (significantly decreased in tumors)-could robustly differentiate disease recurrence, whereas the authors found lipid profiles were more discriminatory for disease recurrence in addition to discriminating tumor versus NAT. To determine if MALDI-MSI could delineate protein expression patterns associated with tumor grade, Stella et al.examined 14 samples from 13 patients with various intratumor histological grades. The resulting spectra showed a clear separation of tumor and NAT regions via PCA, in addition to distinguishing normal cortex tissue from normal medulla tissue. Only G1 tumors could be robustly separated from G4 tumors. Further investigation of the features that could discriminate G1 and G4 tumors found the identified peaks were derived from VIM, which was elevated in G4 tumors and histone subunits H2A and H4, which were elevated in G1 tumors. These studies show the complementary results of MSI approaches to current proteomic characterization techniques, with the added benefit of gaining information related to the spatial distribution and localization of molecules of interest across resected tissues samples.
## Multiple rcc histology tissue profiling
With substantial evidence of the disparate genomic backgrounds of the various histological subtypes of RCC, identifying discriminatory protein features would be relevant in confirming pathological annotation, as well as understanding the contrasting prognosis associated with each RCC subtype. Towards this goal, an early report from Lichtenfels et al.first profiled ccRCC and paired NAT samples to identify differentially expressed proteins using a 2-DE approach. A total of 248 proteins were found to be differentially expressed, with three proteins-CALB1, GSN, and FABP3-selected for examination in a TMA panel comprised of ccRCC (n = 40), pRCC (n = 31), chRCC (n = 16), and renal oncocytic lesions (n = 9) with corresponding NATs. CALB1 staining was weak in many of the RCC tissue samples, but robustly detected in oncocytic and NAT regions indicating negative CALB1 expression was a robust marker of malignant renal disease. GSN expression in NAT was primarily detected in distal tubule and collecting duct cells, while proximal tubule and glomeruli cells were negative. In RCC tissues, almost a third of ccRCC and pRCC tissues were negative for GSN, while > 50% of chRCC and oncocytic tissues were positive for GSN. This pattern of nephron cell type specificity for GSN expression seemed to reflect the cells of origin from which chRCC and oncocytomas arise, specifically distal tubule and collecting duct cells. FABP3 showed the most promise for discriminating the various RCC histological subtypes, with FABP3 expression highest in chRCC and oncocytic RCC lesions, negative in pRCC tissues, and heterogeneous in ccRCC, with approximately 40% of samples showing positive staining. Valera et al.used a very similar experimental design, wherein they utilized 2-DE based approach to identify the differential protein patterns associated with ccRCC, pRCC, chRCC, and renal oncocytomas relative to corresponding NATs. The authors highlighted proteins that showed significant differential expression between NATs and ccRCC (HSPB1, TPI1, HBB, APOA1, and PRDX2), chRCC (SOD1, RAD23B, and SERPINA1), and oncocytic lesions (ENOA1), whereas no differentially expressed proteins were identified between pRCC and patient matched NATs. HSPB1 and TPI1 were selected for IHC validation in ccRCC tissues, with the staining pattern concordant with the proteomic analysis results; however, the authors missed the opportunity to further assess the expression of these proteins in other RCC lesions. Considering several of the proteins (ENOA1 and SOD1) have been associated with ccRCC in previous reports, the utility of several of the candidate markers may be negligible for discriminating various RCC histologies. Using an antibody panel composed of MKI67, CRP, CA9, HIF1A and HIF2A, Abel et al.sought to assess the expression patterns of these proteins in a TMA comprising ccRCC (n = 42), pRCC (n = 11), and chRCC (n = 1) tumors and NATs, and link these results to clinicopathological outcomes. Although CA9, HIF1A and HIF2A were elevated in tumors, the observed differential expression was not determined to be significant. CRP was shown to be reduced in tumors, whereas MKI67 was elevated and found to associate with disease recurrence. Similar to the previous report, there was no assessment of differential abundance of this antibody panel across the three RCC histologies; albeit with only one case of chRCC, interpretation of any potential results would be minimal. Oncocytomas and chRCC are both thought to arise from the distal tubule epithelium, and Drendel et al.profiled oncocytoma and chRCC FFPE samples to identify markers that would discriminate between benign and malignant RCC lesions. PCA showed the resulting protein profiles were clearly separated, with 51 proteins and 59 proteins enriched in oncocytoma and chRCC tissues, respectively. Two proteins-LAMP1 and ITGAV-were selected for verification in an independent cohort of oncocytomas and chRCC, with ITGAV1 expression to exhibit strong staining in oncocytomas, whereas LAMP1 staining was robustly detected in chRCC.
Two studies explored the phosphorylation patterns in RCC using antibodies targeting phosphorylated substrates in RCC cell lines, and primary and metastatic tissues. Lin et al.constructed a TMA comprised of 70 ccRCC, 40 pRCC, and 18 chRCC primary tumors, 22 metastatic RCCs, and 24 NAT and probed phospho-substrates associated with mTOR signaling (AKT-S473, mTOR-S2448, and p70S6-T389). Almost all of the malignant samples showed strong positive staining for these various phospho-substrates relative to NATs indicative of constitutive mTOR signaling and providing robust evidence for the rationale of selecting mTOR inhibitors for treatment of RCC. Of note, the authors did acknowledge the mixed efficacy of targeting mTOR in the clinical setting and need to identify other cellular pathways that may contribute to mTOR inhibition resistance. Haake et al.leveraged a pan-tyrosine enrichment approach to identify activate signaling cascades in ccRCC and pRCC tumors and RCC cell lines. Their results showed PTK2 phosphorylation to be ubiquitous across all RCC samples analyzed, with a tyrosine kinase inhibitor screen consisting of 63 compounds showing those targeting PTK2 to have the most robust response in vitro. Interestingly, when evaluating differential expression of phospho-substrates in ccRCC and pRCC tumors, EGFR-Y1197, ERK2-Y187, ERK1-Y204, and TENC1-Y483 were elevated in ccRCC, while DDR1-Y792/6 and PP4B-Y849 were elevated in pRCC. These latter results suggest disparate signaling pathways activated in ccRCC and pRCC, respectively, and provide rationale for exploring targeting these signaling cascades as a therapeutic intervention in these RCC subtypes.
## Biological fluid profiling
Tissue biopsy sampling is a routine procedure that enables clinicians to obtain a small portion of the malignant tissue to investigate the histopathological and molecular features for diagnostic and prognostic information. However, tissue biopsies are considered an invasive procedure and due to the limited sampling area, may not be fully representative of the tumor, which are known to be heterogeneous. Utilizing biological fluids as a liquid biopsy would offer a minimally invasive strategy for repeat sampling to monitor disease progression and possibly be more representative of the molecular features associated with tumorigenesis. With the ultimate goal of identifying candidates for disease diagnosis and early detection in RCC, many studies have applied proteomic approaches to characterize serum/ plasma and urine protein profiles.
## Serum/plasma profiling
Of the multiple biological fluids found in the human body, blood is often considered to be the ideal source for protein candidates. The relatively non-invasive, simple procedure involved in specimen collection would circumvent the challenges and expertise involved in tissue biopsy acquisition, and the network of arteries, veins, and capillaries that come in contact with organs offers a means for proteins that are secreted, shed, or otherwise released by tumor tissues to be circulated. Several inherent challenges in characterizing protein profiles in serum or plasma, include the dynamic range of protein concentration, which spans up to 12 orders of magnitude, as well as the presence of a small number of highly abundant proteins that mask more low abundant proteins. To address this, a variety of strategies have been performed to investigate and identify differentially abundant proteins related to RCC in plasma or serum samples, including immunodepletion to remove highly abundant plasma proteins (i.e. albumin, transferrin, haptoglobin) and protein pre-fractionation. One study used 2-Dimensional Image Converted Analysis of Liquid chromatography mass spectrometry (2DICAL) to profile plasma samples from twenty RCC patients and 20 healthy controls. The 2DICAL strategy is equivocal to shotgun proteomic approaches, and the resulting profiles showed that FN1 was elevated in RCC patient plasma relative to controls. Considering many proteins that are secreted or localized to the plasma membrane are N-linked glycosylated, thus having a higher chance of being shed into the extracellular space, multiple studies have included N-linked glycoprotein enrichment strategies for biomarker detection. Gbormittah et al.characterized global proteome, N-linked glycoproteome, and N-glycome plasma profiles of RCC patients before and after nephrectomy using a combined approach of immunodepletion and lectin enrichment. The author's global proteomic results revealed several proteins (i.e. HSPG2, CD146, VCAM1) associated with metabolic processes, immune response, and various signal transduction pathways were all reduced following surgical intervention. Glycoproteomics identified another subset of proteins that were reduced after nephrectomy, such as APOB, LGALS3BP, and FN1, while glycomics indicated sialylation and high mannose glycan structures were associated with pre-treatment RCC plasma profiles. In another study, the authors focused on delineating alterations in the serum proteome associated with early stage ccRCC using a paired immunodepletion and iTRAQ approach. Pooling serum samples from ten ccRCC patients and ten healthy controls, they identified 30 differentially abundant proteins, with HSC71 (HSPA8) showing greatest abundance difference. The authors then validated HSC71 expression using ELISA, profiling the serum of ccRCC patients, healthy controls, and patients with other urological diseases such as angiomyolipoma of the kidney, benign prostatic hyperplasia, urinary tract infection, and urolithiasis, showing HSC71 was elevated in ccRCC patient serums relative to healthy controls and non-ccRCC patients. An independent study examined the serum of a larger cohort of patients that included twenty-nine early stage ccRCC patients, 20 patients with transitional cell carcinoma, 24 patients with benign kidney neoplasia, 18 healthy controls, and eight patients with prostate cancer. A total of 74 proteins were found to have differential abundance between ccRCC and healthy controls serum samples, and 27 proteins that were differentially abundant between ccRCC and the other three groups. Leveraging the results from the TCGA characterization of ccRCC, the authors sought to link their serum profiles to tissue gene expression profiles, identifying 11 proteins, including C1QB, C1QC, ANXA1, LYZ, S100A9, and SERPINA4, that were differentially abundant in both datasets and were associated with RCC tumor stage and grade. An important caveat of many of these studies is that few have directly linked the resulting plasma/serum profiles to RCC tissue signatures. An exception was an early report that described a combined tumor-plasma proteome analysis on a single patient, examining tumor tissue, NAT, and pre-operative plasma sample. To identify candidates of interest, the authors prioritized proteins that met four criteria: (1) identified in tumor tissue, (2) not identified in NAT, (3) identified in plasma, and (4) higher abundance in tumor relative to plasma. This approach and filtering criteria identified eight proteins of interest-CDH11, PKM (KPYM), VCAM1, CDH5, DDX23, WWC1, CHD4, and NCOA6-with subsequent validation of the presence of CDH5, CDH11, DDX23, and PKM via immunoblotting in the plasma of the same patient and four others. Although samples analyzed were from only one patient in the initial proteomic characterization, the overall experimental design is ideal for the identification of potential candidate protein biomarkers. This approach would not only determine the differential expression of proteins between tumor and NAT, but also identify tumor-related proteins that are detectable in the circulation.
Several reports have focused on examining "biological trash" in circulating profiles, specifically focusing on examining low molecular weight endogenous peptide fragments. Peptidome profiling is thought to offer a higher degree of sensitivity and specificity relative to other approaches, relying on the aberrant biological activity of disease-related proteases, enzymatic reactions, and degradation products. Using C18-functionalized beads to enrich peptide fragments paired with MADLI-TOF data acquisition, Gianazza et al.profiled the serum of eight-five ccRCC patients, 92 controls, and 29 patients with histologically defined non-ccRCC. Clustering patients into three groups: malignant tumors, benign renal masses, and healthy controls, resulted in the identification of 5 peptides that were discriminatory for the three groups. Incorporating the results of subsequent ESI-LC-MS-MS analysis allowed for the identification of the endogenous peptides, revealing peptides from SDPR and ZYX to be decreased and peptides from SRGN and TMSL3 to be increased in the serum of ccRCC patients. Huang et al.also performed peptidome profiling of serum samples derived from RCC patients and healthy controls, identifying 19 peptides that were differentially abundant between the two groups. Of these, four peptides showed high specificity for discriminating RCC and controls, with three trending downward in abundance, and one trending upward. Highlighting one drawback of peptidome profiling, only two of these four peptides could be identified, derived from the proteins CUBN (decreased) and APOA1 (increased), respectively. In an attempt to identify peptidome profiles associated with RCC, Kodera et al.examined plasma samples from RCC patients before and after nephrectomy. Although the authors were able to find a peptide reduced in abundance following surgical intervention and show it was specific to RCC relative to bladder cancer, the authors were unable to annotate the protein it was derived from. In a larger cohort that examined serum samples from 64 healthy controls and 78 ccRCC patient, including 20 that had pre-and post-nephrectomy serum samples, Yang et al.identified 24 peptides that were differentially abundant across all groups. Three peptides were found to up-regulated in ccRCC that then returned to levels similar to healthy controls following nephrectomy, derived from the proteins RBP6, TUBB, and ZFP3. Together, these studies suggest peptidome profiling has the potential to elucidate candidates of interest; however, the minimal overlap between independent studies and lack of validation in independent cohorts suggest this methodology warrants further development.
## Urine profiling
The kidneys function to balance electrolyte levels, regulate blood pressure, as well as filter circulating blood to remove waste, resulting in urine profiles mirroring the physiological status of an individual. Relative to plasma, urine has a narrower dynamic range of protein concentration, and the reduced abundance of proteins, such as albumin, transferrin, and haptoglobin, enables the detection of lower abundant proteins. With respect to malignancies associated with the kidney, prostate, and bladder, urine is a proximal biological fluid that may offer a richer source of proteins of interest relative to blood. Albeit, the composition of urine, which includes proteins, urea, inorganic salts, and other biomolecules, presents a unique challenge for sample processing prior to proteomic characterization; however, a myriad of techniques have been developed to isolate or enrich proteins from urine, including analytical ultracentrifugation, precipitation, ultrafiltration, and tip-based approaches. Sandim et al.examined the urine of ccRCC patients, grouped by prognosis (good versus poor) and healthy controls. Pooling the urine samples of each group and then performing ultrafiltration as a proteomic-compatible preparatory step, the authors applied a multi-faceted approach that included 1-DE, 2-DE, and LFQ to identify differentially abundance proteins across the three sample classifications. Qualitative assessment via 1-DE revealed the proteins CO3, FIBG, MGAT4A, and APO1A to only be detected in ccRCC samples, while CDH13, AMYA, and APOD were only identified in the urine of healthy controls. Quantitative assessment via LFQ or 2-DE profiling revealed the increasing expression of APOA, FN, HP, and MGAT4A in controls, good prognosis ccRCC samples, and poor prognosis samples, with a concordant decrease in abundance of the proteins KNG1, UMOD, APOD, UBC, CD59, and HSPG2. Chinello et al.stratified ccRCC patients based on the degree venous infiltration of RCC tumors, as assessed by Computed Assisted Tomography, generating three distinct groupings: vascular infiltration, renal vein infiltration, and renal vein thrombosis. Three proteins-UMOD, RALA, and CNDP1-displayed decreased expression proportional to RCC infiltration, while 26 proteins (i.e. HP, LUM, CRNN, ANXA2) were increased in abundance in the pooled patient samples. The authors also examined plasma protein profiles from these same patients, and unlike the urine proteome alterations, no proteins showed concordance with the degree of increasing vein infiltration, while two were inversely correlated-APOA1 and K2C1. When interpreting this latter result, the authors acknowledged a high degree of overlap between the datasets derived from different biological fluids in terms of differentially abundant proteins, but the loss of discriminatory proteins for disease severity in plasma samples may be related to the loss of kidney function or architecture during renal oncogenesis. Santorelli et al.also sought to identify urinary proteins that could differentiate disease severity, focusing on alterations in the N-linked glycoproteome profiles. Employing the N-glyco-FASP technique, which uses lectin enrichment of glycopeptides prior to mass spectrometry analysis, the authors examined ccRCC patients with early stage disease (pT1) and late stage disease (pT3), as well as healthy controls. Generating three patient urine pools and precipitating the protein component, the resulting profiles showed a trend of protein expression associated with ccRCC stage. Three proteins-CD97, COCH, and P3IP (PIK3IP1)-showed elevated abundance in the urine of low-stage patients relative to healthy controls, and then increasing abundance in urine from low stage to high stage. Proteins found to be decreased in abundance in low stage ccRCC relative to healthy controls included APOB, FINC, CERU, CFAH, HPT, and PLPT; however, these proteins levels were slightly elevated in the urine of pT3 ccRCC patients, albeit still lower than levels in healthy controls. In the latter study, the authors highlighted the benefit of glycoprotein enrichment, which facilitated the detection of several proteins that might have otherwise not been identified and quantified. Other studies have sought to differentiate the urine protein profiles of ccRCC patients from those of other renal disorders, including patients with oncocytomas or hereditary VHL mutations. Mandili et al.used a 2-DE approach to identify proteins that could discriminate patients with sporadic ccRCC from those with VHL disease (VHLD) and healthy controls, respectively. The authors performed several paired analyses to elucidate the differentially expressed proteins between healthy controls, patients with sporadic ccRCC, and VHLD patients, including subsets of those VHLD patients with and without a history of ccRCC. While the authors identified proteins that differentiated sporadic ccRCC and VHLD patients from healthy controls, as well as sporadic ccRCC patients from VHLD patients with distinct ccRCC diagnoses backgrounds, the authors chose to validate two proteins-A1AT and APOH-which were found to be elevated in the urine of VHLD-ccRCC-positive patients relative to the other three groups. Considering both ccRCC and VHLD share a similar genotype profile (loss of VHL), delineating differences in urine proteome profiles would be challenging. However, the unique expression pattern of the proteins highlighted by the authors (A1AT and APOH) might suggest a disparate progression of diseases between VHLD patients with ccRCC and those with sporadic ccRCC. Another report focused on identifying proteins that would differentiate healthy controls, patients with ccRCC, and patients with oncocytomas. In a discovery phase, the authors pooled urine samples from healthy controls, oncocytomas, and ccRCC patients with progressive disease and non-progressive/early stage disease (pT1a, tumor size ≤ 4 cm).
A total of 131 proteins showed differential abundance between healthy controls and ccRCC patients with early stage disease, while 71 proteins were differentially abundant between ccRCC patients with early stage disease and oncocytomas. The authors sought to verify the abundance profiles of several proteins in an independent cohort using parallel reaction monitoring (PRM), finding concordance of increased abundance of GLRX, CST3, SLC9A3R1, HSPE1, FKBP1A, and EEF1G in early stage patients relative to healthy controls, and increased abundance of C12orf49 and EHD4 in early stage disease urine samples relative to oncocytomas. When comparing early stage urine samples to those from patients with higher staged disease (progressive ccRCC group), five proteins were elevated EPS8L2, CHMP2A, PDCDPI6, CNDP2, and CEACAM1, with authors finding the combined urinary abundance of EPS8L2 and another protein, CCT6A, to have prognostic value in ccRCC.
In addition to investigation of urinary proteome in ccRCC, peptidome profiling has also been examined. With evidence of disparate endogenous peptide profiles between plasma and urine, potentially resulting from the physiological activity of the kidney, exploration of the peptidome in urine is warranted. Frantzi et al.using capillary electrophoresis coupled with mass spectrometry analysis (CE-MS) profiled forty RCC cases and 68 non-diseased controls; the latter group included normal controls, non-RCC reference patients, and patients with pre-disposing factors towards RCC. In this training phase, the authors revealed 86 peptides (40 which were subsequently identified) with differential abundance between the two groups. In a validation cohort that comprised of RCC cases (n = 30), non-diseased controls (n = 46), and additional patient groups consisting of diabetic nephropathy (n = 195), focal segmental glomerulosclerosis (n = 54), membrane glomerulonephritis (n = 65), systemic lupus erythematosus (n = 46), IgA nephropathy (n = 126), vasculitis (n = 121), cardiovascular disease (n = 33), and bladder cancer (n = 219), the authors found their 86 peptide classifier was discriminatory for RCC against all groups except for vasculitis (64% specificity) and bladder cancer (76% specificity). Interestingly, when exploring the identified peptides, the authors found the resulting profiles to be reflective of the physiological dysfunction of the RCC microenvironment, including reduced abundance of > 1.4 kDa collagen fragments due to extracellular matrix remodeling in renal carcinogenesis and increased abundance of peptides from plasma proteins due to renal function dysregulation. Chinello et al.also examined urinary peptidome profiles derived from healthy controls, ccRCC patients, histologically defined non-ccRCC patients, and renal benign masses. In the discovery phase, the authors found 12 urinary peptides that could discriminate patients with malignant disease from those with benign disease. In an independent cohort, the 12 peptide panel displayed 87% specific and 76% sensitivity for discriminating malignant and benign/control patients, with identification showing peptide fragments from the proteins SCTM1, UROM, MEP1A, KPB1, OSTP, and FIBA to be increased in abundance in malignant urine samples. In a follow-up study, the same group sought to link peptidome profiles to clinicopathological features such as grade, stage, and tumor size. Examining urine samples from ccRCC patients and healthy controls, the authors identified 15 peptides that associated with tumor size and 9 that were differentially abundant relative to controls. A total of 26 peptides associated with tumor stage, including 15 that discriminated ccRCC and controls, and 5 peptides that associated with tumor grade, 4 of which were differentially abundant. Subsequent identification of the peptides found that a peptide fragment from C1RL was elevated in the urine of ccRCC patients, but showed decreasing expression in the urine of patients with higher grade disease, while GAPDHS, which was also reduced in ccRCC urine samples, showed increasing abundance in higher grade disease. Several stage associated peptides included fragments from FIBA and NOTCH2, with the latter showing no differential abundance between ccRCC and control urine samples. Overall, urine is a rich source of candidates of interest for RCC, however, the lack of any FDAapproved urinary markers for RCC suggests more work is to be done to prioritize candidates that may offer diagnostic and discriminatory benefit in the clinical setting.
## Profiling other biological fluids and sources
In addition to blood and urine profiling, several other biological sources have been characterized using proteomics. Interstitial fluid not only serves as transport medium for secreted proteins, nutrients and waste materials between cells and capillaries, but tumor interstitial fluid (TIF) could serve as a rich source of candidate markers due to the proximity of this biological fluid to the tumor. Teng et al.profiled the TIF obtained from ten patients with ccRCC and matched NATs. With previous reports have indicated that TIF contains a significant portion of highly abundant plasma proteins, the authors employed immunodepletion prior to analysis, resulting in the identification of 539 proteins, including 138 with differential abundance between ccRCC and NAT TIF samples. GO annotation revealed many of the proteins increased in abundance in ccRCC TIF localized to the plasma membrane, and included proteins previously identified in more distal biological fluids such as urine or plasma/serum. This supports the hypothesis that the TIF proteome is primarily comprised of shed or secreted proteins that are eventually found in the circulation. Eight proteins (NNMT, ENO2, TSP1, CD14, LGALB1, TBG (SERPINA7), ANXA4, and FTH1) were selected for validation, with increased abundance verified via immunoblotting, selective reaction monitoring (SRM), or ELISA in TIF samples or patient-derived serum samples. SRM showed robust concordance of increased expression of all selected protein in TIF samples, and increased abundance of CD14, TBG, and TSP1 (THBS1) in ccRCC serum samples relative to a healthy control serum samples, while ELISA showed elevated abundance of ENO2 and TSP1 in ccRCC patient-derived serum. Using fluid from a renal cyst, Minamida et al.leveraged a 2-DE approach to identify differentially abundant proteins profiled in cyst fluid derived ccRCC tumors compared to cyst fluid derived from NAT. The authors identified over 200 proteins across the samples, and selected only one protein, YWHAB, for verification via immunoblotting due to its previous lack of association with RCC. Increased abundance of YWHAB was robustly observed in ccRCC cyst fluid and the urine from RCC patients, whereas exploration of tumor tissue and serum showed equal abundance. Validation in an independent cohort of urine from RCC patients (n = 89) and healthy controls (n = 76) via ELISA, showed elevated expression in RCC urine samples and a corresponding decrease in the urine of RCC patients following nephrectomy. The authors did mention several caveats regarding the utility of the candidate protein, including the lack of detection of YWHAB in the urine of all RCC patients, detection of YWHAB in the urine of patients with other cancer types, and limitation of the marker as an indicator of malignancy, but not a marker to assess response to treatment.
Extracellular vesicles (EVs) are secreted microvesicles that have been shown to have a role in proximal and distal intercellular communication. EVs are comprised of several classes, including exosomes (30-100 nm) and ectosomes (100-1000 nm), with reports describing disparate mechanisms of cellular release into the extracellular space. Several studies have shown that EVs are functionally active and can modulate recipient cell phenotypes via the transfer of proteins and nucleic acids (mRNA, miRNA, and DNA). Seeking to identify potential EV-based protein candidates for RCC, Raimondo et al.performed a comparative proteomic analysis of exosomes derived from the urine of RCC patients and healthy controls. After verifying the presence of several EV-positive markers and morphological characteristics using electron microscopy, the resulting protein profiles were assessed using a qualitative approach, determining the disparate detection of select proteins in either RCC patient derived exosomes, or healthy control derived exosomes. Ten proteins were selected for validation, with proteins CD10, EMMPRIN, DPEP1, SDCB1, and AQP1 decreased in abundance in RCC-derived exosomes, and CP, MMP9, PODXL, CAIX, and DKK4 increased in abundance. Another report described the development of ex vivo models of ccRCC and NATs, and subsequent enrichment of EVs secreted from both tissue types. In this manner, the authors could circumvent the challenge of delineating tumor-derived EVs from the total populations of EVs in a biological fluid, as well as examine a more clinically relevant model. Proteomics identified proteins that were only detected in one pathological EV condition, as well as 397 proteins that were differentially abundant. The protein, AZU1, showed the highest fold-change between RCC-derived EVs and NAT-derived EVs, and displayed increasing abundance in advanced ccRCC. Validation in the EVs derived from the sera of patients showed elevated levels in ccRCC relative to sera from healthy controls. Together, these studies are representative of the novel insight that can be gained by examining other sources that are not routinely collected in the clinical setting.
## Future directions
As we look to the future, there are still several aspects of RCC-related biology that have yet to be explored at the protein-level, as well as several unmet clinical needs. The continual utilization of established proteomics technologies, and the application of those currently in development will be the first step in addressing some of these areas that were not discussed previously in this review. Despite the multitude of studies examining the protein profiles of ccRCC, studies focusing on rarer histological subtypes are less frequent, and large-scale, deep proteomic characterization of hundreds of RCC cases that include all the various histologies have yet to be carried out. Although it may be of interest to identify the disparate protein profiles associated with different histological subtypes of RCC for diagnostic information, identifying shared cellular pathways could potentially open the door for therapeutic approaches that can be utilized for patients with different RCC etiologies. Additionally, as more and more large-scale studies are performed for distinct cancer types, developing methodologies and bioinformatics approaches that allow for cross-platform comparisons would begin to link the common molecular features of histologically and anatomically distinct cancers. From these results, we can begin to use rationalized, precision-based therapies currently approved for one cancer type as viable therapeutic options in other cancer types. Several reports have already attempted this undertaking using publicly available datasets, including exploration of TCGA somatic, transcriptomics, and RPPA data, CPTAC gene expression data, and peptidome profiles, however, the results of these pan-cancer efforts are still preliminary. One approach that may find more utility as larger and larger sample sets are being investigated is data-independent data acquisition (DIA), which achieves similar, if not more, depth of proteomic characterization, with the added benefits of robust sample-to-sample quantitation and reduced instrument time needed for analysis. Previous reports have used DIA to characterize tissues, serum, and urine from RCC patients, revealing similar profiles of dysregulated protein and cellular pathways expression compared to more traditional proteomic data acquisition approaches (i.e. data-independent data acquisition).
Another area in RCC biology that has yet to be fully examined at the protein level is an assessment of intratumor heterogeneity (ITH). It is well-established that individual tumors are heterogeneous, and previous reports characterizing ccRCC biopsies obtained from the same tumor have revealed genomic alterations thought to be mutually exclusive occurring in subclones in distinct anatomical regions of the tumor. To date, there have been no large scale proteomic assessments of ITH in RCC, thus information related to how individual subclones within the same tumor may influence protein expression is currently unknown. Determination of the heterogeneity of genetic, and subsequent proteomic profiles would aid in our understanding as to why patients develop disease recurrence or resistance to therapeutic intervention, as well as the selective intratumor/microenvironment influences that result in altered protein expression. Moreover, delineation of the degree of protein-level heterogeneity would confirm the validity of using tissue-based biopsies in the context of patient stratification for prognostic information and selection of targeted therapies. IHC is routinely utilized to visualize the abundance and spatial distribution of proteins, however, this technique does not allow for a full characterization of the proteome in a single analysis. As noted previously, tissue biopsies are routinely used for histopathological examination, and these sample sources may also prove useful for proteomic characterization. A more precise technique of tissue sampling is laser capture microdissection (LCM), which enables the isolation and separation of distinct cell types (e.g. epithelial, fibroblasts, immune cells) to reduce the degree of heterogeneity associated with resected tumors. Several "proof of concept" reports have shown the feasibility of pairing LCM and proteomic profiling to gain insight into the biological variations of distinct cell populations, albeit requiring the pooling of multiple samples to obtain sufficient material for analysis. Potentially illustrated in these studies is the major hurdle in ITH-based proteome characterization, specifically the limited amount of material available from individual tumors available for proteomic analysis. As genomic-and transcriptomic-based single cell analyses become more widespread, complementary mass spectrometry-based proteomic approaches that enable a relatively deep proteomic characterization of tissues using minimalsample input or single cells will be further developed, and will be applicable to explore this area of RCC biology.
# Conclusions
Proteomic technologies offer a comprehensive method for characterizing the functional biomolecules that regulate cellular processes, and determination of the aberrant protein expression patterns that are impacted by the disease state. The multitude of studies highlighted in this review are representative of the myriad of proteomic approaches that have been developed and leveraged to gain insight and a deeper understanding of RCC, and reflect the advancement in the field in terms of sample preparation strategies, instrumentation, and integration with other data types. Although there are several unmet clinical related to RCC, including discriminating various subtypes at the protein-level and an assessment of RCC intratumor protein expression, proteomics will continue to offer a complementary, yet robust technology for disease characterization.
|
Novel compound heterozygous mutations of MTHFR Gene in a Chinese family with homocystinuria due to MTHFR deficiency
Background: Homocystinuria due to methylenetetrahydrofolate reductase (MTHFR) deficiency is a rare autosomal recessive disorder. The purpose of this study is to expand the mutation site of the MTHFR gene and provide genetic counseling for this family.Methods:A couple came to our hospital for pre-pregnancy genetic counseling. We collected the family history and detailed clinical information, then performed whole-exome sequencing, and analyzed the pathogenicity of the candidate mutations.Results:We found that the father of the proband had homocystinuria, the proband and his brother had low blood methionine levels at birth, and the brain MRI showed brain dysplasia. The third fetus was found to have a broadened triangle of the bilateral ventricle at 19 weeks of pregnancy. The compound heterozygous variants of c.602 A > C (p.His201Pro) and c.1316T > C (p.Leu439Pro) of the MTHFR gene in the first three fetuses were found by whole-exome sequencing. The heterozygous c.602 A > C variant of the MTHFR gene is a novel missense variant that has been submitted to the ClinVar with Variation ID 992,662.Conclusion:In consideration of the clinical phenotype, family history, and result of genetic testing, we speculated that both patients may have homocystinuria due to MTHFR deficiency. Homocystinuria due to MTHFR deficiency caused by compound heterozygous mutations composed of the MTHFR gene in this family may be associated with cerebral atrophy and cerebral dysplasia. The novel compound heterozygous mutations broaden the mutation spectrum of the MTHFR gene and enhance the application of genetic counseling and carrier screening in rare diseases.
# Introduction
Homocystinuria due to methylenetetrahydrofolate reductase (MTHFR) deficiency (OMIM: 236,250) is a rare autosomal recessive disorder [bib_ref] Clinical pattern, mutations and in vitro residual activity in 33 patients with..., Huemer [/bib_ref] [bib_ref] Mutation update and review of severe methylenetetrahydrofolate reductase deficiency, Froese [/bib_ref]. The Human MTHFR gene (OMIM: 607,093) is located on chromosome 1p36.22 and mainly consists of 12 exons(NM_005957.4) [bib_ref] Mutation update and review of severe methylenetetrahydrofolate reductase deficiency, Froese [/bib_ref]. MTHFR can reduce 5,10-methylenetetrahydrofolate to 5-methylenetetrahydrofolate which acts as a methyl donor for the methylation of homocysteine (Hcy) to methionine (Met) [bib_ref] Structural basis for the regulation of human 510-methylenetetrahydrofolate reductase by phosphorylation and..., Froese [/bib_ref]. The enzymatic reactions require flavin adenine dinucleotide (FAD) as a prosthetic group and nicotinamide adenine dinucleotide phosphate (NADPH) as an electron donor [bib_ref] Mice deficient in methylenetetrahydrofolate reductase exhibit hyperhomocysteinemia and decreased methylation capacity, with..., Chen [/bib_ref]. Both the folate and methionine cycles are essential products for the provision of cellular activity. Folate is the major cellular carrier in single carbon units, which is an essential material for the synthesis of purines and thymidine monophosphate. In the methionine cycle, the methylation of homocysteine generates methionine, which could further convert to S-adenosine methionine as a vital methyl donor for DNA, RNA, and proteins. The folate and methionine cycles intersect with the enzyme of MTHFR. Therefore, the deficiency of MTHFR restrains the methylation of Hcy to Met, resulting in an abnormally high level of total Hcy (tHcy) and a low level of Met in plasma. In this condition, homocystinuria could be caused by the disorder of folate metabolism and its FAD responsiveness [bib_ref] Functional characterization of missense mutations in severe methylenetetrahydrofolate reductase deficiency using a..., Burda [/bib_ref] [bib_ref] Insights into severe 5,10-methylenetetrahydrofolate reductase deficiency: molecular genetic and enzymatic characterization of..., Burda [/bib_ref].
Up to now, there are 135 cases with MTHFR pathogenic gene variants in the Human Gene Mutation Database (HGMD), most of them are missense mutations [bib_ref] Functional characterization of missense mutations in severe methylenetetrahydrofolate reductase deficiency using a..., Burda [/bib_ref]. Patients with homocystinuria due to MTHFR deficiency often suffer from diverse symptoms, like mental retardation, myelopathy, ataxia, and spasm. However, severe early-onset usually develops in infancy, and children with severe homocystinuria due to MTHFR deficiency often show diverse phenotypes, including eating difficulties, hypotonia, neurocognitive disorders, ventricular dilatation, hydrocephalus, brain atrophy, microcephaly, and epilepsy [bib_ref] Phenotype, treatment practice and outcome in the cobalamin-dependent remethylation disorders and MTHFR..., Huemer [/bib_ref].
In this study, we reported a family case with homocystinuria. To determine the likely pathogenic gene variants in the family pedigree, we performed whole-exome sequencing in this study. We speculate that the compound heterozygous variants of c.602 A > C (p.His201Pro) and c.1316T > C (p.Leu439Pro) of the MTHFR gene may be new genetic variants associated with homocystinuria due to MTHFR deficiency.
# Patients and methods
## Patients
This study was conducted in strict accordance with the principles of the Helsinki Declaration. The patients and their families signed consent for using their data, and the work was approved by the ethical review committee of our hospital.
The proband (II-1) was a 2-month-old female infant, who was admitted to the hospital with impaired consciousness, paroxysmal shortness of breath, and vomiting (occurred 2 times within 12 h of observation). Neonatal screening results showed 6.98 µmol/L of Met (reference value 7.5-45 µmol/L), 0.22 µmol/L of Met/Cys (reference value 0.5-4 µmol/L), and 0.03 µmol/L of Met/Leu (reference value 0.06-0.24 µmol/L). The chest radiograph (CR) showed bronchitis. Computed tomography of her head revealed an enlarged bilateral frontotemporal sub-arachnoid space, slightly enlarged lateral ventricles, left lateral ventricle subependymal cyst, and bilateral maxillary sinusitis. Upper gastrointestinal angiography indicated gastric volvulus. The symptoms did not improve after 1 month of treatment, as evidenced by the examination of the head magnetic resonance imaging (MRI) revealing traffic hydrocephalus, bilateral frontotemporal apical space enlargement, and a left lateral ventricle subependymal cyst. After the following treatment for another month, the examination of the head MRI revealed a left ventricle subependymal hemorrhage and cerebral atrophy. The patient eventually died as a consequence of multiple symptoms, possibly including bronchopneumonia-induced type I respiratory failure and atelectasis, communicating hydrocephalus, and cerebral atrophy.
The second child (II-2) was a 2-month-old male infant who was admitted to the hospital after choking on milk for 1 week and showed shortness of breath during the day. CR revealed a patchy-like high-density shadow in both lungs. Head MRI T2WI showed increased white matter signal, widening and deepening of the cerebral cortex, thinning of the cortex, and widening of subdural space, indicating cerebral dysplasia [fig_ref] Figure 1: Clinical features and genetic sequencing data of the patients [/fig_ref]. Inherited metabolic diseases determination report showed 7.93 µmol/L of Met (reference value 8-35 µmol/L), 0.18 µmol/L of Met/Phe (reference value 0.2-0.6), and 242.58 µmol/L of Glu (reference value 45-200 µmol/L). Chromosomal analysis revealed a normal karyotype (46, XY). After one month, the patient died of multiple symptoms, including severe bronchopneumonia, cerebral dysplasia, bronchopulmonary dysplasia, and moderate anemia.
The third child (II-3) was a fetus at 19 weeks of gestation, of whom MRI examination showed bilateral lateral ventricle trigonometry enlargement. The parents of the fetus decided to choose a medical abortion to avoid birth defects.
The fourth male child (II-4) was delivered by cesarean section at 40 weeks of gestation. The newborn weighed 4060 g and was diagnosed with macrosomia. There was no perinatal abnormality (such as hypoxia and asphyxia) at birth. The level of homocysteine in the postnatal peripheral blood was 10.6 µmol/L (reference value 0-15 µmol/L). Four months later, the homocysteine level was found to be elevated to 16.4 µmol/L (reference value 0-15µmol/L) but the levels of complete blood count, vitamin B12, folic acid, and ferritin were all normal. The father of the proband had a slight increase in peripheral blood homocysteine (Hcy: 27.6 µmol/L) and no other obvious clinical manifestations. It was improved after folic acid and vitamin B12 supplementation but returned to a higher Hcy level (Hcy: 52.2 µmol/L) after withdrawal. The proband's mother is in good health condition. The pedigree chart of the family was presented in [fig_ref] Figure 1: Clinical features and genetic sequencing data of the patients [/fig_ref] A.
## Library preparation
Firstly, genomic DNA was extracted from 200µL peripheral blood, using a Qiagen DNA Blood Midi/Mini kit (Qiagen GmbH, Hilden, Germany). About 50 ng of genomic DNA was interrupted to around 200 bp fragments by the enzyme. The DNA fragments were endrepaired by adding one A base at the 3'end.
Secondly, the DNA fragments were ligated with barcoded sequencing adaptors, and ligated fragments (about 320 bp) were collected by XP beads. After PCR amplification, the DNA fragments were hybridized and captured by NanoWES(Berry Genomics, China) according to the manufacturer's Protocol. The hybrid products were eluted and collected, and then subjected to PCR amplification and purification. Next, the libraries were quantified by qPCR, and size distribution was determined using an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). Finally, the Novaseq 6000 platform (Illumina, San Diego, USA) with 150 bp pair-end sequencing mode was used for sequencing the genomic DNA samples of the family. Raw image files were processed using CASAVA v1.82 for base calling and generating raw data [bib_ref] A nonsense variant of ZNF462 gene associated with Weiss-Kruszka syndrome-like manifestations: a..., Zhao [/bib_ref].
# Data analysis
The sample genomes were aligned to the human reference genome (hg19/GRCh37) using the Burrows-Wheeler Aligner tool and PCR duplicates were removed by using Picard v1.57 (http:// picard. sourc eforge. net/). Verita Trekker ® Variants Detection System by Berry Genomics and the third-party software GATK (https:// softw are. broad insti tute. org/ gatk/) were employed for variant calling. Variant annotation and interpretation were performed with ANNOVAR (Wang, et al., 2010) and the Enliven ® Variants Annotation Interpretation System authorized by Berry Genomics.
Annotation databases mainly included:
[formula] (i) [/formula]
# Results
The couple came to our hospital when they were pregnant with their third child. To determine the likely pathogenic gene variants in the family pedigree, we performed whole-exome sequencing in patient II-3 and reanalyzed the whole-exome data of II-1, and II-2. The genetic status of the parents was detected by Sanger sequencing. It revealed compound heterozygous c.602 A > C (p.His201Pro) variant in exon 5 and c.1316T > C (p.Leu439Pro) variant in exon 8 of the MTHFR gene in the patients of II-1, II-2, and II-3. The Sanger sequencing also indicated that the c.602 A > C variant was inherited from the mother and the c.1316T > C variant was inherited from the father [fig_ref] Figure 1: Clinical features and genetic sequencing data of the patients [/fig_ref]. One year later, we performed whole-exome sequencing in patient II-4, and we found he was only a carrier of c.602 A > C. Both identified heterozygous variants were missense mutations, in which the genetic variant of c.602 A > C is a newly reported mutation while the variant of c.1316T > C was previously reported in adults but not in infants with severe homocystinuria due to MTHFR deficiency [bib_ref] The 1316T > C missenses mutation in MTHFR contributes to MTHFR deficiency..., Liu [/bib_ref]. The newly discovered variant sites have been submitted to the ClinVar website, Variation ID: 992,662, https:// www. ncbi. nlm. nih. gov/ clinv ar/ varia tion/ 992662. To further evaluate the pathogenicity of the compound heterozygous mutation, several amino acid missense mutation prediction programs, e.g. SIFT, PROVEAN, PolyPhen-2, Mutation Taster, Revel, and ClinPred were used to predict the deleteriousness of variants. Differently, Polyphen-2 showed the genetic variant of c.602 A > C and c.1316T > C were likely damaging, while SIFT, PROVEAN, and Mutation Taster showed that both heterozygous variants of c.602 A > C and c.1316T > C were damaging. The Revel score and ClinPred score of the variant of c.602 A > C were 0.887 and 0.999, the Revel score and ClinPred score of the variant of c.1316T > C were 0.835 and 0.995, they all show that these two missense mutations are damaging.
According to the ACMG guidelines, the variant c.1316T > C (p.Leu439Pro) should be classified as likely pathogenic by fulfilling the standard PS3 + PM2 + PP3. The rs545086633 SNP (p.Leu439Pro) results in an L439P substitution in MTHFR protein and drastically decreases mutant protein expression by promoting proteasomal degradation (PS3) [bib_ref] The 1316T > C missenses mutation in MTHFR contributes to MTHFR deficiency..., Liu [/bib_ref]. The mutation was not found in the human exome database (ExAC) and the population genome mutation frequency database (gnomAD). The frequency in the reference population 1000 genomes (1000G) was 0.001 (PM2). In silico prediction analyses predicted a deleterious effect on gene product for this variant (PP3).
According to the ACMG guidelines, the variant c.602 A > C (p.His201Pro) should be classified as likely pathogenic by fulfilling the standard PM2 + PM3 + PP1 + PP3. The variant was not found in ExAC, 1000G, and gnomAD (PM2). This variant forms compound heterozygosity (in trans state) with likely pathogenic c.1316T > C variation site (PM3). In the family, two affected family members (II-1 and II-2) with compound heterozygous variants were co-segregation with disease. Bilateral lateral ventricle trigonometry enlargement was also observed in II-3 with compound heterozygous variants. We were unable to observe further phenotypes due to the fetus was induced. While II-4 only carry one variant was healthy (PP1) [bib_ref] Expert specification of the ACMG/AMP variant interpretation guidelines for genetic hearing loss, Oza [/bib_ref]. Multiple software predicted its conservatism, the results showed that the site was evolutionarily conserved and had potential functional effects; the protein function was predicted by SIFT and LRT, and the results showed that it was harmful (PP3).
The MUSCLE (https:// www. ebi. ac. uk/ Tools/ msa/ muscle/) was used to compare the amino acid sequence of the MTHFR gene for multiple species and found two missense variants are highly conserved across different species [fig_ref] Figure 2: Clustal W alignment of MTHFR proteins among the representative species at the... [/fig_ref]. According to the identified MTHFR gene variants, the SWISS-MODEL (https:// swiss model. expasy. org/ inter active) was used to generate the three-dimensional (3D) models of the protein variants for functional structure prediction compared with the known Human MTHFR protein in the current database (https:// swiss model. expasy. org/ repos itory/ unipr ot/ P42898) [fig_ref] Figure 3: Modeling of the Human MTHFR protein [/fig_ref] A, 3B, 3 C, 3D). The comparison suggested that the MTHFR protein variant may lose its FAD-binding domain, resulting in the loss of its binding ability with FAD-during the methylation of Hcy.
# Discussion
There is a wide range of clinical manifestations in homocystinuria due to MTHFR deficiency. However, severe early-onset usually develops in infancy, and children with severe homocystinuria due to MTHFR deficiency often show diverse phenotypes, including eating difficulties, hypotonia, neurocognitive disorders, ventricular dilatation, hydrocephalus, brain atrophy, microcephaly, and epilepsy [bib_ref] Clinical pattern, mutations and in vitro residual activity in 33 patients with..., Huemer [/bib_ref]. In this study, severe homocystinuria due to MTHFR deficiency occurred 2 months after birth for two children, accompanied by initial symptoms, like impaired consciousness, vomiting milk, and shortness of breath. In the development of the disease, the functions of the nervous system and respiratory system of the two patients significantly deteriorated. MRI in the two patients showed hydrocephalus, cerebral dysplasia, brain atrophy, and intracranial hemorrhage, as typical manifestations of severe homocystinuria due to MTHFR deficiency [bib_ref] 10-Methylenetetrahydrofolate reductase deficiency with progressive polyneuropathy in an infant, Tsuji [/bib_ref] [bib_ref] Life-threatening methylenetetrahydrofolate reductase (MTHFR) deficiency with extremely early onset: characterization of two..., Forges [/bib_ref] [bib_ref] Methylenetetrahydrofolate reductase deficiency: importance of early diagnosis, Fattal-Valevski [/bib_ref]. Out of 30 patients with severe homocystinuria due to MTHFR deficiency, Huemer et al. found 15 with brain atrophy, 13 with Leukodystrophy/dysplasia of the brain myelin sheath, and 9 with ventricular enlargement/hydrocephalus in their MRI examination [bib_ref] Clinical pattern, mutations and in vitro residual activity in 33 patients with..., Huemer [/bib_ref]. Surtees et al. demonstrated that demyelination caused by defective methylation of homocystinuria due to MTHFR deficiency could be associated with the deficiency of S-adenosine methionine in cerebrospinal fluid, with the evidence of the remyelination achieved by the recovery of S-adenosine methionine [bib_ref] Association of demyelination with deficiency of cerebrospinal-fluid S-adenosylmethionine in inborn errors of..., Surtees [/bib_ref]. In our study, the proband (II-1) and II-2 were found to have a slight decrease in the blood methionine level during neonatal screening and did not receive any treatment. Consequently, this fact makes doctors hard to determine the onset age of homocystinuria due to MTHFR deficiency for patients (a chance of misdiagnosis) and dismiss the link with possible gene variations in MTHFR. Although a lot of examinations and clinical treatments have been conducted for children, unfortunately, their homocysteine and methionine have not been tested, and the opportunity for diagnosis and symptomatic treatment may be missed. The investigation of a genetic defect in homocystinuria due to MTHFR deficiency could be important for optimizing clinical treatment strategies. For instance, the finding on the MTHFR defects in homocystinuria due to MTHFR deficiency helps to explain why the supplementation of B12 did not improve the deteriorating nervous system for homocystinuria due to MTHFR deficiency patients and the use of CH3-THF cannot make the neurotransmitter normalized, whereas the use of betaine can effectively prevent the death of patients with early-onset of homocystinuria due to MTHFR deficiency and improve the dysplasia symptoms of the nervous system by remethylating of Hcy to Met [bib_ref] Early treatment using betaine and methionine for a neonate with MTHFR deficiency, Nishimoto [/bib_ref] [bib_ref] Titration of betaine therapy to optimize therapy in an infant with 5,10-methylenetetrahydrofolate..., Ucar [/bib_ref].
Human MTHFR is a spherical structural protein consisting of a homodimer with identical subunits, each of which carries an N-terminal catalytic domain (aa 1-356), a C-terminal regulatory domain (aa 363-656), and a short-chain connection domain in the middle (aa 357-362) [bib_ref] Structural basis for the regulation of human 510-methylenetetrahydrofolate reductase by phosphorylation and..., Froese [/bib_ref] [bib_ref] The structure and properties of methylenetetrahydrofolate reductase from Escherichia coli suggest how..., Guenther [/bib_ref]. The N-terminal regulatory domain consists of an 8α/8β Tim Barrel structure and three additional α helices as a conserved catalytic domain that binds 5-methyltetrahydrofolate NADPH and FAD to complete the catalytic reaction. The C-terminal catalytic domain consists of two five-stranded β-sheet structures arranged in the core region, as a regulatory domain that binds to its allosteric inhibitor (S-adenosine methionine) to regulate enzyme activity for sustaining the level of methionine in the cell. Another study reported that severe MTHFR deficiency could be associated with genetic variants of the FAD binding site (e.g. Thr129, Arg157, Ala175, and Ala195) in the human MTHFR protein [bib_ref] Sequence-structure analysis of FAD-containing proteins, Dym [/bib_ref]. In this study, we identified a novel gene variant of c.602 A > C (p.His201Pro) in the family with homocystinuria due to MTHFR deficiency, which results in a missense mutation on exon 5 of the MTHFR gene, and the 201 His is one of the FAD-binding sites of the MTHFR protein. In the prediction analysis of MTHFR protein variants, we found that the alteration of p.His201pro may disable the binding ability of MTHFR to the FAD prosthetic group, thus reducing the enzyme activity. Meanwhile, we also detected another heterozygous variant c.1316T > C (p.Leu439Pro) in MTHFR in this family case. Liu et al. previously demonstrated that severe homocystinuria due to MTHFR deficiency patients with tardive dyskinesia carried two MTHFR deficiency-associated SNPs (rs748289202 p.Arg335His and rs545086633 p.Leu439Pro) on the MTHFR [bib_ref] The 1316T > C missenses mutation in MTHFR contributes to MTHFR deficiency..., Liu [/bib_ref] , and their further investigation confirmed that rs545086633 was responsible for the missense mutation of p.Leu439Pro in the C-terminal regulatory domain of MTHFR protein, and the proteasome degradation caused by the mutation led to suppress the expression of MTHFR protein.
Another noteworthy problem is that the proband's father only carries a heterozygous variant of c.1316T > C in the MTHFR gene but the content of homocysteine in peripheral blood increases slightly. This situation has not been found in previous reports [bib_ref] The 1316T > C missenses mutation in MTHFR contributes to MTHFR deficiency..., Liu [/bib_ref]. It was improved after folic acid and vitamin B12 supplementation but returned to a higher Hcy level after withdrawal. Consistent with previous reports, folic acid and vitamin B12 are not effective in patients with MTHFR deficiency [bib_ref] Titration of betaine therapy to optimize therapy in an infant with 5,10-methylenetetrahydrofolate..., Ucar [/bib_ref].
The couple came to our hospital when they were pregnant with II-3 and received a comprehensive prenatal diagnosis. After reanalyzing the whole-exome data of II-1 and II-2, we found that the functional test of c.1316T > C (p.Leu439Pro) had been published [bib_ref] The 1316T > C missenses mutation in MTHFR contributes to MTHFR deficiency..., Liu [/bib_ref]. The pathogenicity of c.1316T > C (p.Leu439Pro) was changed from a variant of uncertain significance to likely pathogenic, which determined the pathogenicity of the compound heterozygous mutation. Finally, the patient chose to have an abortion (II-3) and successfully obtained a healthy baby (II-4). Therefore, data reanalysis and comprehensive prenatal diagnosis are very important for screening and prenatal diagnosis of rare diseases. In consideration of the clinical phenotype, family history, and result of genetic testing, we speculated that both patients had MTHFR deficiency due to the compound heterozygous variants of c.602 A > C (p.His201Pro) and c.1316T > C (p.Leu439Pro) on the MTHFR gene.
# Conclusion
In conclusion, this study performed extensive gene sequencing analysis with existing biochemical screening and imaging examination in a Chinese family with multiple severe homocystinuria due to MTHFR deficiency cases and found novel compound heterozygous gene variants, c.602 A > C (p.His201Pro) and c.1316T > C (p.Leu439Pro) of the MTHFR gene. The new pathogenic gene variants enrich the mutation spectrum of the MTHFR gene and contribute to improving the diagnosis of homocystinuria due to MTHFR deficiency, considering that homocystinuria due to MTHFR deficiency remains difficult to be determined depending on clinical phenotypes. The outcomes will also benefit the genetic counseling of homocystinuria due to MTHFR deficiency in the future.
[fig] Figure 1: Clinical features and genetic sequencing data of the patients. A The pedigree of the family. B Head MRI of II-2. C MTHFR NM_005957.4 c.602 A > C (p.H201P). D MTHFR NM_005957.4 c.1316T > C (p.L439P) [/fig]
[fig] Figure 2: Clustal W alignment of MTHFR proteins among the representative species at the sites of p.His201 and p.Leu439 [/fig]
[fig] Figure 3: Modeling of the Human MTHFR protein. A p.His201 and FAD. B p.His201Pro. C Modeling of the wild MTHFR protein. D Modeling of the MTHFR protein variant [/fig]
|
Integrated evaluation of telomerase activation and telomere maintenance across cancer cell lines
10In cancer, telomere maintenance is critical for the development of replicative immortality. Using 11 genome sequences from the Cancer Cell Line Encyclopedia and Genomics of Drug Sensitivity 12 in Cancer Project, we calculated telomere content across 1,299 cancer cell lines. We find that 13 telomerase reverse transcriptase (TERT) expression correlates with telomere content in lung, 14 central nervous system, and leukemia cell lines. Using CRISPR/Cas9 screening data, we show 15 that lower telomeric content is associated with dependency of CST telomere maintenance 16 components. Increased dependencies of shelterin members are associated with wild-type TP53 17 status. Investigating the epigenetic regulation of TERT, we find widespread allele-specific 18 expression in promoter-wildtype contexts. TERT promoter-mutant cell lines exhibit 19 hypomethylation at PRC2-repressed regions, suggesting a cooperative global epigenetic state 20 in the reactivation of telomerase. By incorporating telomere content with genomic features 21 across comprehensively characterized cell lines, we provide further insights into the role of 22 telomere regulation in cancer immortality. 23 24 telomere lengths, mutations in the ATRX and DAXX chromatin-regulating factors, and genome 50 instability (Cesare & Reddel, 2010). 51 The readily identifiable nature of telomeric DNA repeats has motivated the development 52 of computational methods for the determination of telomere content from whole-genome 53 sequencing (WGS) and whole-exome sequencing (WES) data (Ding et al., 2014). Recently, 54 such methods were employed to characterize telomere content across tumor sequencing data 55 from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) project, 56 which identified genomic markers of relative telomere lengthening (Barthel et al., 2017; Castel et 57 al., 2019). To gain a greater functional understanding of the landscape of telomere maintenance 58 in cancer, we estimated telomeric DNA content (subsequently referred to as telomere content 59 (Feuerbach et al., 2019)) across a diverse array of human cancer cell lines profiled in the 60 Cancer Cell Line Encyclopedia (CCLE) (Barretina et al., 2012; Ghandi et al., 2019) and 61 Genomics of Drug Sensitivity in Cancer (GDSC) (Yang et al., 2013) projects. We hypothesized 62 that telomere content could be reflective of underlying mechanisms of attrition, maintenance, 63 and repair, which may be reflected in associations with genetic markers. By combining these 64 estimates with a rich set of existing CCLE annotations, we determined genetic, epigenetic, and 65 functional markers of telomere content and telomerase activity across a diverse panel of human 66 cancer cell lines. 67 Results 68 Telomere content across cancer cell lines 69 Telomeric reads can be identified in DNA-sequencing reads using the canonical tandemly 70 repeated TTAGGG motif, and normalized telomeric read counts may provide an accurate 71 estimate of telomere content (Ding et al., 2014). We quantified telomere content across cell 72 lines using WGS and WES data from the CCLE (Barretina et al., 2012; Ghandi et al., 2019) and 73 4 GDSC (Yang et al., 2013) datasets. In particular, we considered 329 cell lines profiled with 74 WGS and 326 with WES in the CCLE, and 1,056 samples profiled with WES in the GDSC, of 75which 55 were non-cancerous matched-normal samples. We note that our estimates state 76 telomeric DNA repeat tract content, which is a normalized measure of telomeric reads in a 77 sample, rather than solely telomere length, because telomere length requires the identification 78 of true telomeric DNA from intrachromosomal, non-terminal telomeric DNA repeat tracts and 79 extrachromosomal telomeric DNA(Feuerbach et al., 2019). To assess the fidelity of our 80 telomere content measurements, we examined the agreement between the telomere content 81 estimates in overlapping cell lines from different sequencing datasets(Supplementary Fig. 1). 82We observed high agreement between the telomere content estimates derived from CCLE 83 WGS and GDSC WES data (r = 0.84, P = 3.7×10 -79 , n = 286) and moderate agreement between 84 CCLE WGS and CCLE WES estimates (r = 0.71, P = 1.5×10 -6 , n = 36). Therefore, we 85 generated a merged telomere content dataset (Supplementary Methods) by combining the 86 normalized log-transformed telomere contents derived from the CCLE WGS and GDSC WES 87 datasets for downstream analyses. 88The overall distribution of telomere content displayed a slight skew (Supplementary 89Fig. 1) towards longer telomeres, perhaps reflective of cell lines dependent upon ALT, a 90 hallmark of which is telomeres of abnormal and heterogeneous lengths(Bryan et al., 1997;91 Heaphy, Subhawong, et al., 2011). We matched a substantial number of cell lines (282 for 92 CCLE WGS, 554 for GDSC WES) with the age of the donor at the time of removal, from which 93 we observed weak negative (vs. CCLE WGS: r = -0.05, P = 0.39; vs. GDSC WES: r = -0.17, P = 94 6.0×10 -5 ) correlations between telomere content and the age of the original donor 95(Supplementary Fig. 2a). Among 1,099 merged CCLE WGS and GDSC WES samples, we 96 found raw telomere content to vary substantially both between (P = 2.0×10 -15 , Kruskal-Wallis H 97 test) and within(Fig. 1)cell lines of different primary sites. Cell lines of hematopoietic origin 98 (namely leukemias and lymphomas, which comprised 156 lines) tended to have higher telomere 99
# Introduction
## 25
Telomeres, repetitive nucleoprotein complexes located at chromosomal ends, are an important 26 component of genomic stability [bib_ref] How telomeres solve the end-protection problem, De Lange [/bib_ref]. As protective chromosomal caps, telomeres 27 prevent potentially lethal end-fusion events [bib_ref] The Stability of Broken Ends of Chromosomes in Zea Mays, Mcclintock [/bib_ref] and mis-processing of 28 chromosomal ends as damaged sites by the DNA repair machinery (De Lange, 2005; . Due to factors such as incomplete DNA replication and oxidative stress, 30 telomeres gradually shorten with successive rounds of cell division [bib_ref] A theory of marginotomy. The incomplete copying of template margin 822 in..., Olovnikov [/bib_ref]. If left 31 unchecked, telomere attrition eventually triggers growth arrest and senescence, and further 32 shortening can lead to acute chromosomal breakage and cell death. Telomere shortening 33 therefore acts as a major obstacle in the course of tumor development , and inhibition of telomere maintenance offers still largely untapped opportunities for 35 targeted cancer therapies . 36
Telomere shortening in embryonic development and in certain adult cell populations is 37 offset by telomerase [bib_ref] Identification of a specific telomere terminal 721 transferase activity in tetrahymena extracts, Greider [/bib_ref] , a ribonucleoprotein enzyme with a core 38 reverse transcriptase, TERT, that lengthens telomeres by catalyzing the addition of TTAGGG 39 nucleotide repeats from an inbuilt RNA template component, TERC . Although telomerase is transcriptionally silenced in the majority of somatic cells, 41 telomerase is reactivated in over 85% of all human cancers (N. W. . 42
Reactivation of telomerase is associated with a diverse set of genomic alterations, the most 43 common of which include highly recurrent mutations in the TERT promoter , comparisons between telomere 125 content and mutations in ATRX and DAXX yielded significant associations only between DAXX 126 alterations and GDSC WES telomere content [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref]. We further repeated 127 association tests with TP53, VHL, and IDH1 as identified previously among TCGA samples 128 , with which we confirm that truncating VHL mutations are associated with 129 reduced telomere lengths [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref] although this may be confounded by the 130 high occurrence of VHL mutations in kidney cell lines. Whereas we found relatively few 131 significant associations between telomere content and mutations, we note that we were limited 132 to an absolute estimate of telomere content as opposed to a relative measure of somatic 133 telomere lengthening, which requires a paired normal sample . 134
Telomere content associates with CST complex dependencies Having explored the association between telomere content and non-perturbative annotations, 136 we next considered whether variations in telomere content could confer or reduce selective 137 vulnerabilities to inactivation of certain genes. In particular, we hypothesized that telomere 138 content may be associated with vulnerabilities to reductions in the levels of telomere-regulating 139 proteins. To reveal such associations, we correlated our telomere content estimates with gene 140 inactivation sensitivities assessed via genome-wide CRISPR-Cas9 (Avana 2017)) and RNAi viability screens (Achilles RNAiand DRIVE . Although we found no dependencies that displayed outlier associations with 143 telomere content in the Achilles RNAi screen [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref] , we discovered that 144 sensitivity to Avana CRISPR-Cas9 knockouts of each of the three CST complex proteins was an 145 outlier association with telomere content estimates computed from both the GDSC WES and 146 CCLE WGS data [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref]. Namely, increased sensitivities to 147 knockout of the CST complex components, which are key mediators of telomere capping and 148 elongation termination [bib_ref] The human CST complex is a terminator of 656 telomerase activity, Chen [/bib_ref] , were correlated with lower telomere content 149 [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref]. Although the CST complex was not assessed in the DRIVE 150 screening dataset, we also found TERF1, a key shelterin component, to be among the positive 151 correlated genes with telomere content in the DRIVE panel [fig_ref] Table 2: P values determined using two-sided Pearson's correlation test [/fig_ref]. 152
Using the CST complex as a seed set, we subsequently queried all dependencies under 153 the premise that associated gene dependencies reflect coordinated functions . 154
Within the Avana panel (n = 757-769), we found significant (FDR < 0.01) outlier associations 155 between the CST complex genes and genes encoding five additional telomere-associating 156 proteins (ACD, POT1, TERF1, TERF2, TINF2). These five additional proteins, together with 157 TERF2IP, comprise the shelterin complex, the protector and regulator of telomere length and 158 topology [bib_ref] Shelterin: The protein complex that shapes and safeguards human 678 telomeres, De Lange [/bib_ref]. Interestingly, whereas the five other dependencies were positively 159 associated with telomere content, TERF2IP displayed a weak negative association 160 [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref] , suggesting that TERF2IP may play a distinct regulatory role in 161
shelterin function compared to the other members. To examine the dependency landscape of 162 the CST complex and these five other telomere-associated proteins, we computed a correlation 163 matrix involving these eight genes, clustering of which yielded two main subgroups: one 164 comprised of the CST complex members, and another of the five other genes [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref]. Despite 165 this separation, POT1 and TINF2 also displayed notable correlations with CST dependencies, 166 possibly serving as the primary mediators of previously-reported functional interactions between 167 the shelterin and CST complexes [bib_ref] The human CST complex is a terminator of 656 telomerase activity, Chen [/bib_ref] [bib_ref] OB fold-containing protein 1 (OBFC1), a 903 human homolog of yeast Stn1,..., Wan [/bib_ref]. 168
Whereas we found strong codependency relationships within this group of eight 169 telomere-associated proteins, we also noticed that certain shelterin members displayed notable 170 codependencies with p53 pathway members such as MDM2, ATM, and TP53 itself 171 . Because sensitivity to perturbation of the p53 pathway is highly 172 associated with TP53 mutations in cancer (McDonald et al., 2017), we asked if these 173 codependency relationships were also associated with hotspot mutations in TP53. In fact, TP53 174 was a significant (FDR < 0.001) outlier when a comprehensive set of hotspot and damaging 175 mutations was compared against sensitivity to ACD and TERF1 dependencies in the Avana 176 panel [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref] and against TERF1 and TINF2 dependencies in the 177 DRIVE panel [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref]. These links between these gene 178 dependencies and TP53 mutation status reprise and extend previous reports of p53-dependent 179 DNA damage responses to TERF1 and TINF2 depletion [bib_ref] Robust DNA damage 831 response and elevated reactive oxygen species in TINF2-mutated..., Pereboeva [/bib_ref]. Taken together, we find that CST and shelterin dependencies are correlated with 181 each other, telomere content, and TP53 mutation status [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref]. 182 were found to be closely associated in the Avana dataset [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref] , with both genes being the 188 top codependency with the other. Although these dependencies were not present as top 189 correlates against telomere content, we reasoned that they may be associated with other 190 genomic markers. In fact, among all genes, PML, which has previously been implicated in the 191 organization and regulation of ATRX and DAXX, was the top gene expression correlate with 192 DAXX dependency [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref]. To further investigate the genomic markers of these 193 dependencies, we compared ATRX and DAXX dependencies to CCLE quantitative proteomics 194 measurements . We found an outlier association between reduced DAXX 195 dependency and increased protein levels of ZMYM3 [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref] , a chromatin-binding histone 196 deacetylase complex member involved in the DNA damage response Although ATRX and ZMYM3 are located on adjacent bands of the X chromosome, ATRX 198 protein levels were not significantly associated with DAXX dependencies. Interestingly, a 199 complementary association of ZMYM3 protein levels and all Avana gene dependencies not only 200 revealed DAXX as the top associate, but also a significant association with TERF1 dependency 201 [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref]. [fig_ref] Table 2: P values determined using two-sided Pearson's correlation test [/fig_ref] and connections between DAXX, ZMYM3, and TERF1 regulation. 203
## Investigation of atrx-daxx dependencies
Patterns and mechanisms of telomerase expression Having thoroughly characterized telomere content and its related dependencies across the 205 CCLE, we next focused on the regulation of TERT transcription. Across 1,019 samples 206 previously profiled with deep RNAseq, we found that hematopoietic cell lines (leukemias, 207 lymphomas, and myelomas) were associated with the greatest mean expression of TERT (P = 208 5.8×10 -26 , two-sided Mann-Whitney U test; we considered TERT promoter status for 503 cell lines. We found that only the C228T 228 (chr5:1,295,228 C>T) mutation was significantly (P = 2.8×10 -5 , two-sided Mann-Whitney U test) 229 associated with an increase in TERT expression . Surprisingly, the 230 mean level of TERT expression in monoallelic contexts was only slightly lower than that of 231 biallelic contexts, with less than a 1.5-fold difference between the groups (P = 0.03, two-sided 232
Mann-Whitney U test). Given that cells with biallelic TERT expression do so with twice the 233 transcriptional source sites as those with monoallelic TERT expression, this reduced difference 234 may be a consequence of the effects of the TERT promoter mutation in producing particularly compared to a previous report using CCLE WGS data . Out of these 157 241 cell lines, 87 express TERT from a single allele. Moreover, of these 157 cell lines, 129 have a 242 sequenced promoter, with which we confirm that promoter mutations unanimously drive 243 monoallelic expression . Our expanded set of cell lines also reveals 244 several new tissues of origin in which TERT is monoallelically expressed without a mutant 245 promoter, such as hematopoietic cell lines . This high proportion of 246 TERT monoallelic expression then led us to ask whether there are genomic alterations aside 247 from promoter mutations that could lead to ASE. Under the assumption that such alterations 248 may also induce ASE in a larger region than a promoter mutation, we determined ASE status in 249 consequence of TERT promoter mutations. To address this hypothesis, we performed a 296 genome-wide search for CpG islands (CGIs) with significant differences in methylation levels in 297
TERTp mutant cell lines compared to TERTp wild-type ones. If TERT hypomethylation were a 298 downstream consequence of TERT promoter mutations, then we would expect TERT 299 hypomethylation to be an isolated event, and thus there would be few CGIs outside the vicinity 300 of TERT with methylation levels correlated with TERTp mutant status. Surprisingly, we instead 301 found a broad genome-wide distribution of CGIs that were hypomethylated in TERTp mut samples 302 relative to TERTp WT samples [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref]. Moreover, when correlated with a panel of global histone 303 modification levels, we found that TERTp mutants exhibited increased levels of 304
H3K9ac1K14ac0 and H3K9ac1K14ac1 marks [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref] , which have been suggested as marks 305 of transcriptionally active chromatin [bib_ref] Multivalent engagement of 848 chromatin modifications by linked binding modules, Ruthenburg [/bib_ref]. Likewise, when H3K9ac1K14ac0 306 levels were compared against a genome-wide panel of CGI ASM levels, the TERT CGI 307 (chr5:1,289,275-1,295,970) was the top correlate [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref]. 308
To better understand the distribution of these TERTp mut -hypomethylated CGIs, we 309 1,000 TERTp mut -hypomethylated CGIs [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref] , we found significant (FDR < 0.0001) and 312 robust 10-fold enrichment for polycomb repressive complex 2 (PRC2)-repressed regions [fig_ref] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency [/fig_ref] 5f) previously characterized in several cell lines (HepG2, GM12878, HeLa-S3, K562, and 314 HUVEC). Beyond these top 1,000 hypomethylated CGIs, CGIs overlapping with PRC2-315 repressed segments were broadly hypomethylated in TERTp mut cell lines and accounted for 316 nearly all of the previously observed skew towards hypomethylation [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref]. Interestingly, the 317 enrichment of PRC2 segments was much smaller (around 3.5-fold) in the remaining profiled cell 318 line, H1-hESC. Against ENCODE ChIP-seq peak region sets, we also found significant overlap 319 with the H3K9me3 and H3K27me3 heterochromatin marks [fig_ref] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions [/fig_ref]. Furthermore, we also 320 observed a moderate 2-fold (P = 5.1×10 -24 , Fisher's exact test) enrichment for regions within ten 321 megabases of most telomeres, consistent with previous reports that PRC2-repressed and 322
H3K27me3-marked regions are enriched in telomeric and subtelomeric regions . The enrichment of these hypomethylated regions among telomere-proximal regions 324 may also be indicative of a recently-reported telomere position effect, which has been shown to 325 affect the chromatin accessibility of the TERT locus (W. . Among 337 13,547 CGIs, we again found an enrichment of hypomethylation of PRC2-overlapping CGIs 338 , although this was less prominent than previously noted in the CCLE. 339 LOLA enrichment analysis for TERTp mut -hypomethylated CGIs in the TCGA likewise confirmed 340 significant enrichments (FDR < 0.0001) of PRC2-repressed regions and associated histone 341 modifications as the top enriched region sets . However, the fold-342 enrichment was less (about 5-fold) than that observed in the CCLE and did not display any 343 significant enrichment in telomere-proximal regions (P = 0.70, Fisher's exact test). Although the 344 skew towards hypomethylation in TERT promoter mutants among these TCGA samples was 345 weaker than in the CCLE samples, this may be the result of the more heterogeneous nature of 346 these primary TCGA samples as well as the differences in coverage between the Illumina 450k 347 array and RRBS. 348 In addition to telomere content, we also investigated the landscape of telomere 373 maintenance mechanisms, namely mechanisms of TERT reactivation, across cancer cell lines. 374
The enrichment of TERT promoter mutations in certain tissues has inspired several 375 explanations, and our findings in both the CCLE and TCGA suggest a specific epigenetic 376 signature that may underlie this unique pathway of telomere maintenance. We found that in 377
TERT promoter mutants, CpG islands were preferentially hypomethylated in PRC2-repressed 378 regions located near telomeres, which may relate to previous reports of a long-range telomere When multiple read groups were present in a sample, telomere content was computed as a 424 mean of the individual read group estimates weighted by the total read count per group. 425
Whereas we found decent agreement between overlapping samples in CCLE WGS and GDSC 426 WES, we found a comparatively weak correlation between both sets and the CCLE WES 427 estimates . Therefore, we excluded the CCLE WES telomere content 428 estimates from subsequent analyses. 429
In comparing the CCLE WGS and GDSC WES estimates, we also noticed a batch effect 430 resulting in two clusters of GDSC WES estimates. To identify and correct this batch effect 431 across all GDSC WES estimates, we observed that these batches were distinguished by 432 frequencies of reads containing exactly 4, 5, and 6 telomeric motifs. We then ran a k-means 433 clustering on these read frequencies to estimate the clusters across all GDSC WES samples, 434 which were subsequently adjusted by re-centering the mean of one cluster (after applying a z-435 scored log-transformation) to match the mean of the other. 436
We also attempted to use Telseq to estimate telomeric repeat-containing RNA (TERRA) 437 expression across 1,019 RNA-seq samples from the CCLE. However, because the majority of 438 these samples were found to contain little or no reads containing telomeric reads, TERRA 439 capture was determined to be too low for any meaningful analysis. We also considered processed methylation estimates available on the CCLE data portal, 463 namely the TSS 1kb upstream estimates as well as the promoter CpG cluster estimates, which 464 we correlated against telomere content estimates. For these annotations, we filtered out regions 465 with a standard deviation of less than 0.05. 466
## Genomic and transcriptomic markers
Results of TERT and TERC expression associations, as well as telomere content 467 associations, are available in [fig_ref] Table 2: P values determined using two-sided Pearson's correlation test [/fig_ref] To identify codependencies with the CST complex members, we employed an iterative 478 approach to identify highly ranked correlations. In particular, starting with a seed set of genes 479 (the base case of which was the CST complex), we searched for codependencies between two 480 genes x and y under the criteria that the r 2 association between the two is among the top five for 481
x vs. all other genes, and among the top five for y vs. all other genes as well. We recursively 482 applied this method four times, which added the five shelterin components ACD, POT1, TERF1, 483 TERF2, and TINF2 to our gene set. To construct the clustered correlation matrix in [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref] To rank and visualize the codependencies shown in ,b and the 499 dependency-mutation associations shown in [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref] ,e, we used a signed q-value approach. We 500 first transformed the raw false discovery rates by taking the negative of the base-10 logarithm, 501 and we then applied a sign to this transformed value as determined by the direction of the 502 codependency (the sign of the correlation coefficient) or dependency-mutation association 503 (negative for greater sensitivity in mutants, and positive otherwise). 504
Dependency analyses results are available in . 505
## Characterization of allele-specific tert expression
506 Allele-specific expression may be detected by looking for discordant counts of reads mapping to 507 single-nucleotide polymorphisms (SNPs) in DNA-sequencing vs. RNA-sequencing reads 508 . In particular, allele-specific expression is evidenced by the biased 509 frequency of a single allele of a heterozygous SNP in RNAseq reads compared to that of DNA-510 sequencing reads. To assess TERT expression in the context of allele-specificity, we examined 511 cell lines for which DNA (WES or WGS) and RNA (RNAseq) sequencing data were available. 512
To identify heterozygous anchor SNPs, we considered mutations in the TERT gene body called 513 using Mutect 1.1.6 (Cibulskis et al., 2013) with default settings. We then applied a filter for 514 mutations with at least eight reads supporting both the reference and alternate alleles that 515 passed the Mutect quality control filter (i.e. classified as PASS). To force call the matching allele 516 frequencies in RNA, we processed the matching aligned RNAseq reads using the 517 ASEReadCounter tool provided in GATK 3.6 (Van der Auwera et al., 2013) with arguments -518 minDepth 8, --minBaseQuality 16, --minMappingQuality 255, and -U 519
## Allow_n_cigar_reads. 520
We then used these RNA and DNA allele frequencies to classify cell lines as monoallelic 521
and biallelic expressors of TERT as well as two neighboring genes, SLC6A19 and CLPTM1L. In 522 particular, we examined the odds ratio derived from a binary contingency table with the two sets 523 of categories being the context (DNA vs. RNA) and the allele (reference vs. alternate) of the 524 read counts. To account for edge cases where the denominator of the odds ratio was zero, we 525 added a pseudocount of 0.5 to each category before computing the odds ratio. We then 526 denoted MAE lines as those having an odds ratio computed using the major allele as the 527 denominator of greater than five. In instances where there were multiple informative SNPs, we 528 considered only the SNP with the greatest supporting total RNA-seq read count. In cases where 529 the same SNP was detected across multiple sources (for instance, in both CCLE WES and 530 WGS), we considered the source with the greatest coverage of the SNP. 531
Allele-specific calls for TERT, SLC6A19, and CLPTM1L are described in 532 . 533
## Genome-wide allele-specific methylation analysis
## 534
To characterize and compare CpG-level ASM around the TERT genomic region, we utilized 535 RRBS data generated by the CCLE [fig_ref] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status [/fig_ref]. Mapped BAM files were downloaded 536 from the CCLE FireCloud workspace, and ASM levels for each CpG pair were estimated using 537 the allelicmeth command from the MethPipe package . Within each sample, 538 we first included only CpG pairs with a minimum coverage of eight reads. Next, among all 928 539 cell lines, CpG pairs included in less than 5% of these samples were excluded. 540
## 541
To estimate ASM, we employed a strategy similar to the original MethPipe ASM pipeline. 542
For each pair of CpGs, we considered the four combinations of methylation states between the 543 two CpGs: methylated-methylated (mm), methylated-unmethylated (mu), unmethylated-544 unmethylated (uu), and unmethylated-unmethylated (mm). For semi-methylated CpG pairs not 545 subject to ASM, we would expect high and relatively equal frequencies of the mu and um pairs, 546
whereas for ASM CpG pairs, we would expect the allele bias to result in high mm and uu counts 547 and low um and mu counts. To quantify this imbalance, we used the mean square contingency 548 coefficient (Φ) with a pseudocount of 0.5. Namely, for each CpG pair, we computed 549
[formula] Φ =̇ * ̇−̇ * √ (̇+̇)(̇+̇)(̇+̇)(̇+̇) 550 [/formula]
where ̇= + 0.5, ̇= + 0.5, ̇= + 0.5, and ̇= + 0.5. ASM CpG pairs 551 therefore had a positive Φ, whereas non-ASM pairs had a Φ of around 0. We rounded negative 552 Φ values to 0. Before computing these imbalance values, we excluded CpG pairs with a 553 methylation level of less than 0.1 or greater than 0.9 on either CpG, so as to filter out CpG pairs 554 that were likely to be fully methylated or completely demethylated. 555
We first examined the ASM levels of the TERT locus, which considered as the TERT 556 gene body as well as the flanking five kilobase regions. For these methylation estimates, we 557 excluded CpGs with less than 25% valid ASM estimates. We segmented these CpGs into five 558 To identify genome-wide methylation events indicative of TERT promoter mutations, we 563 searched for correlates with TERT promoter mutation status among average methylation levels 564 of CpG islands (CGIs). CpG island annotations were downloaded from the UCSC genome 565 browser at http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cpgIslandExt.txt.gz. To 566 filter out low-coverage CpG islands, we considered only CpG islands with at least eight CpG 567 sites. Methylation levels per island were then estimated by taking the mean across all CpGs 568 profiled within the island. Using the same filtering parameters, we also computed mean ASM 569 estimates across these CGIs. 570
TERT locus methylation estimates are described in . DepMap data are available at online portals as previously described. 618 Telomere content and related genomic features across human cell lines. Cell lines were grouped by cancer type and ordered by telomere content within each type, and are displayed such that each column represents a cell line. Telomere content measurements reflect combined z-scored estimates derived from CCLE WGS and GDSC WES with means for samples with telomere content estimates from both sources. Relative copy number values are shown as log2(relative to ploidy + 1) -1. Cell lines shown are filtered such that annotations for telomere content, TERT and TERC RNA-seq expression, TERT and TERC copy number, and ATRX and DAXX mutation status are all available, and each cancer type is represented by at least ten cell lines (n = 683 cell lines total). RNA expression estimates are in terms of log2(TPM+1). CNS: central nervous system; PNS, peripheral nervous system; UADT, upper aerodigestive tract. Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussian-based asymptotic approximation. *P < 0.05, **P < 0.01, n.s, not significant; two-sided Mann-Whitney U test. . b, Kernel density distributions of rank-biserial correlations between CGI methylation levels for PRC2-overlapping regions and non-PRC2-overlapping regions. A negative correlation indicates that a CGI is hypomethylated in TERTp mut cell lines relative to TERTp WT ones, and a positive correlation indicates the opposite. PRC2 regions were sourced from the HepG2 segmentation. c, Rank-biserial correlations between TERTp status (mutant or wild-type) and global histone modification levels (n = 302-475). Significance determined by twosided Mann-Whitney U test. d, Pearson correlation levels between global H3K9ac1K14ac0 levels and ASM imbalance of CGIs (n = 261-884). e, H3K9ac1K14ac0 levels are significantly increased in TERTp mutants. Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussian-based asymptotic approximation. *P < 0.01, n.s, not significant; two-sided Mann-Whitney U test. f, LOLA core set enrichment analysis of CGIs hypomethylated in TERTp mut cell lines reveals enrichment of PRC2-repressed regions. g, LOLA ENCODE Roadmap region enrichment analysis of CGIs hypomethylated in TERTp mut cell lines reveals enrichment of H3K9me3 and H3K27me3 regions. Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussian-based asymptotic approximation. *P < 0.05, two-sided Mann-Whitney U test against WT/silent category; n.s, not significant. b, Volcano plots of Pearson correlations and false discovery rates (q values) of associations between merged telomere content estimates and several profiling datasets. Sample sizes listed in Supplementary [fig_ref] Figure 4: Allele-specific methylation of the TERT locus is indicative of both promoter mutation... [/fig_ref]. Transcriptomic associations between TERT, TERC, and telomere content. a and b, Pearson correlations between log2(TPM + 1) levels of TERT mRNA and telomere content within tissue subtypes in the CCLE WGS and GDSC WES datasets, respectively. CNS, central nervous system; PNS, peripheral nervous system; UADT, upper aerodigestive tract. c, Associations between total TERT (ENSG00000164362.14) mRNA, full-length TERT (ENST00000310581.5) mRNA, minus-beta TERT (ENST00000296820.5) mRNA, TERC (ENSG00000270141.2) RNA, and z-scored log2-transformed telomere content estimates in the CCLE WGS and GDSC WES datasets. mRNA expression measured as log2-transformed TPMs with a pseudocount of +1. d, Associations between exon inclusion levels of TERT exons 7 . c, Selected correlations between CCLE WGS and GDSC WESderived telomere content and Avana dependencies of CST and shelterin complex members. . TP53 mutation status and shelterin member dependencies. a, Codependencies of CST and shelterin complex members in the Avana CRISPR-Cas9 dataset as measured by Pearson correlation and the associated two-sided P value (n = 710-769 cell lines). q-values are shown for correlations between each gene indicated on the x-axis and all other genes, with qvalues transformed and ranked by the sign of the correlation. b, Repeat of codependency analysis in (a), but for DRIVE RNAi dependencies (n = 88-386 cell lines). c and d, comparison of dependencies of select members (Avana and DRIVE, respectively) with respect to TP53 hotspot mutation status. Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussian-based asymptotic approximation. P values determined by two-sided Mann-Whitney U test. , and mRNA expression as measured in log2(TPM + 1) from RNAseq (c). Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussianbased asymptotic approximation. *P < 0.01, two-sided Mann-Whitney U test against wild-type (WT) values; n.s, not significant. d, Correlation between TERT gene body methylation and mRNA expression in promoter-wildtype (left) and promoter-mutant (right) cell lines. e, Frequencies of TERT promoter mutations across different tissue subtypes. . Global methylation changes associated with TERT promoter mutations. a, Methylation at the cg11625005 CpG probe and TERT mRNA expression across TCGA samples annotated by TERTp status (n = 1,553). b, Comparison of cg11625005 methylation levels in TERTp mutants and wild-type TCGA samples with TERT mRNA expression greater than 1 (as indicated in the horizontal line in (a). Boxes, interquartile range (IQR); center lines, median; whiskers, maximum and minimum or 1.5 × IQR; notches, 95% confidence interval of bootstrapped median using 1,000 samples and a Gaussian-based asymptotic approximation. P = 1.5 × 10 -26 , two-sided Mann-Whitney U test. c, Kernel density distributions of rank-biserial correlations between CGI methylation levels for PRC2-overlapping regions and non-PRC2overlapping regions. A negative correlation indicates that a CGI is hypomethylated in TERTp mut cell lines relative to TERTp WT ones, and a positive correlation indicates the opposite. PRC2 regions were sourced from the HepG2 segmentation. d, LOLA core set enrichment analysis of CGIs hypomethylated in TERTp mut samples reveals enrichment of PRC2-repressed regions. e, LOLA ENCODE Roadmap region enrichment analysis of CGIs hypomethylated in TERTp mut samples reveals enrichment of H3K9me3 and H3K27me3 regions. f, Genomic distribution of CGIs hypomethylated in TERTp mut CCLE cell lines. Shaded regions denote 10Mb chromosome ends. . Telomere content estimates and sample information. [fig_ref] Table 2: P values determined using two-sided Pearson's correlation test [/fig_ref]. Telomere content and transcriptomic associations. . Unsupervised dependency associations. . Allele-specific expression calls. . TERT locus methylation. . Genome-wide methylation analysis in the CCLE. . Genome-wide methylation analysis in TCGA.
# Figures
# Supplementary tables
[fig] 235: robust monoallelic expression (Huang et al., 2013) or expression of TERT from multiple sites all 236 of the same allele (Rowland et al., 2019). 237 To further explore allele-specific expression (ASE) patterns of TERT, we employed an 238 ASE-calling pipeline (Supplementary Methods) and determined TERT allele-specific 239 expression status for 157 cell lines(Supplementary Fig. 7a), an increase of 69 cell lines 240 [/fig]
[fig] 379: position effect (W. Kim et al., 2016; Yuan et al., 2019) and of TERT expression necessitating 380 specific chromatin states in promoter-wildtype and mutant samples (Salgado et al., 2019). 381 Considering that normal tissues typically exhibit particularly low methylation of the TERT 382 promoter (Salgado et al., 2019; Stern et al., 2017) and that PRC2 occupies the inactive allele in 383 TERT promoter mutants (Stern et al., 2017), our genome-wide signature may relate to the latter 384 part of the two-step mechanism proposed for TERTp mutation-driven telomerase upregulation 385 (Chiba et al., 2017). Moreover, epigenetic mechanisms have been shown to produce synergistic 386 effects with driver mutations in tumor evolution (Tao et al., 2019). Besides reflecting a direct 387 cooperation with TERT expression, this signature raises the possibility that the "memory" of 388 short telomeres may be preserved through these telomere-proximal hypomethylated regions. It 389 may also be indicative of the stemness of cell lines, which has been proposed as a major factor 390 in the proliferative advantage of TERT promoter mutations (Chiba et al., 2015). Future studies 391 will be necessary to elucidate the nature of this epigenomic signature, how it impacts the 392 regulation of telomerase expression, and the complexities of TERT expression beyond binary 393 measures of allele-specificity (Rowland et al., 2019). Furthermore, incorporation of telomere 394 content into studies using cancer cell lines may help improve our understanding of sensitivities 395 to drug or genetic perturbations across cell lines. 396 Through our analysis, we show relevant markers of telomere-associated protein 397 function, patterns of TERT reactivation across cancers, and epigenetic determinants of TERT 398 promoter status. We detail various features of telomere regulation and dysfunction in cancer, 399 and we provide a substantial addition of new features to a well-characterized set of cell lines. By 400 doing so, we complement molecular studies of telomeres in parallel studies across the GTEx 401 (Demanelis et al., 2019) and TCGA (Barthel et al., 2017), providing a valuable resource that will 402 guide additional studies on the roles and functions of telomeres in cancer. 403 Acknowledgements 404 We thank Dr. Elizabeth Blackburn (UCSF) and Dr. Thomas Cech (University of Colorado 405 Boulder) for providing insightful comments and feedback on the manuscript. We thank Dr. Hani 406 Goodarzi (UCSF) for his generosity in providing storage and computing resources. 407 Author contributions 408 M.G. and F.W.H. conceived the studies. M.G. obtained initial telomere content estimates. K.H. 409 performed all subsequent data acquisition and analysis. K.H. and F.W.H. organized figures and 410 tables. K.H. wrote the paper, and M.G. and F.W.H. oversaw the analyses and commented estimates were computed using Telseq (Ding et al., 2014) with the default 418 settings. Telseq records the frequencies of reads containing various frequencies of the 419 canonical TTAGGG telomeric repeat, and then normalizes this number of telomeric repeats 420 using a GC-adjusted coverage estimate and the average chromosome length. 421 Telomere content was estimated for WGS and WES samples in the CCLE (Ghandi et 422 al., 2019) as well as WES samples in the GSDC (Yang et al., 2013) using the default settings. 423 [/fig]
[fig] Figure 2: Telomere-binding protein dependencies are associated with telomere content and TP53 mutation status. a, Pairwise plot of Pearson correlations between dependencies of all genes in the Avana dataset and CCLE WGS telomere content (x-axis, n = 192-210 cell lines) and GDSC WES telomere content (y-axis, n = 395-416 cell lines) estimates. b, Pairwise plot of significance levels of correlations shown in (a) with correction for multiple hypothesis testing. c, Pairwise Pearson correlation matrix between Avana dependencies among CST members and five shelterin components (n = 757-769 cell lines;Supplementary Table 3). d, Associations of CST and shelterin member Avana dependency scores with damaging and hotspot mutations (n = 756-767 cell lines). For each gene dependency, mutation associations are shown ranked by signed q-value, with sign indicating the direction of association (negative for greater sensitivity in mutants, positive for greater sensitivity in wild-type cell lines). P values determined using twosided Pearson's correlation test. e, Associations of shelterin member DRIVE dependency scores with damaging and hotspot mutations (n = 372-375 cell lines;Supplementary Table 3)under the same scheme used in (d). f, Network schematic of the co-dependency matrix shown in (c) and annotated with association with telomere content or TP53 mutation status. [/fig]
[fig] Figure 3: Transcriptomic and proteomic determinants of DAXX dependency. a, Comparison of DAXX and ATRX dependency levels as measured in the Avana dataset. b, Correlations of DAXX Avana dependencies with genome-wide RNAseq gene expression estimates (n = 566 cell lines). c, Correlations of DAXX Avana dependencies with all profiled protein expression estimates (n = 4-262 cell lines). Two ATRX isoforms (UniProtKB: P46100 and P46100-6) are also highlighted. d, Correlations of ZMYM3 protein levels with all Avana dependencies (n = 242-263 cell lines). P values determined using two-sided Pearson's correlation test. [/fig]
[fig] Figure 4: Allele-specific methylation of the TERT locus is indicative of both promoter mutation status and allele-specific expression. a, Heatmap of CpG methylation levels along the TERT locus, sorted in order of mean methylation levels along the upstream 5kb region. TERT gene expression levels are also indicated for each cell line. Each column represents a cell line (n = 451), and each row represents a CpG pair (n = 208) sorted from the 5' to 3' direction along the TERT sense strand. White blocks indicate missing ASM values. Cell lines with unavailable ASM values for at least half of TERT locus CpGs were excluded. b, ASM levels of TERT locus subregions in cell lines are indicative of TERTp status and allele-specific expression. [/fig]
[fig] Figure 5: TERT promoter mutations associate with genome-wide decreased methylation of PRC2-repressed regions. a, Pairwise plot of median CGI methylation levels in TERTp mut cell lines (n = 21-83; Supplementary Table 6) versus TERTp WT cell lines (n = 95-410, Supplementary [/fig]
[fig] Figure S3: Associations between telomere content and cell line characteristics. a, Distributions of z-scored log2-transformed telomere content estimates from merged CCLE WGS and GDSC WES estimates, stratified by mutations in TP53, ATRX, DAXX, IDH1, and VHL. [/fig]
[fig] Figure S5: Telomere content and telomere protein dependencies. a, Scatterplots of zscored log2-transformed telomere content estimates from the CCLE WGS (left) and GDSC WES (right) datasets against sensitivity to members of the CST complex measured in the Avana dataset. b, Distribution of Pearson correlations between telomere content estimates from CCLE WGS and GDSC WES datasets and all gene dependencies in the Avana CRISPR-Cas9 dataset (vs. CCLE WGS: n = 192-210 cell lines, vs. GDSC WES: n = 395-416 cell lines; Supplementary [/fig]
[fig] Figure S7: TERT ASE and promoter mutations. a, Distribution of TERT allele-specific expression across tissue subtypes. CNS, central nervous system; UADT, upper aerodigestive tract. b, Distribution of TERT allele-specific expression and promoter mutations across tissue subtypes. c, Contingency table of TERT allele-specific expression and TERT promoter mutation status. P value determined by Fisher's exact test. d, Distribution of RNA-DNA allele ratios of the major expressed alleles in TERT, with matched TERT promoter mutation status also indicated. The dotted line at a ratio of five indicates the cutoff for classifying samples as MAE as opposed to BAE. Displayed allele ratios were capped at a maximum of 25 for viewability. [/fig]
[fig] Figure S8: TERT promoter mutations, methylation, and gene expression. a and b, Distributions across different promoter mutation types of TERT gene body methylation (a), gene body ASM (b) [/fig]
[table] 440: Cell lines were annotated with sample descriptors from the CCLE data portal 441 (Cell_lines_annotations_20181226.txt, https://portals.broadinstitute.org/ccle/data). Harmonized 442 sample information, telomere content estimates, and other matched annotations are available in 443 [/table]
[table] Table 2: P values determined using two-sided Pearson's correlation test. [/table]
|
Noise in the intensive care unit and its influence on sleep quality: a multicenter observational study in Dutch intensive care units
Background: High noise levels in the intensive care unit (ICU) are a well-known problem. Little is known about the effect of noise on sleep quality in ICU patients. The study aim is to determine the effect of noise on subjective sleep quality.Methods: This was a multicenter observational study in six Dutch ICUs. Noise recording equipment was installed in 2-4 rooms per ICU. Adult patients were eligible for the study 48 h after ICU admission and were followed up to maximum of five nights in the ICU. Exclusion criteria were presence of delirium and/or inability to be assessed for sleep quality. Sleep was evaluated using the Richards Campbell Sleep Questionnaire (range 0-100 mm). Noise recordings were used for analysis of various auditory parameters, including the number and duration of restorative periods. Hierarchical mixed model regression analysis was used to determine associations between noise and sleep.Results: In total, 64 patients (68% male), mean age 63.9 (± 11.7) years and mean Acute Physiology And Chronic Health Evaluation (APACHE) II score 21.1 (± 7.1) were included. Average sleep quality score was 56 ± 24 mm. The mean of the 24-h average sound pressure levels (L Aeq, 24h ) was 54.0 dBA (± 2.4). Mixed-effects regression analyses showed that background noise (β = − 0.51, p < 0.05) had a negative impact on sleep quality, whereas number of restorative periods (β = 0.53, p < 0.01) and female sex (β = 1.25, p < 0.01) were weakly but significantly correlated with sleep. Conclusions: Noise levels are negatively associated and restorative periods and female gender are positively associated with subjective sleep quality in ICU patients.
# Background
Recently, hospital noise and its potential negative influence on patient outcome has gained widespread attention among caregivers [bib_ref] Excessive noise in intensive care units, Darbyshire [/bib_ref] [bib_ref] The intensive care unit was so noisy I couldn't sleep, Hinton [/bib_ref]. Noise is generally expressed as sound pressure in decibels (dB), whereby often a correction is made for the frequency of the sound (called "A-weighting") to account for the relative loudness of the sound as perceived by the human ear. Noise levels in ICUs have been found to be beyond acceptable levels with average daytime sound pressure levels of around 60 A-weighted decibels (dBA) and peak levels > 90 dBA, the equivalent of standing next to a highway [bib_ref] An investigation of sound levels on intensive care units with reference to..., Darbyshire [/bib_ref] [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref]. Even more relevant, nighttime sound pressure levels are only slightly lower with averages of around 50 dBA. These sound pressure levels clearly exceed those of 35 dBA recommended by the World Health Organization (WHO) for nighttime in hospitals. In the ICU, different factors contribute to high sound pressure levels, including a large number of alarm-generating monitoring equipment, use of mechanical ventilators and around-the-clock activities by staff members [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref] [bib_ref] Sleep in the critically ill patient, Weinhouse [/bib_ref] [bib_ref] Contribution of the intensive care unit environment to sleep disruption in mechanically..., Gabor [/bib_ref]. Excessive noise may cause multiple auditory and non-auditory effects, of which sleep disturbances are thought to be the most deleterious [bib_ref] Auditory and non-auditory effects of noise on health, Basner [/bib_ref]. Sleep disturbances occur frequently in the ICU, characterized by an increase in stage 2 sleep and a decrease in stage 3 and rapid eye movement (REM) sleep [bib_ref] Quantity and quality of sleep in the surgical intensive care unit: are..., Friese [/bib_ref] [bib_ref] Characterisation of sleep in intensive care using 24-hour polysomnography: an observational study, Elliott [/bib_ref]. Up to now, only a few studies have studied the potential relationship between excessive noise and disturbed sleep in a real ICU setting. Studies using polysomnography in ICU patients demonstrated that between 11 and 24% of arousals are caused by environmental noise [bib_ref] Sleep quality in mechanically ventilated patients: comparison of three ventilatory modes, Cabello [/bib_ref] [bib_ref] Abnormal sleep/ wake cycles and the effect of environmental noise on sleep..., Freedman [/bib_ref]. Although polysomnography is considered the gold standard for evaluating sleep, it is labor intensive and a burden for ICU patients. Moreover, it is notoriously difficult to interpret and may not adequately reflect subjective sleep [bib_ref] Sleep in the intensive care unit, Pisani [/bib_ref] [bib_ref] When a gold standard isn't so golden: lack of prediction of subjective..., Kaplan [/bib_ref]. Furthermore, studies have been performed in a small number of patients and in single centers, thus limiting the generalizability of the findings. Also, the acoustic parameters that have been analyzed are conventional measures of sound level, indicative of physical changes in the sound field. To further quantify the effects of noise on humans, advanced parameters, such as loudness and restorative periods, defined as periods of relative quietness, may be more useful. We therefore set up a prospective, multicenter observational study aimed to determine the association between various acoustic parameters and subjective perceived sleep quality in ICU patients.
# Methods
## Study design
This was a prospective observational multicenter study in the ICUs of five teaching hospitals and one university medical center in the Netherlands (CinicalTrials.gov. number NCT01826799). These hospitals were Jeroen Bosch Ziekenhuis, 's Hertogenbosch (JBZ), Radboud University Medical Center, Nijmegen (RadboudUMC), Gelre Ziekenhuizen, Apeldoorn (Gelre), Isala Klinieken, Zwolle (Isala), Onze Lieve Vrouwe Gasthuis, Amsterdam (OLVG) and Ziekenhuis Gelderse Vallei, Ede (ZGV). Sound recording equipment was installed in 2-4 patient rooms in every participating ICU. Characteristics of the participating ICUs and rooms were collected, such as year of construction, layout, number of beds, level, population (e.g. surgical or cardiothoracic) and number of patients per room.
## Patients
ICU patients aged ≥ 18 years, admitted to one of the equipped rooms, were eligible after 48 h of admission to the ICU. Patients were not included if they were unable to understand Dutch or were unable to be assessed for sleep quality, defined as either a Richmond Agitation and Sedation Scale (RASS) score of − 2 or less or presence of delirium. Delirium detection was based on the Confusion Assessment Method for the ICU (CAM-ICU) three times daily.
Before study recruitment and enrolment, each patient and/or relative was given a full explanation of the study. Since this was an observational design and no actual sound recording was made inside the rooms that could be retraceable to individual patients, the need for informed consent was waived by the regional medical ethical committee (registration number MJ504, Medisch-Ethische Toetsing Onderzoek Patienten en Proefpersonen (METOPP), Tilburg, The Netherlands). Patient characteristics and demographics including relevant previous medical history, admission diagnosis, severity of illness score (expressed by the Acute Physiology And Chronic Health Evaluation (APACHE-)II score) and length of stay in the ICU were collected. Patients' data were entered into a web-based electronic case record form by the research nurses of the participating hospitals and were only accessible by the investigators.
## Sound measurements
A measurement European conformity (CE)-marked microphone (M23, Earthworks Inc., Milford NH, USA) connected to a laptop or PC, was placed in 2-4 patient rooms, above the patients' head (2.1~2.4 m from the floor). Only if the patient became eligible were the sound data used for analysis. The data were coded and stored on a hard disk and were de-identified to prevent identification of persons by members of the project group.
Based on previous research, the primary measures of interest for comparing the different acoustic conditions between the six hospitals were the A-weighted time-averaged sound pressure level (L Aeq ) and the 10th percentile (L 90 ) sound pressure level (A-weighted on fast-response mode), which is an estimate of background noise [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref]. Furthermore, the occurrence rate of loudness peaks per hour, and the number and the average duration of restorative periods were recorded. A restorative period was defined as a continuous time interval of at least 5 min during which the sound pressure level (SPL) did not exceed the predefined threshold of 17.7 dBA above the L 90 [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref]. The threshold of 5 min was chosen to be the minimum time required for a patient to be able to go to sleep after a disturbance, and has been reported in the literature previously, while the 17.7 dB minimum relative sound level has been shown to be the average level needed to see an arousal in polysomnographic measurements of (healthy) persons exposed to ICU noise during sleep [bib_ref] The influence of white noise on sleep in subjects exposed to ICU..., Stanchina [/bib_ref] [bib_ref] Characterizing noise and perceived work environment in a neurological intensive care unit, Ryherd [/bib_ref].
The number of restorative periods per hour and their average duration were calculated. Longer restorative periods provide more opportunity for undisturbed sleep. The detection of loudness peaks was based on the psycho-physiological model by Chalupper and Fastl to ensure that the impact of peak sounds was assessed based on the auditory perception and expressed in units of sone [bib_ref] Dynamic loudness model (DLM) for normal and hearing-impaired listeners, Chalupper [/bib_ref]. Sone defines the sound level with respect to how the frequency sensitivity of the human ear changes with level, with us being less sensitive to lower frequency sounds at low sound levels, giving a metric that is perceptually more accurate yet harder to compute. A doubling of the sone level is equal to a doubling of the perceived loudness, unlike all of the other sound level measures reported in this paper, which are based on the logarithmic decibel scale. The rate was calculated either including all peaks or only peaks that had a minimum level of 10 sone, which is equivalent to a noise peak of at least 73 dB at 1 kHz. We calculated values for each parameter for three time periods: the whole day, day time (7 a.m.-11 p.m.), and night (11 p.m.-7 a.m.). Secondary measures of interest are presented and discussed in Additional file 1. Noise data were analyzed using Matlab version R2017a (The Mathworks Inc., Natick, MA, USA).
## Sleep assessment
Patients' sleep was evaluated using the validated Richards Campbell sleep questionnaire (RCSQ) [bib_ref] Measurement of sleep in critically ill patients, Richards [/bib_ref]. This 5-item questionnaire is used to evaluate different aspects of sleep, namely perceived sleep depth, sleep latency, number of awakenings, efficiency and time awake. Each item is rated on a visual analog scale (VAS) (0-100 mm), whereby higher scores indicate better sleep. The mean of the scores on these 5 items represents the overall RCSQ score. Usually, one item, regarding whether the noise level is disturbing for sleep is also part of the questionnaire [bib_ref] Patients' sleep in an intensive care unit-patients' and nurses' perception, Frisk [/bib_ref] [bib_ref] Patient-nurse interrater reliability and agreement of the Richards-Campbell sleep questionnaire, Kamdar [/bib_ref] and therefore this item was added to the questionnaire. The RCSQ has proven to be a valid, non-invasive tool for sleep perception in the ICU [bib_ref] Measurement of sleep in critically ill patients, Richards [/bib_ref]. A Dutch translation of the RCSQ was created and validated according to the principles of good translation. Sleep evaluation was started after patients were identified as eligible, and was continued for a maximum period of five nights. The RCSQ was filled in by the patient at around 7 a.m.. If the patient was not able to fill in the RCSQ, no score was recorded.
# Statistical analysis
Data were compared using Student's t test and proportions were compared using the chi-square test. To determine associations between variable noise parameters and sleep quality, exploratory hierarchical mixed-model regression analyses were performed specifying random intercepts for rooms in hospitals and for patients and selecting the best-fitting model using an automated model selection procedure based on the Akaike information criterion (AIC) [bib_ref] AIC model selection using Akaike weights, Wagenmakers [/bib_ref] [bib_ref] On the behaviour of marginal and conditional {AIC} in linear mixed models, Greven [/bib_ref]. We based the calculation of p values on Satterthwaite estimated degrees of freedom and carried out the analyses on the data from the nighttime recordings between 11 p.m. and 7 a.m. The goodness of fit of the model was calculated based on the method described by Nakagawa and Schielzeth [bib_ref] A general and simple method for obtaining R 2 from generalized linear..., Nakagawa [/bib_ref]. All statistical analyses were performed using SPSS version 20 (SPSS, IBM) and R (version 3.4.1, R Foundation for statistical computing, Vienna, Austria).
# Results
A total of 71 patients fulfilled the criteria between April 2013 and August 2015 and were included in this study. Data on seven patients were removed from the final analysis due to missing audio data. Baseline characteristics of the remaining 64 patients can be found in [fig_ref] Table 1: Patient characteristics on inclusion [/fig_ref]. On average, patients were 63.9 ± 11.7 years old and 48 (68%) patients were male. Most participating ICUs had single-bed rooms and the daily visiting routine occurred at similar times (see [fig_ref] Table 2: Hospital characteristics [/fig_ref].
## Sleep quality of the patients
No sleep evaluation was registered for five patients; finally, 151 nights of sleep were evaluated (mean 2.4 nights/patient). Average total sleep quality was 56 ± 24 mm and was not significantly different between the participating hospitals [fig_ref] Table 3: Results from the sleep evaluation per site and of all sites [/fig_ref]. Based on the additional question of the RCSQ, noise was considered quite disturbing with an average VAS score of 34 mm, whereby a lower score indicates more disturbance. In 64 of 151 nights (42%), patients provided an answer to which noise they found was the most disturbing factor during their sleep. Patients found that monitor/equipment alarms were the most disturbing to sleep (28/64) followed by other (n = 21), staff speech (n = 9), and other staff activities (n = 6).
## Noise levels in participating icus
The mean of the 24-h average sound pressure level (L Aeq, 24h ) was 54.0 ± 2.4 dBA, with no significant differences between day and night (seeand . The L 90 was 38.1 ± 4.0 dBA on average. Restorative periods occurred on average 1.2 times per hour during the day, increased to 2.4 during the night (p < 0.0001), whereby also the average duration of the restorative period significantly increased from 11.4 min during the day to 14.1 min during the night (p < 0.001, see. Loudness peaks with a minimum magnitude of 10 sone occurred 23.1 times per hour during the day, significantly decreasing to 6.0 times per hour during the night (p < 0.0001).
## Association between noise and sleep
Female patients rated their sleep quality on average 1.2 points higher than men (p < 0.01; see and Additional file 1). The number of restorative periods per hour during the night also significantly positively contributed to sleep quality: with every additional restorative period (per hour), sleep quality significantly increased by 0.53 points (p < 0.01). Higher levels of the background noise (L 90 ) significantly decreased sleep quality ratings by 0.51 points (p < 0.05) . We identified conditional R 2 glmmðcÞ of 0.2; regression model diagnostics did not highlight violations of model assumptions. Note that we did not incorporate the additional RCSQ item, regarding which noise source was most disturbing for sleep, as the predictor or outcome variable in the regression analysis because we did not gather a sufficient number of data points to enable us to make reliable conclusions. A more elaborate description of the analyses and model diagnostics is provided in Additional file 1.
# Discussion
In this prospective multicenter, observational study, we showed that background noise was negatively associated with sleep, while gender (female) and the number of restorative periods were positively correlated with sleep quality. Patients' perceived sleep quality was poor and did not differ between participating hospitals. Overall, noise levels were consistently above the values (< 35 dB L Aeq day and night; < 40 dB L AFmax night) recommended by the WHO. This is the first study to evaluate a link between subjective sleep quality and objective noise parameters in ICUs. Sleep disturbances are very common in ICU patients [bib_ref] Sleep in the intensive care unit, Pisani [/bib_ref]. As a risk factor for sleep disturbances [bib_ref] Sleep in the intensive care unit, Pisani [/bib_ref] , noise in ICUs is ubiquitous and is mainly caused by staff activity, machines and alarms [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref].
Research on the relationship between noise and sleep arousals in ICU patients, thereby using polysomnography, shows that noise peaks > 80 dBA are associated with arousal from sleep and that noise is responsible for 11 to 24% of the total number of arousals [bib_ref] Sleep quality in mechanically ventilated patients: comparison of three ventilatory modes, Cabello [/bib_ref] [bib_ref] Abnormal sleep/ wake cycles and the effect of environmental noise on sleep..., Freedman [/bib_ref] [bib_ref] Environmental noise as a cause of sleep disruption in an intermediate respiratory..., Aaron [/bib_ref]. More subjectively, patients themselves consider noise in the ICU to be disturbing to sleep [bib_ref] Experiences of critically ill patients in the ICU, Hofhuis [/bib_ref].
We found that a higher L 90 led to a moderate decrease in sleep quality. L 90 is considered to be indicative of the background noise level, generated by, for example, air conditioning or computer ventilators. The average nighttime value of L 90 was 38.1 dBA, indicating that levels of background noise even exceed the threshold for average sound pressure levels stated by the WHO. Because the participating hospitals were different in design and layout, considerable differences were found in L 90 between hospitals. Given that within-hospital L 90 was not different between day and night, the observed differences cannot be attributed to differences in procedures and staff movement but must be due to the building characteristics. This finding underlines the importance of taking building properties into account when designing a new ICU. Average day and nighttime noise levels were comparable with others studies, whereby differences between day and night were only marginal [bib_ref] An investigation of sound levels on intensive care units with reference to..., Darbyshire [/bib_ref] [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref]. This is the first study showing a positive association between restorative periods and better sleep. Restorative periods occurred on average only 2.4 times per hour during the night, and the average duration of a restorative period was 14.1 min, which is indicative of the high number of peak noises. Since restorative periods are most frequently ended by high-level noisy events due to staff activity or speech [bib_ref] Analysis of the soundscape in an intensive care unit based on the..., Park [/bib_ref] , interventions aimed at reducing staff-generated noise appears to be a reasonable and achievable goal to improve sleep quality in critically ill patients. Additionally, since noise coming from alarms and monitors were found to be the most disturbing noise sources in our qualitative analysis, nighttime alarm modification may also be of additional value in improving sleep quality in critically ill patients.
Female patients expressed that they had better sleep compared to the male patients. Gender differences in subjective sleep quality in ICU patients have not been reported previously in the literature, however, studies in the general population generally indicate worse subjective sleep quality in women, compared to men [bib_ref] Exploring gender difference in sleep quality of young adults: findings from a..., Fatima [/bib_ref]. Interestingly, in a large study on the effects of traffic noise on objective sleep and subjective sleep quality in healthy L Aeq A-weighted time-averaged sound pressure level, dBA A-weighted decibel in ICU, L 90 10th percentile sound pressure level (A-weighted on fast-response mode), Peak10S hourly rate of loudness peaks of at least 10 sone people, there were larger effects of noise on objective sleep parameters in men compared to women, whereas no clear differences were found in subjective sleep quality. Future research on gender differences in sleep quality should further elucidate this finding. In addition to pure physical properties of sound, we also analyzed the impact of noise peaks on human perception (also called loudness) by using a validated model [bib_ref] Dynamic loudness model (DLM) for normal and hearing-impaired listeners, Chalupper [/bib_ref]. Loudness peaks with a magnitude of at least 10 sone occurred significantly less during the night than during the day. It is noteworthy that the nighttime hourly rate of these peaks differed substantially between the participating hospitals, varying from 2.5 to 8.8.
Without a source-specific analysis, it is difficult to pinpoint the source of these loud peaks and what causes this difference between participating hospitals.
Some limitations need to be addressed. First, we used the RCSQ as a measure of sleep quality instead of polysomnography. Although RCSQ is validated for the use with ICU patients, subjective evaluation of sleep quality is subject to different forms of bias, including recollection and response bias, which may specifically be true for ICU patients who are recovering from critical illness. However, the results of the sleep evaluations are in line with previous studies and the repetitive design of up to five measurements helps in reducing the impact of outliers. Moreover, the RCSQ is easily applicable and interpretable and is therefore easier to use in daily practice in contrast to polysomnography. Second, we included a relatively small group of ICU patients who were awake and able to communicate, thus selecting only a subgroup of less severely ill patients, which makes it difficult to generalize the findings of this study to the whole ICU population. Although sleep evaluation in sedated and/or delirious patients remains difficult, interventions aimed at improving sleep quality by addressing noise may also have beneficial effects in this other patient group. Third, we did not take the long-term outcomes of the patients into account. Therefore, negative effects on sleep may not necessarily lead to worse outcomes. However, other studies have clearly demonstrated an association between sleep deprivation and the development of delirium in ICU patients, which has a multitude of negative long-term consequences [bib_ref] Bench-to-bedside review: delirium in ICU patients -importance of sleep deprivation, Weinhouse [/bib_ref]. Moreover, even brief periods of sleep deprivation in the general population can have long-term negative consequences in immune and cognitive function and in hypertension and obesity [bib_ref] Sleep deprivation as a neurobiologic and physiologic stressor: allostasis and allostatic load, Mcewen [/bib_ref].
# Conclusions
Associations between various noise parameters and subjective sleep quality were found in this multicenter study, confirming the negative consequences of noise on the sleep quality in ICU patients and thereby strengthening the usefulness of noise-reducing strategies. Sleep quality in general was poor and did not differ between participating ICUs. Noise levels were high and periods of relative quietness occurred only rarely. Increasing the number of nighttime restorative periods appears to be a reasonable goal for improving patients' sleep.
[fig] APACHE: Acute Physiology And Chronic Health Evaluation, SOFA Sequential Organ Failure Assessment [/fig]
[table] Table 1: Patient characteristics on inclusion; a n = 62 due to missing data in 2 patients [/table]
[table] Table 2: Hospital characteristics [/table]
[table] Table 3: Results from the sleep evaluation per site and of all sites; average scores per item of the Richards Campbell sleep questionnaire (RSCQ) and the average overall RCSQ score are expressed as mean mm (SD) [/table]
|
Association of Infants Exposed to Prenatal Zika Virus Infection With Their Clinical, Neurologic, and Developmental Status Evaluated via the General Movement Assessment Tool
[bib_ref] Prechtl's method on the qualitative assessment of general movements in preterm, term..., Einspieler [/bib_ref] [bib_ref] An early marker for neurological deficits after perinatal brain lesions, Prechtl [/bib_ref] [bib_ref] A systematic review of tests to predict cerebral palsy in young children, Bosanquet [/bib_ref] [bib_ref] accurate diagnosis and early intervention in cerebral palsy. Advances in diagnosis and..., Novak [/bib_ref] [bib_ref] The predictive validity of general movements -a systematic review, Burger [/bib_ref] [bib_ref] Predictive validity of Prechtl's method on the qualitative assessment of general movement:..., Darsaklis [/bib_ref] [bib_ref] Prechtl's method on the qualitative assessment of general movements in preterm, term..., Einspieler [/bib_ref] [bib_ref] An early marker for neurological deficits after perinatal brain lesions, Prechtl [/bib_ref] [bib_ref] Sensitivity and specificity of general movement assessment for diagnostic accuracy of detecting..., Morgan [/bib_ref] [bib_ref] Prechtl's method on the qualitative assessment of general movements in preterm, term..., Einspieler [/bib_ref]
# Results (additional)
The motor optimality score, MOS Neither country of origin nor ethnicity (defined by the parents of the child and the investigators) affected the MOS of the control infants: infants born in South America (median=26; IQR=24-26) did not differ from infants born in North America or Europe (median=26; IQR=24-28; p=0.505); Caucasian infants (median=26; IQR=24-28) did not differ from non-Caucasians (median=26; IQR=24-26; p=0.685). Overall, the median MOS of the control group was with 26 within the optimal range . These results emphasised once again 6 that endogenously generated movement patterns are independent of ethnicity, environmental settings and care-giving procedures.
The number of normal movement patterns (co-occurring with fidgety movements) ranged from 2 to 14 (median 5) in control infants, from 0 to 10 (median 3) in the Rio de Janeiro cohort, and from 0 to 2 in infants with microcephaly (median 0; . The most frequently occurring normal movement patterns in control infants were fidgety movements (100%), foot-to-foot contact (62%), visual scanning (57%), smiling (38%), hand-to-mouth contact (36%), wiggling-oscillating arm movements (32%), hand-to-hand contact (32%), legs lift (32%), and kicking (30%). Each of these movement patterns occurred significantly less often in infants prenatally exposed to acute maternal infection with rash who did not develop microcephaly (p values ranging from 0.037 to 0.001). The few age-specific movement patterns observed in infants with microcephaly were smiling (14%), and foot-to-foot contact (6%) or side-to-side movements of the head (6%). A small proportion of control infants exhibited a few abnormal movement patterns. Among those were long lasting and/or repetitive tongue protrusion (8%) and abnormal foot-to-foot contact (i.e. both feet frequently striking each other on the tibial side without any plantar contact; 5%). Abnormal tongue protrusion also occurred in 14 infants of the Rio de Janeiro cohort (18%) and in nine infants with microcephaly (26%). Apart from tongue protrusion, the most frequently occurring abnormal movement patterns in the Rio de Janeiro cohort were long-lasting wigglingoscillating movements of arms, neck and/or legs (33%) and abnormal eye movements, such as strabismus or slow pursuit (9%). Abnormal eye movements were also seen in 24 infants with microcephaly (69%). In addition, they showed circular arm movements (46%), long lasting stiff head anteflexion (40%), en bloc trunk rotations (31%), long-lasting kicking (29%), and/or monotonous side-to-side movements of the head (26%). Ten infants (29%) with microcephaly were almost hypokinetic in the lower limbs, which also occurred in two infants (3%) of the Rio de Janeiro cohort who did not develop microcephaly.
Seventy-five percent of all control infants had four normal postural patterns such as head in midline, symmetric body posture, absence of asymmetric tonic neck posture, and variable finger postures, whereas 75% of the infants from the Rio de Janeiro cohort had only three normal postural patterns of the main manuscript). Infants with microcephaly had only one normal postural pattern.
The majority of infants with microcephaly (86%) could not maintain the head in midline, whereas no difference regarding this item was observed between the Rio de Janeiro cohort and controls (29 vs. 19%; of the main manuscript). Long-lasting extensions of arms and/or legs were also almost equally often observed in the Rio de Janeiro cohort and the controls. Long-lasting extension of the legs, however, occurred in 91% of infants with microcephaly (p<0.001). A persistent asymmetric tonic neck posture, hyperextension of the neck and/or trunk, and spreading of toes merely occurred in infants with microcephaly of main manuscript). None of the infants with microcephaly had variable finger postures; 13 infants (37%) were constantly fisting and the remaining 22 infants (63%) had rarely occurring monotonous finger movements. External rotation and abduction of the hip hardly occurred in infants with microcephaly but was present in 26% of the Rio de Janeiro cohort. A crampedsynchronised movement character (i.e. stiff limb and trunk muscles contract almost simultaneously and then relax almost simultaneously) 6 occurred only in infants with microcephaly (19/35; 53%). A monotonous movement character occurred in all three groups; it was present in 27% of the control infants, in every second infant of the Rio de Janeiro cohort, and in almost all infants with microcephaly of main manuscript). Tremulous movements occurred in only one infant of the Rio de Janeiro cohort and in 13 infants with microcephaly (37%). In any case the monotonous, jerky and/or stiff movement character observed in the control group did not correlate with the Bayley-III scores at 12 to 30 months as their scores were in the normal range. The described irritability and tremulousness e.g.,10,11 in ZIKV-exposed non-microcephalic infants was not confirmed in our study. The explanation might be that previous observations referred mainly to neonates whereas we included three to five-month-olds.
Only two of 35 infants showed normal foot-to-foot contact and normal side-to-side movements of the head; another five infants socially smiled. Apart from these few normal patterns, infants with microcephaly showed abnormal eye movements (strabismus, nystagmus), long lasting stiff head anteflexion, en bloc trunk rotations, circular arm movements, monotonous and long lasting (mainly unilateral) kicking or no leg movements at all; if the legs were not moving they were kept in full extension. A cramped-synchronised movement character was observed in every second infant with microcephaly. Such rigid movements with almost simultaneous contraction of at least two limbs and the trunk followed by an almost simultaneous relaxation is known to be highly predictive for severe spastic cerebral palsy, 1 especially if it persists beyond two months after term. Finally, none of the infants with microcephaly exhibited variable finger postures. This is in line with previous observations about the association between variable finger postures and a normal cognitive development. [bib_ref] The general movements assessment helps us to identify preterm infants at risk..., Einspieler [/bib_ref]
|
Spontaneous rupture of pyonephrosis leading to pyoperitoneum
A B S T R A C TForniceal rupture after pyonephrosis can lead to retroperitoneal abscess but extension into peritoneal cavity is unusual. We present a case of 50 year old lady who presented with acute abdomen, CT scan of abdomen showed features suggestive of left pyelonephritis and large intraperitoneal collection. Laparotomy was planned on suspicion of intraperitoneal source of sepsis. Peroperatively no intraperitoneal source of sepsis was found. Retrograde pyelography showed forniceal rupture at lower pole of kidney extending to peritoneal cavity without any obstructing stone or lesion. This case highlights an unusual presentation of pyonephrosis with peritonitis and pyoperitoneum.
# Introduction
Acute abdomen is a common clinical presentation in emergency department. Stratifying patients into those requiring surgery and those to be managed conservatively is an important aspect of initial evaluation. [bib_ref] Surgical risk model for acute diffuse peritonitis based on a Japanese nationwide..., Nakagoe [/bib_ref] Diagnosis of peritonitis is mainly clinical, supported by imaging studies to confirm the diagnosis and find out underlying cause. [bib_ref] The role of computed tomography in the preoperative assessment of gastrointestinal causes..., Reginelli [/bib_ref] In cases of peritonitis patient is usually evaluated for an intraabdominal source of sepsis and if found should undergo surgical treatment. Extroperitoneal pathologies e.g. pancreatitis and pyelonephritis may also present with signs of peritonitis, however these are managed conservatively as there is no direct extension into peritoneal cavity. We present a case of forniceal rupture after pyonephrosis leading to direct extension into peritoneal cavity causing pyoperitoneum and acute abdomen.
## Case presentation
A 50 year old, female with no known comorbids presented to ER with severe abdominal pain and vomiting for last one month. Pain started in epigastric region and was radiating towards umbilicus. There was no exacerbating or relieving factor and no history of dyspepsia, burning micturition or lower urinary tract symptoms. She also reported high grade intermittent fever during previous month. Past medical and surgical history was unremarkable. At presentation she was in sepsis with pulse of 102-beats/min, blood pressure of 90/53 mmHg and temperature of 101°F. The abdomen was diffusely tender with some guarding and rigidity more marked in left upper quadrnat.
Baseline investigations showed TLC count of 25.5 × 109 with 97% neutrophils and Procalcitonin of 24.67ng/ml. Serum amylase, Lipase, Creatinine and liver function tests were within normal limits. Urinalysis was positive for nitrites and leucocyte esterase and microscopy showed > 20 WBC/HPF. Urine Culture later revealed > 103 and < 104 CFU/ml of Klebsiella Pnemoniae.
Ultrasound showed swollen left kidney with moderate hydronephrosis and some echogenic material within lower pole calyx. As clinical examination was suggestive of generalized peritonitis, Contrast enhanced Computed tomography (CT) scan was carried out which showed swollen Left kidney with marked perinephric stranding. Additionally a perinephric collection was noted abutting left psoas muscle and extending anteriorly to duodenojejunal flexure. High-density fluid collection was also noted in rectouterine pouch, extending anterior to urinary bladder with peritoneal enhancement suggesting pyoperitoneum [fig_ref] Figure 1: CT abdomen showing intraperitoneal collection [/fig_ref].
As pyelonephritis usually leads to perinephric abscess with extension limiting to retroperitoneum, this intraperitoneal collection was unusual. So an additional intra-abdominal source of sepsis such as perforated bowel or diverticulitis was suspected and a laparotomy was performed both to drain pus collection and rule out intraperitoneal source of sepsis.
Urology and General Surgery teams were involved and after optimization, patient was planned for laparotomy. Intraoperatively purulent collection was found throughout the peritoneal cavity and 1 L of pus was drained. This intra-abdominal collection was found to be ex- tending to left perinephric space which was also evacuated. No additional intraperitoneal source of pyoperitoneum was identified. However, on retrograde pyelography forniceal rupture was found at left lower pole calyx with contrast extravasating in perinephric space extending to peritoneal cavity . A Double J stent was placed.
Postoperatively patient recovered well. She received IV fluids and IV antibiotics till complete resolution of sepsis and discharged on 6thpostoperative day after removal of abdominal drains. She was followed in clinic 2 weeks after surgery and DJ stent was removed after 6 weeks. She has remained asymptomatic till last follow up. Repeat ultrasound showed no hydronephrosis or collection. As no pathological cause of pyelonephritis including diabetes mellitus, voiding dysfunction or renal stones was identifiable and pus culture was negative for any unusual pathogens, she has been advised general precautions to prevent recurrent UTI.
# Discussion
Peritonitis and pyoperitoneum usually originate from intraperitoneal pathology. However retroperitoneal infections can also lead to irritation of peritoneum producing signs and symptoms of peritonitis e.g. pyelonephritis and pancreatitis. Nevertheless in such cases the mainstay of management is conservative management, along with treatment of any underlying pathology, for instance ureteric stones in case of pyelonephritis or gallstones in cases of pancreatitis. Abdominal exploration with laparotomy or laparoscopy is only required if some additional pathology is suspected or there is direct involvement of intraperitoneal structures. In patients with pyelonephritis and forniceal rupture, the pus collection is usually limited to retroperitoneum as there is no anatomical extension with general peritoneal or pelvic cavity. Yet direct extension into peritoneal cavity from retroperitoneal infection without any inciting event has been reported previously. Shifti and Bekele reported a case of ruptured pyonephrosis leading to peritonitis in a patient with pelviureteric junction obstruction leading to cortical thinning. [bib_ref] Generalized peritonitis after spontaneous rupture of pyonephrosis: a case report, Shifti [/bib_ref] Similarly, Hendaoui presented similar case of intraperitoneal extension in patient with non-functioning kidney due to urolithiasis. 4 However, our patient had a good functioning kidney with no parenchymal loss. Spontaneous forniceal rupture is known to occur in patients with pyelonephritis 5 specially if there is an obstructing lesion. However, pyoperitoneum is an unusual sequelae of such retroperitoneal infection.
# Conclusion
This unusual presentation of pyonephrosis suggests that retroperitoneal infections can rarely lead to intraperitoneal extension and rupture of pyonephrosis can potentially occur even in absence of any obstructing stone or lesion. Peritoneal lavage and exploration is still warranted in such cases to rule out concomitant intraperitoneal pathology.
## Fig. 2.
Retrograde pyelography showing forniceal rupture at lower pole (black arrow) and contrast extension into retroperitoneum and peritoneal cavity (asterisk *).
[fig] Figure 1: CT abdomen showing intraperitoneal collection (A and B, black arrow) along with perinephric collection (C, asterisk*). I.K. Jalbani, et al. Urology Case Reports 26 (2019) 100928 [/fig]
|
Addressing Health Equity Through Action on the Social Determinants of Health: A Global Review of Policy Outcome Evaluation Methods
Outcome(s)Intermediate: access to drinking water, electricity and sanitation; healthcare expenditure per capita; number of basic healthcare facilities, public hospital, physicians, nurses and midwives; national health insurance coverage; percentage of skilled deliveries, prenatal visits, contraception and vaccination prevalence; vaccination; antennal care Long term: neonatal, infant and under-5, maternal mortality; proportion of stunting and malnutrition; children with diarrhoea and pneumonia; pregnant women with anaemia Inequity Dimension Education + Income -wealth quintiles + Place of residence -urban/ruralSummary(1) Study compared the access to drinking water, electricity and sanitation; healthcare expenditure per capita; number of basic healthcare facilities, public hospital, physicians, nurses and midwives; national health insurance coverage; percentage of skilled deliveries, prenatal visits, contraception and vaccination prevalence; neonatal, infant and under-5, maternal mortality rates; proportion of stunting and malnutrition in children; pregnant women with anaemia before and after the Millennium Development Goals (following the evolution of healthcare system); comparing between rural/urban areas, mother's education level, wealth quintiles, and public and private section (2) Study measured the spatial disparity by comparing the percentage of antennal care, delivery assistance, contraception, vaccination, employed women, GDP per capita, children with diarrheal and suspected pneumonia, stunting children, young illiteracy, need in medical staff; under-5 mortality rate a ross 13 regions. Burstrom et al. 2010
geographical distribution of each population in need to measure geographical diversity of disadvantage populations;
(3) measuring the equity of infrastructure provision using mean targeting ratios comparing the tertiles of clustering of the disadvantaged population by black and minority ethnic groups, children, socio-economic disadvantage and retired household.
## Emery, fleisch & mcintyre 2013
Legislated changes to federal pension income in Canada will adversely affect low income seniors' health
## Policy
Federally-funded retirement benefits (raising age of eligibility from 65 to 67)
## Policy area
Social Protection and Welfare
## Policy target upstream
## Geographic scope
Localised ( Inequity Dimension Education + Health -poor health + Sex Summary 1) Study measured the associations between employment rates and various degree of flexible, deregulated labour market; different relative generosity and entitlement of welfare benefits (decommodification); and greater investment and emphasis on active labour market policies by comparing the age-standardised employment rates between healthy population with those with longstanding illness that was not limiting; or those with limiting longstanding illness, stratified by sex, education level and countries The impact of tobacco control policies on smoking among socioeconomic groups in nine Policy Different tobacco control policies
## Policy area environment/living conditions
## Policy target midstream
## Geographic scope
Cross-National Comparison (N=9)
## Dataset(s)
Repeated cross-sectional; Multiple dataset; 1990-2007; Secondary data
# Analysis
Descriptive analysis: graphical analysis + statistical summaries; Inferential analysis: fixed effects regression model
## Outcome(s) intermediate: smoking prevalence
## Inequity dimension education + occupation + sex
Summary 1) Study compared the country trends in the price and non-price measures over time 2) Study compared the trends in age-standardised smoking prevalence by education in each country for men and women over time 3) Study measured the association between price (popular cigarette and cheapest cigarette price) and non-price related tobacco control policies, and smoking, comparing the odds ratio between man and women stratified by education level and occupation
## Mcnamara 2015
Trade liberalization, social policies and health: An empirical case study
## Policy
Social protective policies of welfare states and labour markets during trade liberalization (focus on employment changes of textile and clothing sector)
## Policy area social protection and welfare
## Policy target upstream
## Geographic scope
Cross-National Comparison (N=32)
## Dataset(s)
## Inequity dimension age + education + occupation
## Summary
Study measured the association between the orientation of public gender equality policies and gender equalities in health by 1) comparing the prevalence ratios and age-standardised prevalence of poor self-perceived health among women and men, stratified by country and country typology of family policies (dual earner, traditional-central, traditional-southern, market oriented, contradictory) 2) comparing the prevalence ratio of poor self-perceived health according to gender, stratified by age, born in low-income country, education level, marriage status and employment status
## Summary
Study measured the association between welfare reforms and the probability of tobacco smoking; binge alcohol consumption; mental health; medical access; affordability of medical access; and routine Pap test and mammography screening between single mothers and the "control" group (married mothers, single nonmothers, and married nonmothers); between single mothers and a synthetic control group; and between employed and unemployed single mothers
[formula] Korenbrot, Kao & Crouch 2009 [/formula]
Funding of tribal health programs linked to lower rates of hospitalization for conditions sensitive to ambulatory care
## Policy
## Changes in indian health service funding models of tribally operated health programs
## Policy area
Health System Management/Health Insurance
## Policy target upstream
## Geographic scope
Localised -California, US
## Dataset(s)
Longitudinal; Multiple dataset; 1998-2002; Secondary data (linked hospital records)
# Analysis
Descriptive analysis: statistical summaries; Inferential analysis: generalised estimating equations regression model + multilevel regression model
## Outcome(s)
Long-term: Hospitalization rate for ambulatory-care sensitive conditions
## Inequity dimension education + income -poverty + occupation -unemployed
## Summary
Study measured the association between higher funding of Tribally Operated Health Programs and reduced rate ratio of the hospitalisations for ambulatory care sensitive conditions, comparing among federal funding and medical price index, eligible population, socioeconomic status and resources, disparities in occupation, education, and household differences between indigenous and white population, Indian gaming revenue
## Observational -longitudinal/repeated cross-sectional
## Summary
Study measured the association between health insurance reform and the likelihood of having insurance coverage at each time point
# Analysis
Descriptive analysis: graphical analysis + statistical summaries; Inferential analysis: fixed effects regression model + difference-in-difference-in-difference regression model
## Outcome(s)
Intermediate: economic index (years of completed education, ratio of family income to the federal poverty line, and family wealth); health services utilization; medical debt Long-term: composite chronic condition index (high blood pressure, diabetes, heart disease/heart attack, and obesity)
## Inequity dimension age: youth + income
## Summary
(1) Study measured the association between exposure (duration and timing) to Medicaid during early childhood on economic attainment and adult health stratified by income and predicted participation (2)
## Summary
(1) Study measured the association to which investment was allocated to need by comparing the investment amount per household to their income deprivation decile (2) Study measured the association between the mental and physical health and the level of investments by comparing the mean scores of mental and physical health with lower, medium and higher level of investments (3) Study measured the association of level of investment and change in the mean score of mental and physical health, comparing between the pre-and post-scores, stratified by 3 investment groups of lower, medium and higher level.
## Inequity dimension place of residence
## Summary
(1) Study measured the trend of association between public health care allocations and measures of disease burden (antenatal HIV prevalence and infant mortality), income per capital and race over time (2) Study measured the association between public health allocations and burden of disease (antenatal HIV prevalence, infant mortality and crude death rate),
## Interventional -pre-experimental
## Aghajanian et al. 2007
## Impact of rural health development programme in the islamic republic of iran on rural-urban disparities in health indicators
## Policy
Community-based primary health care (PHC) system (started early 80s, expanded in 90s)
## Policy area
Health System Management/Health Insurance
## Policy target upstream
## Geographic scope
Localised
# Analysis
Descriptive analysis: graphical analysis + statistical summaries + principal components analysis; Inferential analysis: chi-square analysis + hierarchical regression model + meta-regression model
## Outcome(s)
Long-term: prevalence of dental trauma in 12-year-old schoolchildren
## Inequity dimension place of residence -disadvantaged area
## Summary
Study measured the differences in prevalence of dental traumas between areas that were demonstrated to be "betteroff" (top 50% of areas) after the Healthy Cities Project implementation relative to the areas that were demonstrated to be "worse-off" (bottom 50% of areas)
# Analysis
Descriptive analysis: graphical analysis + statistical summaries; Inferential analysis: change-on-change regression model
## Outcome(s)
Intermediate: frequency of bus travel, active travel, car travel; transport poverty; transport injury; assault; car dependence; access to education, training, independent travel; effects on older citizens Long-term: physical/mental health Inequity Dimension Age -youth + Income -deprivation + Race/Ethnicity
## Summary
(1) Study compared bus, car, and active transport use; road injuries; and assaults pre-and post-policy intervention between youth and adults (2) Study compared distances travelled by youth according to journey purpose by decile of deprivation pre-and post-policy intervention (3) Summary 1) Study measured the association between the cycling initiative interventions and change in prevalence of cycling; walking; driving; and taking public transport to work, comparing the pre-and post-intervention absolute and relative changes in the intervention group with matched comparison groups; unfunded comparison groups; non-London national comparison groups 2) Study measured the association between small-area deprivation and prevalence of cycling to work, comparing the percentage-point change and ratio of change in the intervention group with matched comparison groups; unfunded comparison groups; non-London national comparison groups stratified by deprivation indices
## Kearns & mason 2015 policy
Housing investment, transformational area regeneration programme, and relocation due to demolition and rebuilding
## Policy area environment/living conditions
## Policy target upstream
## Regeneration, relocation and health behaviours in deprived communities
## Geographic scope
Intra-National Comparison: Glasgow, UK
## Summary
Study assigned samples into 3 groups based on their current location, residential circumstances and receiving regeneration in their living areas (since 2008): those living in non-regeneration area and not receiving any improvements (nonmovers); those who have remained living in the same dwellings but received regeneration (remainers); and those who have moved out of the regeneration areas into a social-rented dwellings (outmovers). Study compared the odd ratios of each outcomes in the 3 comparison groups: between remainers and outmovers relative to nonmovers; and outmovers relative to remainers, each comparison stratified by dwelling quality, neighbourhood quality, sex, education level, occupation, long-standing illness, and household types
## Summary
(1) Study modelled the estimated changes in incidence, prevalence and case-death rates of 16 tobacco-related diseases from the 4 interventions (2) Study modelled the years of life saved, adjusting for disability from the change in disease rates (3) Study modelled the cost of cigarettes, factoring in travel cost by estimating change in distance and time travelling to reduced and location of retail outlets (3) Study measured change in disease rate factoring in price elasticities from changes in smoking behaviour (4) Study modelled QALY and cost-savings, comparing between business-as-usual-model and the 4 interventions, stratified by different population groups
# Meta-analysis
## Ogilvie et al. 2008
The harvest plot: a method for synthesising evidence about the differential effects of interventions.
## Policy
## Population-level tobacco control interventions
## Policy area environment/living conditions
## Policy target mixed
## Geographic scope
Not specified
## Dataset(s)
Studies which assessed the effects of any type of population-level tobacco control intervention and had reported effects stratified by at least one demographic or socio-economic characteristic irrespective of study design, methodological quality or outcomes measured (N=85)
# Analysis
Descriptive analysis: graphical analysis + statistical summaries; Meta-analysis: Harvest plot
## Outcome(s)
Intermediate: change in awareness, change in attitude, change in perceived availability Long-term: change in self-reported smoking status Inequity Dimension Age + Education + Income + Occupation + Race/Ethnicity + Sex
## Summary
Study systematically reviewed the results of studies on the effects of population-level tobacco control interventions on changes in awareness, attitude, perceived availability and self-reported smoking status stratified by income, occupation, education, gender, ethnicity, and age
[table] 69: that are food insecure stratified by main source of income to forecast how the policy change may affect food insecurity in affected populations as an indicator of economic security and physical/mental health [/table]
|
Regulation of the human TRAIL gene
# Introduction
TNF-related apoptosis-inducing ligand (TRAIL) was initially identified by its sequence homology with other tumor necrosis factor (TNF) family members. 1,2 TRAIL has received considerable attention, primarily due to its tumor-selective apoptosisinducing capability demonstrated in several human cancer cell lines. [bib_ref] TRAIL-R2: a novel apoptosis-mediating receptor for TRAIL, Walczak [/bib_ref] TRAIL is expressed in a variety of human fetal and adult tissues including the spleen, thymus, prostate, small intestine and placenta. [bib_ref] Identification and characterization of a new member of the TNF family that..., Wiley Sr Sr [/bib_ref] Contrary to other TNF-family members, membranebound TRAIL is conditionally expressed in immune cells such as natural killer (NK) cells, B cells, monocytes and dendritic cells following cytokine stimulation. [bib_ref] Human dendritic cells mediate cellular apoptosis via tumor necrosis factor-related apoptosis-inducing ligand..., Fanger [/bib_ref] [bib_ref] Monocyte-mediated tumoricidal activity via the tumor necrosis factorrelated cytokine, TRAIL, Griffith [/bib_ref] [bib_ref] Human B cells express functional TRAIL/Apo-2 ligand after CpGcontaining oligodeoxynucleotide stimulation, Kemp [/bib_ref] [bib_ref] Natural killer (NK) cell-mediated cytotoxicity: differential use of TRAIL and Fas ligand..., Zamai [/bib_ref] Intracellular stores of TRAIL have also been found in polymorphonuclear neutrophils [bib_ref] Interferon-activated neutrophils store a TNF-related apoptosis-inducing ligand (TRAIL/Apo-2 ligand) intracellular pool that..., Cassatella [/bib_ref] [bib_ref] Neutrophil stimulation with Mycobacterium bovis bacillus Calmette-Guerin (BCG) results in the release..., Kemp [/bib_ref] [bib_ref] Neutrophil-derived TNF-related apoptosis-inducing ligand (TRAIL): a novel mechanism of antitumor effect by..., Koga [/bib_ref] [bib_ref] Tumor necrosis factor-related apoptosis-inducing ligand: a novel mechanism for Bacillus Calmette-Guérin-induced antitumor..., Ludwig [/bib_ref] [bib_ref] IFNalpha-stimulated neutrophils and monocytes release a soluble form of TNF-related apoptosis-inducing ligand..., Tecchio [/bib_ref] that are released after a variety of stimuli. [bib_ref] Neutrophil stimulation with Mycobacterium bovis bacillus Calmette-Guerin (BCG) results in the release..., Kemp [/bib_ref] [bib_ref] TNF-related apoptosis-inducing ligand (TRAIL) is expressed throughout myeloid development, resulting in a..., Simons [/bib_ref] [bib_ref] Identification of the mycobacterial subcomponents involved in the release of tumor necrosis..., Simons [/bib_ref] At physiological conditions, TRAIL is capable of binding to four distinct transmembrane receptors in humans: the proapoptotic death receptors DR4 and DR5 or the two decoy receptors DcR1 and DcR2. Both TRAIL and its receptors form homotrimers in a binary complex upon ligand binding. [bib_ref] 2.8 A resolution crystal structure of human TRAIL, a cytokine with selective..., Cha [/bib_ref] [bib_ref] Triggering cell death: the crystal structure of Apo2L/TRAIL in a complex with..., Hymowitz [/bib_ref] In mice, TRAIL-R is the only cognate death receptor. [bib_ref] Molecular cloning and functional analysis of the mouse homologue of the KILLER/DR5..., Wu [/bib_ref] TRAIL is a member of the TNF superfamily that induces tumorselective cell death by engaging the pro-apoptotic death receptors DR4 and DR5. The antitumor potential of the TRAIL pathway has been targeted by several therapeutic approaches including recombinant TRAIL and TRAIL-receptor agonist antibodies among others. Interest in sensitizing tumor cells to TRAIL-mediated apoptosis has driven investigations of TRAILreceptor gene regulation, though regulation of the TRAIL gene has been less studied. Physiologically, TRAIL serves as a pro-apoptotic effector molecule in the immune surveillance of cancer that is conditionally expressed by immune cells upon stimulation via an interferon-response element that was identified in early studies of the TRAIL gene promoter. Here, we map the TRAIL gene promoter and review studies of TRAIL gene regulation that involve several modalities of gene regulation including transcription factors, epigenetics, singlenucleotide polymorphisms and functionally distinct isoforms. that are responsible for basal TRAIL gene transcription. Increased TRAIL gene transcription has been reported in response to several cytokines and in some cases have been ascribed to transcription factors [fig_ref] Figure 3: Molecules that alter human TRAIL gene transcription [/fig_ref] with consensus binding sequences found in the TRAIL gene promoter [fig_ref] Figure 4: Sequence analysis of the human TRAIL gene promoter [/fig_ref].
Interferons. Interferons are cytokines that are intimately involved in the immune response in a variety of response by causing immune cell recruitment and/or activation by several mechanisms. Luciferase reporter experiments found that interferon-γ (IFN-γ) was capable of upregulating TRAIL gene promoter activity by 2-fold in a region between -165 and -35, 52 which provided a molecular explanation for a previous report that found rapid induction of TRAIL following incubation with IFN-γ. 5 STAT1 and IRF1 are thought to directly mediate the effects of IFN-γ on the TRAIL promoter. IFN-γ has also been shown to induce FasL-and TRAIL-dependent apoptosis in lung cancer cells [bib_ref] Potentiation of Fas-and TRAILmediated apoptosis by IFN-gamma in A549 lung epithelial cells:..., Kim [/bib_ref] and is responsible for IL18-and TLR3-induced TRAIL expression in NK cells. [bib_ref] Synergy between TLR3 and IL-18 promotes IFN-γ dependent TRAIL expression in human..., Tu [/bib_ref] There are reports that suggest type I interferons (IFN-α and -β) induce the TRAIL gene more strongly that IFN-γ. [bib_ref] Genomic organization and transcriptional regulation of human Apo2/TRAIL gene, Gong [/bib_ref] IFN-α and -β potently induce TRAIL in CD4 + and CD8 + peripheral blood T cells following CD3-stimulation as well as Jurkat T cells. [bib_ref] Type I interferons (IFNs) regulate tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) expression..., Kayagaki [/bib_ref] IFN-α has also been shown to induce TRAIL in macrophages [bib_ref] Distinct functions of IRF-3 and IRF-7 in IFN-alpha gene regulation and control..., Solis [/bib_ref] and in lymphoma cells in a JNK-dependent manner. [bib_ref] Requirement of c-Jun NH2-terminal kinase activation in interferon-alpha-induced apoptosis through upregulation of..., Yanase [/bib_ref] IFN-β induces TRAIL in colorectal cancer cell lines by a Stat-1dependent mechanism. [bib_ref] Stat1-dependent induction of tumor necrosis factor-related apoptosis-inducing ligand and the cell-surface death..., Choi [/bib_ref] Taking the evidence together, IFN-α, -β, -γ have been shown to induce TRAIL gene transcription, though the robustness of TRAIL induction among them seems to be context-dependent.
NFAT. The human TRAIL gene is also positively regulated by the transcription factor NFAT, which is activated by calcineurin-mediated dephosphorylation that occurs during T-cell activation. Wang et al. investigated the role of NFAT regulation of the TRAIL gene promoter due to the two putative binding sites they previously noted along with three other newly found potential NFAT binding sites. [bib_ref] Isolation and molecular characterization of the 5'-upstream region of the human TRAIL..., Wang [/bib_ref] [bib_ref] NFATc1 regulation of TRAIL expression in human intestinal cells, Wang [/bib_ref] Among the five NFAT family members, NFATc1 was by far the most potent positive regulator of TRAIL gene transcription. Interestingly, NFAT-dependent TRAIL gene upregulation was not affected by deletion of their putative binding sites in TRAIL gene promoter luciferase reporters but instead was abrogated by deletion of the -165 to -35 region. Chromatinimmunoprecipitation and electrophoretic mobility shift assay (EMSA) experiments revealed that NFATc1 antagonizes SP-1 binding to the TRAIL gene promoter. This reports highlights a mechanism that immune cells such as cytotoxic T cells might T cells in response to cytokines, particularly interferon-gamma that possesses a response element in the TRAIL gene promoter. [bib_ref] Genomic organization and transcriptional regulation of human Apo2/TRAIL gene, Gong [/bib_ref] TRAIL-knockout mice or zebrafish do not display any gross developmental defects, [bib_ref] Delineation of the cell-extrinsic apoptosis pathway in the zebrafish, Eimon [/bib_ref] [bib_ref] Increased susceptibility to tumor initiation and metastasis in TNF-related apoptosis-inducing ligand-deficient mice, Cretney [/bib_ref] though the mice are more susceptible to carcinogen-induced sarcomas and metastasis. [bib_ref] Increased susceptibility to tumor initiation and metastasis in TNF-related apoptosis-inducing ligand-deficient mice, Cretney [/bib_ref] TRAIL-R knockout mice develop normally but have an enlarged thymus and decreased radiation-induced apoptosis in several tissues. [bib_ref] DR5 knockout mice are compromised in radiation-induced apoptosis, Finnberg [/bib_ref] TRAIL-knockout mice also do not induce thymocyte apoptosis, which is deregulated in autoimmune diseases. [bib_ref] Defective thymocyte apoptosis and accelerated autoimmune diseases in TRAIL-/-mice, Lamhamedi-Cherradi [/bib_ref] Clinical observations have also suggested a role for TRAIL in autoimmune diseases as patients with systemic lupus erythmatosus or multiple sclerosis have elevated serum levels of soluble TRAIL. [bib_ref] TNF-related apoptosis inducing ligand (TRAIL) as a potential response marker for interferon-beta..., Wandinger [/bib_ref] [bib_ref] Soluble TRAIL concentrations are raised in patients with systemic lupus erythematosus, Lub-De Hooge [/bib_ref] TRAIL has also been implicated in cardiovascular problems such as atherosclerosis 49,50 and diabetes. 51
## Regulation of the human trail gene
The human TRAIL gene is tightly regulated, potentially due to its considerable apoptotic potential and its involvement in some immune responses. The first report to clone the TRAIL gene promoter characterized a ~1.6 kB promoter region, which is 97 bp upstream of the start of translation. [bib_ref] Isolation and molecular characterization of the 5'-upstream region of the human TRAIL..., Wang [/bib_ref] This report identified several putative transcription factor binding sites in the TRAIL gene promoter: NHF3, GKLF, AP-1, CEBP, NFAT, GATA and interferon-γ-activated sequence (GAS), GSP1, GSP2 and GSP4. Sequential deletions of the human TRAIL gene promoter driving luciferase reporters in Caco-2 cells suggested that the regions between -1371 to -819 and -165 to -35 contain critical elements . Crystal structure of TRAIL:DR5 complex. Homotrimeric DR5 (gray, light pink,and yellow) bound to homotrimeric TRAIL (cyan, green and magenta). A single zinc atom (green) is found in the center of the complex. The figure was generated using PyMOL software with Protein Data Bank (PBD) accession number 1D4V. [bib_ref] Structure of the TRAIL-DR5 complex reveals mechanisms conferring specificity in apoptotic initiation, Mongkolsapaya [/bib_ref] and/or attenuation of T-cell activation to prevent autoimmunity. NFκB is also responsible for constitutive TRAIL expression in human T-cell leukemia virus type I (HTLV-1)-induced leukemia, though other concomitant effects of NFκB seem to cause resistance to TRAIL-mediated apoptosis. [bib_ref] Resistance to Apo2 ligand (Apo2L)/tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-mediated apoptosis and..., Matsuda [/bib_ref] In accordance with a positive regulatory role for NFκB with respect to the TRAIL gene, the prostaglandin 15d-PGJ2 represses TRAIL gene transcription by inhibiting the binding of NFκB to the NFκB binding site 1 [fig_ref] Figure 4: Sequence analysis of the human TRAIL gene promoter [/fig_ref]. This study also identified the transcription factor HSF-1 as a negative regulator of the TRAIL gene that is involved in 15d-PGJ2-induced repression of TRAIL gene transcription. This negative regulation of HSF-1 is completely dependent on its DNA binding domain and is in contrast to its positive regulation of the FasL gene. [bib_ref] The cyclopentenone-type prostaglandin 15-deoxy-delta 12,14-prostaglandin J2 inhibits CD95 ligand gene expression in..., Cippitelli [/bib_ref] Cnb, the regulatory subunit of calcineurin, was recently shown to active NFκB-mediated TRAIL expression through direct binding to the integrin CD11b. [bib_ref] The calcineurin B subunit induces TNF-related apoptosis-inducing ligand (TRAIL) expression via CD11b-NF-κB..., Su [/bib_ref] P53. Two potential p53 binding sites in the TRAIL gene promoter were identified following the observation that p53-inducing chemotherapies such as 5-fluorouracil and doxorubicin elevated TRAIL promoter activity. [bib_ref] TNFSF10 (TRAIL), a p53 target gene that mediates p53-dependent cell death, Kuribayashi [/bib_ref] One of these binding sites at the -630 position was shown to mediate p53-induced TRAIL promoter activity using luciferase reporter assays. Radiation has also been reported to induce TRAIL expression and it has utilize to upregulate TRAIL and underscores the need to directly test binding sites to accurately delineate transcriptional regulation. These observations also suggested that SP-1 might negatively regulate the TRAIL gene promoter.
SP-1. The study that described NFAT-mediated regulation of the TRAIL promoter also found that SP-1 represses TRAIL gene transcription, at least in human intestinal cells. [bib_ref] NFATc1 regulation of TRAIL expression in human intestinal cells, Wang [/bib_ref] However, a recent study reported a positive regulation of TRAIL gene transcription by SP-1 in vascular smooth muscle cells that appears to involve SP-1 phosphorylated at Thr453. [bib_ref] TRAIL promotes VSMC proliferation and neointima formation in a FGF-2-, Sp1 phosphorylation-,..., Chan [/bib_ref] Interestingly, phosphorylation of SP-1 at Thr453 and Thr739 by p38 mitogen-activated protein kinase appears to be essential for its positive regulation of the VEGF gene. [bib_ref] Heme oxygenase-1/ carbon monoxide induces vascular endothelial growth factor expression via p38..., Lin [/bib_ref] Future studies need to further examine if these differences are due to the difference in cell type or if this phosphorylation event modulates SP-1-mediated TRAIL gene regulation. The mechanism of HDACi-induced upregulation of TRAIL gene transcription appeared to involve SP-3. [bib_ref] Tumor-selective action of HDAC inhibitors involves TRAIL induction in acute myeloid leukemia..., Nebbioso [/bib_ref] NFκB. Inhibition of the prosurvival transcription factor NFκB in Jurkat T cells and primary T lymphocytes revealed that NFκB positively regulates TRAIL gene expression in a manner that depends on the NFκB binding site 1 [bib_ref] Disruption of NF-kappaB signaling reveals a novel role for NF-kappaB in the..., Baetu [/bib_ref] [fig_ref] Figure 4: Sequence analysis of the human TRAIL gene promoter [/fig_ref]. NFκB also upregulates FasL and together, these proapoptotic ligands may be responsible for tumor-cell elimination in immune surveillance . Apoptotic signaling induced by TRAIL. TRAIL initiates cell death by binding to the proapoptotic death receptors DR4 or DR5 that colocalizes their intracellular death domains. This clustering recruits the Fas-associated death domain (FADD) and pro-caspase-8 that results in its activation through autocatalytic cleavage. In type I cells, activate caspase-8 directly activates caspases-3, -6 and -7 to trigger the extrinsic cell death pathway. In type II cells, active caspase-8 cleaves Bid to a truncated form, tBid, which subsequently interacts with proapoptotic Bcl-2 family members Bax and Bak. This interaction leads to permeabilization of the mitochondrial membrane and release of cytochrome c. Cytosolic cytochrome c then combines with Apaf-1 and ATP to form the apoptosome that activates caspase-9 to trigger apoptosis through the caspase cascade.
heat shock has been linked to protection from TRAIL-mediated cell death. [bib_ref] Heat shock protects HCT116 and H460 cells from TRAIL-induced apoptosis, Ozören [/bib_ref] The MAPK pathway. Cells transformed with HRAS G12V were reported to have silenced expression of the TRAIL gene due to hypermethylation of CpG islands in the TRAIL gene promoter that lie ~2,000 bp upstream of the start of transcription. [bib_ref] Transformation-dependent silencing of tumor-selective apoptosis-inducing TRAIL by DNA hypermethylation is antagonized by..., Lund [/bib_ref] Despite this silencing, TRAIL could still be induced by IFN-γ in these transformed cells and the silencing could be reversed with decitabine, a DNA methyltransferase inhibitor. Oncogenic mutations in KRAS during colon cancer progression are a common event in colon cancer. [bib_ref] Detection of high incidence of K-ras oncogenes during human colon tumorigenesis, Forrester [/bib_ref] [bib_ref] Prevalence of ras gene mutations in human colorectal cancers, Bos [/bib_ref] [bib_ref] Genetic alterations during colorectal-tumor development, Vogelstein [/bib_ref] This finding may explain the silencing of TRAIL expression that has been noted in colon cancer, particularly during the progression from adenoma to carcinoma. [bib_ref] Expression of TRAIL (TNF-related apoptosis-inducing ligand) and its receptors in normal colonic..., Koornstra [/bib_ref] Interestingly, oncogenic RAS sensitizes transformed cells to TRAIL-mediated apoptosis, potentially through MEKdependent upregulation of DR4 and DR5. [bib_ref] Oncogenic Ras sensitizes normal human cells to tumor necrosis factor-alpha-related apoptosis-inducing ligand-induced..., Nesterov [/bib_ref] Thus, concomitant silencing of TRAIL expression may be required to prevent induction of TRAIL-mediated apoptosis during Ras transformation that commonly occurs in tumorigenesis.
Other inducers. A recent report found that pigment epithelium-derived factor (PEDF) induced the expression of TRAIL on the surface of macrophages. [bib_ref] Pigment epithelium-derived factor (PEDF) promotes tumor cell death by inducing macrophage membrane..., Ho [/bib_ref] Further analysis revealed that PEGF also induces peroxisome proliferator-activated receptor-gamma (PPARγ), which binds to the TRAIL promoter to upregulate transcription at a PPAR-response element (PPRE). This represents yet another mechanism utilized by immune cells to induce the expression of TRAIL. Other molecules have been reported to induce TRAIL gene transcription but the underlying direct transcriptional mechanism has not been elucidated. Lipopolysaccharide (LPS), which is a component of been hypothesized that TRAIL mediates the bystander effects observed following radiation. [bib_ref] TRAIL induction by radiation in lymphoma patients, Unnithan [/bib_ref] [bib_ref] Role of tumor necrosis factor-α and TRAIL in high-dose radiation-induced bystander signaling..., Shareef [/bib_ref] The role of p53 in radiationinduced TRAIL expression should be explored since it is well accepted that ionizing radiation robustly induces p53.
FOXO. Overexpression of Foxo3a in prostate cancer cells was found to induce TRAIL gene transcription by gene expression profiling. [bib_ref] FOXO proteins regulate tumor necrosis factor-related apoptosis inducing ligand expression. Implications for..., Modur [/bib_ref] In silico analysis revealed a binding site in the TRAIL gene promoter between -121 and -138 that was validated by luciferase reporter construct mutations and EMSAs to be completely responsible for Foxo3a-induced TRAIL gene promoter activity. This study also revealed that TRAIL expression is significantly decreased in metastatic prostate cancer. Foxo3a-dependent TRAIL upregulation has been recently described as the apoptotic mechanism responsible for memory B-cell loss that results from chronic HIV infection. [bib_ref] Loss of memory B cells during chronic HIV infection is driven by..., Van Grevenynghe [/bib_ref] Unexplored binding sites. Clearly there are several putative transcription factor-binding sites within the TRAIL gene promoter that have been unexplored. It is worth noting that only a few of the examined binding sites for a given transcription factor have turned out to be functionally important, as in the case of NFκB, AP-1 and SP-1, which have multiple putative binding sites in the TRAIL gene promoter. For example, Oct-1 is involved in regulating several housekeeping genes and has a putative binding site on the TRAIL promoter. Interestingly, HDACi have been found to induce Gadd45 expression through Oct-1. [bib_ref] p53-independent induction of Gadd45 by histone deacetylase inhibitor: coordinate regulation by transcription..., Hirose [/bib_ref] This observation along with the report that HDACi induces TRAIL gene transcription suggests that the direct examination of Oct-1 binding to the TRAIL gene promoter may be worth further investigation. The role of the heat shock elements in the proximal region of the TRAIL gene promoter may also be worth investigation as NFκB and the authors note the possibility that truncated TRAIL may have intracellular activity. Future studies should directly gram-negative bacteria, has been reported to induce TRAIL gene transcription at higher concentrations,though an early report did not observe TRAIL-induction by LPS at lower doses. [bib_ref] Monocyte-mediated tumoricidal activity via the tumor necrosis factorrelated cytokine, TRAIL, Griffith [/bib_ref] Regulation of TRAIL activity by isoforms. The human TRAIL gene spans ~20 kb and contains five exons and four introns that contain typical splice acceptor-AG/GT-splice donor consensus sites at their boundaries. [bib_ref] Genomic organization and transcriptional regulation of human Apo2/TRAIL gene, Gong [/bib_ref] [bib_ref] RNA splice junctions of different classes of eukaryotes: sequence statistics and functional..., Shapiro [/bib_ref] The first exon encodes for the 21 amino acid transmembrane domain and the 17 amino acid cytoplasmic domain. Exons 4 and 5 encode for the amino acids in the extracellular domain that are responsible for the interaction of TRAIL with its receptors. Exon 5 also encodes for the C-terminal amino acids along with containing the 3'-UTR and poly-A tail.
TRAIL is well known for its potent and cancer-selective apoptotic activity. However, it seems that this activity is unique to only one specific isoform of TRAIL among the nine variants that have been reported to date [fig_ref] Figure 5: Genomic structure of human TRAIL variants [/fig_ref]. The first report of TRAIL isoforms identified three variants that involved variable inclusion of exons 2 and 3: the full length TRAILα, TRAILβ that lacks exon 3, and TRAILγ that lacks exons 2 and 3. [bib_ref] TRAIL-beta and TRAIL-gamma: two novel splice variants of the human TNF-related apoptosis-inducing..., Krieg [/bib_ref] Computational analysis of exons 2 and 3 revealed that both exons were flanked by consensus splice donor and splice acceptor sequences that are involved in post-translational alternative splicing. Interestingly, TRAILα and TRAILβ where localized to the cytoplasm whereas TRAILγ was associated with the nuclear and cell surface membrane. Sequence analysis of TRAIL mRNA in granulosa tumor cells identified TRAILδ, which lacks exons 3 and 4. [bib_ref] Cisplatinmediated sensitivity to TRAIL-induced cell death in human granulosa tumor cells, Woods [/bib_ref] The truncated TRAIL isoforms do not induce apoptosis like the full-length TRAIL (TRAILα) and it has been suggested that these variants are negative regulators of their full-length counterpart. No TRAIL variants other than the full length TRAIL have been reported in murine cells.
Recently, seven alternatively spliced TRAIL truncated variants were identified that were incapable of potentiating apoptosis: AK, E2, E3, E4, DA, BX424 and BX439. [bib_ref] Novel transcript variants of TRAIL show different activities in activation of NF-κB..., Wang [/bib_ref] All of these isoforms contain common N-terminal sequences and possess the transmembrane helix but vary in the C-terminal region. The DA isoform lacks exon 3 and encodes for the same protein as TRAILβ and BX424 lacks exons 3 and 4, yielding the same protein as TRAILδ. BX439 completely lacks exons 2-4 but possesses the same exons 1 and 5 as the full length TRAIL. AK and DA contain a unique exon not shared by any other reported isoforms. E2, E3 and E4 contain exons 1-2, 1-3 and 1-4, respectively, with an extended sequence at C-terminus. All of these variants, however, activated changes in TRAIL mRNA levels or with multiple sclerosis as a diagnostic or prognostic marker. The authors of the study note that SNP(1) is in an AP1 binding site, though they do not comment on other binding sites predicted at the other SNP sites. Interestingly, both SNP3 and SNP4 each lie in p53 response element half-sites that was previously shown to be important for TRAIL gene upregulation in response to chemotherapy. [bib_ref] TNFSF10 (TRAIL), a p53 target gene that mediates p53-dependent cell death, Kuribayashi [/bib_ref] While these SNPs were not associated with basal TRAIL levels in this population, it would be interesting to determine if these SNPs affect p53-induced TRAIL gene expression and response in chemotherapy-treated cancer patients.
A SNP at the -716 position of the TRAIL gene promoter was identified but not associated with prostate cancer. Other TRAIL gene promoter SNP associations have been identified in other disease settings but have not been evaluated for its functional effects on TRAIL gene transcription such as -1525/-1595 in fatty liver disease. [bib_ref] sTRAIL levels and TRAIL gene polymorphisms in Chinese patients with fatty liver..., Yan [/bib_ref] SNPs have been identified in coding regions such as position 1595 in exon 5 being linked with multiple sclerosis. [bib_ref] TNF-related apoptosis inducing ligand (TRAIL) gene polymorphism in Japanese patients with multiple..., Kikuchi [/bib_ref] A small cohort study of healthy volunteers found five SNPs in TRAIL exons: three in the 3'-UTR at 1525, 1588 and 1595 whereas the other two were in in exon 1 and two at positions 192 and 912. [bib_ref] Three polymorphisms in the 3' UTR of the TRAIL (TNFrelated apoptosis-inducing ligand)..., Gray [/bib_ref] These two polymorphisms in the coding region do not alter the encoded amino acid sequence. Another SNP study in bronchial asthma patients found five SNPs in the TRAIL gene with one SNP being in a coding region at 825 and five SNPs examine the ability of these various TRAIL isoforms to bind to TRAIL receptors through in vitro studies.
The abundance of these isoforms varied significantly across the tested cancer cell lines with THP-1 human leukemia cell line and BT-325 human glioma cell line expressing most of these variants. A notable exception among the cell line panel was the human colon cancer cell line HCT116 that only expressed the full-length isoform.
SNPs in the TRAIL gene. Due to the altered expression of TRAIL in various disease settings, there have been multiple efforts to identify single nucleotide polymorphisms (SNPs) in various patient populations. SNP analysis of peripheral blood samples found that having a T instead of a C at position -723 was significantly associated with sporadic breast cancer and decreased TRAIL mRNA levels (SNP1) [bib_ref] Functional implication of TRAIL -716 C/T promoter polymorphism on its in vitro..., Pal [/bib_ref] [fig_ref] Figure 4: Sequence analysis of the human TRAIL gene promoter [/fig_ref]. Luciferase reporter assays in cell lines indicated that this mutation from C to T indeed repressed TRAIL transcriptional activity. In silico analysis found that this mutation is predicted to create an SP-1 binding site. The authors propose that SP3 is negatively regulating this SP1 site, though future studies will need to validate the mutation-induced putative binding site and directly evaluate this possibility.
Another SNP analysis of the TRAIL gene promoter revealed four SNPs that were highly polymorphic in healthy individuals. [bib_ref] Identification and functional characterization of a highly polymorphic region in the human..., Weber [/bib_ref] However, these SNPs were not significantly associated with
# Conclusion
The role of TRAIL as an effector molecule in the immune system and its apoptotic potential is reflected in the regulation of the TRAIL gene. Regulation of the TRAIL gene appears tightly controlled by transcriptional mechanisms that respond to interferon stimulation or are involved in immune cell activation as well as transcription factors that are tumor suppressors such as p53 and Foxo3a. Assimilating our current knowledge of TRAIL gene regulation, it is clear that the regulation is multimodal and can be highly context-dependent. The emerging evidence on TRAIL isoforms and TRAIL gene SNPs add other layers of complexity to how cells can tune the TRAIL gene to alter its effects. Mapping transcription factor binding sites and SNPs in the TRAIL gene promoter revealed several overlapping sites that should be examined in future studies. A detailed understanding of TRAIL gene upregulation will be better defined how TRAIL is utilized by the immune system and how it is altered in diseases. Such studies also have the potential to yield drug targets that can harness the antitumor potential of TRAIL. being in the 3'UTR at 1053, 1202, 1438, 1501 and 1508. [bib_ref] Association studies of 33 single nucleotide polymorphisms (SNPs) in 29 candidate genes..., Unoki [/bib_ref] Future studies should further investigate the functional effects of identified SNPs on TRAIL gene transcription and therapeutic response to chemotherapies.
TRAIL expression in disease and physiology. Altered expression levels of TRAIL been noted in several diseases relative to healthy controls. For instance, TRAIL mRNA levels are elevated in patients with multiple sclerosis. [bib_ref] Apoptosis mediators fasL and TRAIL are upregulated in peripheral blood mononuclear cells..., Huang [/bib_ref] Patients with systemic lupus erythematous or multiple sclerosis have elevated serum levels of soluble TRAIL. [bib_ref] TNF-related apoptosis inducing ligand (TRAIL) as a potential response marker for interferon-beta..., Wandinger [/bib_ref] [bib_ref] Soluble TRAIL concentrations are raised in patients with systemic lupus erythematosus, Lub-De Hooge [/bib_ref] These clinical observations support a critical role for TRAIL in preventing autoimmune disorders, which is in line with the observation that TRAIL-knockout mice cannot induce thymocyte apoptosis. [bib_ref] Defective thymocyte apoptosis and accelerated autoimmune diseases in TRAIL-/-mice, Lamhamedi-Cherradi [/bib_ref] Breast cancer patients with brain metastases have downregulated TRAIL mRNA levels. [bib_ref] Genes that mediate breast cancer metastasis to the brain, Bos [/bib_ref] TRAIL gene expression silencing has also been noted in metastatic prostate cancer [bib_ref] FOXO proteins regulate tumor necrosis factor-related apoptosis inducing ligand expression. Implications for..., Modur [/bib_ref] and colon cancer. [bib_ref] Expression of TRAIL (TNF-related apoptosis-inducing ligand) and its receptors in normal colonic..., Koornstra [/bib_ref] A role for TRAIL has been implicated in particular types of cellular differentiation such as colonic epithelial cells through a reciprocal expression relationship with PKCε. [bib_ref] TRAIL up-regulation must be accompanied by a reciprocal PKCε down-regulation during differentiation..., Gobbi [/bib_ref] TRAIL-receptor signaling has also been implicated in the DNA damage response to radiation as well as the late effects of radiation. [bib_ref] DR5 knockout mice are compromised in radiation-induced apoptosis, Finnberg [/bib_ref] [bib_ref] TRAIL-R deficiency in mice promotes susceptibility to chronic inflammation and tumorigenesis, Finnberg [/bib_ref]
[fig] Figure 3: Molecules that alter human TRAIL gene transcription. Interferons (IFN) activate TRAIL gene transcription through ISRE and IRFE sequences in the promoter region. Mutant HRAS (G12V) silences TRAIL gene expression through hypermethylation of CpG islands in the TRAIL gene promoter. Green arrows indicate activating relationships and the red lines indicate inhibitory relationships. [/fig]
[fig] Figure 4: Sequence analysis of the human TRAIL gene promoter. Putative binding sites indicated by highlights above the appropriate sequences. Nucleotides contained in two or more putative binding sites are highlighted in red. Binding sites that have been empirically demonstrated to affect TRAIL promoter activity in experiment are bolded. Vertical lines below the sequence indicate SNPs. TRAIL sequence obtained from accession number AF178756. [/fig]
[fig] Figure 5: Genomic structure of human TRAIL variants. Exonic sequences of the full-length TRAIL are shown in green whereas novel sequences are shown in yellow. [/fig]
|
Evaluating Crossbred Red Rice Variants for Postprandial Glucometabolic Responses: A Comparison with Commercial Varieties
Consumption of white rice predisposes some Asian populations to increased risk of type 2 diabetes. We compared the postprandial glucometabolic responses to three newly-developed crossbred red rice variants (UKMRC9, UKMRC10, UKMRC11) against three selected commercial rice types (Thai red, Basmati white, Jasmine white) using 50-g carbohydrate equivalents provided to 12 normoglycaemic adults in a crossover design. Venous blood was drawn fasted and postprandially for three hours. Glycaemic (GI) and insulin (II) indices, incremental areas-under-the-curves for glucose and insulin (IAUC ins ), indices of insulin sensitivity and secretion, lactate and peptide hormones (motilin, neuropeptide-Y, orexin-A) were analyzed. The lowest to highest trends for GI and II were similar i.e., UKMRC9 < Basmati < Thai red < UKMRC10 < UKMRC11 < Jasmine. Postprandial insulinaemia and IAUC ins of only UKMRC9 were significantly the lowest compared to Jasmine. Crude protein and fiber content correlated negatively with the GI values of the test rice. Although peptide hormones were not associated with GI and II characteristics of test rice, early and late phases of prandial neuropeptide-Y changes were negatively correlated with postprandial insulinaemia. This study indicated that only UKMRC9 among the new rice crossbreeds could serve as an alternative cereal option to improve diet quality of Asians with its lowest glycaemic and insulinaemic burden.
# Introduction
Conventional rice breeding programs have long focussed on developing new rice varieties with improved yield-associated traits and micronutrient capacity to meet food security and nutritional needs of developing countries in Asia [bib_ref] Nutritional enhancement of rice for human health: The contribution of biotechnology, Bhullar [/bib_ref]. However, epidemiologic evidence suggests that greater white rice intake was associated with significantly higher odds of developing metabolic syndrome and type 2 diabetes in Asian Chinese and Japanese populations [bib_ref] White rice consumption and risk of type 2 diabetes: Meta-analysis and systematic..., Hu [/bib_ref]. The high glycaemic index (GI) of rice renders this food staple the major contributor of glycaemic load in Asian diets. However, disparities in GI values exist even for the same variety of rice. The glycaemic variability can be largely attributed to the inherent starch characteristics of specific cultivars, although within a given rice variety, the mode of post-harvesting processing and at-home preparation also bear considerable influence on starch digestibility [bib_ref] Prospects in development of quality rice for human nutrition, Se [/bib_ref] [bib_ref] A systematic review of the influence of rice characteristics and processing methods..., Boers [/bib_ref].
Rice improvement programs have not considered the development of varieties targeting the prevention and management of type 2 diabetes because the genetics of GI remain unclear. Natural polymorphisms in the starch biosynthesis related genes, such as granule bound starch synthase I (Waxy), branching enzyme I and glucose 6-phosphate translocator could potentially modulate the glycaemic potencies of rice by changing its amylose content and retrogradation rate. A large-scale phenotyping of 235 rice varieties found that Waxy gene was the strongest predictor for GI variability. Therefore, lowering the GI characteristic of this dietary staple via breeding technologies could improve the glycaemic burden and diet quality of rice-based diets in Asians [bib_ref] Prospects in development of quality rice for human nutrition, Se [/bib_ref]. To date, new white rice varieties with lower GI values have been produced through marker-assisted breeding [bib_ref] Glycemic index of a novel high-fiber white rice variety developed in India-A..., Mohan [/bib_ref] or by increasing resistant starch content via genetic modification [bib_ref] Postprandial glycaemic and insulinaemic responses to GM-resistant starch-enriched rice and the production..., Li [/bib_ref] [bib_ref] High amylose white rice reduces post-prandial glycemic response but not appetite in..., Zenel [/bib_ref].
In Malaysia, three new transgressive variants with red pericarp grain were derived from advanced backcrosses between a wild rice accession, Oryza rufipogon Griff. IRGC105491 and a Malaysian high-yielding rice cultivar, Oryza sativa L. subsp. indica cv. MR219 [bib_ref] Development and evaluation of advanced backcross families of rice for agronomically important..., Sabu [/bib_ref]. According to a Distinctiveness, Uniformity and Stability test for evaluating New Plant Variety status (Protection of New Plant Varieties Act 2004), UKMRC9 and UKMRC10 were different but UKMRC9 and UKMRC11 were similar for 58 morphological and physiological traits. Experimental field trials and physicochemical analyses have confirmed the superiority of these red rice variants in yield potential, resistance against blast disease and antioxidant properties compared to MR219 [bib_ref] Transgressive variants for red pericarp grain with high yield potential derived from..., Bhuiyan [/bib_ref] [bib_ref] Red pericarp advanced breeding lines derived from Oryza rufipogonˆOryza sativa: Physicochemical properties,..., Fasahat [/bib_ref]. Karupaiah et al. [bib_ref] A transgressive brown rice mediates favourable glycaemic and insulin responses, Karupaiah [/bib_ref] previously experimented with UKMRC9 and found it had low glycaemic and insulinogenic properties, but these favorable traits were lost upon polishing.
In this study, postprandial glucometabolic evaluations were applied to a wider range of red rice variants (UKMRC10, UKMRC11) and compared against imported specialty rice varieties, namely Basmati white, Jasmine white and Thailand red rice. As secondary outcomes, we have evaluated the postprandial effect of these rice types on plasma lactate and peptide hormones. The inclusion of lactate was in light of recent observations relating elevated fasting plasma lactate with various metabolic aberrations, such as insulin resistance, type 2 diabetes and reduced oxidative capacity [bib_ref] Comprehensive review on lactate metabolism in human health, Adeva-Andany [/bib_ref]. Therefore, the key questions addressed in this study were:
## ‚
Do the related crossbred red rice variants reflect similar glycaemic and insulin indices (II) with UKMRC9?
‚ What is the relationship between nutrient content and cooking characteristics of the six rice types with the GI and II characteristics?
‚ Does consumption of rice with varying GI values have a role to play in modulating postprandial insulin sensitivity, pancreatic β-cell function and peptide hormones?
# Materials and methods
## Test and reference food
The wild parent Oryza rufipogon Griff. (IRGC105491) and Oryza sativa L. subsp. indica cv. MR219, a Malaysian high-yielding rice cultivar were backcrossed to produce advanced breeding lines (BC 2 F 5 and BC 2 F 6 generations) where transgressive variants with red pericarp grain and higher grain yield were detected [bib_ref] Development and evaluation of advanced backcross families of rice for agronomically important..., Sabu [/bib_ref]. The three red pericarp genotypes tested in this study are registered as New Plant Variety under the Ministry of Agriculture, Malaysia, namely UKMRC9 (PBR 0032), UKMRC10 (PBR 0033) and UKMRC11 (PVBT041/09). Additionally, Basmati white rice (IndiaGate, KRBL Ltd., Hyderabad, India), Thailand red rice (NutriRice, Jasmine Food Corp. Pte. Ltd., Kuala Lumpur, Malaysia) and Jasmine white rice (Rambutan AAA, Jasmine Food Corp. Pte. Ltd., Kuala Lumpur, Malaysia) were the commercial varieties tested. All red rice varieties were tested in dehusked form whilst Basmati and Jasmine rice were polished. For each rice type, different packaged samples were mixed together as a single sample for homogeneity, before re-packaging into 500 g portions and then storing at 2˘1˝C.
Fifty grams of unflavoured dextrose monohydrate were dissolved in 250 mL water and served as the reference food (Glucolin™, The Boots Company, Nottingham, UK). Glucose standard was tested twice at the first and last feeding sessions to obtain an average value for the determination of plasma glucose and insulin standards [bib_ref] Glycaemic index methodology, Brouns [/bib_ref].
## Chemical composition of rice
Rice samples were analyzed for crude protein (method 981.10), fat (method 991.36), moisture (method 950.46), ash (method 923.03) and total dietary fiber (method 991.43) content by the Association of Official Analytical Chemists methods. Total carbohydrate content was calculated by proximate difference. Total energy content was calculated by multiplying the nutrient content with Atwater conversion factors 4, 9 and 4 kcal/g for protein, fat and carbohydrate respectively. Amylose content was determined using the iodine colorimetric method [bib_ref] A simplified assay for milled-rice amylose, Juliano [/bib_ref]. Total phenolic content was evaluated using the Folin-Ciocalteu method [bib_ref] Red pericarp advanced breeding lines derived from Oryza rufipogonˆOryza sativa: Physicochemical properties,..., Fasahat [/bib_ref].
## Rice preparation for postprandial testing
The rice was cooked to completion in an electronic rice cooker (Panasonic model SR-WN36, Osaka, Japan) using a consistent rice-to-water ratio of 1:2 (w/w). Rice samples were portioned and weighed when cooled to room temperature according to 50 g available carbohydrate content.
## Subject recruitment and screening procedures
Twelve non-obese subjects (5 men and 7 women), aged 21-40 years and without history of chronic disease(s) were enrolled into the study. Exclusion criteria included pregnant or nursing women or subjects receiving pharmacotherapy that would interfere with glucose metabolism, smokers, consuming alcohol or on low-calorie diets. Eligible participants underwent baseline medical, dietary practices and blood chemistry screening. Their baseline characteristics were as follows (mean˘SD): age = 23.2˘1.4 years; body mass index = 22.1˘3.1 kg/m 2 ; fasting plasma glucose = 5.020
.32 mmol/L; fasting plasma insulin = 6.17˘2.07 mU/L. None of the subjects were insulin resistant as their individual homeostatic model assessment of insulin resistance (HOMA-IR) scores were below 2.6.
## Experimental protocol
A crossover design was adopted with all subjects completing eight postprandial evaluations on separate mornings with a one-week washout period between rotations. Before each test rotation, subjects refrained from strenuous physical activity or sport games and maintained their customary dietary intake for 48 h prior to testing days. On testing days, fasted subjects reported to the feeding laboratory between 07:00 and 08:00 h. Subjects rested for 15 min prior to blood collection. Fasting blood samples (0 min) were drawn before subjects consumed 50 g carbohydrate equivalents of test rice with 250 mL of plain water within 10 min. Subjects continued resting until all postprandial blood samplings at 15, 30, 45, 60, 90, 120 and 180 min were completed. The study protocol was approved by the Research Ethics Committee of Universiti Kebangsaan Malaysia (registration No.: NN-069-2012) and written informed consent was obtained from all subjects prior to study commencement.
## Blood sampling, processing and storage procedures
Fasting and postprandial venous blood samples were obtained via antecubital venipuncture, with the subject's arm alternated for each sequential blood draw. Blood samples were collected into three evacuated Vacutainer ® tubes containing ethylenediaminetetraacetic acid (0.117 mL of 15% EDTA), lithium heparin or sodium fluoride-potassium oxalate, respectively (Becton Dickinson Vacutainer, Franklin Lakes, NJ, USA). The tubes were centrifuged for 10 min at 3000 rpm and collected plasma was subsequently aliquoted, snap-frozen using liquid nitrogen and then stored at´80˝C for later analysis.
## Biochemical analyses
(i). Plasma glucose: Plasma glucose concentrations (mmol/L) were quantified using a Roche Modular P800 (Roche Diagnostics, Tokyo, Japan) automated analyzer by the enzymatic hexokinase method [bib_ref] Die enzymatische bestimmung von glucose und fructose nebeneinander, Schmidt [/bib_ref]. The assay had a detection limit of 0.11 mmol/L and the intra-and inter-assay coefficients of variation (CV) were <2.0%.
(ii) Plasma insulin: Heparinized plasma samples were analyzed for insulin concentrations (mU/L) using electrochemiluminescence immunoassay on the Modular Analytics E170 system (Roche Diagnostics, Tokyo, Japan). The fully-automated assay adopts a solid-phase, two-site, enzyme-labeled immunoassay based on the direct sandwich technique [bib_ref] Assays for insulin, proinsulin(s) and C-peptide, Clark [/bib_ref]. The intra-and inter-assay CVs were <5%, with a lower detection limit of 0.20 mU/L. (iii) Plasma lactate: The plasma L-lactate concentration (mmol/L) was assayed on a Roche Modular P800 analyser (Roche Diagnostics, Tokyo, Japan) using the lactate oxidase method [bib_ref] Test-strip method for measuring lactate in whole blood, Shimojo [/bib_ref]. The assay had a detection range between 0.22 and 15.5 mmol/L and inter-assay CV of 2.0%. (iv) Peptide hormones: Plasma concentrations of motilin (EK-045-04), neuropeptide-Y (EK-049-03) and orexin-A (EK-003-30) were determined in duplicate using commercially-available enzyme immunoassay (EIA) kits from Phoenix Pharmaceuticals (Burlingame, CA, USA), as described previously [bib_ref] Acute effects of different glycemic index diets on serum motilin, orexin and..., Wu [/bib_ref]. The enzyme-linked immunosorbent assay (ELISA) was performed according to the manufacturer's protocol and absorbance was read with a Tecan Infinite M200 microplate reader (Tecan Group Ltd., Mannedorf, Switzerland). Plasma concentrations were calculated using four-parameter non-linear logistic curve fitting (Magellan Data Analysis Software v. 311 for PC, Tecan Group Ltd., Mannedorf, Switzerland). The standard curve plots were generated using the five standard concentrations ranged from 0.01 to 100 ng/mL. The coefficients of determination for standard curves were >0.97.
For the motilin EIA, the minimum detectable concentration was 0.14 ng/mL, the intra-and inter-assay CVs were <5% and <9%, respectively. The neuropeptide-Y EIA had a lower detection limit of 0.09 ng/mL, the intra-and inter-assay CVs of <4% and <8.5%, respectively. The orexin-A EIA had a minimum detectable concentration of 0.22 ng/mL with the intra-and inter-assay CVs of <6% and <10%, respectively.
## Outcome measures
(i) Quality control: The mean intra-individual CV for glycaemic response after two 50 g glucose standard loads was 21.3%, which was in concordance with the recommended CV < 30% required for precision and accuracy [bib_ref] Measuring the glycemic index of foods: Interlaboratory study, Wolever [/bib_ref]. (ii) Glucometabolic markers: Kinetic markers of incremental glucose and insulin peaks are defined as maximum increases in plasma glucose and insulin concentrations obtained at any point after a test rice or glucose challenge. Incremental areas-under-the-curves (IAUC), excluding areas beneath fasting values, for plasma glucose, insulin and lactate were calculated geometrically using the trapezoidal method. The GI and II were calculated by dividing the net IAUC generated from the 3 h postprandial plasma glucose-/insulin-timed responses of the test food with that by the standard glucose load (GI and II = 100), with each subject being their own reference. Individual GI or II scores differing from the mean value by >2 standard deviations (outliers) were excluded from the dataset [bib_ref] Measuring the glycemic index of foods: Interlaboratory study, Wolever [/bib_ref].
Matsuda Index " 10000{pG 0ˆI0ˆGmeanˆImean q 0.5
where G 0 = baseline glucose concentration (mg/dL); I 0 = baseline insulin concentration (mU/L); G mean = mean glucose throughout 2 h postprandial (mg/dL); I mean = mean insulin throughout 2 h postprandial (mU/L) IGI were further adjusted by HOMA-IR (IGI/HOMA-IR) and fasting plasma insulin (IGI/FPI) to account for between-subjects variations in insulin sensitivity.
## Statistical analyses
The crossover design allowed each subject to serve as his/her own control for the eight postprandial rotations. All data were assessed for normality using the Shapiro-Wilk test. Data are presented as mean˘standard error of the mean (SEM) unless otherwise stated.
Postprandial changes in plasma glucose, insulin, lactate and peptide hormones for the six test rice were analyzed using the general linear model (GLM) for repeated measures to examine the time, diet and time x diet interaction effects. As the data distribution of peptide hormones across most timed intervals did not meet normality assumption, data were log-transformed prior to GLM analysis. Greenhouse-Geisser correction for degrees of freedom was used when Mauchly's test of sphericity was significant. Bonferroni-corrected post hoc comparisons were used when the main effects were significant. Calculated GI, II, IAUC and insulin kinetics data for each rice type were compared using univariate analysis of variance, followed by Tukey's post hoc test. Magnitude of the significant paired difference was assessed using partial eta-squared (η p 2 ). Bivariate associations were examined using
Pearson's correlation test. Statistical significance was pre-set at p < 0.05. All analyses were computed using SPSS ® for Windows™ applications (Version 16.0; SPSS Inc., Chicago, IL, USA).
# Results
## Proximate composition and cooking characteristics of rice
Proximate nutrient composition of the six test rice varieties is presented in [fig_ref] Table 1: Proximate nutrient composition [/fig_ref]. Crude protein content in UKMRC9 and Basmati were similar to Thai red rice (p > 0.05) but significantly higher compared to UKMRC11 and Jasmine (p < 0.05). Both Basmati and Jasmine had the lowest crude lipid, total dietary fiber, ash, and phenolic content (all p < 0.001). Polished Basmati and Jasmine required relatively shorter duration for cooking to completion, compared to dehusked red rice cultivars, which took more than 40 min.
## Glucometabolic responses
Postprandial glycaemia for the six rice types were not significantly different (p = 0.065, η p 2 = 0.143).
However, a marginal significance for glycaemic response was noted between UKMRC9 and Jasmine (p = 0.056) [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref] , which was also reflected by their IAUC glu (p = 0.06) [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref]. UKMRC9 (GI = 46˘7.7) and Basmati (GI = 50˘5.8) were categorized as low GI, whilst Thai red (GI = 55˘8.6), UKMRC10 (GI = 59˘8.8) and UKMRC11 (GI = 63˘8.6) were categorized as intermediate GI. Jasmine (GI = 77˘7.3) was the only test rice in the high GI category [fig_ref] Table 2: Kinetic markers of postprandial glycaemic and insulin responses [/fig_ref]. GI (p = 0.093), maximum (p = 0.074) and incremental (p = 0.063) glucose peak values were not significantly different between the six rice types. Notably, after consuming Jasmine compared to Basmati, the time taken was significantly greater to reach maximum concentration of plasma glucose (p = 0.021). Postprandial insulinaemic trends of the six test rice were in tandem with their postprandial glycaemic effects [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref] , and diet effects on postprandial insulinaemia were significant (p = 0.013, η p 2 = 0.194). Post hoc comparison indicated UKMRC9 elicited a marginally significant lower insulin generation compared to UKMRC11 (p = 0.083) and Jasmine (p = 0.052) [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref]. For IAUC ins , Jasmine induced the highest postprandial insulinaemia amongst all rice types, which was significant compared to both Basmati (p = 0.032) and UKMRC9 (p = 0.033) [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref]. However, the IAUC ins between the crossbred red rice variants were not significantly different, except for a marginal difference between UKMRC11 and UKMRC9 (p = 0.069).; * One subject was excluded (n = 11) as individual GI and II values >2 standard deviations from the respective mean GI and II scores [bib_ref] Measuring the glycemic index of foods: Interlaboratory study, Wolever [/bib_ref] ; GI, glycaemic index; GLU-C max , maximum concentration of postprandial plasma glucose; GLU-∆ peak , incremental glucose peak; GLU-T max , time taken to reach GLU-C max ; GLU-T ∆0 , time taken for returning of plasma glucose to baseline levels; IGI/FPI, ratio of insulinogenic index to fasting plasma insulin; IGI/HOMA-IR, ratio of insulinogenic index to homeostatic model assessment of insulin resistance; II, insulin index; INS-C max , maximum concentration of postprandial plasma insulin; INS-∆ peak , incremental insulin peak; η p 2 = partial eta-squared, 0.01, 0.06 and 0.14 were used to denote small, moderate and large differences in measured outcomes, respectively. Both UKMRC9 (II = 51˘5.3, p = 0.043) and Basmati (II = 52˘5.3, p = 0.059) had the lowest II values compared to Jasmine (II = 76˘7.1) [fig_ref] Table 2: Kinetic markers of postprandial glycaemic and insulin responses [/fig_ref]. However, the II values of the three crossbred red rice varieties were not significantly different (p > 0.05). A marginal significant difference was observed in the Matsuda index of insulin sensitivity between rice types (p = 0.058). Conversely, postprandial pancreatic β-cell function, as modelled by IGI/HOMA-IR and IGI/FPI, did not differ significantly between rice types (p > 0.05).
Postprandial lactate responses and IAUC lac were not significantly different between six rice types (p > 0.05) [fig_ref] Figure 1: Baseline-adjusted trends [/fig_ref].
## Correlation between nutrient composition, cooking characteristics, gi and ii
Only crude protein (r =´0.357, p = 0.002) and total dietary fiber (r =´0.237, p = 0.047) content correlated negatively with GI values of the six test rice [fig_ref] Table 3: Correlation between nutrient content, cooking characteristics, glycaemic and insulin indices of test... [/fig_ref] , whereas only crude protein was inversely related with II values (r =´0.385, p = 0.001). No significant correlations were observed between cooking characteristics of test rice with the GI and II properties (p > 0.05).
## Postprandial changes in plasma motilin, neuropeptide-y and orexin-a
Log-transformed values corrected to baseline were reported for plasma motilin, neuropeptide-Y and orexin-A [fig_ref] Figure 2: Baseline-adjusted trends in postprandial plasma [/fig_ref]. Biphasic secretory responses peaking at 30-and 90-min were observed for motilin and orexin-A, but between-diet effects were not significant for the six test rice (p = 0.804 for motilin; p = 0.162 for orexin-A). Contrarily, biphasic neuropeptide-Y responses were only evident after consuming UKMRC9, Basmati and Jasmine. Pairwise comparisons demonstrated that postprandial neuropeptide-Y trend for Thai red differed significantly from UKMRC9, Basmati and Jasmine (all p < 0.01) (Supplementary [fig_ref] Table 1: Proximate nutrient composition [/fig_ref].
# Discussion
## Moderators of gi
This study confirmed the low GI characteristic of UKMRC9, as the mean GI values derived from current and previous studies [bib_ref] A transgressive brown rice mediates favourable glycaemic and insulin responses, Karupaiah [/bib_ref] were similar (46 vs. 51). However, despite originating from the same gene pool [bib_ref] Transgressive variants for red pericarp grain with high yield potential derived from..., Bhuiyan [/bib_ref] and having similar amylose content, we found that both UKMRC10 and UKMRC11 did not exhibit low GI properties compared to UKMRC9. Wang [bib_ref] The amylose content in rice endosperm is related to the post-transcriptional regulation..., Wang [/bib_ref] was the first to attribute the Waxy gene as the regulator of amylose synthesis in rice and Fitzgerald et al.proposed that this gene was associated with GI variations in 235 rice samples. However, in view that these crossbred red rice variants had similar amylose content, this study does not corroborate the Waxy gene-amylose content link to GI potential of rice.
The glucose-raising potential of high-amylose rice varieties is reported to be lower than rice with greater amylopectin content [bib_ref] High amylose white rice reduces post-prandial glycemic response but not appetite in..., Zenel [/bib_ref] [bib_ref] The effect of apparent amylose content and dietary fiber on the glycemic..., Trinidad [/bib_ref]. The compact linear chain of amylose has been hypothesized to delay digestion by amylase enzymes in the human gastrointestinal tract, permitting a sustained release of glucose into bloodstream [bib_ref] Rice varieties with similar amylose content differ in starch digestibility and glycemic..., Panlasigui [/bib_ref]. We did not find any significant relationship between amylose content and GI of test rice, in concordance with some studies [bib_ref] A transgressive brown rice mediates favourable glycaemic and insulin responses, Karupaiah [/bib_ref] [bib_ref] Rice varieties with similar amylose content differ in starch digestibility and glycemic..., Panlasigui [/bib_ref] [bib_ref] Glycaemic index of some commercially available rice and rice products in Great..., Ranawana [/bib_ref] , but not all [bib_ref] High amylose white rice reduces post-prandial glycemic response but not appetite in..., Zenel [/bib_ref] [bib_ref] The effect of apparent amylose content and dietary fiber on the glycemic..., Trinidad [/bib_ref]. The lack of association observed in this study could be attributed to the relatively narrow range of amylose content studied (18%-23%), contrasting with other research comparing the GI values of rice cultivars across a wide range of amylose categories (0%-2% vs. >20%) [bib_ref] High amylose white rice reduces post-prandial glycemic response but not appetite in..., Zenel [/bib_ref] [bib_ref] The effect of apparent amylose content and dietary fiber on the glycemic..., Trinidad [/bib_ref]. However, this study found that the amylose content of Jasmine was higher (23%) compared to those commonly reported, ranging from 12% to 17%. Cooked rice starches with higher amylose content undergo retrogradation upon storage, causing an increase in resistant starch content and lowering of rapidly digestible starch content [bib_ref] Impact of amylose content on starch retrogradation and texture of cooked milled..., Yu [/bib_ref] [bib_ref] Digestibility and physicochemical properties of rice (Oryza sativa L.) flours and starches..., Zhu [/bib_ref]. This could perhaps explain the lower GI value of Jasmine (=77) observed in this study compared to those typically reported as >90. However, Chiu and Stewart [bib_ref] Effect of variety and cooking method on resistant starch content of white..., Chiu [/bib_ref] discovered that postprandial glycaemia did not differ significantly in healthy adults after consuming two rice samples with distinctly different resistant starch content (0.20 vs. 2.55 g/100 g). This concurs with the observation that GI variability exists within each category of amylose content. It appears that amylose content alone may not be a good predictor of starch digestibility rate and post-meal glycaemia. The current research direction has evolved to investigating the modulatory roles of fine molecular structures of amylose and amylopectin in rice starch digestibility [bib_ref] The importance of amylose and amylopectin fine structures for starch digestibility in..., Syahariza [/bib_ref].
Additionally, the lower postprandial glycaemic and insulin responses after rice consumption observed in this study were partly mediated by the higher protein and fiber content in rice. Both crude protein and fiber content accounted for about 13% and 6%, respectively, of the GI variations of the six test rice. Physical entrapment of the rice starch granules by the spherical protein bodies located in the sub-aleurone layer may hinder and delay hydrolysis of endosperm starch. Furthermore, in vitro studies have demonstrated that the intact bran layer serves as a physical barrier that delays the access of amylase enzymes and gastric acid during starch hydrolysis [bib_ref] In vitro digestibility and physicochemical properties of milled rice, Dhital [/bib_ref] [bib_ref] Physical changes in white and brown rice during simulated gastric digestion, Kong [/bib_ref]. The GI-increasing effect has been previously reported when both dehusked and polished versions of the same rice type were fed to the same group of subjects [bib_ref] A transgressive brown rice mediates favourable glycaemic and insulin responses, Karupaiah [/bib_ref] [bib_ref] Blood glucose lowering effects of brown rice in normal and diabetic subjects, Panlasigui [/bib_ref].
## Glucometabolic responses
This study found that the time taken to reach postprandial peak glucose values for Jasmine (high GI) averaged 47.5 min compared to UKMRC9 and Basmati (low GI), which occurred at 37.5 min and 35.0 min, respectively. Moreover, the time needed for glycaemia to return to baseline levels was longer for Jasmine (136.5 min), compared to Basmati (99.6 min) and UKMRC9 (85.8 min), implying that the glucose clearance for high-GI rice required longer duration than those with a low-GI characteristic. These observations paralleled the findings by Brand-Miller et al. [bib_ref] Glycemic index, postprandial glycemia, and the shape of the curve in healthy..., Brand-Miller [/bib_ref] , whereby postprandial glycaemic curves do not differ between low and high GI foods, except for the magnitude of glucose excursions. Indeed, GI values of the six test rice in this study explained 37% and 49% of the variations in postprandial absolute and incremental glucose peaks, respectively. This has important clinical implications, as consumption of high GI foods would therefore translate into greater magnitude of post-meal glycaemic spikes and prolonged glycaemia. This metabolic milieu has been implicated in the pathophysiology of atherogenesis, systemic inflammation and type 2 diabetes [bib_ref] Impact of postprandial glycaemia on health and prevention of disease, Blaak [/bib_ref].
GI (r =´0.267, p = 0.024) and II (r =´0.391, p = 0.001) of all test rice correlated negatively with the Matsuda index of insulin sensitivity. Furthermore, although not statistically significant, the large effect size observed in the differences in insulin sensitivity after consuming rice with varying glycaemic impact suggests that rice GI evaluations should factor in postprandial insulin responses. Consumption of lower GI foods or diet at the expense of hyperinsulinaemia may not be metabolically favorable as it has been cited as one of the pathophysiological contributors to the development of chronic diseases, including diabetes [bib_ref] Relationships among insulin resistance, type 2 diabetes, essential hypertension, and cardiovascular disease:..., Reaven [/bib_ref]. Chronic fluctuations in postprandial glycaemia and insulinaemia were proposed to increase circulating non-esterified fatty acids and reduce the number of insulin receptors, ultimately contributing to insulin resistance [bib_ref] Glycemic index: Physiological significance, Esfahani [/bib_ref]. Notably, we found that ingestion of rice with differing GI values did not significantly affect postprandial pancreatic β-cell function. This implies that consumption of a single carbohydrate-rich food per se did not significantly alter postprandial insulin secretory capacity.
The glucose standard stimulated the highest early-phase insulin secretion (0-30 min) and was closely followed by UKMRC11, but was not evident in either UKMRC9 or UKMRC10. The early-phase lactate secretion also followed this trend. Surge in plasma glucose concentrations and greater insulin demand after 15 min postprandially is possibly attributed to the release of rapidly digestible starch fractions. Alternately, acute plasma lactate elevations (0-60 min) may contribute to delayed glucose and insulin clearance by peripheral tissues. We found that IAUC ins for the rice evaluations was weakly and positively correlated with IAUC lac as well as incremental and maximum concentrations of plasma lactate (data not shown). Indeed, in vitro and in vivo rodent studies have elucidated the suppressive effects of increased plasma lactate on glycolytic enzymes, which led to decreased insulin-stimulated glycolysis and sustained postprandial insulinaemia [bib_ref] Lactate downregulates the glycolytic enzymes hexokinase and phosphofructokinase in diverse tissues from..., Leite [/bib_ref] [bib_ref] Plasma lactate levels increase during hyperinsulinemic euglycemic clamp and oral glucose tolerance..., Berhane [/bib_ref].
Generation of peptide hormones in response to human physiological digestion of rice alone has never been explored. Peptide hormones, namely motilin, neuropeptide-Y and orexin-A have been investigated in response to GI effects of mixed meals [bib_ref] Acute effects of different glycemic index diets on serum motilin, orexin and..., Wu [/bib_ref]. Wu et al. [bib_ref] Acute effects of different glycemic index diets on serum motilin, orexin and..., Wu [/bib_ref] reported that a low GI breakfast reduced the secretion of orexin-A but significantly stimulated motilin secretion, without marked effects on neuropeptide-Y secretion. However, we could not discern any significant differences in all three peptide hormones responses attributed to GI variations of test rice. Fluctuations in plasma motilin and orexin-A observed in this study would perhaps be explained by the form in which carbohydrates were administered, i.e., solid (rice) versus liquid (glucose standard). This suggests that consumption of rice with varying GI values may not significantly alter the gastrointestinal motility as well as satiety response.
Both early and late phases of post-meal neuropeptide-Y changes were negatively correlated with postprandial insulinaemia. Insulin enters the brain through the blood-brain barrier and serves to regulate feeding behaviour and metabolism in humans [bib_ref] Insulin in the brain: There and back again, Banks [/bib_ref]. The early-phase insulin spikes (0-30 min) elicited by UKMRC10 and UKMRC11 could have suppressed neuropeptide-Y secretion, resulting in a relatively unchanged secretory pattern throughout the 2 h postprandial period. This concurs with a cell line study, which observed that insulin inhibits neuropeptide-Y neuronal activity in the hypothalamic arcuate nucleus, consequently reducing the neuropeptide-Y secretions [bib_ref] Insulin suppresses ghrelin-induced calcium signaling in neuropeptide Y neurons of the hypothalamic..., Maejima [/bib_ref]. In contrast, another cell line study elucidated the inhibitory effects of this hormone on insulin secretions via the G-protein coupled receptor pathway [bib_ref] Neuropeptide Y and somatostatin inhibit insulin secretion through different mechanisms, Schwetz [/bib_ref] , suggesting that bi-directional relationship exists between these two hormones. The higher neuropeptide-Y responses observed after UKMRC9 and Basmati consumption could perhaps explain the lower postprandial insulin trends. However, whether or not these elevations would trigger physiological feeling of hunger in humans remains to be explored.
# Conclusions
Amongst the three crossbred red rice variants and the comparator Thai red rice, only UKMRC9 facilitated the most desirable glucometabolic responses, particularly the acute postprandial insulin sensitivity. Since rice is a significant cereal option for most Asians, replacing white rice with a red rice displaying the characteristics of UKMRC9 becomes a critical factor in lowering dietary glycaemic load and insulin surge patterns attributed to the aetiology of metabolic syndrome. The incorporation of culturally acceptable, high-quality staple foods in substitution for refined grains is in line with recent international dietary guidelines. A robust intervention trial would serve to answer if there is/are any health-bearing benefits of replacing white rice with a low-GI, polyphenol-rich red rice on glucometabolic markers among Malaysians, particularly those with diabetes or at high risk for diabetes.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6643/8/5/308/s1, [fig_ref] Table 1: Proximate nutrient composition [/fig_ref] : Postprandial plasma motilin, neuropeptide-Y and orexin-A responses (ng/mL) at baseline and 2 h post-consumption of test rice and glucose standard (mean˘SEM).
[fig] Figure 1: Baseline-adjusted trends (mean ± SEM) in postprandial plasma (a) glucose; (b) insulin and (c) lactate responses for six test rice and glucose standard, and incremental area-under-the-curve (IAUC) for postprandial (ai) glycaemia; (bi) insulinaemia and (ci) lactataemia. § Mean values bearing the same alphabets were not significantly different (p > 0.05, univariate analysis of variance followed by Tukey's post hoc test, ns = not significant); ηp 2 = partial eta-squared, 0.01, 0.06 and 0.14 were used to denote small, moderate and large differences in measured outcomes, respectively. [/fig]
[fig] Figure 2: Baseline-adjusted trends in postprandial plasma (a) motilin (MTL); (b) neuropeptide-Y (NPY) and (c) orexin-A (ORXA) responses for six test rice and glucose standard. Note: Horizontal bars = smallest and largest values; Lower band = 25th percentile; Upper band = 75th percentile; (×) = median; (•) = outlier and ( ) = extreme values defined as 1.5 × (Q3 − Q1) below 25th or above 75th percentiles. [/fig]
[table] Table 1: Proximate nutrient composition (% by dry weight basis) † and cooking characteristics of test rice. Values are expressed as mean˘standard deviation with each variety analyzed in duplicate samples (n = 2). Values in the same column not superscripted by the same letter are significantly different, p < 0.05 (univariate analyses of variance, followed by Tukey's post hoc test); ‡ Raw and cooked rice weights were based on 50 g available CHO content; § Cooking time was recorded from the time the electric rice cooker was switched on to the time it automatically turned off. CHO, carbohydrate; TDF, total dietary fiber; TPC, total phenolic content. [/table]
[table] Table 2: Kinetic markers of postprandial glycaemic and insulin responses. Values are expressed as mean˘SEM. Mean values within the same column superscripted by different letters were significantly different ( § p < 0.05, univariate analysis of variance with Tukey's post hoc test; ns = not significant between rice types); 2 GI values were categorized as low (<55), intermediate (55-70) and high (>70) [/table]
[table] Table 3: Correlation between nutrient content, cooking characteristics, glycaemic and insulin indices of test rice. [/table]
|
A Study of the Political Culture of the Ming and Qing Agribusiness Societies under the Imperial Environment
In the Ming and Qing dynasties, while the social life, economic structure, and even cultural atmosphere underwent profound changes, the dominant aspect of political culture was still deduced along the traditional track of the imperial system. e prominent characteristics of China's social form in the Ming and Qing dynasties are the development of the commercial economy and the enhancement of social freedom; the developed trend of common people's culture; and the sustainable development of the centralized monarch-bureaucratic-aristocratic system, which constitute a self-consistent pattern. is paper on the political culture of political spirit, political value, political thought, the four dimensions, the specific period, specific social community, public power setting and operation, and the perspective of the Ming and Qing social form of political culture, this study helps people to the imperial environment of Ming and Qing rural political culture and understanding and exploration.
# Introduction
e study of political culture should include the investigation of the setting and operation of public power in a specific period and a specific social community from the four dimensions of political spirit, political value, political thought, and political ecology [bib_ref] Study on improving the hydrophilicity of coal by a biosurfactant-producing strain screened..., Liu [/bib_ref]. Before the Ming and Qing dynasties, the main political structure of the central government is shown in [fig_ref] Figure 1: Centralized [/fig_ref].
However, in the Ming and Qing dynasties, while social life, economic structure, and even cultural atmosphere underwent profound changes, the dominant aspect of political culture was still deduced along the traditional track of the imperial system. e centralized institutions of the Ming and Qing dynasties are shown in [fig_ref] Figure 2: Centralized structure of the Ming and Qing dynasties [/fig_ref]. e prominent characteristics of China's social form in the Ming and Qing dynasties are the development of the commercial economy and the enhancement of social freedom, the developed trend of common people's culture, and the sustainable development of the centralized monarch-bureaucratic-aristocratic system, which constitute a self-consistent pattern. is selfconsistent pattern shows the potential of the imperial rural business society to continue to extend under the prosperity of the commodity economy [bib_ref] A study of forecasting tennis matches via the Glicko model, Yue [/bib_ref]. e conflict between China and the west in the mid-19th century was not the meeting of two societies at different stages of the same historical process, but the intersection of two civilization systems in different historical processes in the spatial and temporal pattern of the Asian continent [bib_ref] e impact of digital technology on the transformation of political culture of..., Aitymbetov [/bib_ref].
In the rural and commercial societies of the Ming and Qing dynasties, the political culture in the imperial environment is mainly reflected in the political spirit. Political thought is the rational understanding of political phenomenon on the basis of political experience and political perceptual understanding, and it is the abstract and logical explanation of the internal causal relationship of political phenomenon and its development law. Political thought is a systematic, complete, and rigorous political thinking achievement and ideology, usually composed of political thought, political belief, and political theory.
Institutions embody the rigid statutes of the dominant groups in society to keep society in a certain state, among which the most public power and political attributes are the basic institutions of the state [bib_ref] Populism in the political culture of Turkey: the foreign policy dimension, Shlykov [/bib_ref]. Human history has shown a variety of state institutions, including various monarchies, feudal systems, constitutional monarchies, republics, presidential systems, cabinet systems, parliamentary systems, and so on, as well as a variety of mixed and variant forms. e construction of all these basic institutions of the state reflects the expectations, concepts, and demands of the dominant people in a given social community regarding the establishment and operation of public power [bib_ref] Political culture of the Russian students: value-based, figurativesymbolic and behavioural aspects, Evgenyeva [/bib_ref]. e spirit of politics, as a fundamental area of political culture research, is to look behind the rigid structure of the basic institutions of the state to understand their expectations, ideas, and demands to achieve an understanding of the inner rationale and the real and potential tendencies of the political system of concern, as well as an understanding of the universal meaning of a specific political system.
In the Ming and Qing dynasties, China was not a legal or constitutional state, so the basic institutions of the state at that time were not entirely defined by the texts of laws and regulations, but rather by a combination of texts of regulations and the derivation of governance. Specific changes occurred frequently, but the fundamental regulation was consistent [bib_ref] e understanding and perception of the political culture of the modern mongols, Oidov [/bib_ref]. e axis of the public power structure was the emperor-bureaucrat-county system. is system, gobbled up in the earring states, was customized in the Qin and Han dynasties, and in the Yuan dynasty, there were provinces as a large regional hierarchy between the central government and local administration, which continued unchanged until the Ming and Qing dynasties. In contrast to the colorful changes in the economic and cultural fields, the aforementioned basic structure of the Ming and Qing dynasties has a long history and shows no definite signs of change in a practical sense. In the sense of ideology, the clearest movement was the criticism of the privatization of the monarchy expressed by Huang Zongxi in the early Qing dynasty in his Ming Yi Zuoren. Huang's proposal, however, could only reach the level of co-rule of monarchs and ministers and did not consider or propose the abolition of the monarchy, nor could it propose a constitutional system of monarchy [bib_ref] A numerical study of the solar modulation of galactic protons and helium..., Song [/bib_ref]. Moreover, Huang's book became taboo before it was published and did not see the light of day until the late Qing dynasty, and did not have a substantial impact on the trend of the political spirit of the Chinese state in the Qing dynasty. With regard to the bureaucratic-provincial-county system, the Ming and Qing dynasties practiced a centralized system of states and regions under the provincial authority in the core areas: the thirteen prefectural administrations in the Ming dynasty and the eighteen provinces in the Qing dynasty, without change [bib_ref] Cytogenetic characteristics of childhood acute lymphoblastic leukemia: a study of 1541 Chinese..., Meng-Meng [/bib_ref].
is inheritance of the basic system from previous eras indicates a tendency for the political system to remain "unchanged," which in turn indicates that the dominant public power in the Ming and Qing dynasties did not take into account the need for the political system to change with the times, as shown in [fig_ref] Table 1: Changes in the state power system in the Ming and Qing dynasties [/fig_ref].
e system of state power in the Ming and Qing dynasties was not fixed, but it was not in the category of fundamental institutions.
As for the late Qing dynasty, there were many institutional changes in a substantive sense, but already after the external challenges and comprehensive crisis. erefore, to read the spiritual qualities of the Chinese public power system in the Ming and Qing dynasties, imperial power politics, bureaucratic scholar politics, and aristocratic politics are three keywords [bib_ref] Framing the nation and collective identities. political rituals and cultural memory of..., Pavlakovi [/bib_ref].
## Political values
e so-called political values refer to social members' conceptions about the meaning and contingency of political systems, political life, and political phenomena. e social and political value of rural agriculture and commerce in Ming and Qing dynasties mainly includes the three aspects shown in [fig_ref] Figure 3: Social and political value of rural business in the Ming and Qing... [/fig_ref]. Such conceptions determine at a deep level the attitudes and psychology of social members toward the established public power settings and their operational states. Such attitudes and psychology may be expressed in verbal and textual ways, or in nonverbal and textual ways during various opportunities and occasions of life practice; for example, in all societies, the members of the society expect the government to organize to keep the society free from external forces, the government that cannot do this will surely perish; the members of all political societies expect the government to bring order, the government that cannot bring order will be replaced; most members of all political societies expect their governments to maintain social justice, although in fact it is difficult for society to be completely just, it is difficult to maintain a society without justice. Many political conflicts throughout history have been not only conflicts of interest but also conflicts of values. From the perspective of political values and the basic attitudes of members of society toward political life, the following issues in particular need to be considered in the Ming and Qing dynasties [bib_ref] College and the "culture war": assessing higher education's influence on moral attitudes, Broi [/bib_ref].
## National consciousness.
National consciousness is one of the common psychological qualities and one of the national characteristics. It refers to the psychological state embodied in the characteristics of national culture in the process of formation and development. What members of society think the state is, how the state is related to their own existence, and what is the basis and shape of their identification with the state are all elements of political values. e state consciousness of the mainstream people in most of the Ming and Qing dynasties was the consciousness of the subjects [bib_ref] Chinese excellent traditional culture study on the path of integrating the ideological..., Liu [/bib_ref]. e core of this consciousness was obedience to the imperial court, which was the "Emperor of Heaven." If the court's governance led to the people's unhappiness, the people might take the risk of rebellion; if the court completely lost its virtue and power, the walls would fall down and the people would push, and they would establish or change to a new dynasty; in case of foreign invasion, if the court could function normally, they would participate in the court to defend the country from the enemy. In the Ming dynasty, all of the aforementioned situations occurred, and the people's attitude toward the state did not differ significantly from that of previous generations. In the late Ming and early Qing dynasties, the Qing army entered and moved southward, and while achieving the reunification and cultural integration of the national community, it introduced forced changes in culture and customs, which not only met with resistance from the remnants of the Ming dynasty but also provoked direct resistance with the active participation of a large number of southerners [bib_ref] A study on English translation tactics of the report on the work..., Mu [/bib_ref]. After the suppression of the large-scale armed resistance movement against the Qing dynasty, the activities of anti-Qing secret societies continued unabated until the late Qing revolution.
is situation reflects the more complex entanglement between national identity and national cultural identity and the conflicting interests of groups. During this period, the enlightenment of the national view of the state may have occurred with reference to the idea of defining the state by cultural identity expressed by Huang Zongxi, Gu Yanwu, and Wang Fuzhi in the early Qing dynasty, but comparing the early Qing dynasty with the early Yuan dynasty, it is not certain whether a new qualitative situation of national consciousness really occurred at the level of popular and popular psychology in the early Qing dynasty. At the same time, the mainstream of the Qing people, and even the gentry and business class, actually gradually accepted the legitimacy of the Qing dynasty. Otherwise, it would be impossible to account for the nearly 270 years of the Qing dynasty's national movement. is process, because the change of regime in the Ming and Qing dynasties was accompanied by national conquest, made the process of general acceptance of the early Qing regime slower than the same process in the early Ming dynasty. e gradual social acceptance of the Qing dynasty was due to the prevailing subject consciousness of the people under its rule as the primary factor. A population dominated by the subject consciousness was politically passive, with survival as its basic demand, and lacked the capacity for collective behavior, tending not to confront the real rulers when possible [bib_ref] Integrating meaningful selfhood into the sociological study of political languages: blending mead's..., Fietz [/bib_ref]. e measures taken by the Qing government to adapt to the culture of the middle Kingdom and to collect the hearts and minds of the people were the second factor at play. e administrative effectiveness of the early Qing government in contrast to the late Ming government and the effectiveness of its governance was the third factor of action e abolition of the prime ministerial system in the early Ming dynasty, a movement leading to the strengthening of imperial power 2 e cabinet system formed in the Ming dynasty, which is the abolition of the prime ministerial system around the functioning of imperial power gradually occurred after the compensation system, the end of the Ming dynasty has failed to exist in name only until the Qing dynasty 3 e establishment of the Qing dynasty's pre-fan court and the implementation of land conversion in the southwest were implemented in response to the expansion of the Qing dynasty's direct control over geographic space and were adaptations of the existing centralized power and provincial system in the interior to the outer edge of regional management e rise of a new national consciousness after 1840 was a huge difference from any previous era, but it could not occur without the challenge of western colonialism and the impact of western ideas, so it does not prove that it occurred within the logic of the Ming-Qing imperial peasant-merchant society itself. To sum up, if we want to inquire into the changes in state consciousness during the Ming and Qing dynasties, we should focus on the time of the Qing rather than the Ming. Although the Qing and Ming dynasties are referred to as "empires" in many researchers' vocabulary, especially in western scholarly works, the nature of the Ming dynasty is lighter and the Qing dynasty is clearer, the Ming dynasty was essentially a local state, while the Qing dynasty united the vast majority of the territory and people of the traditional Chinese civilization.
is change must have had some impact on the state consciousness of the nobility and the scholarly officials, and probably had a similar impact on the state consciousness of the people. Whether such effects were closer to a "modern" nation, state consciousness in any sense of the word than to the state consciousness of previous generations requires further analysis [bib_ref] A study on urban forms of prefectures in the Ming and Qing..., Xie [/bib_ref].
## E concept of political legitimacy.
Political legitimacy refers to the legitimacy or legitimacy of the government based on the principle recognized by the public. To put it simply, to what extent the government is regarded as reasonable and moral by the citizens. Politics is a system of coercive power that operates to gain legitimacy, and coercive power itself can only constitute an integral part of the state management function on the basis of gaining social legitimacy identity, otherwise, it is only order subversive power. e political power, the supreme ruler, and the major decrees are legitimate under what circumstances, and the political behavior of which people is legitimate, which constitute the bottom line of the operation of the public order of the political system. However, the term "legitimacy" here is a discourse of ideas and psychological tendencies, mainly referring to the orthodoxy, legitimacy, and rationality of political rule, rather than whether it conforms to the attributes stipulated in legal texts. e legitimacy of state power in the Ming dynasty was repeatedly affirmed by the Ming rulers and ministers, the core of which was first, to accept the fate of heaven; second, to sweep away the Hu customs; and third, to return the four seas to the heart. is is repeatedly seen in the various imperial edicts and documents of the early Ming dynasty. To simplify it, it means the will of heaven, the hearts of the people, and cultural inheritance. e first two, certainly entirely traditional in nature, with the restoration of Chinese clothing as the basis for the legitimacy of the Ming dynasty's statehood, had the novelty of the times. In China, before the Yuan Dynasty, although there was a nomadic regime established by nomads, there was no unified nomadic regime.
e rule of the Yuan Dynasty may fundamentally change the direction of Chinese cultural evolution. erefore, the founding of the Ming dynasty had the significance of "restoration" in terms of the cultural circle of the central plains [bib_ref] Study on the drilling performances of a newly developed CFRP/invar cocured material, Zhang [/bib_ref]. is gave the early Ming dynasty an additional layer of legitimacy. However, as a result, the Ming dynasty also became a localized state, with less control over the peripheral regions outside the traditional Han areas. Such affirmations by the Ming government had certain repercussions in the popular psyche. Traces of this can be seen in the gradually increasing literature of the common people from the mid-Ming period onward. e social recognition of the legitimacy of Qing rule encountered a much greater obstacle than that of the Ming.
In order to remove this obstacle, the rulers of the early Qing dynasty, in addition to declaring that they would be the first to accept the fate of heaven and that the four seas would return to their hearts, announced policies such as crushing the "rogue thieves" for the Ming dynasty, consecrating the Ming tombs, and recruiting Confucian scholars and thinkers to serve in the government, inherited the basic institutions and laws of the Ming dynasty, accepted the upper-class culture of the central Plains, and also, by launching a written prison, suppressed the forces that refused to agree with the legitimacy of the Qing dynasty. ey also suppressed the forces that refused to agree with the legitimacy of the Qing dynasty by initiating written jails, etc., and very seriously reevaluated the relationship between rulers and subjects and between China and yi [bib_ref] Brand construction of Chinese traditional handicrafts in the we-media era-a case study..., Ou [/bib_ref]. In the long-term process of moving from general resistance to mainstream identification with Qing rule, force and wai-yi tactics were the main factors initially; gradual stabilization of people's livelihoods and gradual cultural integration were the main factors later. is suggests that the establishment of regime legitimacy in the early Qing dynasty was, in fact, not fundamentally different from the logic of earlier eras. However, it is still important to see that the revolutions that took place during the late Qing dynasty, including the failed Taiping heavenly kingdom and the partially successful Xinhai revolution, all raised the banner of hair storage or anti-Manchu, and the psychology of the Qing dynasty as a conquering ruler has never completely disappeared, meaning that the legitimacy achieved by the Qing dynasty has been limited. e majority of the public thought that the government's rule (including the threat of force) was improper, so the concept of political legitimacy was restricted in the Qing dynasty.
Closely related to political legitimacy is the notion of political rationality, both have a significant influence on political culture, that is, the sense about whether political initiatives are in line with basic political value concepts. For example, the early Ming and early Qing dynasties both had written prisons, which were never endorsed as reasonable in the public psyche or even in the official historical accounts afterwards. e eunuchs' control of the imperial government also never gained public approval. e late Ming emperors sent mine supervisors and tax ambassadors and the court to suppress the Donglin party, which had aroused public anger. Qing officials and honest officials, both in the Ming and Qing dynasties, became a group of people expected and praised by the public. However, the political concepts that can show these phenomena can still be found in the political and cultural traditions of the previous era, and it is difficult to determine, which have been new political and cultural factors.
## E sense of political participation.
Political participation is a great progress in China's political democracy and also an important embodiment of the concept of "peopleoriented" in the field of political life. is is an important concept in Almond and others' analysis of political culture in modern democracies. Ming China was not a democracy, and thus members of society were certainly less enthusiastic about participating directly in politics than the societies Almond examined. Even so, any functioning political system involves the question of whether and to what extent and in what manner members of society participate. In times of nonrevolution, turmoil, war, and frenzy, the enthusiasm of people's political participation are proportional to the public nature of the operation of political power. If the degree of Chinese political participation in the Ming and Qing dynasties was higher than in previous eras, it may mean that the openness and public nature of this social system increased, and if it did, it tended to be more modern in nature.
e main group of people who directly participated in politics in the Ming dynasty were intellectuals, whose preparation stage was to study and join the imperial examinations; and after the imperial examinations, they joined the government and thus entered the national political trend. Another way of political participation for this group of people is to lecture in the field. ere were lectures in the middle and late Ming dynasty, some of which were lectures by intellectuals with the status of a scholar, and some were lectures by intellectuals in cloth without the status of a scholar. At that time, the so-called lectures, there is no knowledge of science as it is called today, are sincere in ruling the country and the world. Whenever it comes to current affairs, it has the nature of political participation [bib_ref] e dominants of contemporary political culture and the transformation of political governance, Tumanov [/bib_ref]. e sense of political participation among members of society in the early Qing dynasty was not more positive but more negative compared to that of the Ming dynasty. is is certainly related to the ethnic conflicts that developed during the Ming-Qing transition, as well as to the longer and harsher writing prison that lasted in the early Qing dynasty. Under the influence of the literary prison, the average number of Jinshi admitted in the Ming and Qing dynasties changed greatly, as shown in [fig_ref] Figure 4: Changes of the average annual number of Jinshi in the Ming and... [/fig_ref]. At the level of the scholarly class, the Qing dynasty did not have the private lectures that used to be prevalent in the Ming dynasty, where commenting on current affairs was an important element, nor did it have the phenomenon of confrontation between rulers and ministers on a very large-scale. e party movement of the scholarly class ceased, and was replaced by the secret society of the civil society, which was more prevalent than in the Ming dynasty. e former was open and thus reflected political participation based on recognition of the existing system; the latter was secret and thus reflected political hostility based on nonrecognition of the existing system. It was only after the mid-19th century that the situation changed radically.
## Political ideology
Political thought is related to political values, ideological, and political education value is a utility relationship between the needs of subjects and the attribute of the object. Socialist core values, as the guidance and foundation of ideological and political education value, play an irreplaceable role in promoting the realization of ideological and political education value. As human essence, ideological and political education practices the socialist core values, and the two are dialectically unified in the process of the realization of ideological and political education value. But political thought is more often embodied in explicit claims expressed through language and texts than in political ideas embodied in emotions, attitudes, and behaviors that are either clear or vague. Generally speaking, to grasp the basic face of political thought in an era, one needs to examine the discourses of thinkers in that era.
erefore, the history of political thought always carries the implication of elite history. e history of ideas is the core part of cultural history, and political thought is also the core part of political culture, and its significance in the study of political culture is self-explanatory. e two most accomplished researchers in the study of traditional Chinese political culture, Yu Yingshi and Liu Zehwa, as mentioned earlier, have paid special attention to the elements and clues that can constitute "ideas" in their research, as shown in [fig_ref] Table 2: Some political ideas with new ideas in Ming dynasty [/fig_ref].
In the political thought of the Ming and Qing dynasties, the ideas that gained practical status were, first of all, the political outlook of authoritarian imperialism, which was prominently reflected in the statements of Zhu Yuanzhang in the Ming dynasty and the Kangxi, Yongzheng, and Qianlong emperors in the Qing dynasty, including the edicts and the chapter and verdicts [bib_ref] Still an outsider? a preliminary study on the political culture of ethnic..., Yu [/bib_ref]. e second is the peopleoriented political doctrine of later Confucianism. e aforementioned monarchs and Confucian scholars and thinkers in politics were in general agreement on the basic institutions of monarchy, bureaucracy, centralization, and limited aristocracy, but diverged on the issue of the limits and constraints of imperial power. e scholars and thinkers of the Ming dynasty had been trying to keep imperial power regulated to a certain extent, seeking to expand the scholars' voice in the political operation of the state; the Qing dynasty's scholars and thinkers had been trying to expand their power to a certain extent. In the Qing dynasty, the scholars abandoned this effort and converted to the absolute authority of imperial power. According to such basic clues, Zhu Yuanzhang was the first representative of imperialist thought in the Ming dynasty, and the subsequent emperors were all on the path of their practice, but rarely had new ideological expressions of the nature of invention and elaboration. e three emperors of the Qing dynasty, Kangxi, Yongzheng, and Qianlong, are the representatives of Qing style imperialism. ey did put forward some political ideas with new meaning based on their admiration of Zhu Yuan zhang's political thought and political action.
In order to more clearly view the role and development of political thought in the Ming and Qing dynasties, it is Journal of Environmental and Public Health necessary to introduce the study of Confucian political thought in the Song and Ming dynasties.
It is the fact that Confucian political thought in the Song and Ming dynasties was people-oriented in terms of value goals, i.e., it took the state of people's livelihood as the measure of political legitimacy and rationality rather than the interests and wishes of the monarch; in terms of institutional thought, it was based on the monarchical-bureaucratic political base, i.e., it fully recognized the monarchy and bureaucracy, as well as a limited aristocracy. On this basis, the political thinkers of the scholars in the early and mid-Ming dynasty did not put forward any particularly innovative ideas based on the political thought of the Song dynasty but only made some technical discussions from the perspective of how to effectively practice the Confucian political philosophy. In the late Ming and early Qing dynasties, Gu Yanwu, Huang Zongxi, and Wang Fuzhi represented the culmination of political system introspection and political philosophy criticism. Of course, their thoughts are still Confucianists in the basic direction, but in response to the lessons of the fall of the Ming Dynasty, they all more sharply advocated the people-oriented political principle, and more fiercely attacked the absolute power of the monarch, and put forward the idea of restricting the monarch's power. However, they could not and did not go beyond the basic pattern of monarchy and bureaucracy. After the Qing dynasty, the independent consciousness of the scholars quickly encountered the impact of the trend of absolute monarchical power, the late Ming and early Qing dynasty political criticism and political introspection stifled, independent political ideas such as flowers and water, so the political thought of the scholars can only be attached to the emperor's political rhetoric. Such a situation extended until the middle of the 19th century. As a result, the political thought of the Ming and Qing dynasties had hardly made any new progress. As for the basic ideas of so-called "modern" politics, such as "democracy," "freedom," and "rights," and as for the so-called "modern" political ideas, such as "democracy," "freedom," "rights," and "rule of law," they could not be developed in the original political concept in the same direction, and could not become clear ideological tools without a more profound transformation of the general shape of society or interaction with the outside world.
## Political ecology
e term 'political ecology' is more specific. I mainly involve only two relevant levels here: one is the orientation of the dynasty and the other is the nature of the imperial power. It is the decline of imperial power, the power of scholar-officials, and the retreat of the sense of responsibility as the basic clue.
Political ecology refers to the general atmosphere, states, and deduction tendencies of a specific political community unfolding in its operation. e so called political history is in fact the course of deduction constituted by the succession of particular states, the state of each particular moment being the basis of the state of the next moment, and all political states being the consequences of the deduction of their previous states. e total state of a functioning political body contains the greater possibility of some political phenomena occurring in the near future, and the lesser possibility or impossibility of some other political phenomena occurring. 1 e rhetoric about the Chinese and the Japanese as one family, as represented in the Danyi Juezhu. is idea was adapted to the expanded geographic, ethnic, and demographic scale of the multi-ethnic state, arguing for the legitimacy of Qing rule, and expressing the idea of the rationality of a multi-ethnic state 2 e unprecedented esteem for the absolute rationality of imperial rule, which is intertwined with the practice of Manchurian ministers calling themselves minions to the emperor. is shows a tendency to see from the perspective of political thought from the strengthening of monarchical dictatorship to absolute strengthening.
## Journal of environmental and public health
Political ecology is naturally closely related to political institutions, ideas, and values, and at the same time is political from the perspective of the integrated dynamics of political operation, and thus does not cover institutions, ideas, and values, and deserves special scrutiny. Looking at the history of the Ming and Qing dynasties from the perspective of political ecology, we can not only see the continuity of the two generations but also the fracture of the two generations, as shown in . e most prominent the continuity is obviously the basic dynamics of the political enhancement of imperial dominance. e abolition of the prime minister in the 13th year of Hongwu, who had already existed during the pre-Qin period and had been the main person responsible for the daily administration of the central government since the Qin and Han dynasties, was a huge change. Huang Zongxi's claim that "the absence of good governance in the Ming dynasty began with the dismissal of the prime minister by the high emperor" was not a false statement, but it transformed the system, in which the emperor mainly handled major decisions and left the day-to-day administration to the bureaucratic system into one, in which the emperor was the "sole arbiter" of both decisions and administration, thus instrumentalizing the bureaucracy of the scholar and the emperor's personal will to a greater extent. If we combine this change with the action of dividing the kings in the early Ming Dynasty, which led to a short upsurge of aristocratic politics, then the strengthening of imperial power based on families in the early Ming Dynasty paved the way for the political ecology of the whole Ming Dynasty. However, the Ming dynasty, imperial power was not continuously strengthened, and the cabinet system actually provided a certain degree of institutional opportunities for the political expansion of the scholar, thus creating a complex entanglement and resistance between the two tendencies of imperial power absolute and scholarly discourse in the political history of the Ming dynasty. e admonition of Zhengde's southern tour, the great Rites, the protest against Zhang Juzheng's seizure of love, the east forest's discussion of politics, and the antimining taxation supervision, all have such implications. When investigating the political ecology of the Ming Dynasty, Mr. Yu Yingshi tried to investigate the relationship between imperial power and literature, but he only grasped the comparison of Zhu Xi's and Wang Yangming's thoughts, and did not pay attention to the above practice. He believed that scholars in the Ming Dynasty had retreated from pursuing "inner sage" and "outer king" to pursuing "inner sage," which was an oversight. In the early Qing dynasty, imperial power was substantially strengthened for the second time.
Compared with the Ming dynasty, the important "fracture" during this period is mainly the Qing dynasty in response to the extension of the multi-ethnic state system, bringing "the first Chong Manchurian" ethnic rule color and more active management of frontier areas. is situation, from the point of view of laying down the map of modern China and promoting the integration of various ethnic groups in the Chinese civilization sphere, is of great historical progress, but from the point of view of political and cultural improvement, it is retrograde. e political science involvement is too complex to discuss briefly. Moreover, the resurgence of aristocratic politics in the early Ming dynasty soon came to an end, while the resurgence of aristocratic politics in the early Qing dynasty continued for a long time and still constituted a realistic system until the late Qing dynasty.
is in turn reflects the differences between the politics of the Ming and Qing dynasties.
Based on the above general observation and looking at the specific level, the most prominent situation at the temple politics level is that the active politics of the scholars in the Ming dynasty was lost until the Qing dynasty. ere was active political criticism, party movements, and even confrontation between rulers and ministers in the temple. is was the last period of active scholarly politics in imperial China. en, after the Qing dynasty, there was no longer a group of scholars to resist the imperial order, and the instrumental significance of the scholars in the temple politics increased. us, the whole temple politics level, forming a period of great prosperity of imperial politics.
is great prosperity of imperial politics was in fact a facet of the Kang, Yong, and Qian dynasties. e politics of the scholars did not directly imply good politics, but the biggest drawback of the politics of the imperial era that could not be eradicated was the absolutization of the emperor's power. e tendency of the emperor's power to be absolute constituted a daily check. Technically, this would have increased the cost of politics, but without such a check, the social costs of absolute imperial power would have been even more profound. However, if modern researchers attach the Ming scholars' demand for restraint of imperial power to the political demand for "democracy," it is difficult to get the point. e political aspirations of the Ming and Qing scholars, even in their most radical manifestations, were not democratic, but rather monarchical-bureaucratic politics based on people's livelihoods. e concept of the foundation of democracy, as shown in modern Western history, has not been sorted out into an ideological discourse in traditional Chinese political culture, and thus cannot be brought together into a logic and theory of thought. e modern democratic system is not a hazy feeling or tendency, but a systematic institutional pattern and profound political and cultural state, which must be developed with the help of rigorous and grand theories. e scholars of the Ming and Qing dynasties did not have this ability, and other people did not have this ability. Moreover, throughout the Ming and Qing dynasties, it is impossible to see which group of people has the possible tendency to continuously enhance this ability. V. social formation of Ming and Qing dynasties in the perspective of political culture.
First, the Chinese imperial system had a high degree of tolerance for the development of the commodity economy. e rapid development of the commodity economy during the Ming and Qing dynasties prompted the imperial state system to make a number of policy adjustments, and such policy adjustments led to structural changes, such as the shift from a fiscal system based on physical goods to one based on money. Regardless of the complexity of the evolutionary process, these adjustments and changes boiled down to a basic rapprochement between the imperial system and the new economic situation of the growth of the commoditymoney economy, thus presenting the peak of simultaneous prosperity of the commodity-money economy and the imperial system. is suggests that the Chinese imperial system was not in direct opposition to the general development of the commodity-money economy, but had more room for accommodation. e hypereconomic coercion of the imperial system constituted some limitations on the highly developed commodity economy, but this was only one side of the function of the imperial system; the other side was that the imperial system was also able to provide some favorable conditions for the development of the commodity economy, at least compared to the hierarchical and feudal political system. e imperial system's social control and ability to coordinate and act in a unified manner in response to external challenges exceeded that of the hierarchically divided system, requiring a large geographic deployment of manpower and materials, and thus some degree of compliance with the market economy. e basic features of the imperial era, such as a unified currency, transportation facilities, social security covering a large geographical area, and the existence of metropolises, also constituted the conditions for a developed commodity economy and market.
us, the Chinese imperial system was not susceptible to a process of disintegration in the context of the general development of the commodity economy, nor was the natural economy of self-sufficiency a fundamental feature of the system. Secondly, the evolution of political culture in the Ming and Qing dynasties has always been entangled between change and unchanged, and there is no new political culture trend of breaking through the imperial ideology. Some innovative political ideas and concepts stimulated by the development of commodity economy are dissolved in the framework of the imperialist ideology, or shaded by the imperialist ideology. e reason for this may be related to the development that the commodity economy has not reached a sufficient level. [fig_ref] Figure 5: Economic [/fig_ref] shows the economic development trend of ancient China, but more certainly, because the political culture is rooted in a deeper level of tradition.
is era even the most radical thinkers, also must adopt the concept of Confucianism to think about real problems, guided by the connotation of Confucianism, and the most valuable Confucian political orientation, and imperial system has formed in the long-term historical evolution of many ways, so not enough to provide system update systematic ideas.
# Conclusion
From the perspective of politics and culture, this paper intends to analyze the spirit and psychological characteristics of public power operation in China in the Ming and Qing dynasties, in order to deepen the understanding of the comprehensive characteristics and evolution tendency of the Ming and Qing dynasties. And from the perspective of political culture, this paper presents this pattern and makes a preliminary discussion, so as to discuss the characteristics of social form and trend of the Ming and Qing dynasties in the future.
## Data availability
e labeled dataset used to support the findings of this study is available from the corresponding author upon request.
## Conflicts of interest
e author declares that there are no conflicts of interest. : Characteristics of the secondary reinforcement of imperial power. [bib_ref] Study on improving the hydrophilicity of coal by a biosurfactant-producing strain screened..., Liu [/bib_ref] rough the prison of words, crackdown on cronies, etc., to break the sense of autonomy of the scholars and drive away the bureaucratic tools and even slavery mentality 2 Resurrect the aristocratic politics more substantially, with the participation of the Manchu aristocracy in politics as a complement to the imperial power, offsetting the institutional status of the Shih Taishu bureaucracy 3 e emperor was diligent and the use of secret folding, court mail, and other management methods, directly weakened the voice of the scholar, to achieve the effective control of the imperial power over the entire decision-making and administrative process
[fig] Figure 1: Centralized [/fig]
[fig] Figure 2: Centralized structure of the Ming and Qing dynasties (the emperor directly commanded the six departments). [/fig]
[fig] Figure 3: Social and political value of rural business in the Ming and Qing dynasties. [/fig]
[fig] Figure 4: Changes of the average annual number of Jinshi in the Ming and Qing dynasties. [/fig]
[fig] Figure 5: Economic [/fig]
[table] Table 1: Changes in the state power system in the Ming and Qing dynasties. [/table]
[table] Table 2: Some political ideas with new ideas in Ming dynasty. [/table]
|
Women’s Lives Matter—The Critical Need for Women to Prioritize Optimal Physical Activity to Reduce COVID-19 Illness Risk and Severity
# Introduction
The World Health Organization (WHO) declared coronavirus disease 2019 (COVID-19) a pandemic on 11 March 2020. To control community spread of the COVID-19 pathogen (severe acute respiratory syndrome coronavirus 2, SARS-CoV-2), most countries around the world issued mandates and guidelines for social distancing, which included staying~two meters (~six feet) apart from other peopleand not gathering in groups . Unfortunately, a negative consequence of such necessary measures to control the spread of an infectious disease is that, it creates barriers to engaging in sufficient amounts of physical activity (PA), thus predisposing societies to a "pandemic of physical inactivity". Reduced PA, coupled with reduced social interaction and changes in work and living arrangements due to the COVID-19 pandemic, has negatively impacted the health and wellness of individuals and communities. The negative effects of social distancing and isolation range from mental health concerns, such as anxiety and depression, to disturbances in physical health in the form of metabolic changes, increased adiposity, and multisystem deconditioning (e.g., negative changes in the cardiopulmonary, neuromuscular, and musculoskeletal systems). According to the United Nations and The World Economic Forum, data from the Ebola and Zika epidemics indicate that, during epidemics, . Health benefits of an active lifestyle. Being physically active rather than sedentary has multiple beneficial effects and improves overall health and wellness. Due to positive modulatory effects on multiple physiological systems, being physically active during the COVID-19 pandemic can be beneficial for healthy women, as well as for women who might have asymptomatic or uncomplicated COVID- We acknowledge that the gender and biological sex of a person are related but not synonymous. In this paper, the terms "women/females" and "women's health" refer to biological females and their specific health considerations, respectively. However, the information presented here may be useful for both biological females and males, as well as individuals of diverse genders. We also clarify that in this review, where appropriate, we have used the more inclusive term physical activity (PA) rather than the specific term "exercise" because exercise is a subtype of PA, which must meet certain precise criteria. flexibility, and neuromotor control. PA that covers these four domains has been shown to produce local effects within cardiac and skeletal muscle, as well as systemic effects in all other physiological systems in the body in both women and men. Specifically, the local and systemic benefits of PA are relevant to maintaining an optimal body mass index (BMI), possessing better insulin sensitivity, achieving a healthy blood lipid profile, and avoiding high blood pressure, which collectively reduces the risk of heart disease-the leading cause of death globally according to the WHO. Since there is a link between PA load (the combination of activity intensity and duration) and illness risk, it is important not to perform high loads of unaccustomed PA during the COVID-19 pandemic. For the majority of women in the population whose physical ability level is not at the level of elite athletes, moderate PA is likely to be beneficial while unaccustomed, intense PA is likely to be harmful. Each individual should gauge for themselves, under the advice of suitable healthcare professionals, what their "just right" level of PA is (Green Zone in . For elite athletes, while setting PA load, the risk of illness must be weighed against the benefit of being able to return to a competitive level when COVID-19 restrictions are lifted. Furthermore, healthy elite athletes who are accustomed to high PA loads as part of systematic training have a lower risk of illness even with high PA loads. (B) Since the self-reported rating of perceived exertion (RPE) correlates well with VO2 max and HR, RPE is a simple yet reliable tool for gauging the intensity of PA and for adjusting PA load as needed. flexibility, and neuromotor control. PA that covers these four domains has been shown to produce local effects within cardiac and skeletal muscle, as well as systemic effects in all other physiological systems in the body in both women and men. Specifically, the local and systemic benefits of PA are relevant to maintaining an optimal body mass index (BMI), possessing better insulin sensitivity, achieving a healthy blood lipid profile, and avoiding high blood pressure, which collectively reduces the risk of heart disease-the leading cause of death globally according to the WHO. Since there is a link between PA load (the combination of activity intensity and duration) and illness risk, it is important not to perform high loads of unaccustomed PA during the COVID-19 pandemic. For the majority of women in the population whose physical ability level is not at the level of elite athletes, moderate PA is likely to be beneficial while unaccustomed, intense PA is likely to be harmful. Each individual should gauge for themselves, under the advice of suitable healthcare professionals, what their "just right" level of PA is (Green Zone in . For elite athletes, while setting PA load, the risk of illness must be weighed against the benefit of being able to return to a competitive level when COVID-19 restrictions are lifted. Furthermore, healthy elite athletes who are accustomed to high PA loads as part of systematic training have a lower risk of illness even with high PA loads. (B) Since the self-reported rating of perceived exertion (RPE) correlates well with VO2 max and HR, RPE is a simple yet reliable tool for gauging the intensity of PA and for adjusting PA load as needed. Since there is a link between PA load (the combination of activity intensity and duration) and illness risk, it is important not to perform high loads of unaccustomed PA during the COVID-19 pandemic. For the majority of women in the population whose physical ability level is not at the level of elite athletes, moderate PA is likely to be beneficial while unaccustomed, intense PA is likely to be harmful. Each individual should gauge for themselves, under the advice of suitable healthcare professionals, what their "just right" level of PA is (Green Zone in . For elite athletes, while setting PA load, the risk of illness must be weighed against the benefit of being able to return to a competitive level when COVID-19 restrictions are lifted. Furthermore, healthy elite athletes who are accustomed to high PA loads as part of systematic training have a lower risk of illness even with high PA loads. (B) Since the self-reported rating of perceived exertion (RPE) correlates well with VO 2 max and HR, RPE is a simple yet reliable tool for gauging the intensity of PA and for adjusting PA load as needed.
## Mental health benefits of pa
To assess the impact of COVID-19 on women, vast sex-disaggregated data will have to be collected and analyzed. Based on recent COVID-19 reports and experience from past MERS and SARS outbreaks, it is known that women face specific risks due to social environments, norms, and unequal power relations, making them highly vulnerable to psychological and physical distress. During long periods of social isolation, it is common for people to experience symptoms of depression, such as sadness, loss of interest or pleasure, feelings of guilt or low self-worth, disturbed appetite or sleep, tiredness, poor concentration, and suicidal thoughts. In particular, women are twice as likely to develop anxiety disorders and mental health crises when compared to men, presumably because of the effect of sex hormones such as estradiol and progesterone. During COVID-19-related quarantines, data suggest that there has been an increase in mild depression, stress, and anxiety reported in women. These changes in mental health might manifest as stress-related eating or using psychoactive substances (e.g., drugs, alcohol, and nicotine).
Longitudinal and cross-sectional studies have demonstrated the positive impact of regular PA on depression. Specifically, PA undertaken before an emotionally stressful stimulus reduces the magnitude of immediate stress and the consumption of unhealthy foods. What is highly encouraging is that PA reduces sleep disorders, anxiety, and depression, and psychoactive substance cravings, even after a single session. Such improvements in mental health and cognition have been linked to changes in the prefrontal area of the cerebral cortex. Acute PA (>60% maximum oxygen uptake) releases beta-endorphin (an endogenous peptide), which modulates pain, reduces stress, activates reward and pleasure areas in the brain, and stabilizes mood and behavior through its agonistic effects on opioid receptors. Due to mandates and guidelines for social distancing during the COVID-19 pandemic, mental health professionals have been required to use telemedicine to provide consultation and prescribe cognitive, behavioral, and pharmacotherapies to treat mental health conditions. Given that in-person psychotherapy might be challenging in terms of scheduling due to social distancing, the therapeutic use of PA to improve mental health is even more appropriate.
## Physical health benefits of pa
The American College of Sports Medicine recommends paying attention to four domains of PA, namely cardiorespiratory fitness, muscular fitness (strength and endurance), flexibility, and neuromotor control. PA that covers these four domains has been shown to produce local effects within cardiac and skeletal muscle, as well as systemic effects in all other physiological systems in the body in both women and men. Specifically, the local and systemic benefits of PA are relevant to maintaining an optimal body mass index (BMI), possessing better insulin sensitivity, achieving a healthy blood lipid profile, and avoiding high blood pressure, which collectively reduces the risk of heart disease-the leading cause of death globally according to the WHO.
Cardiorespiratory fitness (CRF) and the maximal intensity of PA that a person is capable of performing are biomarkers of cardiovascular health. The gold standard measure of CRF is maximal oxygen consumption (VO 2 max), which is the product of cardiac output and the arteriovenous oxygen concentration difference during increasingly demanding PA. In more informal clinical or at-home settings, CRF and PA intensity may be assessed as a rating of perceived exertion (RPE) on a scale of 6-20 or 0-10. Since VO 2 max, heart rate (HR), and RPE (on a 6-20 scale) are well correlated, it is possible for a person to assess their PA intensity based on RPE and to use RPE as a guide to engage in optimal PA based on their own capacity. The promising aspect of using RPE to adjust PA intensity is not only its simplicity but also its ability to account for HR changes, which may be caused by cardiovascular medications.
When compared to age-matched men, women have lower VO 2 max levels due to physiological factors, such as reduced ventricular ejection fraction, hemoglobin concentration, muscle mass, and higher body fat percentages. Even though women have lower CRF compared to men prior to menopause, they have a lower risk of mortality from cardiovascular disease, possibly because hormones such as estradiol and progesterone play protective roles. However, since those at highest risk for COVID-19 complications are over the age of 65 years, women >65 years who contract COVID-19 are most likely to be postmenopausal and not have the benefit of premenopausal cardioprotection. Other factors that negatively impact COVID-19 outcomes are hypertension, cardiovascular disease, diabetes, and obesity. Since these preexisting conditions are positively impacted by CRF, optimizing cardiorespiratory function through PA might most likely be beneficial in the context of preventing COVID-19 complications in women. Furthermore, since better CRF correlates with optimal functioning of the immune system and its inflammatory responses, it is likely that improved CRF might even have direct benefits in the context of COVID-19.
While excessive PA can be detrimental to health in untrained individuals, several studies have shown that moderate PA has a modulatory effect on the immune system and inflammation. Depending on regularity, type, duration, and intensity, PA can have proor anti-inflammatory downstream effects. The balance between these opposing effects is important because immune responsiveness determines whether PA is beneficial or detrimental (e.g., improperly dosed PA can result in muscle injuries or, even worse, rhabdomyolysis and renal failure). Therefore, there is a dose-response relationship between PA and health outcomes. PA can induce changes in peripheral blood cell numbers, granulocyte activity, natural killer (NK) cells, lymphocytes, and plasma cytokine profiles, which correlate with improvements in outcomes of physical health. Angiotensin converting enzyme 2 (ACE2), which is a plasma membrane protein, acts as an entry point for SARS-CoV-2 into host cells and also undergoes changes with PA that might confer a protective effect on the organ systems affected by COVID-19. However, unaccustomed, intense, and prolonged PA can cause tissue damage, impair the ability of the immune system to respond appropriately to an immune challenge (due to lymphopenia), trigger excessive inflammation, and even result in immunosuppression. The effects of excessive PA (e.g., prolonged and repetitive high-intensity activity)can result in physiological changes that resemble sepsis, albeit with milder symptoms. The benefits of optimal PA, however, might not just improve overall health and wellness in women, but might also have direct benefits related to decreasing COVID-19 morbidity and mortality that go beyond the natural biological advantages of the female sex in the context of COVID-19. Thus, it could be argued that, optimal PA might be one of the most effective strategies for women and for society in general to stay healthy during the COVID-19 pandemic-indeed, there is already evidence supporting this notion. One retrospective observational study of over 48,000 patients found that those who were more physically active in the two years preceding COVID-19 infection had reduced odds for hospitalization and death due to COVID-19. Another study of over 76,000 adults found that those who engaged in regular strength training and aerobic PA were less likely to become infected with COVID-19, and those that were infected were less likely to die.
The emerging evidence suggests that patients who develop COVID-19 complications have an abnormal immune response, which includes lymphocytopenia (in~83%), thrombocytopenia (in~36%), leukopenia (in~33%), and elevated levels of c-reactive protein (CRP, in~58%). Additionally, elevated pro-inflammatory cytokine levels, reduced interferon-γ (IFN-γ) levels, and reduced CD4 + and CD8 + T cells suggest that the immune system is dysregulated in COVID-19 with a positive correlation between severity of symptoms and the extent of dysregulation. Many months after COVID-19 was considered a pandemic, promising therapeutic and prophylactic pharmacological agents received emergency authorization, but a cure per se has not yet been established. In this regard, PA as a nonpharmacological modality that can help to enhance the immune and musculoskeletal systems if performed safely at an optimal intensity and duration.
There is a strong association between the type of PA and benefits to the immune system. PA, such as Pilates training performed for 180 min per week, during two weeks of acute PA, improves the innate immune response in adult women, as detected by increased NK cell lytic activity and decreased monocyte chemotactic protein-1 (MCP-1). Bicycle ergometry performed for six minutes at 55% of VO 2 max or for 30 min at 11.11 km/h increases the number of leukocytes (by~36%), granulocytes (by~29%), lymphocytes (by~46%), and monocytes (by~68%) in circulating blood. An acute bout of PA increases circulating concentrations of CD4+ lymphocytes (by 30-40%) and CD8+ lymphocytes (by 90-105%) in peripheral blood. Moderate intensity PA reduces toll-like receptors (TLR), TLR2 (by~35%), TLR4 (by~25%), and IL-6 (by~20%). After moderate treadmill aerobic training or resistance training that was performed three times per week for three months, blood concentrations of pro-inflammatory markers TNF-α, IL-2, IL-4, and CRP were reduced in women. Moderate PA and improved CRF reduce CRP levels and might, therefore, be beneficial for patients with COVID-19. Moderate PA can also help reduce tissue oxidative stress, which in turn reduces inflammation. Thus, due to immune system modulating effects, moderate PA during the COVID-19 pandemic might be beneficial for both healthy women as well as women with asymptomatic COVID-19 infection.
IL-6 is a cytokine that has a dual role, in that, it exerts pro-inflammatory effects when released by inflammatory cells and anti-inflammatory effects when released by skeletal muscle. During PA, contracting muscles release IL-6 into the circulation, which acts as an endocrine signal and exerts positive effects on multiple target tissues. IL-6 release from skeletal muscles is linked to glycogen depletion in muscle, which is in contrast to what is observed in COVID-19 and related diseases, where IL-6 elevation is a result of injury and inflammation in infected cells/tissues. PA-induced elevation in blood IL-6 levels is transient and returns to resting levels usually within a few hours after PA, whereas IL-6 elevation may persist for many days with tissue injury and inflammation.
## The specific role of mucosal immunity and immunoglobulin a (iga) in protection against respiratory infections and symptoms
It is well known that the mucosal immune system provides resistance to the upper respiratory tract infection (URTI), primarily through airway secretory immunoglobulin A (abbreviated as SIgA or S-IgA; sometimes referred to as salivary IgA and abbreviated s-IgA when measured in saliva). SIgA represents one of the body's first lines of defense against URTI through its capacity to inhibit pathogen colonization, bind antigens for transport across epithelial barriers, and neutralize viruses. It is now well established that high PA loads (e.g., marathon running) tend to decrease SIgA levels and, thus, render individuals more susceptible to upper respiratory illness (URI) and upper respiratory symptoms (URS), while moderate PA loads tend to increase SIgA levels, thus providing a first line of defense against URS.
The role of PA levels on SIgA is relevant to COVID-19 since emerging data suggest a link between SIgA and SARS-CoV-2. Ejemel and colleagues found that an IgA form of an antibody raised against the SARS-CoV-2 spike protein showed superior target binding and virus neutralization when compared to its IgG counterpart. Since mucosal SIgA exists mostly in a dimeric form, Wang and colleagues compared the neutralizing effects of monomeric and dimeric forms of anti-SARS-CoV-2 IgA and found that the dimeric form is 15-fold more effective in neutralizing SARS-CoV-2 than the monomeric form and is also several-fold more effective than anti-SARS-CoV-2 IgG. Sterlin and colleagues studied samples from patients with COVID-19 and found the following: anti-SARS-CoV-2 IgA antibody levels rise and fall earlier than IgG antibody levels; IgA preparations were more effective than IgG preparations in neutralizing SARS-CoV-2 pseudovirus; anti-SARS-CoV-2 IgA levels positively correlated with virus neutralization; and anti-SARS-CoV-2 IgAs in bronchoalveolar lavage preparations were more effective in pseudovirus neutralization than compared to their IgG counterparts-all suggesting that IgA-based mucosal immunity likely plays a role in countering SARS-CoV-2. The relevance of IgA-mediated immunity relative to vaccine-induced protection against COVID-19 is, thus, obvious. However, PA-induced SIgA changes, which correlate with protection against URI, and URS must also be emphasized since all individuals might not respond in the same manner with respect to vaccines (e.g., immunocompromisation and immunosenescence)and because vaccine eligibility and supplies are affected by individual, social, political, and economic factors. Several studies investigating the elderly have demonstrated that SIgA levels and secretion rates increase with many weeks to months of moderate intensity PA, which includes both strength and endurance training, thus suggesting that the effects of immunosenescence could somewhat be countered by consistent PA in this population. Although the positive effects of PA on SIgA and the benefits of SIgA in defense against URI and URS are known, at this time, it is unknown as to whether or not PA specifically improves mucosal immunity against SARS-CoV-2 in either women or men.
## Women, pa, and post-acute sequelae of sars-cov-2 infection (pasc)
COVID-19 infection rates appear to be similar between females and males in young, asymptomatic populations, as well as in older symptomatic cohorts around the world. Mortality rates from acute COVID-19 infections are higher in males, while chronic illness (i.e., PASC also known as "long COVID" or "long haul COVID") rates are higher among females. PASC is associated with symptoms, such as physical and cognitive fatigue, breathing difficulty, gastrointestinal disturbances, and changes in mental health, which can persist for many months after COVID-19 infection. The debilitating consequences of PASC underscore the need for both men and women to engage in regular, moderate PA in order to maximize mental, physical, and immunological health. However, the health and societal burden from chronic functional impairments associated with PASC appear to fall disproportionately on women, which may have potentially devastating downstream effects on individuals, families, and societies far beyond the acute phase of SARS-CoV-2 infection.
The potential for long-term illness following acute SARS-CoV-2 infection is supported by longitudinal studies on survivors of SARS-CoV-1 infection, which was responsible for the original SARS outbreak of 2003. In one cohort of 233 survivors hospitalized in Hong Kong, 40% reported the persistence of at least one psychiatric illness, while 40.3% reported chronic fatigue based upon a survey conducted four years after acute SARS illness. Furthermore, healthcare workers were at increased risk for psychiatric symptoms (odds ratio 3.24), while females were overrepresented as study participants (70.4%). Clinical interviews (performed on 181 of 233 survivors) revealed that 46.2% of participants with persistent psychiatric symptoms remained unable to work at the 4-year follow-up. Only 3.3% had a prior history of psychological disturbances before SARS, and they still had ongoing psychological symptoms four years after acute illness, which included post-traumatic stress disorder (in 54.5%), depression (in 39%), somatoform pain disorder (in 36.4%), panic disorder (in 32.5%), and obsessive-compulsive disorder (in 15.6%).
Both SARS-CoV-1 and SARS-CoV-2 are beta coronaviruses, which are positive-sense, single-stranded RNA viruses, and enter host cells through ACE2. The respective spike (S) proteins of SARS-CoV-1 and SARS-CoV-2, which decorate the surface of viral particles and give these viruses with their characteristic solar corona-like appearance, share~75% sequence homology. The~25% difference in the S protein between SARS-CoV-1 and SARS-CoV-2 could be responsible for differences between SARS and COVID-19 (e.g., symptomatic virus shedding from the lower airways in SARS-CoV-1 versus asymptomatic virus shedding from the upper airways in SARS-CoV-2), and thẽ 75% homology could explain why both diseases are highly contagious and are associated with high case fatality in people ≥50 years). Based on post-SARS-CoV-1 infection data, the potential for similar lingering symptoms following COVID-19-particularly in women-appears to be high and exacerbated by the sustained numbers of new SARS-CoV-2 infections around the globe.
The potential for debilitating fatigue, psychiatric illness, and neurological complaints following COVID-19 infection is physiologically supported by laboratory studies. Translocation of the S protein from SARS-CoV-2 from the systemic circulation into the brain occurs via adsorptive transcytosis across the blood-brain barrier (BBB) in murine models. Additionally, in vitro studies suggest that SARS-CoV-2 can replicate within neuronal cells. Collectively, it appears that SARS-CoV-2 infection, through direct and indirect effects on the brain and other neural tissues, may cause a variety of neurological and psychological manifestations that are common in PASC (e.g., fatigue and "brain fog"). The emergence of PASC highlights the effects of sex as a biological variable in COVID-19. Females mount a stronger innate, cellular, and humoral immune response to viral infections but are at higher risk for chronic autoimmune and immunogenic disorders. Thus, although men are at a higher risk than women for severe illness and death from acute COVID-19, women are at a greater risk for chronic COVID-19 illness due to PASC.
It is important to note that post-viral fatigue is not specific to SARS and COVID-19, as chronic fatigue syndromes are well described following infections caused by influenza viruses (H1N1), Epstein-Barr virus, Ebola virus, and West Nile virus. The common theme of post-viral fatigue, regardless of the pathogen, is that it is more frequent, severe, and prolonged in women than in men. For this reason, physiciansand scientistsare linking the pathogenesis and clinical signs and symptoms of PASC with a similar disabling condition known as myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). ME/CFS also disproportionately affects women (especially White/Caucasian women) and was first described as "yuppy flu" in the 1980s due to its poorly understood psychological and physical (e.g., fatigue) manifestations. The pathophysiology of ME/CFS is characterized by autoimmunity and low-grade inflammation resulting from elevated oxidative and nitrosative stress (O&NS), mitochondrial dysfunction, and activation of pro-inflammatory pathways.
The early signs and symptoms of PASC mimic ME/CFS, and emerging studies confirm that females are overrepresented in cohorts with lingering post-COVID symptoms. What is most concerning, however, is the growing scientific recognition that persistent fatigue, neurological manifestations, and PA intolerance occur independent of symptom severity and age. Recent reports document PASC in~51% of individuals in a cohort of 43 COVID-19-positive college-students (96% female) with mild symptoms, as well as in five children between the ages of 9 and 15 (80% female). Since PASC can affect females from a diverse range of populations (e.g., healthcare and services sector workers, primary caregivers to dependent children and others, teachers, and school-age children), it is of great population health and socioeconomic concern.
A worrisome hallmark of ME/CFS is PA intolerance, wherein even slightly excessive PA appears to exacerbate symptomatology or precipitate a relapse into chronic fatigue. Curiously, overtraining syndrome mimics both ME/CFS and PASC, suggesting that overlapping neuro-inflammatory, autoimmune, and/or autonomic pathophysiological processes might be at play. Thus, the true paradox of PA and COVID-19 is that although mild to moderate regular PA may prevent or attenuate morbidity and mortality from COVID-19, once infected with SARS-CoV-2 and PASC develops, the positive health benefits of PA may be negated. Future longitudinal investigations are required to further dissect the effects of sex as a biological variable in the effect of PA as a preventive and/or remedial measure against acute and chronic COVID-19 illness. At present, it appears that in untrained individuals and in individuals with compromised PA tolerance, intense and fatiguing PA in the context of acute or chronic COVID-19 illness may be detrimental, but mild to moderate PA that is adjusted based on RPE might be beneficial for optimizing mental, physical, metabolic, and immune health. As with ME/CFS, the most practical strategy to work through PASC might be to balance PA with intentional rest, avoid fatigue, pace daily routines, and resist the urge to "push oneself physically" on good days. Additionally, due to the gravity of acute COVID-19 and PASC, it is highly recommended that women receive prophylactic pharmacological therapy against acute COVID-19 as early as possible (i.e., vaccination when eligible) and to continue to follow nonpharmacological physical risk mitigation measures (distancing, masking, hygiene, and self-isolation when sick) as part of maintaining an active lifestyle that prioritizes optimal PA.
## Recommendations for staying active during a the covid-19 pandemic
The WHO campaign "Be active and Stay Healthy at home", in accordance with the PA Guidelines for Americans, recommends performing adequate PA to improve the following: physical fitness (cardiorespiratory and muscular fitness), cardiometabolic fitness (blood pressure, lipid profile, and glycemic control), bone health, cognitive outcomes and mental health, balance, and flexibility. For adults between 18 and 64 years, 150-300 min of moderate intensity PA or at least 75-150 min of vigorous PA throughout the week is recommended. Pregnant healthy women should undertake at least 150 min per week of moderate intensity cardiorespiratory (i.e., aerobic) PA in order to increase or maintain CRF, optimize BMI, and reduce the risk and severity of postpartum depression. If pregnant women were accustomed to vigorous aerobic PA before pregnancy, they may continue that level of PA during pregnancy. The recommendation for girls between 6 and 17 years is to perform an average of 60 min of moderate to vigorous PA per day. For older adults, it is suggested that they engage in moderate intensity PA for >3 days each week and to undertake up to 300 min of PA per week in order to enhance functional capacity and prevent falls.
The American College of Sports Medicine (ASCM) has published specific guidelines on how to remain physically active during the COVID-19 pandemic. There is no recommendation at this time to limit PA during acute uncomplicated COVID-19. However, in light of what is known (and has been discussed in preceding sections of this paper) about the effects of PA on the immune system, it seems logical that unaccustomed and intense PA may not be advisable in order to avoid overwhelming the immune and other physiological systems. Furthermore, when COVID-19 is suspected or confirmed, it is necessary to monitor symptoms (mainly difficulty breathing and reduced oxygen saturation measured with a finger pulse oximeter) and assiduously follow physical risk mitigation measures (distancing, masking, hygiene, and self-isolation) in order to avoid complications and reduce community spread of SARS-CoV-2 through aerosolized viral particles.
Regular, moderate intensity PA provides numerous mental and physical health benefits to women in the context of COVID-19 or otherwise . However, the complex roles played by women in society can render it challenging for them to be motivated to consistently make PA a priority. Women are likely to adhere to regular PA routines when there is social support, while men may rely more on competition to keep them motivated. While there is no doubt that performing optimal levels of PA during the COVID-19 pandemic might be challenging, receiving prophylactic pharmacological therapy against serious complications as early as possible (i.e., vaccination when eligible) and continuing to follow nonpharmacological physical risk mitigation measures (distancing, masking, hygiene, and self-isolation when sick) make the goal of consistently engaging in moderate-intensity PA quite achievable.
## Covid-19 vaccination and its relevance to women's health and maintaining an active lifestyle
As of 26 July 2021, a conservative estimate of total COVID-19 cases was~194 million people, of which four million people had died, thus placing the worldwide case fatality rate at~2% (about one death for every 50 confirmed cases). When COVID-19 was declared a global pandemic in 2020, clinicians and scientists around the world desperately looked to find therapeutics that could be repurposed to reduce the rapidly rising number of COVID-19 deaths. In less than a year, through the concerted and concurrent efforts of health agencies, scientists, clinicians, industry partners, and research volunteers worldwide, many vaccine candidates were developed and tested. Vaccines that passed rigorous preclinical testing (testing in animals) and phased clinical trials (testing in humans) and that were deemed safe (i.e., extremely rare serious side effects and adverse events) and effective (i.e., reduced the probability of infection and serious illness) by multiple regulatory agencies were granted emergency use authorization. Emerging data indicate that the widely administered BNT162b2 vaccine (mRNA technology; manufactured by: Pfizer, New York, NY, USA, and BioNTech, Maintz, Germany) and ChAdOx1 nCoV-19 vaccine (adenoviral vector technology; manufactured by: Oxford University, Oxford, UK, and AstraZeneca, Cambridge, UK) are effective at reducing infectionand hospitalization, even against new and highly contagious SARS-CoV-2 variants (e.g., the Delta strain). Thus, it can be concluded that the best method to prevent hospitalization and death from COVID-19 would be to receive one of the vaccines that have been recognized by a reputable health agency, such as the WHO. The unprecedented ability to receive protection against COVID-19 hospitalization and death has unfortunately been undermined by rampant misinformation regarding COVID-19coupled with vaccine inequity and ineligibility worldwide. Receiving a vaccine as soon as possible when eligible will help women engage more safely in PA due to the reduced risk of acute COVID-19 infection and complications if exposed to SARS-CoV-2.
With regards to COVID-19 vaccination and women's health, data from the United States collected during the first month of the vaccine rollout when only mRNA vaccines were available showed that more women (61.2%) received vaccination compared to men. However, a greater proportion of women also reported side effects or adverse events (78.7%) after receiving a vaccine. Even though women report side effects or adverse events more frequently after receiving an mRNA COVID-19 vaccine, the protection rendered against acute COVID-19, with or without PASC, far exceeds the transient side effects. Further investigation is warranted on sex-specific, post-vaccination symptomatology, immunological responses, and the risk of breakthrough infection and transmission.
However, at this time there are no scientific data to support concerns regarding derangements in menstruation, fertility, childbearing capabilities, or an increased risk with respect to pregnant women or the developing fetus following vaccination. It could be argued that the safety and adverse event profiles of the authorized vaccines are even better than some of the commonly used over-the-counter medications, such as nonsteroidal antiinflammatory drugs (NSAIDs; e.g., drugs that end with the suffix -profen, -proxen, -oxicam, and -fenac), that are taken for musculoskeletal pain. Gaining vaccine confidence in women might have widespread global health benefits due the pivotal role they play in the health and wellness of families through nurturing and caregiving for dependent children and others. Nonetheless, women must be allowed to make independent and informed decisions regarding receiving COVID-19 vaccination in consultation with their healthcare providers-this must be based on best medical practices and not based on misinformation and societal pressure. Finally, women should be able to rest and slowly ramp up PA based on how their body responds to COVID-19 vaccination. Some nations, such as New Zealand, have implemented leave policies for individuals who might develop a rare adverse reaction following COVID-19 vaccination(Supplementary Materials-personal correspondence from Mr. Moses Benjamin, Allied Health Director, Auckland District Health Board, New Zealand).
Despite the protection against COVID-19 hospitalization and death offered by vaccines, the WHO is requesting nations where vaccination rates are high to continue to follow nonpharmacological physical risk mitigation measures, such as social distancing, wearing proper facemasks, following good sanitation and hygiene practices, and getting tested and self-isolating when sick. The need for continued physical risk mitigation measures even after vaccination is supported by mathematical models, which suggest that, even with perfect vaccine acceptance scenarios, it would likely take many months to a year for community spread to consistently remain at low levels that are not of concern. The risk of rare breakthrough infections (i.e., vaccinated individuals that are infected with SARS-CoV-2), the possibility of new viral variants emerging in unvaccinated and vaccinated individuals due to the inherent biology of coronaviruses, and the fact that only a few countries currently have enough doses to vaccinate their populationscollectively validate the WHO's abundance of caution and related recommendations. Since the health of individuals in any part of the world has an impact on global health, all nations must show solidarity with the rest of the world and follow the WHO's recommendations in order to aggressively vaccinate their populations, share unused vaccine doses with other countries, and continue to follow physical risk mitigation measures in order to complement worldwide vaccination efforts. The slogan of COVAX, the WHO-led alliance for global equitable access to COVID-19 vaccines, sums it best: "with a fast-moving pandemic, no one is safe, unless everyone is safe".
# Conclusions
PA during the COVID-19 pandemic is a double-edged sword for women since mild to moderate PA (based on RPE) may be beneficial, but unaccustomed and intense PA could increase illness risk. Moderate PA may enhance immune and other physiological functions, but intense PA is best avoided by untrained individuals because it may trigger maladaptive physiological responses, rendering people more susceptible to acute and chronic COVID-19 complications. Although SARS-CoV-2 is likely to infect women and men at similar rates, sex-specific behavioral and physiological responses may alter the clinical trajectory of COVID-19, e.g., higher risk of acute illness complications in men but higher incidence and severity of PASC in women may occur. From a mental health perspective, it is clear that women, as caregivers, are disproportionally overburdened by mental health crises. Depression, emotional stress, anxiety, eating disorders, and psychoactive substance cravings are reduced by regular PA and, therefore, should be encouraged in order to improve both mental and physical health. Pandemic precautions must, however, be followed diligently to keep oneself safe and to minimize community spread of SARS-CoV-2. Moving forward, investigations on the influence of sex hormones on PA-induced immunomodulation may identify physiological responses that may be protective against COVID-19 (and offer therapeutic targets). Due to the novelty of SARS-CoV-2 in humans, comprehensive clinical studies, follow-up cohort assessments, and analyses of data in a sex-disaggregated manner are needed for elucidating the effects of preventive interventions (e.g., PA) when pandemic precautions are in effect. Finally, as part of prioritizing an active lifestyle, it is essential that women receive prophylactic pharmacological therapy against serious complications as early as possible (i.e., vaccination when eligible) and continue to follow nonpharmacological physical risk mitigation measures (social distancing, masking, hygiene, and self-isolation when sick). Such healthy behaviors will contribute to personal, family, community, and global health and wellness, and will ultimately accelerate exiting the COVID-19 pandemic.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10.339 0/ijerph181910271/s1, File S1: Personal correspondence from Moses Benjamin, Allied Health Director, Auckland District Health Board, New Zealand, regarding New Zealand's leave policy for employees who experience side effects and/or adverse events following COVID-19 vaccination.
Author Contributions: All authors contributed equally to the conceptualization and writing of this paper. The original idea to write a paper linking women's health, COVID-19, and PA was from K.P. Institutional Review Board Statement: Not applicable since this is a review paper and did not involve data collection from humans or animals.
Informed Consent Statement: Not applicable since this is a review paper and did not involve data collection from humans. Data Availability Statement: Data sharing not applicable. No new data were created or analyzed in this study. Data sharing is not applicable to this article.
|
The STAT3 Target Gene TNFRSF1A Modulates the NF-κB Pathway in Breast Cancer Cells1
The transcription factor STAT3 is activated inappropriately in 70% of breast cancers, most commonly in triple negative breast cancer (TNBC). Although the transcriptional function of STAT3 is essential for tumorigenesis, the key target genes regulated by STAT3 in driving tumor pathogenesis have remained unclear. To identify critical STAT3 target genes, we treated TNBC cell lines with two different compounds that block STAT3 transcriptional function, pyrimethamine and PMPTP. We then performed gene expression analysis to identify genes whose expression is strongly down-regulated by both STAT3 inhibitors. Foremost among the down-regulated genes was TNFRSF1A, which encodes a transmembrane receptor for TNFα. We showed that STAT3 binds directly to a regulatory region within the TNFRSF1A gene, and that TNFRSF1A levels are dependent on STAT3 function in both constitutive and cytokine-induced models of STAT3 activation. Furthermore, TNFRSF1A is a major mediator of both basal and TNFα-induced NF-κB activity in breast cancer cells. We extended these findings to primary human breast cancers, in which we found that high TNFRSF1A transcript levels correlated with STAT3 activation. In addition, and consistent with a causal role, increased TNFRSF1A expression was associated with an NF-κB gene expression in signature in breast cancers. Thus, TNFRSF1A is a STAT3 target gene that regulates the NF-κB pathway. These findings reveal a novel functional crosstalk between STAT3 and NF-κB signaling in breast cancer. Furthermore, elevated TNFRSF1A levels may predict a subset of breast tumors that are sensitive to STAT3 transcriptional inhibitors, and may be a biomarker for response to inhibition of this pathway.
# Introduction
Breast cancer is the most common cancer among women, and the second leading cause of cancer deaths in women, accounting for over 230,000 new cases and 40,000 deaths per year in the US, and for 517,000 deaths in 2015 worldwide. One approach to developing novel therapies for cancer is to identify molecular pathways that are activated inappropriately in this disease. The transcription factor STAT3, which regulates genes controlling proliferation, survival, and pluripotency, is normally activated rapidly and transiently in response to hormones and growth factors [bib_ref] STAT5 Outcompetes STAT3 To Regulate the Expression of the Oncogenic Transcriptional Modulator..., Walker [/bib_ref] [bib_ref] Distinct roles of STAT3 and STAT5 in the pathogenesis and targeted therapy..., Walker [/bib_ref]. However, in approximately 70% of primary breast cancers, STAT3 is activated constitutively, and directly contributes to the pathogenesis of this disease [bib_ref] Identification of a Genetic Signature of Activated Signal Transducer and Activator of..., Alvarez [/bib_ref] [bib_ref] Reciprocal effects of STAT5 and STAT3 in breast cancer, Walker [/bib_ref].
Under basal conditions, STAT3 resides in the cytoplasm in an inactive conformation. It becomes activated by phosphorylation on a critical tyrosine residue, tyrosine 705, by cytokine-receptor-associated tyrosine kinases, like Janus kinases (JAKs), growth factor receptors with intrinsic tyrosine-kinase activity, or non-receptor protein tyrosine kinases. Upon activation by tyrosine phosphorylation, STAT3 forms active dimers that translocate into the nucleus and www.neoplasia.com Volume 20 Month 2018May 2018 pp. 489-498 489
bind to DNA at cognate binding sites in the regulatory region of target genes. STAT3 then regulates the expression of key target genes involved in a variety of cellular processes, including proliferation, differentiation, migration, survival, and angiogenesis [bib_ref] Dangerous liaisons: STAT3 and NF-κB collaboration and crosstalk in cancer, Grivennikov [/bib_ref] [bib_ref] Nuclear translocation of STAT3 and NF-κB are independent of each other but..., Martincuks [/bib_ref].
Since STAT3 is largely dispensable in normal cells but essential for the survival of malignant cells, STAT3 inhibition can have a high therapeutic index. A number of approaches have been taken to inhibit STAT3 therapeutically, including kinase inhibitors and dimerization antagonists, though these may display off-target or non-specific effects [bib_ref] The JAK2 inhibitor AZD1480 potently blocks Stat3 signaling and oncogenesis in solid..., Hedvat [/bib_ref] [bib_ref] Discovery of JSI-124 (Cucurbitacin I), a selective janus kinase/signal transducer and activator..., Blaskovich [/bib_ref] [bib_ref] A low-molecular-weight compound discovered through virtual database screening inhibits Stat3 function in..., Song [/bib_ref] [bib_ref] Structure-based design of conformationally constrained, cellpermeable STAT3 inhibitors, Chen [/bib_ref]. Using transcription-based screening assays, it has been possible to identify a number of novel inhibitors of STAT3 function [bib_ref] A chemical biology approach to developing STAT inhibitors: molecular strategies for accelerating..., Nelson [/bib_ref]. As STAT3 inhibitors are now being introduced into clinical trials (https://clinicaltrials.gov/ct2/show/NCT01066663), an emerging question is how to identify those cancers most likely to respond to STAT3 inhibition. One approach would be to identify direct STAT3 target genes whose expression is reproducibly inhibited by STAT3 transcriptional inhibitors. Such a gene product might also serve as a pharmacodynamic marker to monitor response to STAT3-directed therapy, and might reveal important novel aspects of STAT3 signaling in cancer cells.
To address this question, we began by identifying STAT3 target genes whose expression is inhibited by two different compounds that specifically block STAT3-dependent gene expression. We then focused on one of these STAT3 targets, which is also unique in that it links STAT3 signaling with another oncogenic transcription factor, NF-κB, in both breast cancer cell lines and primary human breast cancers.
# Materials and methods
## Cell lines
MDA-MB-468 cells (from Myles Brown, Dana-Farber Cancer Institute) and BT549 cells (from Kornelia Polyak, Dana-Farber Cancer Institute) were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum. SK-BR-3 cells (received from Lyndsay Harris, Dana-Farber Cancer Institute) were maintained in RPMI containing 10% fetal bovine serum. Cells were passaged for less than 3 months after thawing. All cells were maintained in a humidified incubator at 37°C with 5% CO 2 , and were authenticated by short tandem repeat DNA profiling.
Microarray SK-BR-3 cells were pre-treated with PMPTP (4-[4-(phenylmethyl) piperidin-1-yl] thieno pyrimidine; 5 μM), pyrimethamine (5 μM), or DMSO vehicle for 1 hour then stimulated with 10 ng/mL of LIF for 90 minutes. Total cellular RNA was isolated using Trizol, and then further purified using a Qiagen RNeasy Mini kit (Valencia, CA). RNA quality was evaluated on a NanoDrop 8000 spectrophotometer (Thermo Fisher Scientific). 1.5 μg of total RNA was submitted for gene expression profiling on the Human Gene ST 2.0 array (21,094 genes; Affymetrix) by the Dana-Farber Cancer Institute Molecular Biology Core Facilities. Raw data expressed as CEL files were normalized using Expression Console software (Affymetrix) and differential gene expression between drug-treated and control samples investigated using Transcriptome Analysis Console software (Affymetrix). Genes exhibiting ≥1.2-fold change were considered up-regulated by LIF compared to unstimulated cells.
## Cytokine stimulation
Cells were stimulated with 10 ng/mL interleukin (IL)-6 (Peprotech, Rocky Hill, NJ), 10 ng/mL TNFα (Peprotech), or 10 ng/mL LIF (EMD Millipore, Billerica, MA). Cells were stimulated for 15 minutes for whole protein analyses (immunoblotting), 30 minutes for cellular fraction protein analysis and ChIP analysis, 90 minutes for mRNA analyses, and 6 hours for luciferase reporter assays.
## Chromatin immunoprecipitation (chip)
ChIP was performed as previously described [bib_ref] Isolation of unique STAT5 targets by chromatin immunoprecipitation-based gene identification, Nelson [/bib_ref]. Briefly, cells (1.5 x 10 7 ) were fixed in 1% formaldehyde for 10 minutes, sonicated in 15 second pulses using a Fisher Scientific Sonic Dismembranator Model 500 PDQ on setting 15, and lysates were immunoprecipitated overnight at 4°C with an antibody for STAT3 (sc-482, Santa Cruz Biotechnology). Quantitative PCR was performed using the indicated primers , and signal detected was normalized to input and compared to a non-binding region.
## Transfection of expression constructs
Cells (5 × 10 5 cells per well in a 6-well plate) were seeded, and the following day were transfected using Lipofectamine 2000 (Invitrogen) with 1 μg of TNFR1 (pBMNZ-neo-Flag-TNFR1 L380A (from Martin Kluger), Addgene plasmid # 43949), or an empty vector as a control. 48 hours after transfection, the cells were stimulated with TNFα as described above and total protein lysates were obtained.
## Rna interference
## Immunoblot analyses and nuclear fractionation
Cells (5 × 10 5 cells per well in a 6-well plate) were lysed on ice for 15 minutes in RIPA lysis buffer (Boston BioProducts, Boston, MA) with phosphatase and complete protease inhibitors (Roche). Immunoblots were probed with antibodies to TNFR1 (21574-1-AP, Proteintech, Rosemont, IL), p65 (3033, Cell Signaling Technology Inc., Danvers, MA), PARP (9542, Cell Signaling), and tubulin (T5168, Sigma-Aldrich). Cellular fractionation was performed according to the manufacturer's protocol (Active Motif Nuclear Extract Kit Cat. No. 40010; Carlsbad, CA). Band intensity was quantitated using ImageJ software (National Institutes of Health).
## Luciferase reporter assays
Cells (5 × 10 4 cells per well in a 24-well plate) were reverse transfected using Lipofectamine RNAiMAX (Invitrogen) with 10 nM of small interfering RNA (siRNA) targeting TNFRSF1A#1, TNFRSF1A#2 or TNFRSF1A#3, or non-targeting siRNA Control. The following day, 1 μg of an NF-κB-dependent luciferase reporter (Stratagene) was transfected into the cells in combination with 0.1 μg Renilla luciferase transfection control (Promega) using Lipofectamine 2000 (Invitrogen). 24 hours after transfection, the cells were stimulated for 6 hours with TNFα, then lysed and quantitated by a dual-luciferase assay (Promega), and read on a Luminoskan Ascent luminometer (ThermoLab Systems, Helsinki, Finland). NF-κBdependent luciferase production was normalized to Renilla luciferase values.
In the case of ectopic TNFRSF1A expression, cells (5 × 10 4 ) were seeded, and the following day were transfected (using Lipofectamine 2000; Invitrogen) with 0.5 μg TNFR1 plasmid (pBMNZ-neo-TNFR1 L380A, Addgene, Boston, MA), or an empty vector control, and 0.5 μg of an NF-κB-dependent reporter (NF-κB-luciferase) in combination with 0.05 μg Renilla luciferase transfection control (phRL-TK-luc; Promega). 24 hours after transfection, the cells were stimulated for 6 hours with TNFα, and analyzed as described above.
## Mrna expression analyses (rt-pcr)
Total cellular RNA was isolated using Qiagen RNeasy Mini kits. RNA quality was evaluated on a NanoDrop 8000 spectrophotometer (Thermo Fisher Scientific), and reverse transcribed with TaqMan (Applied Biosystems, Foster City, CA) to generate cDNA. Quantitative polymerase chain reaction (qPCR) was performed in quadruplicate using Power SYBR master mix (Applied Biosystems) on a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems). Specificity of amplification was confirmed by melt curve analysis. Cycle threshold (Cτ) values for target genes were normalized to the endogenous reference gene GAPDH, and the fold change was determined by dividing the expression in each sample by that of the unstimulated control sample. Primer sequences (Supplementary were designed from the UCSC genome browser reference mRNA sequences using Primer3.
## Analysis of primary breast cancer data
Data on the phosphorylation of STAT3 on tyrosine 705 (STAT3_PY705) and gene expression were downloaded from the Cancer Genome Atlas (TCGA) breast invasive carcinoma dataset from cBioportal on December 12, 2016. Two other breast cancer microarray datasets were downloaded from the Gene Expression Omnibus (GSE5460 [bib_ref] Identification of a Genetic Signature of Activated Signal Transducer and Activator of..., Alvarez [/bib_ref] and GSE6861). Phosphorylation of STAT3 on tyrosine 705 (STAT3_PY705) on dataset GSE5460 was published previously [bib_ref] Identification of a Genetic Signature of Activated Signal Transducer and Activator of..., Alvarez [/bib_ref].
Gene-set enrichment analysis on datasets GSE5460 and GSE6861 was performed using xapps.gsea.Main from the Broad Institute of MIT and Harvard (http://software.broadinstitute.org) [bib_ref] Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles, Subramanian [/bib_ref] [bib_ref] PGC-1 [alpha]-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human..., Mootha [/bib_ref] , based on STAT3 [bib_ref] Identification of a gain-offunction STAT3 mutation (p.Y640F) in lymphocytic variant hypereosinophilic syndrome, Walker [/bib_ref] and NF-κB [bib_ref] NF-kappa B genes have a major role in Inflammatory Breast Cancer, Lerebours [/bib_ref] expression signatures, as well as Hallmarks gene sets collection (v6.1) from the Broad Institute of MIT and Harvard (http://software.broadinstitute.org) [bib_ref] Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles, Subramanian [/bib_ref] [bib_ref] PGC-1 [alpha]-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human..., Mootha [/bib_ref].
## Statistical analyses
Results are presented as ±SD. Two-tailed Student t tests for paired samples were performed with GraphPad Prism 6 software (La Jolla, CA). Values of P b .05 were considered significant (*, P b .05; **, P b .01; ***, P b .001; ****, P b .0001).
# Results
## Tnfrsf1a is down-regulated by stat3 transcriptional inhibitors in breast cancer cells
We initially focused on identifying key target genes of STAT3 that mediate its oncogenic effects and could serve as biomarkers for therapy targeting STAT3. To do this, we utilized two pharmacological inhibitors of STAT3 transcriptional function, pyrimethamine and PMPTP, that do not disrupt upstream events such as phosphorylation or nuclear localization at the low concentration used in the current study (5 μM) [bib_ref] A chemical biology approach to developing STAT inhibitors: molecular strategies for accelerating..., Nelson [/bib_ref] [bib_ref] Pyrimethamine inhibits adult polycystic kidney disease by modulating STAT signaling pathways, Takakura [/bib_ref]. While these compounds are specific for STAT3, in that they do not block the transcriptional activity of highly related transcription factors, they may not bind directly to STAT3, and their precise mechanism of action is still being determined. We reasoned that gene expression changes occurring in response to both of these compounds were most likely due to effects on STAT3 transcription rather than to effects on unrelated pathways.
To optimize the dynamic range of gene expression, we initially used SK-BR-3 breast cancer cells, which lack constitutive STAT3 phosphorylation, but show robust induction of STAT3 phosphorylation in response to the cytokine leukemia inhibitory factor (LIF) [bib_ref] STAT5 Outcompetes STAT3 To Regulate the Expression of the Oncogenic Transcriptional Modulator..., Walker [/bib_ref]. We first used gene expression profiling to identify genes induced after 90 minutes of LIF stimulation. We identified 3227 LIF-up-regulated genes, defined as those whose mRNA levels were induced by at least 1.2-fold in LIF-stimulated versus unstimulated, vehicle-treated cells. We then compared the induction of these genes in cells pre-treated for one hour with PMPTP or pyrimethamine, defining genes reduced by STAT3 inhibitors as those showing a decreased induction of at least 20% compared to control cells. PMPTP reduced the induction of 1924 LIF-up-regulated genes while pyrimethamine reduced the induction of 1714 LIF-up-regulated genes, with an overlap of 1140 genes whose induction was reduced by both drugs [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref].
To restrict our analysis to the LIF-up-regulated genes most likely to be directly regulated by STAT3, we intersected the LIF-up-regulated genes with genes demonstrating direct STAT3 DNA binding by ChIP-Seq. This yielded 196 genes, from which we focused on the gene that showed the maximal repression with both inhibitors, TNFRSF1A (Tumor Necrosis Factor Receptor Superfamily Member 1A, TNFR1), which was inhibited by greater than 95% by both compounds.
## Stat3 regulates the expression of tnfrsf1a in breast cancer cells
Given the identification of TNFRSF1A as a STAT3 target gene that is down-regulated by STAT3 transcriptional inhibitors, we wished to validate and further elucidate the relationship between STAT3 and TNFRSF1A. ChIP-seq identified a peak of STAT3 binding within the first intron of this gene (Supplementary [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref] , and subsequent sequence analysis revealed three canonical STAT3 binding sites within approximately 1 kb in this region (Supplementary [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref]. This proximity of STAT3 binding sites is significant, given the cooperativity of STAT3 DNA binding at tandem sites [bib_ref] Impact of the N-terminal domain of STAT3 in STAT3-dependent transcriptional activity, Hu [/bib_ref]. We next wished to confirm that STAT3 could bind to this regulatory region of TNFRSF1A in an inducible manner. We performed directed ChIP to determine the binding of STAT3 to this region in SK-BR-3 cells in the presence or absence of LIF treatment. Treating cells with LIF for 30 minutes led to a 20-fold increase of STAT3 binding [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref] , suggesting that this is a functional site.
We next focused on the regulation of TNFRSF1A mRNA by STAT3. Triple negative breast cancers (TNBCs) characteristically display constitutive phosphorylation of STAT3, and the viability of these cells is inhibited by STAT3 inhibitors [bib_ref] Distinct roles of STAT3 and STAT5 in the pathogenesis and targeted therapy..., Walker [/bib_ref]. To determine the functional interaction between STAT3 and TNFRSF1A in triple negative breast cancer cells, we used RNA interference to silence STAT3 in MDA-MB-468 and BT549 cell lines, which have constitutive activation of STAT3. Knockdown of STAT3 reduced the expression of TNFRSF1A mRNA and protein by greater than 50% in both cell lines [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref].
We next considered the possibility that STAT3 regulates TNFRSF1A expression in breast cancer cells that do not display constitutively active STAT3, but in which STAT3 activation can be induced by cytokine stimulation. We examined the breast cancer cell Neoplasia Vol. The STAT3 Target Gene TNFRSF1A Modulates the NF-κB Pathway Egusquiaguirre et al.
The STAT3 Target Gene TNFRSF1A Modulates the NF-κB Pathway Egusquiaguirre et al. line SK-BR-3, which lack constitutive STAT3 phosphorylation, but in which STAT3 can be specifically activated by cytokine stimulation, as may occur in the tumor microenvironment. Treatment with IL-6 induced approximately a two-fold increase in TNFRSF1A expression in these cells. However, this effect was completely abrogated when STAT3 was depleted by RNA interference [fig_ref] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells [/fig_ref]. Taken together, these findings indicate that TNFRSF1A is a direct transcriptional target of STAT3 in both constitutive and cytokine-activated breast cancer systems.
## Stat3-modulated tnfrsf1a expression regulates nf-κb activity
Having identified TNFRSF1A as a key target gene of STAT3 in breast cancer cells that is down-regulated by STAT3 transcriptional inhibitors, we next focused on the biological function of this protein.
TNFRSF1A is the key cell surface receptor for the cytokine TNFα. When TNFα binds to TNFRSF1A, it induces activation of the transcription factor NF-κB, which also promotes survival and proliferation in breast cancer [bib_ref] Targeting of nuclear factor κB pathways by dehydroxymethylepoxyquinomicin, a novel inhibitor of..., Matsumoto [/bib_ref] [bib_ref] Hyperactivated NF-κB and AP-1 transcription factors promote highly accessible chromatin and constitutive..., Ndlovu [/bib_ref]. Furthermore, malignant cells themselves can produce TNFα, thereby enhancing the activation of NF-κB pathway [bib_ref] Dangerous liaisons: STAT3 and NF-κB collaboration and crosstalk in cancer, Grivennikov [/bib_ref] , and TNFα in the breast cancer microenvironment can enhance tumorigenesis [bib_ref] TNF-α in cancer treatment: molecular insights, antitumor effects, and clinical utility, Van Horssen [/bib_ref]. To prevent excessive signaling through both of these pathways, cells have negative feedback mechanisms to minimize co-activation of STAT3 and NF-κB, although these processes can be subverted in breast cancer pathogenesis [bib_ref] STAT3 Induction of miR-146b Forms a Feedback Loop to Inhibit the NF-κB..., Xiang [/bib_ref]. Given this association of TNFα with NF-κΒ activation, we hypothesized that STAT3-driven TNFRSF1A expression regulates NF-κB transcriptional activity in TNBC cells. To test this hypothesis, we transfected MDA-MB-468 or BT549 TNBC cells with a luciferase reporter gene under the control of an NF-κBregulated promoter. When the cells were treated with any one of three distinct siRNAs targeting TNFRSF1A, there was a decrease in basal NF-κB activity [fig_ref] Figure 2: TNFRSF1A expression modulates NF-κB activity [/fig_ref]. Even more significantly, the increase in NF-κB activity induced by TNFα was completely abrogated by knockdown of TNFRSF1A [fig_ref] Figure 2: TNFRSF1A expression modulates NF-κB activity [/fig_ref]. To determine whether this role for TNFRSF1A extended to endogenous NF-κB-regulated genes, we assessed the mRNA expression of well-annotated NF-κB target genes, including IL-8, A20, BIRC3 and IL-6. Knockdown of TNFRSF1A completely suppressed the TNFα-induced expression of each of these genes in both TNBC cell lines [fig_ref] Figure 2: TNFRSF1A expression modulates NF-κB activity [/fig_ref].
TNFRSF1A ultimately controls NF-κB activity by regulating the release of transcriptional NF-κB subunits from inactive complexes in the cytoplasm, to allow nuclear translocation and DNA binding. The p65 subunit of NF-κB (RelA) is the most abundant transcriptionally active form of NF-κB in breast cancer cells. Therefore, we analyzed the effect of TNFRSF1A expression on the nuclear localization of p65 in TNBC cells. Knockdown of TNFRSF1A led to a prominent decrease of TNFα-induced p65 nuclear localization in both cell lines, suggesting that TNFRSF1A regulates NF-κB transcriptional activity through regulation of nuclear accumulation of p65.
To further elucidate how modulation of TNFRSF1A expression affects NF-κB activity, we increased expression of this protein by transfecting an expression construct (TNFR1) into MDA-MB-468 and BT549 cells [bib_ref] Targeting of tumor necrosis factor receptor 1 to low density plasma membrane..., D'alessio [/bib_ref]. We first verified that increased TNFRSF1A protein levels could be detected in MDA-MB468 and BT549 whole cell lysates 24 h and 48 h after transfection, relative to the empty vector [fig_ref] Figure 4: TNFRSF1A expression modulates NF-κB activity [/fig_ref]. We then analyzed integrated NF-κB transcriptional activity using an NF-κB-dependent luciferase reporter construct. Increased expression of TNFRSF1A led to an increase in both basal NF-κB activity, as well as that detected after stimulation with TNFα [fig_ref] Figure 4: TNFRSF1A expression modulates NF-κB activity [/fig_ref]. To determine whether increased TNFRSF1A expression also affected expression of endogenous NF-κB-regulated genes, we used RT-PCR to quantitate the mRNA of well-annotated NF-κB target genes, including IL-8, A20, BIRC3 and IL-6. TNFRSF1A overexpression led to an increased TNFα-induced expression of each of these genes in both TNBC cell lines [fig_ref] Figure 4: TNFRSF1A expression modulates NF-κB activity [/fig_ref]. Taken together, these findings suggest that modulation of TNFRSF1A expression is a major regulator of NF-κB transcriptional function in TNBC cells.
## Tnfrsf1a expression correlates with stat3 activation in primary breast cancers
Having identified TNFRSF1A as a STAT3-regulated gene that links STAT3 and NF-κB signaling in breast cancer cell lines, we next wished to determine if this relationship occurs in primary human breast tumors. Using data from The Cancer Genome Atlas (TCGA) and from the breast cancer microarray dataset GSE5460 [bib_ref] Identification of a Genetic Signature of Activated Signal Transducer and Activator of..., Alvarez [/bib_ref] , we first segregated breast cancers based on the activating tyrosine phosphorylation of STAT3. If TNFRSF1A is a key STAT3 target gene in primary breast cancers, we predicted that TNFRSF1A expression would be higher in tumors with activated STAT3. Indeed, we found significantly higher mRNA levels of TNFRSF1A in breast tumors with phosphorylated STAT3 [fig_ref] Figure 5: TNFRSF1A expression correlates with STAT3 activation in primary breast cancers [/fig_ref]. As the phosphorylated-tyrosine epitope of STAT3 may be labile when analyzed from primary tumors, a STAT3 gene expression signature can be used as an independent measure of transcriptionally active STAT3. Thus, we used gene set enrichment analysis (GSEA) to determine the relationship between a STAT3 gene expression signature [bib_ref] Identification of a gain-offunction STAT3 mutation (p.Y640F) in lymphocytic variant hypereosinophilic syndrome, Walker [/bib_ref] and TNFRSF1A levels. We found a strong correlation between the presence of a STAT3 gene expression signature and TNFRSF1A mRNA expression [fig_ref] Figure 5: TNFRSF1A expression correlates with STAT3 activation in primary breast cancers [/fig_ref] , further suggesting that TNFRSF1A follows STAT3 activation in primary breast cancers. . TNFRSF1A depletion decreases nuclear NF-κB (p65) localization. MDA-MB-468 and BT549 cells were transfected with siRNA targeting TNFRSF1A or a non-targeting control, and then stimulated with TNFα. Nuclei were isolated, and p65 (RelA) was quantitated by immunoblot and normalized to PARP, which served as a loading control for nuclear protein.
If TNFRSF1A is a key mediator of NF-κB activity within primary breast tumors, then we would predict that higher expression of TNFRSF1A would correlate with increased NF-κB transcriptional activity. To test this hypothesis we examined the relationship between TNFRSF1A mRNA levels and an NF-κB gene expression signature using GSEA [bib_ref] NF-kappa B genes have a major role in Inflammatory Breast Cancer, Lerebours [/bib_ref]. We found that an NF-κB gene expression signature was highly correlated with high TNFRSF1A expression [fig_ref] Figure 6: TNFRSF1A expression correlates with NF-κB activation in primary breast cancers [/fig_ref] and B). Taken together, these data support the hypothesis that the STAT3regulated gene TNFRSF1A links STAT3 activity and NF-κB activity in both breast cancer model systems and primary breast cancers.
# Discussion
Oncogenic transcription factors like STAT3 lie at the convergence points of many upstream pathways that may be activated by mutation in cancer cells. Since these transcription factors can often be inhibited in normal cells with minimal consequences, STAT3 inhibition holds the potential for having a high therapeutic index. However, because STAT3 regulates many genes simultaneously, it is critical to identify a cohort of STAT3-regulated genes that can serve as predictive markers for response to STAT3 inhibitors.
Since STAT3 is activated constitutively in approximately 70% of breast cancers [bib_ref] Identification of a Genetic Signature of Activated Signal Transducer and Activator of..., Alvarez [/bib_ref] , and particularly among TNBC tumors, we chose to focus on genes regulated by STAT3 in this tumor type. TNBCs are particularly aggressive tumors, which are more common in young women and among African-American women [bib_ref] Triple-negative breast cancer in African-American women: disparities versus biology, Dietze [/bib_ref]. As TNBCs lack expression of estrogen receptor and progesterone receptor, they are not susceptible to endocrine-based therapeutic strategies, and chemotherapy remains the primary systemic treatment [bib_ref] Triplenegative breast cancer: challenges and opportunities of a heterogeneous disease, Bianchini [/bib_ref]. Thus, STAT3 inhibitors have significant potential for therapeutic benefit in this disease that is otherwise challenging to treat. To ensure examination of direct STAT3 targets, we identified genes with a proximal STAT3 binding site based on ChIP-seq and then focused on genes whose expression was inhibited by two Neoplasia Vol. The STAT3 Target Gene TNFRSF1A Modulates the NF-κB Pathway Egusquiaguirre et al.
different inhibitors of STAT3 transcriptional function in two different TNBC cell lines. From this integrated approach, we identified TNFRSF1A as a key STAT3-regulated gene. TNFRSF1A is one of the major transmembrane receptors for TNFα. When TNFα binds to TNFRSF1A, it induces activation of NF-κB, a collective name for a family of transcription factors consisting of five proteins: RelA/p65, RelB, c-Rel, p50 and p52 [bib_ref] TNF-α in cancer treatment: molecular insights, antitumor effects, and clinical utility, Van Horssen [/bib_ref] [bib_ref] The NF-κB family of transcription factors and its regulation, Oeckinghaus [/bib_ref]. A heterodimer formed by p65 and p50 subunits is the most abundant form and comprises the majority of NF-κB transcriptional activity [bib_ref] Mechanism of κB DNA binding by Rel/NF-κB dimers, Phelps [/bib_ref] [bib_ref] The Regulatory Logic of the NF-κB Signaling System, Hoffmann [/bib_ref]. NF-κB is a crucial regulator of the expression of genes involved in control of innate and adaptive immune responses, inflammation, and cancer progression [bib_ref] Regulation and function of NF-κB transcription factors in the immune system, Vallabhapurapu [/bib_ref]. Under basal conditions, NF-κB (p65-p50 heterodimer) is maintained in an inactive state in the cytoplasm, through interaction and binding to inhibitor of kappa B (IκB) proteins. TNFα binding to TNFRSF1A leads to a cascade of phosphorylation events culminating in the phosphorylation and subsequent proteasome-mediated degradation of IκBα [bib_ref] Shared Principles in NF-κB Signaling, Hayden [/bib_ref]. This uncovers a nuclear localization signal (NLS) of NF-κB, which directs the transcription factor to the nucleus, where it binds to promoter regions to regulate the expression of target genes involved in cell survival, proliferation, migration, invasion, and epithelial-mesenchymal transition (EMT) (http://www.bu.edu/nf-kb/gene-resources/target-genes/) [bib_ref] Shaping alternative NF-kappaB-dependent gene expression programs: New clues to specificity, Natoli [/bib_ref]. A negative feedback loop leads to re-synthesis of NF-κB-dependent IκBα proteins, leading to NF-κB export from the nucleus via CRMI-dependent nuclear export [bib_ref] Nuclear localization of I kappa B alpha promotes active transport of NF-kappa..., Arenzana-Seisdedos [/bib_ref] [bib_ref] Duration of nuclear NF-κB action regulated by reversible acetylation, Chen [/bib_ref].
Like STAT3, NF-κB is often aberrantly or constitutively active in many human malignancies, playing a role in the regulation of the apoptosis-proliferation balance in tumor cells [bib_ref] NF-κB in cancer: a marked target, Karin [/bib_ref] [bib_ref] NF-kB in development and progression of human cancer, Dolcet [/bib_ref]. Continuous activation of NF-κB is caused either by mutational activation of upstream signaling molecules or in response to extracellular stimuli within the tumor microenvironment [bib_ref] NF-[kappa]B in cancer: from innocent bystander to major culprit, Karin [/bib_ref].
Under physiologic conditions, both STAT3 and NF-κB are tightly regulated, and are transcriptionally active for only minutes to hours after a cytokine stimulus. However, both of these transcription factors can be active constitutively in cancer, thereby driving gene expression underlying the malignant phenotype. The STAT3 and NF-κB pathways interact on multiple levels. A key NF-κB target gene mediating the acute phase response is IL-6, also a potent STAT3 stimulus. In addition, there is evidence that STAT3 and NF-κB cooperate via physical interactions at certain promoters [bib_ref] Dangerous liaisons: STAT3 and NF-κB collaboration and crosstalk in cancer, Grivennikov [/bib_ref] [bib_ref] IKK/NF-κB and STAT3 pathways: central signalling hubs in inflammation-mediated tumour promotion and..., Bollrath [/bib_ref]. Finally, there are negative feedback loops preventing constitutive activation of both signaling pathways. For example, the microRNA miR-146b is a prominent STAT3-regulated gene that suppresses activation of NF-κB [bib_ref] STAT3 Induction of miR-146b Forms a Feedback Loop to Inhibit the NF-κB..., Xiang [/bib_ref]. However, the miR-146b locus is frequently silenced by methylation in breast cancer, allowing for simultaneous activation of both pathways. Thus, the functional interaction between these pathways is complex and dynamic.
In this study, we identified TNFRSF1A as a STAT3-dependent gene both in constitutive and cytokine-induced systems of STAT3 activation. This adds an additional point of interaction between the STAT3 and NF-κB pathways, one that promotes simultaneous activation of both. Although this interaction was discovered in breast cancer cell lines, there is strong evidence from gene expression analyses that it holds true in primary breast cancers as well [fig_ref] Figure 5: TNFRSF1A expression correlates with STAT3 activation in primary breast cancers [/fig_ref].
STAT3 represents a high value target in breast cancer and other tumors. A number of approaches have been taken to inhibit this transcription factor, including targeting the SH2, DNA binding, or Nterminal domains [bib_ref] Distinct roles of STAT3 and STAT5 in the pathogenesis and targeted therapy..., Walker [/bib_ref]. STAT3 inhibitors with novel mechanisms of action, like pyrimethamine, are now in clinical trials. In this era of precision medicine, it is becoming more important to identify patients most likely to benefit from a STAT3 inhibitor. One approach is to use immunohistochemistry (IHC) to identify tumors with tyrosinephosphorylated, and presumably nuclear, STAT3. However, phosphorylated epitopes may be labile. In addition, since the phosphorylated tyrosine of one STAT3 monomer is bound to the SH2 domain of its STAT3 binding partner, it is necessary to use antigen retrieval techniques to successfully detect this phosphorylation by IHC. Thus, elevated TNFRSF1A expression levels could serve as a more feasible biomarker for detecting functional STAT3 activation. Monitoring changes in TNFRSF1A expression in cancer patients receiving STAT3 inhibitor treatment could also serve as a useful pharmacodynamic marker to titrate the dose of inhibitor to optimal STAT3 inhibitory activity.
# Conclusions
In summary, TNFRSF1A is a direct STAT3 target gene that also regulates NF-κB activity in TNBC cell lines. Evidence from primary breast cancers suggests that a similar relationship is present in human tumors as well. In addition to being a connecting node between these two oncogenic pathways, TNFRSF1A may be an important biomarker for novel drugs that are being developed to target STAT3.
[fig] Figure 1: STAT3 regulates the expression of TNFRSF1A in breast cancer cells.(A) Two STAT3 inhibitors, PMPTP and pyrimethamine, both decrease expression of 1140 genes in SK-BR-3 cells stimulated with LIF, of which 196 genes, including TNFRSF1A, show direct STAT3 binding by ChIP-seq. (B) SK-BR-3 cells were unstimulated or stimulated with LIF to activate STAT3, and ChIP was performed for the STAT3 binding site in the TNFRSF1A gene. (C) MDA-MB-468 and BT549 cells, which display constitutive STAT3 activation, were transfected with siRNA targeting STAT3 or a non-targeting control. They were then analyzed by qRT-PCR for expression of TNFRSF1A mRNA (top; n = 3) and by immunoblot for TNFRSF1A protein (images, middle; quantitation, bottom. Representative of 3). (D) SK-BR-3 cells were transfected with siRNA targeting STAT3 or a nontargeting control. They were then stimulated with IL-6, after which RNA was harvested and analyzed by qRT-PCR for expression of TNFRSF1A (top; n = 3). Immunoblotting was used to assess the degree of knockdown of STAT3 (and phosphorylated STAT3; bottom). [/fig]
[fig] Figure 2: TNFRSF1A expression modulates NF-κB activity. (A) MDA-MB-468 and BT549 cells were transfected with the indicated siRNA targeting TNFRSF1A or a non-targeting control. Cells were left untreated or stimulated with TNFα, and were then analyzed by a luciferase reporter assay for NF-κB-dependent transcriptional activity (n = 3). MDA-MB-468 (B) and BT549 (C) cells were transfected with siRNA targeting TNFRSF1A or a nontargeting control. Cells were then stimulated with TNFα and analyzed by qRT-PCR for expression of endogenous NF-κB target genes ( [/fig]
[fig] Figure 3: TNFRSF1A depletion decreases nuclear NF-κB (p65) localization. MDA-MB-468 and BT549 cells were transfected with siRNA targeting TNFRSF1A or a non-targeting control, and then stimulated with TNFα. Nuclei were isolated, and p65 (RelA) was quantitated by immunoblot and normalized to PARP, which served as a loading control for nuclear protein. [/fig]
[fig] Figure 4: TNFRSF1A expression modulates NF-κB activity. MDA-MB-468 and BT549 cells were transfected with TNFR1 plasmid (encoding TNFRSF1A) or an empty vector. They were then stimulated with TNFα and analyzed by(A) immunoblot for the expression of TNFR1 (TNFRSF1A) in whole cell lysates (with tubulin serving as a loading control), (B) luciferase reporter assay for NF-κB-dependent transcriptional activity (n = 3), and (C) qRT-PCR for expression of endogenous NF-κB target genes (normalized to GAPDH; n = 3). [/fig]
[fig] Figure 5: TNFRSF1A expression correlates with STAT3 activation in primary breast cancers. (A, B) TNFRSF1A mRNA levels in breast cancers were compared based on STAT3 phosphorylation in the samples. Differences were analyzed by Student's unpaired t-test with Welch's correction. The data in (A), where positive staining (pSTAT3+) was defined as a z score N0.5 (n = 70), and negative staining (pSTAT3-) was defined as a z score b−0.5 (n = 128), are from The Cancer Genome Atlas (TCGA) Breast Cancer provisional dataset (n = 403 patients). The data in (B) are from the breast cancer microarray dataset GSE5460 (129 patient samples). (C, D) Gene set enrichment analysis (GSEA) was performed on two breast cancer microarray datasets; GSE5460 (129 patient samples) (C) and GSE6861 (161 patient samples) (D).In both, the 50 samples with highest and lowest TNFRSF1A mRNA levels were compared for expression of a STAT3 gene expression signature. Statistical significance was defined as FDR q value b0.25 and normalized P value b.05. [/fig]
[fig] Figure 6: TNFRSF1A expression correlates with NF-κB activation in primary breast cancers. (A, B) GSEA was performed on two breast cancer microarray datasets; GSE5460 (129 patient samples) (A) and GSE6861 (161 patient samples) (B). In both, the 50 samples with highest and lowest TNFRSF1A mRNA levels were compared for expression of an NF-κB gene expression signature. Statistical significance was defined as FDR q value b0.25 and normalized P value b.05. [/fig]
|
Chromosome looping in yeast
ong-range chromosome organization is known to influence nuclear function. Budding yeast centromeres cluster near the spindle pole body, whereas telomeres are grouped in five to eight perinuclear foci. Using live microscopy, we examine the relative positions of right and left telomeres of several yeast chromosomes. Integrated lac and tet operator arrays are visualized by their respective repressor fused to CFP and YFP in interphase yeast cells. The two ends of chromosomes 3 and 6 interact significantly but transiently, forming whole chromosome loops. L For chromosomes 5 and 14, end-to-end interaction is less frequent, yet telomeres are closer to each other than to the centromere, suggesting that yeast chromosomes fold in a Rabl-like conformation. Disruption of telomere anchoring by deletions of YKU70 or SIR4 significantly compromises contact between two linked telomeres. These mutations do not, however, eliminate coordinated movement of telomere (Tel) 6R and Tel6L, which we propose stems from the territorial organization of yeast chromosomes.Abbreviations used in this paper: 2D, two-dimensional; 3D, three-dimensional; ARS, autonomously replicating sequence; Chr, chromosome; IF, immunofluorescence; MSD, mean square displacement, NE, nuclear envelope; r c , radius of confinement or spatial constraint; SPB, spindle pole body; Tel, telomere.The online version of this article contains supplemental material.
# Introduction
Long-range chromosome organization is thought to influence nuclear function, yet little is known about how chromosomes fold in their natural states, and what forces or constraints produce recognizable patterns of chromosome positioning (for review see [bib_ref] Functional architecture in the cell nucleus, Dundr [/bib_ref] [bib_ref] The dynamics of chromosome organization and gene regulation, Spector [/bib_ref]. Two general types of interphase organization have been described. One is the polarized Rabl configuration, with centromeres and telomeres at opposite poles of the nucleus (for review see [bib_ref] The dynamics of chromosome organization and gene regulation, Spector [/bib_ref] , and the second a domain organization in which individual chromosomes occupy discrete territories that generally do not overlap [bib_ref] Non-random radial higherorder chromatin arrangements in nuclei of diploid human cells, Cremer [/bib_ref]. The Rabl configuration was initially observed in rapidly dividing embryonic and salivary gland nuclei of salamanders , and is prominent in both Drosophila melanogaster [bib_ref] Spatial organization of chromosomes in the salivary gland nuclei of Drosophila melanogaster, Hochstrasser [/bib_ref] and plant cells (wheat, rye, barley; for review see [bib_ref] The architecture of interphase chromosomes and nucleolar transcription sites in plants, Shaw [/bib_ref].
However, in Arabidopsis [bib_ref] Interphase chromosomes in Arabidopsis are organized as well defined chromocenters from which..., Fransz [/bib_ref] and in most mammalian cells [bib_ref] Inheritance of gene density-related higher order chromatin arrangements in normal and tumor..., Cremer [/bib_ref] [bib_ref] The dynamics of chromosome organization and gene regulation, Spector [/bib_ref] this polarized chromosomal organization, which results naturally from the pulling forces of the anaphase spindle, degrades after telophase. Thereafter, individual chromosomes tend to occupy distinct zones called territories, with variable nearest neighbors. Despite this irregularity in chromosome-chromosome contacts, one can detect tissue-specific distributions of chromosomes in mammalian cells [bib_ref] Tissue-specific spatial organization of genomes, Parada [/bib_ref]. Moreover, their relative radial position correlates with gene density on human chromosomes in both lymphocytes and transformed cells [bib_ref] Differences in the localization and morphology of chromosomes in the human nucleus, Croft [/bib_ref]. Nonetheless, the forces that determine chromosome position in interphase nuclei are unknown and no specific mutations have been reported that alter their spatial juxtaposition.
One means for the positioning of chromosomes may be the anchorage of specific chromosomal elements. It is well established in unicellular organisms like yeasts, Plasmodia and Trypanosoma , that telomeres are grouped in clusters at the nuclear envelope (NE; [bib_ref] Cell cycle-dependent specific positioning and clustering of centromeres and telomeres in fission..., Funabiki [/bib_ref] [bib_ref] Plasmodium telomeres: a pathogen's perspective, Scherf [/bib_ref]. This organization is well-characterized in Saccharomyces cerevisiae , where there are five to eight discrete perinuclear foci, each containing five to seven telomeres . Moreover, yeast centromeres cluster near the membrane-embedded spindle pole body (SPB; [bib_ref] Centromere position in budding yeast: evidence for anaphase A, Guacci [/bib_ref] [bib_ref] Yeast nuclei display prominent centromere clustering that is reduced in nondividing cells..., Jin [/bib_ref] [bib_ref] The positioning and dynamics of origins of replication in the budding yeast..., Heun [/bib_ref] [bib_ref] Longrange compaction and flexibility of interphase chromatin in budding yeast analysed by..., Bystricky [/bib_ref]. Theoretically, these clustering events are compatible with a Rabl-like arrangement for yeast interphase chromosomes [bib_ref] A polymer model for large-scale chromatin organization in lower eukaryotes, Ostashevsky [/bib_ref] , in which centromeres and telomeres would be found at opposite poles. However, no imaging study to date has specifically tagged both right and left telomeres of a given yeast chromosome to formally demonstrate a Rabl configuration. A recent cross-linking study suggests that subtelomeric regions of the budding yeast chromosome (Chr) 3 interact preferentially in living cells [bib_ref] Capturing chromosome conformation, Dekker [/bib_ref]. Because Chr 3 is a small yeast chromosome which bears unusual GC-rich isochores [bib_ref] GϩC content variation along and among Saccharomyces cerevisiae chromosomes, Bradnam [/bib_ref] and active and silent mating type loci, it was unclear whether other yeast chromosomes would yield similar cross-linking results.
Here, we directly examine the organization of chromosomes in vegetatively growing yeast cells, exploring the relationship of this organization to mechanisms that anchor telomeres at the NE. Exploiting two different bacterial repressor proteins with high affinity for integrated operator site arrays ( lac op or tet op arrays), we analyze chromatin structure in vivo at high resolution with live fluorescence microscopy [bib_ref] Visualizing chromosome dynamics with GFP, Belmont [/bib_ref]. Past studies have used such tools to examine the dynamics of individually tagged chromosomal loci, revealing rapid and constant, yet spatially constrained movements. Typical loci shift position frequently (0.1-0.5 m/s) within restricted subnuclear volumes [bib_ref] Interphase chromosomes undergo constrained diffusional motion in living cells, Marshall [/bib_ref] [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref] , characteristics that seem to be conserved from yeast to man. Yeast centromeres and telomeres, on the other hand, move within more tightly restricted zones and remain near the nuclear periphery [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref] [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. By using GFP derivatives fused to different bacterial repressors (i.e., lacI and tetR), we are able to use similar techniques to study the global folding of chromosomes in vivo, avoiding artifactual transinteractions between arrays of like repressor molecules [bib_ref] Functional dissection of in vivo interchromosome association in Saccharomyces cerevisiae, Aragon-Alcaide [/bib_ref]. We further examine the positions of differentially tagged telomeres relative to subnuclear landmarks such as the SPB, the nucleolus and NE.
The right and left telomeres of several budding yeast chromosomes interact frequently, but not stably. The interaction is most pronounced for two small chromosomes, Chr 3 and Chr 6, which have relatively short chromosomal arms of roughly equal length, yet the resulting Rabl-like organization is demonstrated for two other larger chromosomes. Telomeretelomere interactions are compromised in cells lacking the proteins involved in perinuclear anchoring, namely yKu and Sir4p. Nonetheless, even in strains lacking these tethers, the movement of telomeres on opposite chromosome arms is coordinated, which is not observed for telomeres of unlinked chromosomes. We suggest that centromere anchorage and telomere-telomere interactions, together with the general compaction of chromatin into a 30-nm fiber [bib_ref] Longrange compaction and flexibility of interphase chromatin in budding yeast analysed by..., Bystricky [/bib_ref] , determine chromosome position in the yeast interphase nucleus.
# Results
## Nuclear polarity is maintained throughout interphase
To examine chromosome positioning in yeast, we first identified fixed points of reference within the nucleus from which cell-autonomous measurements could be made. The yeast nucleolus, which represents a single large domain of rDNA transcription and processing, occupies a distinct subnuclear territory, adjacent to the NE [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] A, CFP-Nop1). Immunofluorescence (IF) on fixed cells maps the nucleolus to a zone opposite the SPB, an integral NE structure that serves as the microtubule-organizing center in yeast [bib_ref] Higher order structure is present in the yeast nucleus: autoantibody probes demonstrate..., Yang [/bib_ref] [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. Time-lapse imaging has shown that the SPB movement is constrained to very small volumes over 5-min intervals, moving far less than a telomere or centromere [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref]. However, it was not clear when the SPB becomes positioned opposite the nucleolus, nor how long this arrangement persists. To visualize this organization in living yeast cells, we have fused a component of the SPB (Spc42) and Nop1, an abundant RNA-binding nucleolar protein, to GFP alone or to CFP in combination with a GFP-Nup49 fusion that labels nuclear pores [bib_ref] Dynamics of nuclear pore distribution in nucleoporin mutant yeast cells, Belgareh [/bib_ref]. Cells were subjected to live time-lapse imaging at 12-s intervals for GFP alone (Video 1, available at http://www.jcb.org/cgi/content/ full/jcb.200409091/DC1), or at 3-min intervals for CFP and GFP in combination with capture of the transmission channel using a scanning confocal microscope (Video 2, available at http:// www.jcb.org/cgi/content/full/jcb.200409091/DC1). Visual inspection confirmed that normal cell growth was not impaired during or after imaging.
Consistent with IF results, the crescent-shaped nucleolus [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] A, CFP-Nop1, red) is found at one end of the nucleus, directly opposite the SPB [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] A, 0-and 45-min frames, CFP-SPB in white). Remarkably, the SPB focus is reproducibly positioned on a vector that can be drawn from the nucleolus toward the emerging bud, indicating that nuclear and cellular polarities are linked [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. As cells advance to late G2 phase and the nucleus elongates into the daughter cell, the duplicated SPBs separate, and one migrates back toward the nucleolus in the mother cell. These cells traverse mitosis rapidly [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] min; Video 2), at which point the two SPBs are found at opposite ends of the extended nucleus and the nucleolus spans the length of the spindle. In early telophase, the duplicated nucleolus splits in two, assuming symmetrical positions in mother and daughter nuclei and in G1 phase, the nucleus rotates slightly such that the nucleolus is again localized opposite the SPB [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. Thereafter, the nucleolus remains stably positioned opposite both the SPB and the future or actual site of bud emergence throughout G1 and S phase [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. Statistical support for this observation, comes from scoring SPB position in cells arrested in late G1 phase: in Ͼ 80% of the cases the SBP falls within 5 Њ of a perpendicular line extending from the nucleolus to the bud neck [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. We conclude that the nucleolus maintains a position opposite the SPB, which itself maintains a fixed position throughout interphase.
Further evidence that the nucleus does not rotate continuously in G1 phase, is based on GFP-Nup49 FRAP experiments [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. We irreversibly photobleached the nuclear pore fluorescence within the NE and monitored fluorescence recovery at time intervals relevant to those used to monitor chromatin dynamics of interphase chromatin (i.e., 1.5-10-s intervals over several minutes). If the nucleus were turning rapidly, we would expect to see the bleached zone move from the plane of focus. This does not occur [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. Instead we observe a slow diffusion of pore fluorescence inwards from the edges of the bleached zone, beginning at ف 80 s [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. We conclude that the global orientation of the interphase nucleus in yeast is quite stable, not only with respect to the SPB, but also with respect to cytoplasmic structures. Nuclear landmarks such as these can thus be used to monitor relative position of chromosomal tags, and rotation of the nucleus can be ruled out as a source of chromatin mobility.
## Juxtaposition of right and left telomeres at the nuclear periphery
Previous studies have shown that yeast telomeres are enriched near the nuclear periphery in G1-and S-phase cells, both when detected individually or through repeat sequences [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. Nonetheless yeast telomeres are dynamic, shifting irregularly along the NE and occasionally into the nucleoplasm. To explore spatial relationship of pairs of telomeres in vivo we have differentially tagged the two ends of chromosomes 3, 5, 6, and 14, within the most distal unique sequences, such that subtelomeric repeats remain unaltered [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref]. We measured distances separating the lac op and tet op insertions, visualized by the binding of CFP-or YFP-fusions to the bacterial repressors, on three-dimensional (3D) confocal stacks of intact cells [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref]. The distributions of 3D measurements ( n ϭ 60-160 for each telomere pair) are plotted in [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] (C and D), and the mean distances between tagged sites are summarized in [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. At a given moment, the left and right telomeres of Chr 3 and 6 coincide or are immediately adjacent to each other (separation in 3D ϭ 0.2 Ϯ 0.2 m) in 35-40% of the cells measured. Telomere separation for these two chromosomes is clearly skewed to small distances: Ͼ 75% of the intra-telomere 3D measurements are under 0.8 m . This is in contrast to the separation of two peripheral but unlinked telomeres (5L and 14R; or 6L and 14L), which follows a near Gaussian distribution around 1 m [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref]. Indeed, if two telomeres on the same chromosome were to have no bias toward interaction, the distribution of distances should be Gaussian over a range from 0.1 to 2 m, depending on the compaction ratio of the chromatin and the length of chromosomal arms. Separation distances for right and left telomeres of Chr 5 and Chr 14 are also biased toward values Ͻ 0.8 m, but unlike Chr 3 and Chr 6, telomeres are immediately adjacent or superimposed in only ف 12% of cells.
## Chromosomes fold back on themselves in interphase
We next analyzed the relationship of telomere pairs to the centromere by combining the double-tagged chromosomes with staining for the SPB [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref] , A-C). Elsewhere we have established that all centromeres cluster within 200-300 nm of the SPB [bib_ref] Longrange compaction and flexibility of interphase chromatin in budding yeast analysed by..., Bystricky [/bib_ref]. By measuring the 3D distance between two telomeric spots and the distance between each telomere and the SPB [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref] , A-C; [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] , available at http:// www.jcb.org/cgi/content/full/jcb.200409091/DC1), we deter- mine the long-range organization of the chromosome and calculate an angle ␣ that subtends telomere separation. Again, right and left telomeres of Chr's 3 and 6 are frequently juxtaposed ( Ͼ 30% at Ͻ 0.2 m, 60-70% at Ͻ 0.6 m), whereas the telomere (Tel) 5R and Tel5L separation exhibits greater variability [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref]. Importantly, we note that right and left telomeres are almost always more closely juxtaposed to each other than either is to the SPB [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref]. This argues for a fold-back structure that is dominant for Chr's 3 and 6, and statistically significant for Chr 5 (see below).
By triangulation we determined the angle ␣ at between right and left chromosome arms, using the SPB signal as the apex. The distribution of these angles is summarized in [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref] E. Mean angle values for each chromosome are 31 Њ Ϯ 32 Њ for Chr 3, 38 Њ Ϯ 26 Њ for Chr 6, and 44 Њ Ϯ 29 Њ for Chr 5. This large variability is inherent to the dynamic nature of telomeres and does not represent different subpopulations (see below). It is noteworthy, however, that among the three chromosomes studied, very few angles are Ͼ 90 Њ and none are Ͼ 110 Њ , and ف 50% of Chr 3 and Chr 6 arms meet at angles Ͻ 30 Њ . If telomeres were on 1 m long arms randomly distributed on the surface of a sphere around a fixed point (the SPB), the subtending angles would have Gaussian distribution around 60 Њ . We can conclude, therefore, that the fold-back organization of Chr's 3 and 6 is statistically significant, reflecting right and left telomere interaction. Chr 5 appears also nonrandomly folded ( Ͼ 70% of the angles are Ͻ 60 Њ ), although Tel5R-5L interactions are less frequent. Finally, the average distance separating Tel14L and 14R [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref] is less than the one separating Tel 5L and 5R, arguing that Chr 14 also assumes a Rabl-like organization.
## Right and left telomere interactions are favored by perinuclear constraints
Two parameters may influence telomere-telomere interaction: the length of chromosome arms and their association with the NE. Indeed, the arms of Chr 3 and 6 are both short and of nearly equal lengths (3R/3L ϭ 115 kb/200 kb and 6R/6L ϭ 122 kb/148 kb), which is not true for either Chr 5 or 14. However, short, equal arm length is not alone sufficient to favor interaction of chromosome ends: the chromosomal arms of Tel 5L and Tel 14R are also short and of equal length (152 and 150 kb, respectively), yet these ends are separated on average by ف 1 m [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref].
Thus, chromosome arm length probably only favors telomeretelomere interaction when the arms are physically linked. We next examined whether the efficiency with which each telomere is found at the GFP-Nup49-tagged NE, correlates with the efficiency of their interaction in trans. We scored telomere position relative to three equal zones of the nucleoplasm, focusing on the peripheral-most zone, which has a width of only 0.184 times the radius [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] , available at http:/ /www.jcb.org/cgi/content/full/jcb.200409091/DC1). For all except Tel5R, we monitor a significant enrichment in this zone, with the following hierarchy: [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. Only Tel5R has a near-random distribution in G1-phase cells. Similarly, nontelomeric loci, such as MAT a , which sits in the middle of Chr 3, or origins of replication located 73 or 437 kb from the nearest telomere (autonomously replicating sequence [ARS] 607 or ARS1, respec- Individual telomere position reflects their distribution among three zones of equal surface (see [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] ; values Ͼ33% in zone 1 represent enrichment in the peripheralmost zone. n is the number of G1-phase cells analyzed. t test was performed to compare distributions with a random distribution (P values for 95% confidence level are shown). Individual telomere dynamics were analyzed by 2D confocal live microscopy as described in Figs. 4 and 5. The average velocity of each telomere is obtained by dividing the total path length by the total time period. The diffusion constant (D ϭ MSD/⌬t) is proportional to the initial slope of the abs MSD plot (1.5-s interval) and the radius of constraint is determined from its plateau and the formula maximal MSD ϭ 4/5 (r c ) 2 . a From [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. b From [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref]. c r c for Tel 14L is determined at ⌬t ϭ 60 s, because the tendency to move horizontally along the NE distorts the r c value for this telomere.
[formula] Tel14R Ͼ 5L Ϸ 6R Ͼ 14L Ϸ 3R Ϸ 6L Ͼ 3L [/formula]
tively), are either randomly distributed or depleted from the periphery [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. Although the well-paired telomeres (those of Chr 3 and 6) tend to be perinuclear, from these measurements one can draw no simple correlation between the efficiency of NE interaction and telomere interaction.
## Two color time-lapse imaging reveals constraints on telomere movement
Every measurement on a fixed cell is, of course, a snapshot of a dynamic chromosomal state, and even telomere-telomere interactions are not static. To monitor directly how stable telomere interactions are, we used live time-lapse imaging to follow the relative movement of differentially tagged telomeres. Up to 250 sequential two-channel (CFP-YFP) confocal images were acquired at 1.5-s intervals without detectable impact on cell-cycle progression. For each strain, we analyze 8-12 independent two-dimensional (2D) time-lapse series (totaling 35-58 min each) of G1-phase nuclei, after the tagged foci by adjusting the focal plane. Representative sequences and videos are shown in [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref] and Videos 3-6, available at http://www. jcb.org/cgi/content/full/jcb.200409091/DC1. Projection of the paths taken by the individual telomeres onto one plane shows that movements are not only restricted to a fraction of the total nuclear volume, but that the tracks of the Tel 3R-3L and Tel 6R-6L coincide extensively (see examples from typical videos; [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref]. Tel 5R-5L move in close proximity but with little overlap. The juxtaposition does not arise from the methodology used, because movements of unlinked telomeres (Tel 6L-14L) are distinct and uncoordinated, consistent with measurements at fixed time points [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. By summing all individual steps over the total time and dividing by the period elapsed, we calculate the average velocity of each individual telomere [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. We find that all telomeres except Tel 5L and 5R are significantly less mobile than the tagged centromere-proximal ARS1 locus.
Assuming that chromatin motion resembles a constrained random walk [bib_ref] Interphase chromosomes undergo constrained diffusional motion in living cells, Marshall [/bib_ref] , locus mobility can also be characterized by plotting its mean square displacement (MSD or Ͻ⌬d 2 Ͼ) over increasing time intervals. Unconstrained diffusion gives a linear relationship between increasing time intervals and the square of the distance travelled by a particle during that time, where ⌬d 2 ϭ (d(t)Ϫd(tϩ⌬t)) 2 [bib_ref] Methods for visualizing chromatin dynamics in living yeast, Hediger [/bib_ref]. The MSD curve for chromatin with a spatially constrained diffusion process generally reaches a plateau by ⌬t Ͼ 50s. This analysis is highly robust because ⌬t intervals are pooled from all videos of a given strain.
If we monitor movement as displacement relative to the nuclear center or the nearest point on the NE (d ϭ distance be- tween one fluorescent telomere spot and the center of the nuclear background fluorescence, cf. [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref] , the resulting MSD curve reflects the dynamics of a given locus relative to the nuclear periphery (radial MSD or radMSD; [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref]. RadMSD curves show that the dynamics of telomeres 5L, 6R and 6L are nearly equally restricted relative to the NE, whereas Tel 5R moves without constraint relative to the NE [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref]. The two telomeres of Chr 3 exhibit NE-constrained movement very similar to Chr 6 (unpublished data). By comparing telomere movements and paths, we conclude that path superposition of right and left telomeres correlates positively with constraint relative to the NE, even though precise distance from the NE may vary. Thus, constrained movement relative to the periphery, whether directly at the NE or not, does correlate with contact between telomeres.
## Absolute and relative constraints on telomere dynamics
A more accurate analysis of spatial constraint is based on measurements that reflect the actual distances covered from any one time point to all others (i.e., rather than distances relative to the periphery; , after an alignment of nuclear centers to eliminate background drift. These d values were then subjected to the similar MSD analysis (here called absolute or absMSD) for both telomeres of Chr 3, 5, and 6. When absolute step sizes are the basis of the curve, the radius of confinement or spatial constraint (r c ) determines the plateau of the MSD curve (Ma ϫ MSD). For our geometry, this dependence is Ma ϫ MSD ϭ 4/5 (r c ) 2 (J. Dorn and Neumann, F., personal communication). Solving for r allows us to calculate the radius of confinement from experimental MSD curves. This analysis shows that Tel 5R and Tel 5L are relatively mobile and do not reach a plateau, yet from the radial analysis we know that Tel 5L tracks along the NE [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref]. By contrast, movements of Tel 6R, 6L, 3R, and 3L, show clear spatial constraint and r c values ranging from 0.40 to 0.46 m.
The initial slope of the absMSD plot is proportional to the maximal diffusion constant (D ϭ MSD/⌬t). These slope values confirm that Tel 5R and 5L are more dynamic than other telomeres, with diffusion rates similar to those of the centromere proximal ARS1 locus (6.9 ϫ 10 Ϫ11 cm 2 /s; [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. Although Tel 5R and 5L are more mobile than other telomeres, we show here that they move in a paired manner, by scoring the relative separation of the telomere pairs throughout Ͼ2,000 frames [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref] , available at http://www.jcb.org/cgi/content/full/jcb. 200409091/DC1). Distances separating telomeres derived from time-lapse series confirm the values determined in 2D and 3D at fixed time points [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] : Tel 6R-6L and Tel 3R-3L are tightly juxtaposed, with 32-37% of all distances Յ0.2 m, and Ͼ50% Յ0.4 m. Strikingly, Ͼ60% of the separation values for Tel 5R-5L are Յ0.5 m, whereas the separation of two unrelated telomeres (i.e., Tel 6L-14L) is Ͻ0.2 m in only 5% of all frames. This confirms that Tel 5R-5L are adjacent although rarely interacting.
To quantify the freedom of movement that two telomeres have relative to each other, we plot the change of distances separating the telomeres as a function of t. In this "relative MSD" analysis, d is defined as the distance between two telomeres at any given time point , relative MSD; [bib_ref] Interphase chromosomes undergo constrained diffusional motion in living cells, Marshall [/bib_ref] [bib_ref] Multiple regimes of constrained chromosome motion are regulated in the interphase Drosophila..., Vazquez [/bib_ref]. These MSD plateaus confirm that all telomere pairs tested undergo obstructed diffusion, yet the values for linked telomere pairs are grouped around Ͻ⌬d 2 Ͼ ϭ 0.1-0.14 m 2 . This suggests that two different telomeres move more freely relative to one another than do two identical centromere proximal sites monitored in a diploid cell (for LEU2/Cen3, Ͻ⌬d 2 Ͼ ϭ 0.06 m 2 ; [bib_ref] Interphase chromosomes undergo constrained diffusional motion in living cells, Marshall [/bib_ref]. It is nonetheless noteworthy that even two unlinked telomeres (Tel6L-14L), which are separated by roughly 1 m in the nucleus, show a relative radius of constraint of r c ϭ 0.25 m. From this one can conclude that, independent of their pairing efficiency, telomeres assume fairly fixed positions in interphase nuclei.
## Nuclear order is disrupted in the absence of yku70 or sir4
We have recently established that yeast telomeres are bound at the NE through dual pathways. One requires Sir4 and the other yKu [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref] [bib_ref] Separation of silencing from perinuclear anchoring functions in yeast Ku80, Sir4 and..., Taddei [/bib_ref]. To examine directly whether the observed fold-back organization of chromosomes depends on telomere anchoring, we analyzed the posi- . Looping of short chromosomes correlates with reduced telomere mobility. (A) Absolute MSD calculated using the 2D videos as described in [fig_ref] Figure 4: Live imaging of telomere dynamics [/fig_ref] for telomeres 5R, 5L, 6R, and 6L using d ϭ actual distance from any one time point to all others (see diagram; for t ϭ 1.5-101.5 s) after nuclear alignment. (B) Relative MSD calculated using d ϭ distance between two telomeres at all possible time intervals (see diagram; for t ϭ 1.5-61.5 s), for the indicated pairs of telomeres.
tion and dynamics of Tel 6L and 6R after disruption of either YKU70 or SIR4. In the absence of the yKu complex, Tel 6R is delocalized from the periphery [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref] becoming randomly distributed in the nucleus, whereas Tel 6L anchoring is only slightly diminished . In contrast, sir4 deletion releases Tel 6L, but not Tel 6R . Confirming the redundancy of the anchoring pathways, we note that all telomeres analyzed to date lose their perinuclear position in double sir4 ku70 mutants [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. The mobility of Tel 6R and 6L also increases in these mutants, as monitored by live time-lapse imaging and absMSD analysis . Plateau heights correspond to increases in average r c from 0.38 or 0.43 m in wild-type cells, to 0.5 m in the sir4 mutant and Ͼ0.6 m in yku70 cells.
We next asked whether the relative distance between the two telomeres changes significantly in these mutants. The separation between telomere pairs was monitored for mutant and wild-type cells as a function of time . In both yku70 and sir4 mutants, Tel 6L and 6R show significantly greater separation than in wild-type cells (t test P Ͻ 0.003, yku70, and P Ͻ 0.005 for sir4). In the mutants Ͻ23% of the distances measured are Յ0.2 m, as compared with Ͼ30% in wild-type cells. Because the two arms of Chr 6 are short, a 25% increase of the mean distance between the two telomeres corresponds to a large change in the angle between the two chromatids [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref]. The average angle ␣ increases from 39Њ to 48Њ, which is larger than that observed for Chr 5 in a wild-type strain (44Њ). Because centromere clustering near the SPB is unaffected by either the yku70 or sir4 deletion (unpublished data), we conclude that the fold-back organization of Chr 6, monitored as telomere-telomere proximity, is severely disturbed when either telomere loses its perinuclear anchoring.
In summary, the loss of yKu or Sir4p should make Chr 6 behave like Chr 5 (i.e., one telomere moves freely and the other is anchored; [fig_ref] Figure 8: Schematic representation of a Rabl-like chromosome organization in yeast [/fig_ref]. Therefore, we plotted the relative MSD between Tel 6L and 6R in the mutant strains , to score their loss of coordination. Indeed, the relative MSD plateau for Tel 6R-6L in the yku mutant is higher, similar to that scored for Tel 5R-5L in a wild-type background and consistent with increased mobility of one end . Nonetheless, the plateau is still quite low, as it is in the sir4 background, suggesting that the ends of a given chromosome preserve a territorial inertia even though they interact less frequently.
## Coordinated chromosome dynamics can occur independent of telomere interactions
Do linked telomeres move in a coordinated manner, or simply show constraint relative to each other? To address this we acquired time-lapse videos in 3D (7-image stack of a 300-nm step size) capturing double-tagged telomeres at two wavelengths on the confocal microscope [fig_ref] Figure 7: 3D and two-color fluorescence time-lapse imaging of telomere dynamics [/fig_ref]. Cellular integrity is confirmed by following the imaged cell through the subsequent mitosis. Coordinates of the center of the fluorescent spots were obtained using the IMARIS software, and the nuclear center is interpolated from the YFP-tetR background signal. The nucleus and spot positions for Tel 6L-6R and for Tel 6L-14L were then reconstructed in 3D [fig_ref] Figure 7: 3D and two-color fluorescence time-lapse imaging of telomere dynamics [/fig_ref] , A-C, shown here as projections onto the x, y, and z planes over time). Tel 6L-6R appear frequently, but not always, closely juxtaposed. Even when not juxtaposed, they seem to move in a coordinated fashion, which is not true for 6L and 14L.
The degree to which movement is coordinated can be assessed by a correlation coefficient c (see Materials and methods; no correlation ϭ 0, identical movement ϭ 1). Direction cosines were determined for every vector joining two neighboring points of two separate trajectories, and the mean of Pearson's correlation coefficients (c) in each direction was determined. This was performed both for 2 color 2D and 3D timelapse series. The movements of Tel 6R-6L have a mean corre- . Nuclear order is disrupted in the absence of yKu70p or Sir4p. Mobility, telomere-telomere separation, and telomere anchoring of Chr 6 are compared in wildtype, yku70, and sir4 cells. (A) Positions relative to the NE in wt (gray), yku70 (blue), and sir4 (green) strains of GFP tagged telomeres 6L and 6R mapped to zone 1 (as described in [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] and [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref]. The number of G1phase cells analyzed and the 95% confidence values (P) for the t test between random and test distributions for 6L are: 122, P ϭ 2.5 ϫ 10 Ϫ3 for 6L wt; 81, P ϭ 0.6 for 6L yku70; 57, P ϭ 4.2 ϫ 10 Ϫ5 for 6L sir4; for 6R data see [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. (B) Absolute MSD was calculated using the 2D videos as described in Figs. 4 and 6 (for t ϭ 1.5-61.5 s). (C) Frequencies of distances from 2D timelapse series between the two tagged loci are displayed as a function of 0.2-m intervals (Ϯ0.1). (D) Relative MSD calculated using d ϭ distance between telomeres 6R and 6L at all possible time intervals (for t ϭ 1.5-61.5 s).
lation coefficient of 0.39 in 3D (0.26 in 2D), indicative of closely coordinated movement. Confirming our methodology, we found that two tags on the same telomere gave correlation coefficient of 1 (unpublished data). In contrast, Tel 6L-14L movements show no significant coordination (correlation coefficients of 0.03, for a 3D time-lapse series). Thus, the 6R-6L telomeres move with significant coordination over time, whereas unlinked ends do not.
Similar analysis was performed in strains bearing disruptions of YKU70 or SIR4, which compromises both anchoring and telomere-telomere interactions . Strikingly, however, in yku70 and sir4 mutants the Tel 6R-6L correlation coefficients are ,51.0ف which is still half the coordination detected in wild-type cells. In the case of the yku70 mutant, the 3D timelapse analysis of Tel 6R and 6L trajectories projected onto x, y, and z planes, suggests a low but detectable degree of coordination in the mutants [fig_ref] Figure 7: 3D and two-color fluorescence time-lapse imaging of telomere dynamics [/fig_ref]. We predict that this residual coordination in chromosome dynamics can be attributed to their physical contiguity, i.e., that they represent two ends of a single chromosome. The release of one telomere from the NE and the ensuing drop in telomere interaction nonetheless does lead to a significant increase in unlinked movement.
# Discussion
Using high resolution microscopy techniques on living budding yeast cells we establish that the anaphase polarity of chromosomal organization is maintained in nuclei despite the continuous dynamic movement of interphase chromatin. Furthermore, our analysis of position and movement of multiple pairs of budding yeast telomeres in wild-type and mutant strains, shows that right and left telomeres of Chr 3 and Chr 6 interact in a reversible, but highly significant manner. This is the first study in which contact between specific yeast telomeres has been documented by either fixed or live microscopy. These interactions, coupled with the stable polarized clustering of centromeres near the SPB, provides a direct demonstration that yeast chromosomes can assume a looped, Rabl-like organization. Even Chr 5 and 14, whose telomeres interact less fre-quently, appear to fold-back upon themselves, arguing that a combination of telomere anchoring and trans-interactions contribute to spatial organization [fig_ref] Figure 8: Schematic representation of a Rabl-like chromosome organization in yeast [/fig_ref].
The Rabl-like arrangement that we document in budding yeast persists throughout interphase, until the mitotic spindle actively alters chromosome position. A second documented instance of chromosome clustering involves all telomeres during the "bouquet" stage before pachytene in meiotic prophase [bib_ref] A bouquet makes ends meet, Scherthan [/bib_ref]. In budding yeast this clustering is mediated by a sporulation-specific protein scNdj1 [bib_ref] Meiotic telomere protein Ndj1p is required for meiosis-specific telomere distribution, bouquet formation..., Trelles-Sticken [/bib_ref] , which has no known function in mitotically dividing cells. Similarly, the fission yeast protein spTaz1 mediates meiotic but not mitotic, telomere clustering, a phenomenon that involves telomere anchoring at the SPB [bib_ref] Fission yeast Taz1 protein is required for meiotic telomere clustering and recombination, Cooper [/bib_ref]. In wild-type cells, short chromosomes with equal length arms, such as Chr 6 and Chr 3, form loops through telomere interactions. Other chromosomes fold back less rigidly. This is more pronounced when both telomeres are anchored in the NE, which is not the case for Chr 5. (B) Nuclear order is disrupted by deletion of yku70 as the telomeres of Chr 6 detach from the NE and become more mobile.
## Multiple elements constrain chromatin mobility to help define chromosome position
We provide novel evidence that long-range interactions between telomeres can be altered in vegetatively growing yeast by interfering with the telomere-associated proteins yKu and Sir4p. These same factors are directly involved in the anchorage of yeast telomeres to the NE. Indeed, silencing-incompetent forms of each protein are sufficient to relocate an otherwise internal locus to the nuclear periphery [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref] [bib_ref] Separation of silencing from perinuclear anchoring functions in yeast Ku80, Sir4 and..., Taddei [/bib_ref]. Importantly, the disruption of anchorage at just one end of Chr 6 significantly reduces telomere-telomere interaction [fig_ref] Figure 7: 3D and two-color fluorescence time-lapse imaging of telomere dynamics [/fig_ref]. Correlation analysis of movement in 3D argues that despite their separation, Tel 6R and 6L continue to move in a partially coordinated manner in these yku70 or sir4 mutant cells. Because unlinked telomeres do not behave in a similar fashion, we conclude that not only direct interaction, but the contiguity of the chromosomal fiber influences chromatin movement, even though telomeres are separated by several hundred kilobases and a kinetochore.
It has been questioned whether the notion of chromosome "territories" is appropriate for yeast due to the relatively large r c monitored for individual loci (r c ϭ 0.5-0.65 m) and the small size of the yeast nucleus (nuclear radius ϭ 1 m). The movement we document here indicates that two linked telomeres move in a partially coordinated manner, thus providing a quantifiable parameter for a "chromosomal territory". In contrast to this, a 16-kb ring of chromatin released from its chromosomal context by an inducible recombinase, traverses the nucleoplasm freely and randomly, moving in all directions (r c Ն 0.8 m for the ring vs. 0.6 m for the chromosomal locus; [bib_ref] Sir-mediated repression can occur independently of chromosomal and subnuclear contexts, Gartenberg [/bib_ref]. The unconstrained movement of this ring further stresses the impact of chromatid contiguity both on the relative positioning of linked telomeres and on general chromosome positioning in interphase nuclei. In conclusion, we propose that chromosome position is defined by three types of constraint: the contiguity and compaction of the chromosomal fiber, sites of anchorage to less mobile nuclear landmarks (centromeres to the SPB and telomeres to the NE) and finally, reversible interactions between right and left chromosome ends.
Our data strongly support the looped Chr 3 model proposed from an assay that scores the efficiency of cross-linking in vivo [bib_ref] Capturing chromosome conformation, Dekker [/bib_ref]. Chr 3 is unique among yeast chromosomes in that it carries three homologous mating type loci that participate in a gene conversion event required for mating type switching. Chr 3 also has unique, strongly pronounced GC-rich "isochores" of 30-50 kb [bib_ref] GϩC content variation along and among Saccharomyces cerevisiae chromosomes, Bradnam [/bib_ref] , which are not found on the other chromosomes analyzed here. It is conceivable that the folded structure of Chr 3 reflects its propensity for recombination between MAT (on the right arm) and HML (on the left arm) in MATa cells. However, because Chr 6 forms a whole chromosome loop as efficiently as Chr 3, these Chr 3-specific features are unlikely to be critical for its folding pattern in vivo. We note that the interactions of telomeres on Chr 3 and 6 may well be aided by the fact that these two small chromosomes have similar arm lengths and compaction ratios [bib_ref] Longrange compaction and flexibility of interphase chromatin in budding yeast analysed by..., Bystricky [/bib_ref]. Conversely, one might assume that grossly different chromosome arm lengths limit pairing. Finally, we note that chromatid arm length is not a sufficient criterion to determine stable pairing events, because the telomeres of 5L and 14R do not interact despite the equal length of these chromosome arms.
Extended sequence homology is not critical for telomere pairing Does sequence homology contribute to selective telomeretelomere interactions? It was suggested that transient contact between homologues, or chromosome "kissing" events, would facilitate homology searches in meiotic prophase [bib_ref] Potential advantages of unstable interactions for pairing of chromosomes in meiotic, somatic,..., Kleckner [/bib_ref] [bib_ref] Limitations of silencing at native yeast telomeres, Pryde [/bib_ref]. One might imagine that once in contact, sequence homology could in turn promote more stable interactions in trans through ligand binding. Such trans-interactions have been proposed to facilitate silencing, but also are thought to help coordinate the timing of replication of right and left telomeres in budding yeast [bib_ref] Replication dynamics of the yeast genome, Raghuraman [/bib_ref]. Our study, however, demonstrates that the selective interactions of the Tel 3R and 3L and Tel 6R and 6L does not result simply from sequence homology. Neither pair of telomeres shares any homology other than the universal TG-rich and STR/core X element repeats. Moreover, a pair of telomeres that shares Ͼ90% homology over 16 kb (Tel 6L and 14L), almost never interact despite the presence of highly conserved Y' elements. Consistently, entire chromosomal homology also has little impact on pairing: in a yeast strain that serendipitously contains a duplication of the double-tagged Chr 3 (bearing Tel 3R-tet op and 3L-lac op sequences) the two homologous chromosomes are far apart within the nucleus and each forms a separate fold-back structure (unpublished data).
## Heterochromatin factors anchor telomeres and contribute to transinteraction
In budding yeast, a strong candidate for contributing to telomere-telomere interactions could be silent chromatin itself. Silencing efficiency, like the availability of Sir proteins and telomere-telomere pairing, varies from end to end [bib_ref] Limitations of silencing at native yeast telomeres, Pryde [/bib_ref]. Sir4p, a 174-kD protein bears a COOH-terminal coiled-coil domain that is necessary for homo-dimerization as well as interaction with Sir3, Rap1, and yKu (for review see [bib_ref] The molecular biology of the SIR proteins, Gasser [/bib_ref]. The Sir4 COOH-terminal domain has often been compared with nuclear lamins in higher eukaryotes, although Sir4 requires another perinuclear protein, Esc1p, or yKu to ensure its anchoring . To test rigorously whether or not subtelomeric heterochromatin directly influences pairing, it will be necessary to compare the efficiency of native telomere repression and the efficiency of native telomere interaction systematically in a single strain background. Whereas yku70 mutations compromise telomeric silencing, they do not impair mating type silencing, suggesting that reduced interactions do not significantly disrupt the repressive state.
In analogy to Sir4p, HP1 has been proposed to mediate interactions between repressed domains in higher eukaryotes in trans [bib_ref] KAP-1 corepressor protein interacts and colocalizes with heterochromatic and euchromatic HP1 proteins:..., Ryan [/bib_ref]. However, delocalization of HP1 can occur in mammalian cells without the disruption of the chromocenter [bib_ref] Loss of the Suv39h histone methyltransferases impairs mammalian heterochromatin and genome stability, Peters [/bib_ref]. The elimination of Sir4 and yKu oblate telomere associated silencing, much like spTaz1, which is required in fission yeast both for telomere-associated silencing [bib_ref] spRap1 and spRif1, recruited to telomeres by Taz1, are essential for telomere..., Kanoh [/bib_ref] and meiotic clustering. Surprisingly, however, loss of mitotic clustering is governed by the fission yeast RNAi machinery, not Taz1, and mutation of this ironically does not derepress subtelomeric silencing or perinuclear anchorage [bib_ref] RNA interference machinery regulates chromosome dynamics during mitosis and meiosis in fission..., Hall [/bib_ref]. Therefore, although subsets of heterochromatin components may contribute to long-range chromosomal contacts, the loss of interactions in trans, is not necessarily correlated with changes in repression status. Rather, critical components for telomere-telomere interactions, among which may figure cohesin molecules, may simply associate with heterochromatin, participating to different degrees in both repression and trans-interactions [bib_ref] cis-acting DNA from fission yeast centromeres mediates histone H3 methylation and recruitment..., Partridge [/bib_ref].
## Limited chromosomal mobility can define a territory
Besides specific patterns of chromosome folding and centromere/telomere positioning, we provide evidence that the coordinated movement of the two distal regions of a yeast chromosome is compromised by mutation. We envision this coordinated movement as a sort of "chromosomal inertia," which reflects the tendency of a chromosome to move as one body, even when specific interactions are compromised. Given the mass of a mammalian chromosome, if similar coordination occurs, then this alone could account for the infrequency with which human chromosomes change their territorial distribution (for review see [bib_ref] The dynamics of chromosome organization and gene regulation, Spector [/bib_ref]. Reproducible positioning and limited mobility of chromosomal domains has been documented as well for Drosophila cells [bib_ref] Specific interactions of chromatin with the nuclear envelope: positional determination within the..., Marshall [/bib_ref] [bib_ref] Interphase chromosomes undergo constrained diffusional motion in living cells, Marshall [/bib_ref] [bib_ref] Multiple regimes of constrained chromosome motion are regulated in the interphase Drosophila..., Vazquez [/bib_ref]. We propose that the general inertia of whole chromosome territories in higher eukaryotic cells, may be linked to a phenomenon we quantify here-that of chromosome-wide coordination of constrained movement.
# Materials and methods
Plasmid, strains, and yeast methods Plasmid used to integrate the tet or lac operators and repressors were as described in [bib_ref] Longrange compaction and flexibility of interphase chromatin in budding yeast analysed by..., Bystricky [/bib_ref]. PCR-amplified genomic fragments with the indicated SGD coordinates were used for insertion: 15160-15773 (Tel3L), 294892-295241 (Tel3R), 9645-11059 (Tel5R), 558701-559863 (Tel5L), 16431-17993 (Tel 6L), 256581-256893 (Tel6R), . Unique restriction sites were used to linearize the plasmids for integration, which was verified by colony PCR and pulsed field electrophoresis. LacI-GFP and tetR-GFP or tetR-YFP and the lacI-CFP fusions were introduced by integration of pGVH40 or pGVH30 at the ade2-1 locus. Where indicated GFP-Nup49 fusions were integrated as described previously [bib_ref] The positioning and dynamics of origins of replication in the budding yeast..., Heun [/bib_ref]. Complete yku70 and sir4 deletions were obtained using a PCR-based gene deletion technique [bib_ref] Additional modules for versatile and economical PCR-based gene deletion and modification in..., Longtine [/bib_ref] [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. pBM197 contains a Spc42-CFP fusion for integration at the URA3 gene (a gift from M. Peter, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland). Strains used are listed in [fig_ref] Table I: Average 3D telomere-telomere distances [/fig_ref].
## Microscopy
For live imaging, cultures grown in YPD to 0.2-0.4 ϫ 10 7 cells/ml were imaged on SC agar ϩ 4% glucose patches or in a Ludin chamber at 30ЊC. Initially, attempts to visualize Tel 3L and 3R using CFP-lacI/YFP-tetR failed, thus Tel 3L and 3R were visualized with GFP fusions differentiated by spot size (GA-2195) and later with CFP/YFP (GA-2337). Telomere-telomere distance measurements were compared and found to be identical in both strains. Subnuclear position assignment was performed on 19-image (170-nm step size) stacks of living cells acquired on a microscope (model IX70; Olympus) as described previously [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref] [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref]. IF and time-lapse imaging were performed on the Zeiss LSM510 with a 100x Plan-Apochromat objective (NA ϭ 1.4), and images were acquired in multi-tracking mode using lines at 633, 488, and 543 nm and 10-25% power. 3D stacks on fixed cells were typically 16 slices of 0.2 m or 10 of 0.25 m, whereas live imaging was performed in single tracking mode with closed pinhole (1-1.2 airy units; GFP at 488 nm with 0.1-1.0% transmission; CFP-YFP at 458 nm and 514 nm, with 1-25% transmission) on seven sections of 0.4 m. The maximum speeds for 2D acquisition were 80 ms per image (four averages/ROI 30 ϫ 30) and 1.10s for 3D images (7 sections/4 averages/ROI 30 ϫ 30). After quantification, data were routinely Gauss-filtered to reduce noise for presentation. Chromatic aberration is corrected before image-capture by alignment of 0.1 and 0.2 m Tetraspeck Microsphere signals (Molecular Probes). FRAP of nuclear rim was performed using 50 iterations of 100% power pulses of the 488-nm laser.
## Quantitative analyses of distance, position and dynamics
Distances were measured using the Zeiss LSM510 Confocal software version 2.5. Y/CFP and IF signals were scored on 3D stacks using 40-160 nuclei per point, monitoring nuclear integrity through nucleolar shape and nuclear diameter. Tagged telomere position percentages in zone 1 were compared with a random distribution by t test [bib_ref] Live imaging of telomeres: yKu and Sir proteins define redundant telomere-anchoring pathways..., Hediger [/bib_ref] , with a 95% confidence interval. 2D time-lapse series of GFP or YFP-CFP spots were analyzed with MetaMorph Offline v. 4.6r6 (Universal Imaging). For each strain 8-12 videos from two to three independent cultures were combined and averaged. MSD analysis was performed as described previously [bib_ref] Chromosome dynamics in the yeast interphase nucleus, Heun [/bib_ref] [bib_ref] Multiple regimes of constrained chromosome motion are regulated in the interphase Drosophila..., Vazquez [/bib_ref] with modifications as detailed in Results. The IMARIS software (Bitplane) was used to determine coordinates of the center of the fluorescent spots imaged in 3D over time. Reliable coordinates in 3D could be obtained from 4 out of 12 videos taken for each strain. Videos in 2D were also analyzed. Representations of the trajectories projected onto the three imaged planes were obtained using Mathematica. Direction cosines were determined for every vector, which joins two neighboring points of the trajectories of two separate fluorescent spots (frames analyzed for 3D videos were as follows: n ϭ162 for wt, n ϭ 178 for yku70, n ϭ 70 for sir4). Similar analysis was performed on 2D videos for which the numbers of frames were as follows: n ϭ 1099 for wt, n ϭ 1053 for yku70, n ϭ 829 for sir4, n ϭ 971 for Tel6L-14L. The mean of correlations (Pearson's correlation coefficient, c) in each (x, y, z) direction (or x, y in the 2D videos) was determined. This value (c) expresses the degree of linear relationship between two variables and is equal to the average cross product of the variables in standardized form. Pearson's c values can range between Ϫ1.00 and ϩ1.00, with the latter signifying a perfect positive relationship, whereas Ϫ1.00 shows a perfect negative relationship. The smallest correlation is zero. [fig_ref] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud... [/fig_ref] shows chromosomes 3, 5, and 6 loop back on themselves. [fig_ref] Figure 2: 3D position of telomeres relative to each other in intact cells [/fig_ref] shows the position of telomeres relative to the NE. [fig_ref] Figure 3: Chr 3 and Chr 6 form whole chromosome loops [/fig_ref] shows 2D distances between the telomeres during time-lapse imaging. Videos 1-6. Online supplemental material is available at http://www.jcb.org/cgi/content/ full/jcb.200409091/DC1.
# Online supplemental material
[fig] Figure 1: The yeast SPB and nucleolus are aligned with the site of bud emergence throughout interphase. (A) Selected frames from a Zeiss LSM510 confocal time-lapse series of GA-2253 yeast cells as they progress through G2, mitosis and G1, show the NE (Nup49, green) and the SPB (white) opposite the crescent-shaped nucleolus (Nop1, red). The CFP-SPB signal was substituted digitally with white to facilitate visualization. See also Video 2. d, daughter cell; m, mother cell. (B) A population of cells tagged as in A, showing the relationship of the nucleolus and SPB to the emerging bud (arrows). (C) Schematic representation of interphase nuclear polarity. (D) GFP-Nup49-labeled pores in G1-phase cells (GA-2197) were bleached (white frame) by confocal laser exposure and epifluorescence/phase images were taken at 10-s intervals thereafter. Pores indicated are either immobile (gray arrows) or slowly diffusing (white arrows). Bars, 1 m. [/fig]
[fig] Figure 2: 3D position of telomeres relative to each other in intact cells. (A) CFP-lacI and YFP-tetR fusions allow visualization of the inserted lac op and tet op arrays. (B) Image stacks (x-y planes) of 0.2 m along the z-axis from the Zeiss LSM510 are shown for Tel 5L (CFP, red) and 5R (YFP, green). Bar, 1 m. (C and D) Distances between the two telomeres in strains GA-2337 (3R3L), GA-2201 (6L6R), GA-2199 (5L5R), GA-2468 (14L14R), GA-2757 (5L14R) and GA-2202 (6L14L). Images of intact fixed cells were acquired in 3D, typically taking stacks of 12-16 focal planes of 0.2-m intervals along the z-axis. Distributions of distances are plotted by 0.4-m categories (Ϯ0.2 m). [/fig]
[fig] Figure 3: Chr 3 and Chr 6 form whole chromosome loops. Epi-and IF of G1-arrested haploid cells: (A) GA-2195 (SPB, red; 3L::GFP, green; 3R::GFP, green); (B) GA-2201 (SPB, white; 5L::YFP, green; 5R::CFP, red); (C) GA-2199 (SPB, white; 6L::YFP, green; 6R::CFP, red). An example of four color images with the Nop1 channel in blue is given in the insets of B and C. Bar, 1 m. Shown are maximal projections of 10 0.25-m z-sections. (D) Distance frequencies between the two fluorescent markers (in 0.4 m categories Ϯ 0.2 m) for the telomere pairs of Chr 3 (black diamonds), Chr 5 (blue circles), and Chr 6 (red squares). (E) Distribution of the angles ␣ calculated by triangulation of the 3D measurements for all individual cells examined in G1 and S-phase cells. Schematic representation of a folded chromosome and intervening angle ␣ between two chromatids. [/fig]
[fig] Figure 4: Live imaging of telomere dynamics. A Zeiss LSM510 confocal time-lapse microscopy (2D) was performed on double-tagged Chr 3, 5, 6, and 14 taking frames every 1.5 s, by adjusting the plane of focus when necessary (see time-lapse series as Videos 3-6). (A) Representative sequence of frames taken at 1.5-s intervals in 2D of GA-2337, 3R::CFP (red), 3L::YFP (green). Bar, 1 m. (B) Telomere tracks over time: 100 sequential images from 2D time-lapse series are displayed orthogonally, rotated such that the time axis (z) is horizontal. Top panel: GA-2201 6L::YFP (green), 6R::CFP (red) and bottom panel: GA-2199, 5L::YFP (green), 5R::CFP (red). TetR-YFP also produces the diffuse green background. (C) Examples of telomere tracks over 100 frames of 2D time-lapse videos after alignment of interpolated nuclear centers (YFP, green; CFP, red). The dotted circle represents an idealized nuclear circumference (Ø ϭ 2 m). (D) Radial MSD for telomeres 6R and 6L and 5R and 5L obtained using d ϭ distance between one fluorescent telomere spot and the center of the nuclear background fluorescence for each frame as a function of the time interval (inset, for t ϭ 1.5-101.5 s). [/fig]
[fig] Figure 7: 3D and two-color fluorescence time-lapse imaging of telomere dynamics. [/fig]
[fig] Figure 8: Schematic representation of a Rabl-like chromosome organization in yeast. (A) [/fig]
[table] Table I: Average 3D telomere-telomere distances [/table]
|
Brew temperature, at fixed brew strength and extraction, has little impact on the sensory profile of drip brew coffee
## Supplementary
. Interaction plots of attributes with significant two-way (a, b) or three-way (c) interactions.
|
Progressive multifocal leukoencephalopathy in a patient post allo-HCT successfully treated with JC virus specific donor lymphocytes
Background: Progressive multifocal leukoencephalopathy is a demyelinating CNS disorder. Reactivation of John Cunningham virus leads to oligodendrocyte infection with lysis and consequent axonal loss due to demyelination. Patients usually present with confusion and seizures. Late diagnosis and lack of adequate therapy options persistently result in permanent impairment of brain functions. Due to profound T cell depletion, impairment of T-cell function and potent immunosuppressive factors, allogeneic hematopoietic cell transplantation recipients are at high risk for JCV reactivation. To date, PML is almost universally fatal when occurring after allo-HCT.Methods:To optimize therapy specificity, we enriched JCV specific T-cells out of the donor T-cell repertoire from the HLA-identical, anti-JCV-antibody positive family stem cell donor by unstimulated peripheral apheresis[1]. For this, we selected T cells responsive to five JCV peptide libraries via the Cytokine Capture System technology. It enables the enrichment of JCV specific T cells via identification of stimulus-induced interferon gamma secretion.Results:Despite low frequencies of responsive T cells, we succeeded in generating a product containing 20 000 JCV reactive T cells ready for patient infusion. The adoptive cell transfer was performed without complication. Consequently, the clinical course stabilized and the patient slowly went into remission of PML with JCV negative CSF and containment of PML lesion expansion.Conclusion:We report for the first time feasibility of generating T cells with possible anti-JCV activity from a seropositive family donor, a variation of virus specific T-cell therapies suitable for the post allo transplant setting. We also present the unusual case for successful treatment of PML after allo-HCT via virus specific T-cell therapy.
# Background
Progressive multifocal leukoencephalopathy (PML) is a frequently fatal CNS disorder caused by reactivation of JC virus (JCV). Virus replication in latently infected oligodendrocytes leads to axonal demyelination. Affected individuals frequently present with confusion or seizures but the clinical presentation mainly depends on the extent of demyelination and brain structures involved. Due to profound T-cell depletion and potent immunosuppressive factors, allogeneic hematopoietic cell transplantation (allo-HCT) recipients are at risk for JCV reactivation. PML is almost universally fatal when occurring after allo-HCTand new treatment approaches are highly warranted. Recently, PML attracted attention from the medical and scientific community as a first post allo-HCT
## Open access
Journal of Translational Medicine patient apparently benefited from the adoptive transfer of allogeneic BK virus specific T-cells. BK virus and JCV are genetically related and share a number of immunogenic epitopes. Prior to that, another group had already generated JCV specific T cells from a HLA matched donor. The adoptive transfer of these cells was well tolerated and the patient showed neurological improvement. These two cases argue to the potential of T-cell transfers after allo-HCT, although only selected patients will be in a situation where T-cell transfer is an option. Multiple myeloma (MM) patients treated with the anti-CD38 monoclonal antibody daratumumab (Dara) show an increased risk for infectious complications, potentially attributable to the depletion of NK cells.
Recently, reactivation of hepatitis B including fatalities was reported in Dara-treated patients suggesting a more complex immune suppression by CD38 antibodies also affecting T-cell function (Reactions Weekly (2019) 1747: 1. https ://doi.org/10.1007/s4027 8-019-59642 -3). CD38 is expressed on a subset of activated CD8 + T cells, which may explain a dysfunctional T-cell system in some Daratreated patients. Here, we report the first case of PML after Dara-containing therapy during post-allogeneic transplant relapse. Remarkably, PML was successfully treated using multiple strategies, including the adoptive transfer of JCV specific donor lymphocytes.
## Case description
We describe the case of a 59-year old MM patient with an 11-year treatment history. After failing to achieve a durable remission with seven previous lines of treatment, including an allo-HCT from an HLA-identical family donor 6.5 years earlier, he was currently undergoing his 20th cycle of Dara, pomalidomide, bortezomib, cyclophosphamide and dexamethasone. The patient developed seizures and a cerebral MRI revealed a posterior white matter lesion. This lesion showed a hyperintense T2 lesion in the left temporooccipital region with predominant involvement of the subcortical white matter with subsequent regional progression and increasing permeability of the blood-brain-barrier. Cerebrospinal fluid (CSF) analysis revealed JC virus reactivation with up to 560 copies/dL in two independent examinations. The constellation of clinical and radiological manifestations with JC virus positive CSF met the criteria for PML. Notably, besides JC virus reactivation, the patient also presented with CMV reactivation, persistent parainfluenza type 3 positivity on throat swabs and human papilloma virus driven giant anal condyloma acuminatum, in line with severe disruption of T cell mediated immune surveillance.
In this situation, we stopped multimodal myeloma therapy and initiated treatment with cidofovir and mirtazapine as these drugs have shown some PML activity in anecdotal reports. Furthermore, we administered the last remaining dose of unselected donor lymphocytes containing 1,09 x10 7 CD3 positive cells/kg body weight to facilitate immune reconstitution. Unfortunately, the clinical presentation did not improve and the patient suffered from recurrent focal seizures with subsequent generalization, progressive symptoms of cortical blindness such as fixation issues and restrictions in the ability to perceive unmoving objects as well as hearing impairment. These clinical findings are consistent with occipital PML involvement documented in MRI scans.
# Methods
In an attempt to optimize therapy specificity, we collected lymphocytes from the stem cell donor, an HLAidentical, anti-JC virus-antibody positive family donor by unstimulated peripheral blood leukocyte apheresisto generate a JC virus specific T-cell product. Notably, the donor provided very low numbers of JC virus reactive T cells. One billion (10 9 ) leukocytes, 45% of which were T-cells, served as starting material for selection. JC virus specific T cells were isolated using the cytokine capture system (CCS) technology. It permits the enrichment of interferon gamma (IFNy) secreting CD4 + and CD8 + T cells. For this, mononuclear cells were labeled with a bispecific anti-CD45/anti-IFNγ antibody. They constitute the catch matrix. We then incubated the yet unselected leukocytes with five overlapping JC virus peptide libraries (LT, ST, VP1, 2 and 3, Miltenyi Biotech, Bergisch-Gladbach, Germany). The cells that can recognize the peptides consequently secrete IFNγ, which is then immobilized by the bispecific CD45/anti-IFNy antibody on the cell surface. We considered those JC virus specific. A second superparamagnetic bead-conjugated anti-IFNγ-antibody marks the cells for subsequent immunomagnetic purification.
Further materials used were a CliniMACS Prodigy CCS device (Miltenyi Biotec, Bergisch-Gladbach, Germany), a TS500 tubing set and software version 1.2.0.3148, as described in. cGMP conditions were followed throughout and formally validated quality assays were used for release testing in accordance with the manufacturing permit.
The final product contained a 3:2 mix of CD4 + and CD8 + T-cells, two-thirds of each interferon gamma (IFNγ)-positive, for a total of 20 000 antigen-specific T cells and a dose of potentially allo-reactive, not JC virus responsive T cells of < 200/kg body weight.
# Results
The adoptive cell transfer of JC virus specific T cells was performed without complication roughly 2 months after PML diagnosis. Following these therapies, our patient slowly went into remission of PML with JC virus negative CSF 2.5 months after initial symptoms. Consequently, no further deterioration of vision disorder or sensorineural hearing occurred. We observed containment of PML lesion expansion along with complete regression of contrast agent extravasation 1 month after adoptive transfer. The observed radiological time course with initial progression and then regional containment during therapy further confirm successful PML therapy, although possibly accentuated by concomitant Immune Reconstitution Inflammatory Syndrome (IRIS). This, however, is an observation rarely made with MRI because the prognosis of PML is typically dismal due to limited treatment options.
Lymphocyte differentiation by immunophenotyping showed NK-cell recovery with frequencies between 20 and 60% (Additional file 1: . To monitor frequencies of JC virus specific T cells, we repeatedly collected CSF, which remained cell and virus free for multiple examinations at a lower assay sensitivity of 20 copies/dL. Notably, CMV reactivation as well as HPV activity persisted, the latter highlighted by condyloma acuminatum growth and secondary carcinoma development, ultimately requiring surgical intervention 9 months after adaptive JC virus specific T-cell transfer. Unfortunately, the patient complained about new back pain and the M protein increased during the follow-up period. An FDG PET/CT revealed multiple hypermetabolic lesions throughout the axial skeleton 8 months after PML diagnosis, and we had to resume myeloma therapy. We selected elotuzumab, carfilzomib, lenalidomide, and dexamethasone as salvage regimen to which the patient responded well with very good partial remission (VGPR). Now, 12 months after initial manifestation, focal epilepsy persisted despite antiepileptic therapy, a common neurologic sequelae in PML survivors. However, the patient remains in remission for both PML and myeloma.
# Discussion
We report successful treatment of PML after allo-HCT, using a multimodal approach including withdrawal of immunosuppressive therapy and the adoptive transfer of JC virus specific donor lymphocytes, using the CCS technology. Notably, the donor's JC virus reactive T cells were barely discernible at the level of detection of our initial flow cytometry assay. Had we screened for JC virus reactive T cells, we would have called this donor negative, as our assay is only validated to frequencies of 1:2 000 T cells or better. Assuming recoveries as previously observed with CMV specific T cells with this technology, albeit there for processes with a much greater frequency of targetable cells, the frequency of JC virus reactive T cells in this donor is estimated in the range of 1:10 000. Still, we were able to generate a product of reasonable purity and potentially meaningful cell dose. Our observations support a pivotal role of T-cell therapies and are in line with recent case reports on thirdparty BK virus specific T cellsor JC virus specific T cell transfer. Reportedly, several hundred patients have already received CCS T-cells, albeit not for JC virus but for CMV, EBV and AdV. However, the number of treated patients is still very limited. Larger studies comparing these treatments in a randomized fashion will be needed to definitively conclude on the value of JC virus directed T-cell therapies. Moreover, others and we have not been able to demonstrate JC virus specific T-cell expansion at the lesion site as this is difficult and would require brain biopsies.
To the best of our knowledge, this is the first report on PML occurring after Dara-containing combination therapy. We appreciate that our patient was heavily pretreated and post allo-HCT before Dara-containing therapy, which itself put him at risk for PML. An increased risk for reactivated latent infections such as herpes zoster, herpes encephalitis, or hepatitis B during Dara therapy has been described previously and, at least partly, can be attributed to the depletion of NK cells. Possibly, Dara has introduced another type of immunological sequelae highlighted not only by the development of PML but also by concomitant condyloma acuminatum and CMV reactivation. However, the interplay between innate immunity and T-cell responses during CD38 antibody therapy has yet to be elucidated, and the same holds true for Dara treatment post allo-HCT.
# Conclusion
In summary, we demonstrate feasibility for generating JC virus specific T cells from a seropositive family donor. Finally, we present the unusual case for successful treatment of PML after allo-HCT, supporting recent reports on virus specific T-cell therapy.
|
Drug liking and wanting, not impulsive action or reflection is increased by 4-fluoroamphetamine
Background New psychoactive substances (NPS) are chemical analogues designed to mimic the effects of various classic recreational drugs of abuse including MDMA, LSD, and cannabis. NPS use is associated with concern about the acute and longer-term effects particular substances might have, with abuse and addiction as potential consequences. Impulsivity and sensitivity to the rewarding effects of drugs have been considered as risk factors for drug abuse. In light of the popularity of 4-fluoroamphetamine (4-FA), it is important to assess whether 4-FA can lead to subjective drug liking and wanting, and impulsive behavior, all factors contributing to the abuse likelihood of a substance. Methods A placebo-controlled 2-way crossover study in 12 healthy poly-drug using participants was conducted to test subjective and behavioral effects of 4-FA (100 mg). 4-FA concentrations were determined in serum up to 12 h after administration and two impulsivity tasks and two drug experience questionnaires assessing drug liking and wanting, and good and bad drug effect, were administered between 1 and 11 h post-administration. Results Findings showed that 4-FA did not affect impulsive behavior. Self-ratings of drug liking and wanting and good drug effect were increased 1 h after administration; this effect was absent 11 h after drug intake. Discussion and conclusion To conclude, 4-FA (single dose) increased self-rated liking and wanting, which is known to contribute to the abuse likelihood of a substance; however, it left another factor impulsive behavior unaffected. It has to be noted that the current picture is limited and might change with increased sample size, and/or different 4-FA doses. Clinical trial registration Trial acronym: 4-FA. URL: http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=6164. Registration number: NTR6164 (Dutch clinical trial registry number).
# Introduction
New psychoactive substances (NPS) are chemical analogues designed to mimic the effects of various classic recreational drugs of abuse including MDMA, LSD, and cannabis . Anecdotal evidence by users suggests that NPS have subjective effects comparable to the classical psychoactive substances though these effects have not been substantiated yet in human experimental studies [bib_ref] Novel psychoactive substances (designer drugs): overview and pharmacology of modulators of monoamine..., Liechti [/bib_ref]. Scientific knowledge about NPS's acute effects is needed, seen their exponential increase in availability and number over the last 10 years. Moreover, this surge has gone hand in hand with an increase in use, and emergency department visits due to overintoxication [bib_ref] Five-year trends in self-reported recreational drugs associated with presentation to a UK..., Wood [/bib_ref].
4-Fluoroamphetamine (4-FA) is a prototypical example of an NPS, belonging to the chemical class of phenethylamines . It appeared on the Dutch Bdrug market^between 2007 and 2009 where after reports of acute toxic effects steadily increased to even 16% of all reported cases on large scale events [bib_ref] Acute toxic effects related to 4-fluoroamphetamine, Wijers [/bib_ref]. Although it was first used as an adulterant in drugs such as amphetamine and MDMA, it became a drug of choice, liked by users for its effects [bib_ref] 4-Fluoroamphetamine in the Netherlands: more than a onenight stand, Linsen [/bib_ref]. While it has become a popular drug with a recent survey amongst Dutch partygoers revealing that a quarter of the respondents between 15 and 35 years of age had used 4-FA in the past year, the majority (80%) also states that their 4-FA use remained limited to just a few times.
The effects of 4-FA reportedly range between those of amphetamine, a stimulant, and MDMA, an empathogen [bib_ref] 4-Fluoroamphetamine in the Netherlands: more than a onenight stand, Linsen [/bib_ref] which is in line with the biological profile that is also suggested to be in between that of MDMA and amphetamine. Studies determining the monoamine transporter and receptor binding profile in animals and human tissue showed that 4-FA has relatively more serotonergic transporter action compared to amphetamine. Next to the promotion of norepinephrine (NE) and dopamine (DA) release, it was shown to release serotonin (5-HT) similarly to MDMA [bib_ref] The effects of non-medically used psychoactive drugs on monoamine neurotransmission in rat..., Nagai [/bib_ref] [bib_ref] Monoamine transporter and receptor interaction profiles of novel psychoactive substances: Parahalogenated amphetamines..., Rickli [/bib_ref].
Linked to NPS use is the concern about the acute and potential long(er) term effects the particular substance might have on behavior and cognitive processes [bib_ref] Novel psychoactive substances of interest for psychiatry, Schifano [/bib_ref] with abuse and addiction as potential consequences. It is known that substances with more pronounced action at the DA-transporter (DAT) can have a higher abuse potential compared to substances that increase activity of the 5-HT system [bib_ref] Relationship between the serotonergic activity and reinforcing effects of a series of..., Wee [/bib_ref]. Furthermore, a high DAT-to-serotonin transporter (SERT) ratio is a pharmacological characteristic predicting more stimulant effects and a higher potential for addiction. The DAT/SERT ratio of 4-FA is approximately five times higher than MDMA and seven times lower than damphetamine [bib_ref] Monoamine transporter and receptor interaction profiles of novel psychoactive substances: Parahalogenated amphetamines..., Rickli [/bib_ref]. Drug abuse and addiction are related to impulse control, with substances disturbing impulse control eventually evolving into drug abuse when the behavior becomes driven by drug-cues [bib_ref] Impulsivity as a determinant and consequence of drug use: a review of..., De Wit [/bib_ref] [bib_ref] Insight into the relationship between impulsivity and substance abuse from studies using..., Winstanley [/bib_ref]. Similarly, sensitivity to the rewarding effects of drugs have also been considered as risk factors for drug abuse (Waefer and de Wit 2013).
The most sensitive and reliable measures of abuse likelihood of a substance are self-ratings of drug liking, or the report of how much the user likes the drug [bib_ref] Principles of laboratory assessment of drug abuse liability and implications for clinical..., Carter [/bib_ref]. In addition, the measures wanting, good drug effect, and bad drug effect seem to co-vary with liking [bib_ref] Principles of laboratory assessment of drug abuse liability and implications for clinical..., Carter [/bib_ref]. On the other hand, typical paradigms to assess different components of impulsivity are the matching familiar figures task (MFFT) and the stop-signal task (SST). In the latter SST-paradigm, which tests motor impulsivity or impulsive action, the participant has to respond continuously to stimuli and withhold their pre-potent response to Bstopŝ timuli [bib_ref] On the ability to inhibit simple and choice reaction time responses: a..., Logan [/bib_ref]. In the former MFFT-paradigm, which tests cognitive impulsivity or reflection impulsivity, the participant has to match a target figure to one of the six shown Blook-a-likes^of which five differ from the original one by only a small detail. The skill is to withhold the reaction until the match is found (D'Amour-Horvat and Leyton 2014; [bib_ref] Bright and dark sides of impulsivity: performance of women with high and..., Perales [/bib_ref]. In general, it has been shown that amphetamine (10-40 mg) does not affect impulsive action , though when it does, an enhancing effect was found only in people who either performed badly at baseline [bib_ref] Effects of d-amphetamine and ethanol on a measure of behavioral inhibition in..., De Wit [/bib_ref] [bib_ref] Acute administration of damphetamine decreases impulsivity in healthy volunteers, De Wit [/bib_ref] or when presented with complex stimuli [bib_ref] Effects of d-amphetamine on behavioral control in stimulant abusers: the role of..., Fillmore [/bib_ref]. Likewise, studies with MDMA have shown that in doses ranging from 25 to 125 mg, MDMA either leads to improvement of behavioral inhibition or induces no change [bib_ref] Dose-related effects of MDMA on psychomotor function and mood before, during, and..., Bosker [/bib_ref] [bib_ref] Comparative effects of methylphenidate, modafinil, and MDMA on response inhibition neural networks..., Schmidt [/bib_ref]. In one single study, MDMA led to an increase in impulsive action and impulsive reflection [bib_ref] Effects of acute MDMA intoxication on mood and impulsivity: role of the..., Van Wel [/bib_ref]. The effects of amphetamine and MDMA on subjective measures of drug liking also seem to be similar, with an increase in self-rating up to 3 h after administration of a single dose [bib_ref] Human pharmacology of 3,4-methylenedioxymethamphetamine (Becstasy^): psychomotor performance and subjective effects, Cami [/bib_ref] [bib_ref] Subjective and hormonal effects of 3,4-methylenedioxymethamphetamine (MDMA) in humans, Harris [/bib_ref] [bib_ref] Reinforcing and subject-rated effects of methylphenidate and d-amphetamine in non-drug-abusing humans, Rush [/bib_ref].
In light of the popularity of 4-FA, it is important to assess whether 4-FA can lead to subjective drug liking and impulsive behavior, two risk factors linked to abuse liability of a substance. Given the similarity to amphetamine and MDMA, it was hypothesized that 4-FA would not lead to effects on impulsive action or reflection and would lead to a subjective state of drug liking and wanting, and good drug effect around peak drug concentrations and not after 4-FA plasma concentrations have decreased substantially.
The data presented in this paper are part of a larger project PREDICT ( www.predictnps.eu ) focusing at the safety profile of NPS in humans, in vitro and in silico. Additional data of the present study including the safety profile and neurocognitive effects of 4-FA are published in a separate paper [bib_ref] Safety profile and neurocognitive function following acute 4-fluoroamphetamine (4-FA) administration in humans...., De Sousa Fernandes Perna [/bib_ref].
# Methods
## Study design and treatment
The study was conducted according to a two-way crossover, randomized, counter-balanced, and placebo-controlled design. Treatment was 100 mg of 4-FA or placebo mixed with 100 mL of bitter lemon. The drink was ingested at once. The 100 mg dose was based on a user survey amongst Dutch 4-FA users. The majority (75%) of the users who knew which dose they ingest typically indicated it to be between 50 and 150 mg, the remainder used larger doses. In general, the subjective effects last between 4 and 6 h [bib_ref] 4-Fluoroamphetamine in the Netherlands: more than a onenight stand, Linsen [/bib_ref]. 4-FA has an estimated half-life of 3.7 h in the rat brain [bib_ref] Comparison of 4-chloro-, 4-bromo-and 4-fluoroamphetamine in rats: drug levels in brain and..., Fuller [/bib_ref].
A permit for obtaining, storing, and administering 4-FA was obtained from the Dutch drug enforcement administration. The study was performed in accordance with the Helsinki Declaration of 1975, and its subsequent amendments, and was approved by the Medical Ethics Committee of the Academic Hospital of Maastricht and the University of Maastricht. It was registered in the Dutch Clinical Trial Register (Registration Number: NTR6164).
## Participants
Participants were 12 healthy recreational polydrug users aged 22.3 (± 3.4) years on average (± SD) of whom 7 were male with a mean BMI of 22.9 (± 1.3 SD) and 5 were female with an average BMI of 21.5 (± 2.8 SD). All of them had experience with alcohol use with the units consumed per week ranging from 2 to 20. One participant smoked cigarettes with an average of 15 per day and seven smoked cannabis with an average of 1 Bunit^per week. The use of other drugs was expressed in Blifetime use^; the minimum and maximum (min-max) times used, together with the number of participants with experience (N) is listed per drug (N; min-max): amphetamines (8; 1-32), cocaine (6; 1-26), ecstasy (11; 1-35), 4-FA (5; 1-25), LSD (3; 7-17), and other drugs like mushrooms and ketamine (8; 1-33).
## Procedures
Participants were recruited by means of flyers in the university building, an advert on a research-Facebook page, and by word of mouth. When interested, they were sent the information brochure explaining the background, aims and study procedure, and two questionnaires (medical and drug use history). When they were fully informed, potential questions were answered, and if they fulfilled at first sight of the inclusion criteria, they were invited for a medical screening. When no objections were raised during the physical examination including a standard blood and urine screens and an electrocardiogram (EKG) and participants signed the informed consent, they were included in the study.
Inclusion criteria were previous experience with psychostimulants (≤ 1 time/week) and at least one time during the previous year; age between 18 and 40 years; free from psychotropic medication; healthy based on the assessment of medical history, physical examination, vital signs, EKG (with heart rate 51-100 bpm; lower limit for fit people, 45 bpm), a resting systolic blood pressure 91-140 mmHg, a resting diastolic blood pressure 51-90 mmHg, and the results of the hematology, clinical chemistry, urinalysis, and serology within the reference ranges; normal binocular visual acuity, corrected or uncorrected; absence of any major medical, endocrine and neurological conditions, and normal weight as defined by a body mass index between 19.5 and 28 kg/m 2 ; written informed consent. Exclusion criteria were history of drug abuse or addiction (determined by the medical questionnaire, drug questionnaire, and medical examination); excessive drinking (> 20 alcoholic consumptions a week); pregnancy or lactation; hypertension (diastolic > 90; systolic > 140); current or history of psychiatric disorder (determined by the medical questionnaire and medical examination); liver dysfunction; (serious) side effects to previous psychostimulant use; history of cardiac dysfunctions (arrhythmia, ischemic heart disease,…); simultaneous participation in another clinical trial; being a blood donor; and for women: not using reliable contraceptive.
After study inclusion and prior to the test days, participants were familiarized with the study procedures, tests, and questionnaires. On a test day, participants arrived early in the morning and they were tested for the absence of drugs in urine and alcohol in breath. In case of females, an additional test for pregnancy was conducted in urine. When all tests were negative, participants were given baseline questionnaires, a blood sample was taken and they received a light-standardized breakfast.provides an overview of timing of the questionnaire, the impulsivity tests and blood samples during the test day, which took 12.5 h in total. The test schedule was identical for each test day and each participant. Participants were paid upon completion of the testing periods for their participation.
## Drug experience questionnaires
Sensitivity to drug reinforcement questionnaire
The sensitivity to drug reinforcement questionnaire (SDRQ) asks participants to rate their liking and wanting of 4-FA use during their present condition on a 5-point rating scale (1 = somewhat; 2 = slightly; 3 = moderately; 4 = very; 5 = extremely). The questionnaire is comprised of two questions, BHow pleasant is using 4-FA right now?^and BHow much do you want to use 4-FA right now?^referring respectively to drug liking and drug wanting.
## Profile of mood states
The Profile of Mood States (POMS) [bib_ref] Acute administration of damphetamine decreases impulsivity in healthy volunteers, De Wit [/bib_ref] is a self-assessment mood questionnaire with 72 five-point-Likert scale items on which participants have to indicate to what extent these items were representing their mood. Items are clustered to represent eight basic mood states: anxiety, depression, anger, vigor, fatigue, confusion, friendliness, and elation. From those scales, two composite scales were derived, good drug effect (vigor + friendliness + elation/22) and bad drug effect (anxiety + depression + anger + fatigue + confusion/50).
## Impulsivity tests
## Stop signal test
The current stop-signal test (SST) is adapted from an earlier version of Fillmore and colleagues [bib_ref] Acute effects of oral cocaine on inhibitory control of behavior in humans, Fillmore [/bib_ref] , it assesses impulsive action and it has previously been used in similar research [bib_ref] Effects of acute MDMA intoxication on mood and impulsivity: role of the..., Van Wel [/bib_ref]. It requires participants to make quick key responses to visually presented go signals and to inhibit any response when a visual stop signal (an asterisk) is suddenly presented in one of the corners of the screen. This can occur after one of four fixed delays (50, 150, 250, and 350 ms) after the onset of the go signal. The go signals were four letters presented one at a time for 500 ms in the center of a computer screen. Participants are required to respond to each letter as quickly as possible by pressing on of two response buttons. The computer screen is blank for 1.5 s before the next letter is displayed. This provides a period of 2 s in which the participant can respond to the go signal. A single test consists of 176 trials in which each of the 4-letter stimuli will be presented equally often. A stop signal occurs in 48 trials during a test. Participants are required to withhold any response in case a stop signal is presented. The task lasts about 10 min. Dependent variables are proportion of correct go responses and failed inhibitions on stop trials and corresponding reaction times [bib_ref] On the ability to inhibit simple and choice reaction time responses: a..., Logan [/bib_ref]. The Stop reaction time (stop RT) to stop signal represents the estimated mean time required to inhibit a response.
The method for calculating stop reaction time was taken from the race model of inhibitory control [bib_ref] On the ability to inhibit thought and action. A user's guide to..., Logan [/bib_ref]. This model proposes that the response to stop signals is defined by two parallel processes: execution of a motor action in response to a signal and inhibition of a motor action in response to a stop signal. Crucial to the outcome of the race is the speed of both processes. Response inhibition will fail if the time required to inhibit exceeds the time to complete a motor response at the time of the stop signal.
The speed of the inhibition response cannot be observed directly but can be derived mathematically on the basis of three factors: stop-signal delay, reaction time distribution on go trials, and the probability of successful response inhibitions in stop signal trials. First, reaction times to 128 go trials were rank ordered from shortest to longest. The finishing time of the inhibition response was then determined from the probability of successful response inhibition and the distribution of reaction times. If n percent of the responses on stop-signal trials would be unsuccessfully inhibited (failed inhibitions), then the finishing time would be associated with the nth percentile of the RT distribution. Stop RT was then determined by subtracting the appropriate stop-signal delay from reaction time at the nth percentile of the RT distribution. The resulting values for each stop signal delay were then averaged to yield a single measure of stop reaction time for the test.
## Matching familiar figures test
The matching familiar figures test (MFFT) assesses impulsive reflection, which is the tendency to reflect on the validity of problem solving under the special condition of several possible alternatives. The test involves simultaneous presentation of a target figure positioned on the left of the screen and an array of six alternatives on the right half of the screen, all except one differing in one or more details from the target figure. The participant is asked to select from the alternatives the figure that exactly matches the target as quickly as possible. If the initial selection is incorrect, this is signaled with a beep and subjects are required to give another answer. Each participant is given 2 examples followed by 20 test items. The response latency and number of errors before the correct match are collected per item. The main dependent variables resulting from these measures are the mean latency for first response, the accumulated number of errors made before the correct match, an impulsivity score (I-score), and an efficiency score (E-score). The I-score is a composite index of impulsivity, whereas the E-score reflects the balance between Bfast and accurate^and Bslow and inaccurate.^The I-score is calculated by subtracting the standardized mean latency from the standardized number of errors. The E-score is calculated by adding the standardized mean latency to the standardized number of errors and multiplying the result by − 1 [bib_ref] Bright and dark sides of impulsivity: performance of women with high and..., Perales [/bib_ref].
## Pharmacokinetics
A blood sample (5 mL) was collected at baseline and at regular times after treatment (see. Samples were centrifuged immediately and resulting serum was pipetted into a clean tube and stored at − 20°C until 4-FA concentration determination which took place after study completion. Blood serum (0.5 mL) was diluted with buffer and internal standard solution was added. After liquid-liquid extraction the extract was analyzed using LC-MSMS, with 0.04 ng/mL as the lower limit of quantification.
## Statistical analyses
Questionnaire data and data of the SST was analyzed with repeated measures general linear models (RM GLM) ANOVA with treatment (two levels) and time of measurement (two levels SDRQ, three levels POMS and SST) as within subject factors (SPSS, version 24.0). In case of main effects of time of measurement, Bonferroni-corrected post-hoc tests were conducted. Data of the MFFT was analyzed by means of paired samples t tests since there was only one assessment. The alpha criterion level of statistical significance for all analyses was set at p = 0.05. Partial eta squared (ƞ p 2 ) is reported in case of significant effects in the ANOVA GLM to demonstrate the effect's magnitude, where 0.01 is defined as small, 0.06 as moderate and 0.14 as large. Partial eta squared is based on Cohen's f which defines small, medium and large as respectively 0.10, 0.25, and 0.50 which corresponds to η 2 of 0.0099, 0.0588, and 0.1379 [bib_ref] Eta squared and partial eta squared as measures of effect size in..., Richardson [/bib_ref].
# Results
Pharmacokinetics Mean (± SE) 4-FA serum concentrations were 167.3 ng/mL (±15) at T1, 60′ post-treatment, peaked 2 h after intake (205.4 ng/mL ± 45) and descended over time to 97.2 ng/mL (± 10), 12 h after 4-FA administration (T5 +1).
## Drug experience questionnaires
Sensitivity to drug reinforcement questionnaire RM GLM ANOVA showed statistically significant effects of Treatment, Time and Treatment by Time on both scales of the SDRQ. Ratings of liking were higher after 4-FA compared to placebo (F 1,11 = 26.16; p < 0.001; ƞ p 2 = 0.70), they were highest at T1 compared to T5 (F 1,11 = 22.99; p = 0.001; ƞ p 2 = 0.68), and while the liking ratings remained stable in the placebo condition, they decreased substantially in the 4-FA condition over time (F 1,11 = 13.13; p = 0.004; ƞ p 2 = 0.54) [fig_ref] Figure 1: Mean [/fig_ref]. For wanting the same pattern was observed with higher ratings of wanting after 4-FA compared to placebo (F 1,11 = 19.06; p = 0.001; ƞ p 2 = 0.63), highest ratings at T1 compared to T5 (F 1,11 = 14.73; p = 0.003; ƞ p 2 = 0.57), and while ratings of wanting remained stable in the placebo condition, they substantially decreased in the 4-FA condition over time (F 1,11 = 22.18; p = 0.001; ƞ p 2 = 0.67) [fig_ref] Figure 1: Mean [/fig_ref].
## Profile of mood states
Since one of the POMS sub-scales included in the composite scales displayed a baseline difference between test days, baseline-corrected scores entered the analyses. RM GLM ANOVA showed main effects of treatment (F 1,10 = 7.55; p = 0.02; ƞ p 2 = 0.43) and time (F 2,20 = 17.81; p < 0.001; ƞ p 2 = 0.64) and a treatment by time interaction effect (F 2,20 = 8.12; p = 0.003; ƞ p 2 = 0.45) on good drug effect. The good drug effect was higher after 4-FA compared to placebo; the overall effect was the highest on T1 compared to T3 and T5. The quadratic interaction (F 1,10 = 12.57; p = 0.005; ƞ p 2 = 0.56) between treatment and time demonstrated that while the ratings in the placebo condition were low and decreased slightly over time, the ratings in the 4-FA condition were very pronounced at T1 and steeply decreased from T1 to T3 while remaining at the same low level at T5 compared to T3 [fig_ref] Figure 1: Mean [/fig_ref].
Analyses showed a significant treatment by time interaction effect (F 1,10 = 6.59; p = 0.006; ƞ p 2 = 0.43) on bad drug effect. Post-hoc analyses showed that this was a quadratic effect (F 1,10 = 2.19; p = 0.02; ƞ p 2 = 0.43) with the highest rating showing at T3 for 4-FA compared to the other time-points and placebo; this effect was probably driven by the statistically significant increased levels of fatigue and confusion, two of the sub-scales included in this composite scale bad drug effect which were the highest at this time-point. There was no main effect of treatment (F 1,10 = 2.19; p = 0.02; ƞ p 2 = 0.43) or time on bad drug effect [fig_ref] Figure 1: Mean [/fig_ref].
## Impulsivity tasks
## Stop signal task
One participant was excluded from the analysis because of an absence of responses on go trials on four occasions (three times placebo condition, once 4-FA). Repeated measures GLM ANOVA demonstrated a main effect of time of measurement (F 2,20 = 3.57; p = 0.05; ƞ p 2 = 0.26) and a treatment by time of measurement interaction effect (F 2,20 = 3.52; p = 0.05; ƞ p 2 = 0.26) on the proportion of failed inhibitions. Posthoc tests did not reveal statistically significant differences between separate time of measurements or treatment by time of measurement performances. Inspecting the data visually led to the suggestion that these effects were mainly driven by the high number of failed inhibitions 1 h after 4-FA administration while the number of inhibition failures in the placebo condition was lower and stable over time. The number of failed inhibitions 4 and 8 h after 4-FA administration was comparable to placebo-levels. Analyses did not reveal statistically significant main effects of treatment or time of measurement, or their interaction on proportion of correct go responses, go-RT or stop-RT [fig_ref] Table 2: Mean [/fig_ref].
## Matching familiar figures test
Paired samples t tests did not reveal statistically significant differences between 4-FA and placebo on the dependent variables mean latency of first response (t 1,11 = − 0.64; p = 0.53) and errors (t 1,11 = 0.12; p = 0.9), or the two composite score, impulsivity (t 1,11 = 0.72; p = 0.48) and efficiency (t 1,11 = 0.27; p = 0.79). Mean (± SE) scores after placebo and 4-FA were respectively 13 (1) and 15 (2) for latency in seconds, 4.00 (1.1) and 3.75 (1.5) for total number of errors, 0.2 (0.3) and − 0.2 (0.5) for I-score, and 0.1 (0.4) and − 0.1 (0.5) for E-score.
# Discussion
The present study aimed to assess whether 4-FA elicits risk factors for drug abuse, namely impulsive reflection and action, a subjective state of drug liking and wanting, and good versus bad drug effect. It was hypothesized that 4-FA would not lead to effects on impulsive behavior but would produce a state of drug liking at peak drug concentrations. As expected, findings showed an absence of 4-FA effects on impulsive reflection and action and an increase in self-ratings of drug liking, drug wanting and good drug effect, 1 h after administration and a peak in bad drug effect 4 h after intake. The liking, wanting, and good drug effects were absent 11 h after drug intake. The absence of drug effects on the impulsivity measures was in line with expectations and previous studies with amphetamine, MDMA and cocaine [bib_ref] Dose-related effects of MDMA on psychomotor function and mood before, during, and..., Bosker [/bib_ref] [bib_ref] Effects of d-amphetamine and ethanol on a measure of behavioral inhibition in..., De Wit [/bib_ref] [bib_ref] Acute administration of damphetamine decreases impulsivity in healthy volunteers, De Wit [/bib_ref] [bib_ref] Acute effects of lisdexamfetamine and D-amphetamine on social cognition and cognitive performance..., Dolder [/bib_ref] [bib_ref] Comparative effects of methylphenidate, modafinil, and MDMA on response inhibition neural networks..., Schmidt [/bib_ref]. The average scores during drug and placebo conditions were also comparable to those found in previous drug studies by our group using the same paradigms (e.g., [bib_ref] Effects of acute MDMA intoxication on mood and impulsivity: role of the..., Van Wel [/bib_ref]. Although present findings suggest that a single dose of 4-FA (100 mg) does not induce impulsive behavior, previous studies have shown that personal characteristics, like baseline performance and task-related characteristics like stimulus/ response complexity, can play a role in drug-induced changes. People performing badly at baseline or who were presented with complex stimuli showed an enhancing effect after amphetamine [bib_ref] Effects of d-amphetamine and ethanol on a measure of behavioral inhibition in..., De Wit [/bib_ref] [bib_ref] Acute administration of damphetamine decreases impulsivity in healthy volunteers, De Wit [/bib_ref] [bib_ref] Effects of d-amphetamine on behavioral control in stimulant abusers: the role of..., Fillmore [/bib_ref]. Additionally, stimulants like amphetamine and cocaine have shown to exert enhancing effects on response control in individuals with impulse control problems, like ADHD and drug addiction [bib_ref] Stimulants: therapeutic actions in ADHD, Arnsten [/bib_ref] [bib_ref] Acute effects of oral cocaine on inhibitory control of behavior in humans, Fillmore [/bib_ref] [bib_ref] Cocaine improves inhibitory control in a human model of response conflict, Fillmore [/bib_ref].
Self-ratings of liking, wanting, and good drug effect, all reliable and sensitive indicators of drug abuse likelihood, were increased compared to placebo 1 h after intake. Ten hours later, the self-ratings in the 4-FA condition were indistinguishable from ratings in the placebo condition. Findings demonstrate that a single dose of 4-FA (100 mg) does not lead to craving (wanting) when the good drug effects are subsiding and bad drug effects increasing, suggesting an absence of repeated or compulsive use of this drug at this dose. However, similar to behavioral performance, studies have shown that personal or biological factors like baseline performance or DA receptor availability can play a role in subjective drug experience [bib_ref] The neurobiology and genetics of impulse control disorders: relationships to drug addictions, Brewer [/bib_ref] [bib_ref] Are attention lapses related to d-amphetamine liking?, Mccloskey [/bib_ref]. It was demonstrated previously that participants who performed worse on an attention paradigm-liked amphetamine (20 mg) less and reported smaller increases in wanting compared to participants who exhibited better attentional capacities. It was concluded that participants' attention capacities determined the sensitivity to stimulant-induced effects with worse capacity signaling reduced sensitivity to stimulant-induced euphoria [bib_ref] Are attention lapses related to d-amphetamine liking?, Mccloskey [/bib_ref]. In addition, low baseline measures of D2 receptor availability in nonaddicted people was shown to predict methylphenidate liking and high levels of impulsivity in rats [bib_ref] The neurobiology and genetics of impulse control disorders: relationships to drug addictions, Brewer [/bib_ref].
While the findings of the present study suggest that a single dose of 4-FA (100 mg) does not lead to either impulsive action or reflection, or to liking and wanting of the drug when the peak effects have subsided, it has to be noted that the current picture is limited and might change with increased sample size, including participants with poor baseline impulse control and attention capacity, and/or different 4-FA doses. Furthermore, additional repetitions of self-rated liking and wanting are needed to know whether these feelings are present when for example bad drug effects are high and good drug effects are low, a few hours after drug intake, as this could push the individual to repeated drug use.
[fig] Figure 1: Mean (± SE) ratings of 4-FA liking (a) and 4-FA wanting (b) 1 and 11 h after treatments, and ratings of good drug effect (c) and bad drug effect (d) 1, 4, and 11 h after treatments and corresponding 4-FA serum concentrations [/fig]
[table] Table 2: Mean (± SE) of dependent variables of the SST F-, p-, and partial eta 2 -values of RM GLM ANOVA Mean (± SE) RM GLM ANOVA, main and interaction effects [/table]
|
Pulmonary Rehabilitation in Patients with COVID-19—A Protocol for Systematic Review and Meta-Analysis
# Introduction
The world is now in the middle of the COVID-19 pandemic. Globally, as of 2 August 2022, there have been 575,887,049 confirmed cases of COVID-19, including 6,398,412 deaths, reported to WHO. COVID-19 is an acute infectious illness that primarily affects the respiratory system and lungs, with severe lung damage being the leading cause of mortality [bib_ref] Seroprevalence of SARS-CoV-2 antibodies in people with an acute loss in their..., Makaronidis [/bib_ref]. Given the commonly severe involvement of multiple organs and body functions during COVID-19, other abnormalities may persist after the acute phase has passed, potentially affecting patients' well-being.
These issues can be addressed through pulmonary rehabilitation (PR). PR, defined as a comprehensive non-pharmacological strategy, is often remarkably successful in patients with chronic obstructive pulmonary disease (COPD) [bib_ref] statement: Key concepts and advances in pulmonary rehabilitation, Spruit [/bib_ref]. However, COVID-19 is distinct from other types of lung disease, and understanding the impact of PR on COVID-19 patients is critical to developing better COVID-19 treatments.
According to the 2013 American Thoracic Society (ATS)/European Respiratory Society (ERS) Statement [bib_ref] statement: Key concepts and advances in pulmonary rehabilitation, Spruit [/bib_ref] , PR is "a comprehensive intervention based on a comprehensive patient assessment and patient-tailored treatment, including but not limited to exercise training, education, and behavioral changes. PR aims at improving chronic respiratory, physical, and mental status of patients with disease, and promoting long-term adherence to health care behavior". Since 2015, numerous clinical trials have provided data on safety and clinical outcomes of planned PR models, including home rehabilitation; telerehabilitation [bib_ref] Home-based maintenance tele-rehabilitation reduces the risk for acute exacerbations of COPD, hospitalisations..., Vasilopoulou [/bib_ref] ;
interactive, web-based models [bib_ref] Self-management support using a digital health system compared with usual care for..., Farmer [/bib_ref] ; mixed heart failure/PR models [bib_ref] Home-based telerehabilitation in older patients with chronic obstructive pulmonary disease and heart..., Bernocchi [/bib_ref] ; and so on [bib_ref] Effects of ground-based walking training on daily physical activity in people with..., Wootton [/bib_ref]. The comprehensive assessment of PR mainly includes five areas: exercise capacity, quality of life, dyspnea, nutritional status, and occupational status [bib_ref] Defining Modern Pulmonary Rehabilitation, Holland [/bib_ref].
PR is the cornerstone of the treatment of patients with chronic respiratory diseases [bib_ref] Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary..., Vestbo [/bib_ref]. There is direct evidence that PR can improve exercise capacity, reduce dyspnea, improve health-related quality of life, and reduce hospitalization rates in chronic obstructive pulmonary disease (COPD) patients [bib_ref] Pulmonary rehabilitation for chronic obstructive pulmonary disease, Mccarthy [/bib_ref] [bib_ref] Pulmonary rehabilitation following exacerbations of chronic obstructive pulmonary disease, Puhan [/bib_ref]. There is now growing evidence that PR can improve outcomes in other conditions such as interstitial lung disease, bronchiectasis [bib_ref] Pulmonary rehabilitation in individuals with non-cystic fibrosis bronchiectasis: A systematic review, Lee [/bib_ref] , and pulmonary hypertension [bib_ref] Exercise-based rehabilitation programmes for pulmonary hypertension, Morris [/bib_ref]. Furthermore, education and psychological support in PR can enhance patients' cognition and comprehension of their diseases [bib_ref] Community pulmonary rehabilitation after hospitalisation for acute exacerbations of chronic obstructive pulmonary..., Man [/bib_ref]. Therefore, the extension of PR has great practical significance, and many nations or territories have also issued guidelines and consensus statements for PR nursing in chronic respiratory diseases [bib_ref] Long-term noninvasive ventilation in chronic stable hypercapnic chronic obstructive pulmonary disease. An..., Macrea [/bib_ref] [bib_ref] The occupational burden of nonmalignant respiratory diseases. An official, Blanc [/bib_ref] [bib_ref] Pulmonary Rehabilitation Guidelines for Australia and New Zealand, Alison [/bib_ref] [bib_ref] Chronic respiratory disease lung rehabilitation nursing expert consensus, Gong [/bib_ref].
In recent years, there have been several works discussing the effects of PR on COVID-19 patients. Gonzalez-Gerez et al. [bib_ref] Short-term effects of a respiratory telerehabilitation program in confined COVID-19 patients in..., Gonzalez-Gerez [/bib_ref] found that PR can improve the physical condition, dyspnea, and perceived effort among people with mild to moderate COVID-19 symptoms in the acute stage; however, long-term effects cannot be determined based on the results in the study. A retrospective cohort study [bib_ref] Six-month outcomes and effect of pulmonary rehabilitation among patients hospitalized with COVID-19:..., Dun [/bib_ref] found that PR could be used to promote exercise capacity improvement after COVID-19. Their study, however, can only suggest a potential association between PR and outcomes, not the causal effect of PR, and is highly biased. Conducting high-quality clinical trials in a large number of patients is nearly impossible due to the high contagiousness of COVID-19, and a systematic review and meta-analysis are still lacking. It is still unclear how PR affects COVID-19 patients and whether this impact is affected by the patient's age, disease severity, stage of disease, type of PR, and the program of PR.
A synthesis of the evidence of the association between PR and pulmonary function, exercise capacity, and health-related quality of life would contribute to a better understanding of the relationship between PR and patients with COVID-19. Furthermore, investigating the prognostic value of PR for the above outcomes would aid in better understanding the mechanism of COVID-19 and making better clinical decisions. Unfortunately, most of the current PR recommendations for COVID-19 rely on previous inferences from severe acute respiratory syndrome (SARS) recovery [bib_ref] Pulmonary rehabilitation principles in SARS-CoV-2 infection (COVID-19): A guideline for the acute..., Aytür [/bib_ref].
Stronger conclusions can be drawn from a systematic review of the literature than from any single study, and this protocol will outline the methods and analyses used in a systematic review. The systematic review and meta-analysis will explore whether PR is an effective intervention to improve the prognosis of patients with COVID-19.
# Materials and methods
## Aims
## Primary aim
The aim of this systematic review and meta-analysis protocol is to evaluate the effect of pulmonary rehabilitation (PR) on COVID-19 patients.
## Review questions
What are the effects of PR on COVID-19 patients compared to controls? Which outcomes are significantly influenced by using PR? Whether the effect of PR intervention is different on COVID-19 patients with different disease severity and intervention frequency?
Which types and characteristics of interventions with PR were devised for COVID-19 patients?
Whether PR is an effective intervention for improving the prognosis of patients with COVID-19?
## Design
This review is conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020 [bib_ref] The PRISMA 2020 statement: An updated guideline for reporting systematic reviews, Page [/bib_ref] statement. The systematic review protocol was prospectively registered at the International Prospective Register of Systematic Reviews (PROSPERO, Registration No. CRD42022301418). The intent of the systematic review, as registered in PROSPERO, was to evaluate the effects of PR in patients with COVID-19, and to explore whether PR is an effective intervention for improving the prognosis of patients with COVID-19. If the conditions for effect size merging are met, meta-analysis will exist.
## Inclusion and exclusion criteria
The criteria of this study will be summarized based on the participants, interventions, comparisons, outcomes, and study design (PICOS) schema according to the Cochrane handbook for systematic reviews of interventions.
## Types of studies
Include: All empirical research studies published in peer-reviewed journals that provide comparative quantitative data on our primary outcomes will be considered as eligible, such as randomized controlled trials (RCTs), quasi-RCTs, non-randomized controlled trials, and observational case-control and cohort studies. For relevant but unpublished studies, research groups will be contacted with a request to provide summary data.
Exclude: We will exclude reviews, letters, poster presentations, editorials, case series, and protocols.
## Types of participants
Include: All patients suffering from COVID-19 will be included regardless of stage of disease, severity of illness, sex, age, race, education, and economic status. The diagnostic criteria refer to clinical diagnosis and treatment guidelines issued by the United States [bib_ref] Infectious Diseases Society of America Guidelines on the treatment and management of..., Bhimraj [/bib_ref] and China.
## Types of interventions
Include: A broad range of PR will be included to gain a comprehensive overview of current approaches to PR. Interventions will be included if they match the following definition of PR: "the delivery of rehabilitation services which primarily takes advantage of respiratory training, exercise training, education, and behavioral changes". Respiratory rehabilitation-based exercise training interventions are also incorporated, such as respiratory muscle training, diaphragm training, and traditional Chinese exercise training. Telerehabilitation, face-to-face PR, in-person PR, supervised PR, and unsupervised PR will all be included in the study.
Exclude: Respiratory rehabilitation with the assistance of a respiratory trainer. Exercise training alone that is not based on respiratory rehabilitation, such as endurance and aerobic training.
## Types of outcomes
Include: Outcomes that may be included are listed in [fig_ref] Table 1: Outcomes and Measurements Will be Included in the Review [/fig_ref]. Exclude: Studies that do not contain data on either of the above outcomes.
# Search methods
## Search strategy electronic searches
The review will involve searching the PubMed, Web of Science, Cochrane Library, EBSCO, and CNKI databases from December 2019 to July 2022. Medical subject heading (MeSH) terms will be adopted to search the database, mainly including "COVID-19" AND "Rehabilitation" AND "Pulmonary" AND "Trial". Each database will use subject words and free words to search. All searches will be limited to the English and Chinese languages, but no geographical restrictions will be applied. [fig_ref] Table 2: Search Strategy for PubMed [/fig_ref] lists the search strategies and search words of the PubMed database in detail. The search strategies of other databases will convert the logical operators and search fields accordingly. Different search strategies will be used for different language databases. PubMed is used as an example, and the specific search strategy is detailed in [fig_ref] Table 2: Search Strategy for PubMed [/fig_ref]. Additional Resources
We will also manually search the following resources to identify ongoing or completed clinical trials, such as Google Scholar (http://scholar.google.com), Baidu Scholar (http://xueshu.baidu.com/), Clinical Trials (http://www.clinicaltrails.gov), and the China Clinical Trials Registry (http://www.chictr.org/cn/).
## Study selection
EndNote (Version X9, Clarivate, Philadelphia, PA, USA) will be used for import, grouping, deduplication, adding full text, and so on. After importing the reference into EndNote, we will first filter out the literature by comparing the title, author, year, journal name, volume, page number, and other information to remove duplicate references. Two authors (Y.G., C.N.) trained in evidence-based medicine will conduct a one-by-one review of the titles and summaries of the reference books according to the inclusion and exclusion criteria and remove the references that obviously do not meet the inclusion criteria into the exclusion folder. They will add the full text of the bibliographies, meeting the requirements of the preliminary hearing. The two authors (Y.G., Y.F.) will read the contents of the research design one by one in the full text of the literature, remove the literature that does not meet the requirements to the exclusion folder, and record the reasons for exclusion in Excel. In the screening process, the solution should be discussed first if there is any disagreement. If disagreements still exist, the author (H.H.) will assist in judgment. The literature screening process is shown in [fig_ref] Figure 1: The PRISMA flow diagram of literature screening process [/fig_ref].
## Data extraction
Data will be extracted by two independent reviewers (YG and HH) according to an agreed data extraction form (Appendixes A-E). Any disagreements will be resolved by consensus.
## Data extraction
Data will be extracted by two independent reviewers (YG and HH) according to an agreed data extraction form (Appendices A-E). Any disagreements will be resolved by consensus.
Data to be extracted mainly include bibliographic information (e.g., author, title, year, publication), demographics (e.g., sex, age, sample size), groups (e.g., group name, group description, intervention frequency, intensity, duration, co-interventions), and outcomes (e.g., time points measured/reported, definition, unit of measurement, imputation of missing data). The specific data to be extracted are shown in [fig_ref] Table 3: Variables to be extracted at the full-text stage [/fig_ref]. Authors will be contacted if data are missing or unclear in the selection of articles. If sufficient information cannot be obtained in this way, we will analyze the available data and the potential impact of insufficient data on the study results in the discussion.
## Quality appraisal
The identified trials will be assessed independently by two reviewers (Y.G. and C.N.). For case-control and cohort studies, the risk of bias will be assessed using the Newcastle-Ottawa Scale (NOS). Stars are awarded for each domain, which allows the study to be graded into poor, fair, or good quality. For randomized controlled trials, quality assessment will be carried out with the Cochrane Risk of Bias Tool [bib_ref] RoB 2: A revised tool for assessing risk of bias in randomised..., Sterne [/bib_ref]. Any disagreements will be reviewed by the third reviewer (Y.F.) and resolved by discussion among all reviewers. If the information about the risk of bias in the clinical trial is unclear, we will try to contact the author by email.
## Data synthesis
It is anticipated that the included studies will vary significantly in type and method, though meta-analyses will be conducted if data are available or situations allowed. According to the heterogeneity between studies, the method of data analysis and synthesis will be determined. When we find obvious heterogeneity in the combined data, we will use subgroup analysis, sensitivity analysis, and publication bias to investigate the source of the heterogeneity.
# Meta-analysis
If the heterogeneity between the included studies is within the acceptable range, a meta-analysis of the study results will be conducted. A Chi-square test (χ 2 ) and I2 will be used to analyze the heterogeneity between the clinical trials. If p > 0.1, I2 ≤ 50%, it indicates that the heterogeneity between the clinical trials is within the acceptable range, and a fixed-effect model will be used to analyze the data. If p ≤ 0.1, I2 > 50%, indicating that the heterogeneity between clinical trials is considerable, subgroup analysis will be needed to identify the source of heterogeneity, and the random-effect model will be used to analyze the data. RevMan5.3 software (Cochrane Collaboration, Oxford, UK) will be used to synthesize the study data. Mean difference (MD) or standardized mean difference (SMD) and 95% confidence interval (CI) are used to describe the effect size of continuous data (e.g., spirometry/mL, six-minute walk/min, VO2max/(mL/min/kg), strength/kg, blood pressure/kPa, questionnaire scale scores). The Z test judges the effect size, and it has statistical significance when p ≤ 0.05. The data synthesis results will be presented in the form of forest plots.
## Descriptive review
If the heterogeneity between the included studies is significant, we will make a descriptive analysis of the study results. A Chi-square test (χ 2 ) and I2 will be used to analyze the heterogeneity between the clinical trials. If p ≤ 0.1, I2 > 75%, it indicates that the heterogeneity between the clinical trials is very significant, if the heterogeneity is substantial, we will make a narrative qualitative summary. Study comparisons will be grouped (e.g., severity of disease, sex, age, types of intervention) to answer the research questions and findings will be synthesized based on outcomes. The characteristics of included studies will be presented in a narrative format, as recommended by PRISMA.
# Subgroup analysis
If there is a certain degree of heterogeneity between included clinical trials, subgroup analysis can be used to determine the source of the heterogeneity. The subgroup analysis will be conducted according to age, sex, country, severity of the disease, intervention length, different outcome measurement time points, and different follow-up time points.
# Sensitivity analysis
The purpose of sensitivity analysis is to evaluate the bias variables by eliminating each study one at a time. It will compute the sensitivity of each study over the whole project to determine whether an individual study has a substantial influence on the outcomes.
## Assessment of publication biases
We will assess publication bias by funnel plots for asymmetry when at least 10 trials are available [bib_ref] Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised..., Sterne [/bib_ref]. If the plot is asymmetric and there is no inverted funnel form, there may be publication bias. The causes might be connected to the small sample size, allocation concealment, and insufficient blind method implementation.
# Ethical considerations
As this study is only a systematic review and does not involve human or animal experimentation or personal privacy, ethical approval is not required.
# Results
The results of the systematic review will be published as a peer-reviewed article.
# Discussion
Among patients with pulmonary diseases, rehabilitation helps reduce dyspnea, increase exercise capacity, and improve health-related quality of life [bib_ref] statement: Key concepts and advances in pulmonary rehabilitation, Spruit [/bib_ref]. Therefore, rehabilitation might be a valuable treatment in patients with COVID-19. Patients with COVID-19 often have pathological features, such as pulmonary interstitial or alveolar edema and pulmonary inflammatory lymphoid infiltration, and are prone to acute respiratory distress syndrome (ARDS), causing lung injury [bib_ref] Pulmonary edema in COVID-19 patients: Mechanisms and treatment potential, Cui [/bib_ref]. Patients with severe COVID-19 may experience significant decreases in lung function, potentially requiring mechanical ventilation. Respiratory and circulatory failure are common causes of death among COVID-19 patients [bib_ref] Advanced pulmonary and cardiac support of COVID-19 patients: Emerging recommendations from ASAIO-A..., Rajagopal [/bib_ref]. Although the mechanisms of COVID-19-induced lung injury are still being elucidated, pulmonary rehabilitation (PR) is necessary at any stage in the course of the COVID-19.
Integrated into the individualized treatment of the patient, PR is designed to reduce symptoms, optimize functional status, increase participation, and reduce healthcare costs through stabilizing or reversing systemic manifestations of the disease [bib_ref] American thoracic society/European respiratory society statement on pulmonary rehabilitation, Nici [/bib_ref]. PR has emerged as a cost-effective intervention for managing chronic lung disease. As evidence has shown, PR improves the 6 min walking distance (6MWD), QoL, and respiratory symptoms in patients with chronic obstructive pulmonary disease (COPD) and interstitial lung disease (ILD) [bib_ref] Effectiveness of home-based pulmonary rehabilitation for patients with chronic obstructive pulmonary disease:..., Liu [/bib_ref] [bib_ref] Long-term evaluation of home-based pulmonary rehabilitation in patients with COPD, Grosbois [/bib_ref]. However, not all patients with pulmonary disease benefit from PR to the same degree. In both COPD and ILD patients, studies have shown that PR is not responsive to disease rehabilitation [bib_ref] Clinical predictors of the efficacy of a pulmonary rehabilitation programme in patients..., Vagaggini [/bib_ref] [bib_ref] Success in pulmonary rehabilitation in patients with chronic obstructive pulmonary disease, Scott [/bib_ref].
Currently, studies focusing on PR in patients with COVID-19 are few. Many of the difficult questions about PR have not been answered, such as whether PR is appropriate for all COVID-19 patients, at which stage in the course of COVID-19 should PR be administered, how specifically does PR work against COVID-19 compared to other types of pneumonia, and so on. Although there are some data showing the impressive benefits of PR participation, the effectiveness of rehabilitation has not been systematically summarized yet.
Some international rehabilitation associations provided PR guidelines for COVID-19 patients. For example, the Chinese Association of Rehabilitation Medicine has established different PR programs for patients with light, moderate, severe, and worse symptoms during hospitalization and discharge [bib_ref] Chinese Association of Rehabilitation Medicine; Respiratory Rehabilitation Committee of Chinese Association of..., Zhao [/bib_ref]. The PR guideline from Turkey emphasizes that exercise training is the most effective and compulsory method in achieving the goals of PR [bib_ref] Pulmonary rehabilitation principles in SARS-CoV-2 infection (COVID-19): A guideline for the acute..., Aytür [/bib_ref]. Carda et al. [bib_ref] The role of physical and rehabilitation medicine in the COVID-19 pandemic: The..., Carda [/bib_ref] made suggestions for the PR program according to their clinical experience of COVID-19, and they strongly advised the implementation of teleconsultation and telerehabilitation devices and suggested that patients who had kept negative results of COVID-19 for more than 7 days after their first diagnosis could be given access to PR. The 2020 British Thoracic Societyalso updated the PR guideline for COVID-19 from fatigue, mood disturbances, cognitive function, and support to reopening work. However, most of the current PR guidelines adopted for patients with COVID-19 are based on the experience gained during the response to the SARS epidemic in 2004. Therefore, more evidence is needed to demonstrate the impact of PR and different PR designs on COVID-19 patients. This protocol will represent the first systematic review and meta-analysis on effects of PR in patients with COVID-19. A deep understanding of PR can alleviate the COVID-19 crisis and optimize COVID-19 resource allocation. Therefore, the conclusions of the systematic review will have direct practical implications and clinical relevance.
# Limitations
There will be some limitations to this systematic review. First, the outbreak of COVID-19 is sudden. It is impossible to formulate and implement a large-sample randomized controlled trial in a short period, and the quality of clinical trials may not be high enough, which affects the quality of evidence to a certain extent. Secondly, the different types, frequency, intensity, and duration of PR may cause clinical heterogeneity. Thirdly, this review will only include Chinese and English studies from the literature, which may lead to selective bias. Although there are some limitations, the team members will still carry out this review to provide some references and suggestions for clinical decision-making and further clinical research.
# Conclusions
Pulmonary rehabilitation (PR) is an effective tool of rehabilitation interventions for COVID-19 patients. The findings of this systematic review and meta-analysis can help physical therapists and the general public in actively addressing the challenges posed by COVID-19. Data Availability Statement: Data sharing is not applicable to this article as no new data were created or analyzed in this study protocol.
# Appendix c. methods
## Descriptions as stated in report/paper location in text
Country/Countries
[fig] of 14, Figure 1: Int. J. Environ. Res. Public Health 2022, 19, x FOR PEER REVIEW 6 The PRISMA flow diagram of literature screening process. [/fig]
[fig] Figure 1: The PRISMA flow diagram of literature screening process. [/fig]
[fig] Author: Contributions: Contributions to concept and design: Y.G., J.Y., H.H., C.N. and Y.F.; Data analysis and interpretation: Y.G., C.N., Y.F., X.D. and Y.W.; Drafting the article: Y.G. and Y.F.; Revising critically: Y.G. and J.Y. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research was funded by the National Social Science Foundation of China, grant number 18BTY100. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. [/fig]
[table] Table 1: Outcomes and Measurements Will be Included in the Review.Abbreviation: FVC, forced vital capacity; FEV1, 1 s forced expiratory volume; PEF, peak expiratory flow; VC, vital capacity; VO2max, maximal oxygen uptake; peak VO2, peak oxygen uptake; HGS, handgrip strength; 1RM, one-repetition maximum; BMI, body mass index; MM, muscle mass; LBM, lean body mass; SPOZ, oxygen content of blood. [/table]
[table] Table 2: Search Strategy for PubMed. COVID-19[Mesh] OR COVID-2019[tiab] OR COVID19[tiab] OR COVID-19 pandemic[tiab] OR COVID-19 virus disease[tiab] OR SARS-CoV-2 infection[tiab] OR 2019 novel coronavirus infection[tiab] OR coronavirus disease 2019[tiab] OR 2019-nCoV infection[tiab] OR COVID-19 virus infection[tiab] #2 severe acute respiratory syndrome coronavirus 2[Mesh] OR SARS-CoV-2[tiab] OR 2019-nCoV[tiab] OR 2019 novel coronavirus[tiab] OR COVID-19 virus[tiab] OR COVID19 virus[tiab] OR coronavirus disease 2019 virus[tiab] OR Corona Virus Pulmonary[Mesh] OR pulmonary rehabilitation[tiab] OR lung rehabilitation[tiab] OR pulmonary rehabilitation exercise[tiab] OR pulmonary recovery[tiab] OR lung recovery[tiab] OR lung rehabilitation exercise[tiab] OR pulmonary rehabilitation therapy[tiab] [/table]
[table] Table 3: Variables to be extracted at the full-text stage. [/table]
|
Effect of low salicylate diet on clinical and inflammatory markers in patients with aspirin exacerbated respiratory disease – a randomized crossover trial
Background: Aspirin-exacerbated respiratory disease (AERD) is characterized by eosinophilic rhinosinusitis, nasal polyposis, and bronchial asthma, along with the onset of respiratory reactions after the ingestion of nonsteroidal anti-inflammatory drugs (NSAIDs) or acetylsalicylic acid (ASA). In addition to the therapeutic routines and surgical options available, a low dietary intake of food salicylate has been suggested as adjunctive therapy for this condition. This study aimed to assess the influence of a short-term low salicylate diet on inflammatory markers in patients with AERD and whether that would result in symptomatic improvement. Methods: Prospective study with randomization to either a high or low salicylate diet for 1 week, followed by cross-over to the other study arm. Participants were asked to record their dietary salicylate for each week of the study. Urinary creatinine, salicylate and leukotriene levels were measured at the time of recruitment, end of week one and end of week two and the SNOT-22 questionnaire was filled out at the same time points.Results: A total of seven participants completed the study. There was no statistical difference in the urinary salicylate and leukotriene levels between the two diets; nevertheless, participants on low salicylate diet reported improved SNOT-22 symptoms scores (p = 0.04), mainly in the rhinologic, ear/facial, and sleep dysfunction symptom domains. In addition, these last two domains outcomes were more significant than the minimal clinically important difference. Conclusions: A short-term low salicylate diet may not result in biochemical outcomes changes but seems to provide significant symptomatic relief for patients with AERD. Trial registration: NCT01778465 (www.clinicaltrials.gov)
# Background
Aspirin exacerbated respiratory disease (AERD) is an acquired condition with a median age of onset around 30 years, that consists of bronchial asthma, chronic rhinosinusitis with nasal polyposis and hypersensitivity to acetylsalicylic acid (ASA). In AERD patients, acetylsalicylic acid (ASA) intolerance is a non-allergic hypersensitivity reaction without immunoglobulin E (IgE) involvement. These patients have recalcitrant chronic rhinosinusitis with nasal polyposis (CRSwNP), often requiring frequent surgical intervention, systemic corticosteroid therapy and aspirin desensitization.
The pathophysiology of AERD has not been entirely elucidated; however, it is thought that increased expression of 5-lipoxygenase and leukotriene C4 synthase genes which results in downstream inflammatory activation, including mast cells and eosinophils. Altered metabolism of arachidonic acid results in an imbalance of prostaglandins (PG) and leukotrienes (LT), eliciting increased levels of cysteinyl leukotrienes (CysLT) and decreased levels of prostaglandin E2 (PGE2).
Patient avoidance of non-selective COX inhibitors, medical and surgical procedures to address the underlying bronchial asthma and CRSwNP and ASA desensitization therapy, are standard for AERD management.
Dietary modifications have been used as an alternative to modulating the inflammatory status in patients with chronic diseases, such as CRS. It is estimated that daily intake of dietary salicylate can, occasionally, exceed the equivalent blood salicylate level of patients who take low-dose aspirin (i.e. 81 mg/day); however, salicylate bioavailability in food still warrants further investigation. Nonetheless, the potential influence of dietary components for AERD patients should not be underestimated. Many of these patients are prone to present upper and lower-airway hypersensitivity reactions to alcoholic beverages, in particular beer and red wine. Also, extracts of plants such as poplar, myrtle, willow and meadowsweet, which are rich in salicylates, have been used for millennia due to analgesic and antipyretic effects. Willow and meadowsweet were the first sources of isolated salicylic acid, early in the nineteenth century.
Our group has previously demonstrated that adopting a low-salicylate diet can provide a significant reduction in sinonasal symptoms, along with improved nasal endoscopy scores, in comparison to regular dietary intake. Despite the encouraging results of modifying dietary salicylate intake, the biochemical effect of this intervention has not yet been evaluated with subjective sinonasal symptom correlation.
Therefore, the objective of this study was to evaluate the inflammatory status of AERD patients under both low and high salicylate diets, testing the hypothesis that a low salicylate diet would decrease urinary salicylate levels and inflammatory markers, as well as improve the participants' subjective sinonasal symptoms.
# Materials and methods
This study was a prospective randomized, cross-over study assessing biochemical changes in the levels of urinary salicylates and urinary cysteinyl leukotrienes (CysLT), in relation to dietary salicylate intake in subjects with AERD. This was approved and reviewed by the research ethics board at Western University (London, ON, Canada -REB #103330) and was registered with clinicaltrials.gov (registration #01778465). The Sino-nasal Outcome Test (SNOT-22), categorized into five symptom domains, was used to assess, as a secondary outcome, the participants' symptoms.
Adults, 18 years and older, were assessed in the Department of Otolaryngology -Head and Neck Surgery at Western University. Patients with a history of surgery for chronic rhinosinusitis with nasal polyposis, confirmed asthma, and a documented history of a significant respiratory sensitivity reaction to ASA or ibuprofen were invited to participate in the study. Informed consent was obtained from all individual participants before inclusion. Exclusion criteria included a history of cystic fibrosis, immunodeficiency, recent endoscopic sinus surgery (within the past 6 months), and treatment with a course of oral corticosteroids within 3 months previous to enrollment. None of the enrolled patients were taking leukotriene receptor antagonists.
On the first clinic visit, participants were randomized to high (HS) or low (LS) salicylate diets, using an online randomization tool. They were briefed on a salicylate diet regarding the salicylate contents of foods, and a handbook was given to them comprising a detailed list of the salicylate contents of different foods (Supplement 1). The low-salicylate diet consisted of avoiding foods with high (> 0.5 mg/portion) salicylate content and trying to consume foods with low (0.01 to 0.09 mg/portion) salicylate content for 1 week. Conversely, the high salicylate diet consisted of consuming mainly foods from the high salicylate group during the high salicylate diet week. After 1 week of diet (1st assigned one), patients crossed over to the opposite diet for the following week.
Outcome variables were assessed at the enrollment visit, 1 day after the completion of the first week (1st diet), and 1 day after the second week (2nd diet). At each of these time points, participants filled out the SNOT-22 questionnaire and had urinary samples collected. Additionally, participants kept a diet diary, in which they included a record of their everyday salicylate dietary component intake to help them keep track of their salicylate diet intake and comply with their assigned diets.
## Outcomes
## Urinary biomarkers
The urine samples were aliquoted into five different 1 ml aliquots and stored at − 80°C until processed. Urinary creatinine (Cr) was determined using the standard core laboratory protocol (enzymatic method). After thawing, samples were centrifuged at 3000 rpm for 10 min at 4°C. Cysteinyl leukotrienes were measured in the supernatant of each sample using an ELISA kit (Cayman Chemical Co., Ann Arbor, MI) per manufacturer's protocol. Briefly, 50 μL of supernatant, CysLT acetylcholinesterase conjugated tracer, and CysLT monoclonal antibody were added to each well of a 96-well plate. Following overnight incubation at 4°C, the wells were emptied and washed five times with wash buffer. To develop the assay, 200 μL of Ellman's Reagent was added to each well, and the plate was placed on an orbital shaker in the dark for 90 min. The absorbance was measured at a wavelength of 412 nm. Sample concentrations were calculated according to the manufacturer guidelines.
Urinary Salicylic and Salicyluric acids were measured based on the methods of Baxter et al..
## Sino-nasal outcome test (snot-22)
Study participants responded to a subjective symptom outcome test (sino-nasal outcome test; SNOT 22), a 22-item validated questionnaire that quantifies the severity of sinonasal symptoms (Washington University, St. Louis, MO). Each one of the 22 items in the test are graded from zero to five; the higher the score, the more severe the disease; additionally, it is subdivided into five distinct symptom domain scores: the rhinologic symptoms domain (range: 0-30), extra-nasal rhinologic symptoms domain (range: 0-15), ear/facial symptoms domain (range: 0-25), psychological dysfunction domain (range: 0-35), and the sleep dysfunction domain (range: 0-25).
# Data analysis
The analysis was conducted using SPSS for Windows (version 20.0, Chicago, IL). In this study, the variables are presented as median and interquartile ranges, and those values were used for the hypothesis analysis. Three time points were predetermined for outcomes analysis (e.g. baseline measurements, post HS, and post LS). Furthermore, the effect of the interventions on each variable (Δ) was calculated by subtracting the individual baseline values of each participant from their Post value (i.e. post -baseline). Individual results were then compiled, and a group median was created.
Post diet outcome variables values were compared to their baseline values using the Wilcoxon signed-rank test in order to mimic the comparison between the group intervention to a regular diet. Moreover, the comparisons between the effect of each intervention were made by using the Mann-Whitney test. Both test results are expressed in terms of significance (2-tailed null hypothesis and significance levels p < 0.05) and effect size r, utilizing the equation "r = Z/ √N".
# Results
A total of ten potential participants were screened for this study. After consideration, three of them declined participation before the first week of diet; therefore, seven participants completed this study (5 men and 2 women), average age 55 (SD13.1). The study flow chart details the sequence of the interventions and evaluations throughout the study.
## Urinary biomarkers
Except for creatinine levels, urinary biomarkers of both groups, did not yield statistically significant differences over the study. In the intragroup analysis (i.e. post x baseline), HS presented a significant increase in the urinary creatinine, whereas the LS remained unchanged.
The data variance within group participants was considerably high on both groups, especially on the CysLT deltas. The expression of that outcome, case by case, shows that most of participants (5/7) presented a reduction in the CysLT concentration, regardless of their dietary intake. Despite the absence of statistical significance, the overall CysLT reduction was higher in the LS participants, compared to the HS.
## Sino-nasal outcome test (snot-22)
The participants' nasal symptoms after 1 week on LS demonstrated a statistically significant reduction of 22 points (p = 0.043; effect size r = − 0.53), compared to the baseline, while the group HS did not present significant improvement, on the intragroup analysis. Furthermore, an intergroup analysis of both groups' deltas (e.g. post -baseline) medians values showed that LS reduced significantly 10 points on the nasal symptoms test, whereas HS presented an increase in this variable (p = 0.013; effect size r = − 0.66). Expressed case by case, the difference between interventions on the SNOT 22 outcome is clearly identified where the LS was more effective on reducing the severity of the sinonasal symptoms, except for one participant, in comparison to the HS.
A segmented analysis of the nasal symptoms outcome into domains is presented in. All five domains contributed to the overall result reported on the paragraph above; however, only the Rhinologic (HS, median 2, IQR 7; LS, median − 2, IQR 16; p = 0.017, effect size r = − 0.63) and the ear/facial (HS, median 4, IQR 6; LS, median − 3, IQR 15; p = 0.02, effect size r = − 0.62) symptoms domains were statistically different between both groups.
The analysis of each domains' results regarding the minimal clinically important difference (MCID), according to the literature, showed that on top of the effect presented by the LS group, two domains overcame their respective MCID. The LS group median on the ear/face symptoms domain was − 3 (IQR 15), whereas the MCID for this domain is 1.6, and the sleep disfunction was − 2 (IQR 12), while its MCID is 1.5.
# Discussion
Our results showed that consuming a low salicylate diet for at least 1week resulted in a symptomatic improvement of subjective sinonasal symptoms in patients with AERD. However, the urinary leukotrienes did not correlate with these findings. Salicylates are ubiquitously present in modern diet and strict adherence to diet long-term may be challenging and pose difficulties in achieving a balanced diet. Some patients have commented that they find a less stringent version of the diet more achievable e.g. avoiding foods with high/very high salicylate content. Another successful strategy has been to follow the strict diet as an elimination diet, and add in desired foods after a month to evaluate impact on symptoms. Despite these limitations, a low-salicylate diet could be an adjunctive treatment for patients who have increased frequency and intensity of symptoms and are highly motivated to pursue alternative nonpharmacologic therapy. AERD is among the more challenging presentations of CRSwNP to treat. Most patients require multiple surgeries and multifaceted medical management including topical saline irrigations, topical and systemic corticosteroids and aspirin desensitization. However, apart from requiring expertise in initiating the protocol, up to 30% of the patients on aspirin desensitization protocols report gastrointestinal side effects and discontinue therapy. AERD patients often require multiple revision surgeries with some patients undergoing ten times as many surgeries as non-AERD patients with CRS. Additionally, the interval between surgeries is also decreased for AERD patients compared to non-AERD CRS patients. Monoclonal antibody therapy targeting type II inflammation is a promising development for use in CRSwNP and will change our treatment paradigm. These medications are, however, quite expensive and may not be an option for all patients.
Wood et al.published a systematic review, which showed the average daily dietary salicylate intake for men is 4.42 mg/day and for women is 3.16 mg/day,with even higher values in vegetarians.
Interestingly, in the current study, urinary cys-LTs levels in both low-salicylate and high salicylate diet groups were reduced compared to baseline. Thus, it is possible that the participants already had a significantly high baseline salicylate intake; therefore, the effort to create an enhanced salicylate intake (i.e. high salicylate group), might have been compromised by participants' pre-diet salicylate consumption levels. Additionally, participants might have been unable to alter their diet effectively, despite being advised to increase their dietary salicylate content when on high salicylate diet. Notably, six of the 7 patients continued on a modified low-salicylate diet after the study, as they found significant benefit from it, and were still continuing to follow the diet to some degree 6 months after participation in the research study.
Previous studies have shown that restricting dietary salicylates as an adjunct to treat AERD patients has demonstrated significant improvement in their symptoms. Our group's previous multicenter studywith 30 participants undergoing a low salicylate diet for 6 weeks presented a median reduction of 15 points in the SNOT 22 score. In the present study, with only 1 week under a low salicylate diet, the participants demonstrated an overall reduction in the SNOT-22 scores of 10 points. Both a placebo and Hawthorne effect are realistic possibilities for this effect, but continued observance of the diet (with some allowances) is evidence to the contrary.
The symptomatic improvement presented in our study overcomes the minimal clinically important difference. Asterisks express the Mann-Whitney test with a 2-tailed significance ≤ .05 (MCID) of 9 points, described by Chowdhury et al.in patients with chronic rhinosinusitis undergoing continued appropriate medical therapy. On the other hand, Phillips et al.describe a MCID for medically managed CRS patients of 12, which is higher than the overall median value on the LS group. In this regard, none of these studies specify patients with AERD, making comparison to our study results a challenge. The symptomatic domain analysis of SNOT 22 is performed to segment the quality of life analysis into distinct sections, which can work as references for treatment guidance or therapy effectiveness analysis. For patients with CRS, the psychological and sleep dysfunction domains demonstrate significant associations with productivity loss. As such, there may be potential for the low salicylate diet to help AERD patients improve productivity and their overall quality of life.
The short duration of the diet was chosen based on the half-life of CysLTs and the difficulty with adherence to the restrictiveness of the low salicylate diet. It was hoped that by choosing 1 week, patients would be able to provide maximal adherence to both the HS and LS diets. Additionally, the half-life for salicylates is somewhere between 2 and 30 h. Therefore, a washout period was not included as measurements of both Cys-LTS and SNOT-22 was performed at the end of the week period. It is conceivable that the time period of investigating the diet may have been affected by the half-life of salicylates. The two previous studies investigating salicylate diet in AERD patients had a 6 week time course on the diet, compared to 1 week in the current study. Given the presented subjective symptom outcomes, 1 week under a low salicylate diet already provides AERD patients with overall symptomatic improvement. Measurement of urinary LTS is possible using a variety of techniques. As mentioned, our study utilized an ELISA based methodology as per Baxter et al.. Other reports suggest that utilizing a liquid chromatography followed by tandem mass spectrometry and normalizing to urinary creatinine levels may be a clinically valid technique in the assessment of urinary leukotrienes as a biomarker of inflammation in diseases such as AERD. It is possible that our results would differ using the aforementioned methodology.
Lastly, the low number of patients is also a limitation of this study, but on top of its design as a proof-ofconcept study, recruitment was challenging given the need for three weekly visits after enrollment, and agreement to strictly adhere to the salicylate diet. Future studies should continue to investigate the mechanistic processes of dietary salicylates and their role in AERD. Increased control and recording of food salicylate intake amount, along with periodic measurements of inflammatory biomarkers throughout the dietary period, and a follow-up time after the diet discontinuation, should be included as part of further studies.
# Conclusion
This study confirms the previous findings that a lowsalicylate diet decreases AERD patients' symptoms as measured by SNOT-22. There was reduction in the urinary cys-LTs levels in both groups, with some patients showing a reduction in Cys-LTs while on a LS diet. This was a proof of concept study again demonstrating subjective symptom improvement with reduction of nutritional salicylate intake in the treatment of AERD patients.
## Supplementary information
The online version contains supplementary material available at https://doi. org/10.1186/s40463-021-00502-4.
Additional file 1. Salicylate-Free Diet Food Guide.
|
Related but different: distinguishing postpartum depression and fatigue among women seeking help for unsettled infant behaviours
Background: A growing body of evidence in relatively healthy populations suggests that postpartum depression and fatigue are likely distinct but related experiences. However, differentiating depression and fatigue in clinical settings remains a challenge. This study aimed to assess if depression and fatigue are distinct constructs in women with relatively high fatigue and psychological distress symptoms attending a residential program that assists with unsettled infant behaviour. Methods: 167 women (age: M = 34.26, SD = 4.23) attending a private residential early parenting program completed the Depression Anxiety Stress Scale (DASS21-D), Fatigue Severity Scale (FSS) and self-report sleep variables before program commencement. Confirmatory Factor Analysis examined the associations between depression and fatigue latent factors.Results:A two-factor model of separate but related depression and fatigue constructs provided a significantly better fit to the data than a one-factor model of combined depression and fatigue (p < .001). In the two-factor model, the depression and fatigue latent factors were moderately correlated (.41). Further predictive utility of this two-factor model was demonstrated as both depression and fatigue factors were independently predicted by worse self-reported sleep efficiency.Conclusions: This study provides empirical evidence that for women attending a clinical service with relatively high fatigue and psychological distress, postpartum depression and fatigue remain separate but related experiences. These findings suggest that in women seeking clinical support in the postpartum period, both depression and fatigue need to be carefully assessed to ensure accurate diagnoses, and (b) whilst depression intervention may improve fatigue, targeted fatigue intervention may also be warranted.
# Background
Maternal depression and fatigue symptoms are both prevalent across the first two years after giving birth, with 10 to 20% reporting depressive symptoms and 40 to 60% reporting fatigue symptoms [bib_ref] Mothers' health and work-related factors at 11 weeks postpartum, Mcgovern [/bib_ref] [bib_ref] What happens when fatigue lingers for 18 months after delivery?, Parks [/bib_ref] [bib_ref] Clinical phenotypes of perinatal depression and time of symptom onset: analysis of..., Putnam [/bib_ref]. This may be at least partly due to the under-recognized nature of women's caregiving work and the potential for occupational fatigue associated with the demands of infant caregiving. Within this context, depression and fatigue can share complex bi-directional relationships. Fatigue is one of the most common symptoms of depression and part of the diagnostic criteria for depressive disorders [bib_ref] Assessment of postpartum depressive symptoms: the importance of somatic symptoms and irritability, Williamson [/bib_ref]. Several postpartum studies have reported significant positive univariate associations between depressive and fatigue symptoms within the first 32 weeks postpartum [bib_ref] Meta-analysis of the predictive factors of postpartum fatigue, Badr [/bib_ref]. Depression and fatigue may also predict each other over time: across the first four years postpartum, depressive symptoms have been shown to predict future fatigue levels, and vice versa [bib_ref] Fatigue as a predictor of postpartum depression, Bozoky [/bib_ref] [bib_ref] The impact of fatigue on the development of postpartum depression, Corwin [/bib_ref] [bib_ref] I didn't know it was possible to feel that tired": exploring the..., Giallo [/bib_ref].
## Depression and fatigue in community samples of postpartum women
Given this close relationship between depression and fatigue, there has been a debate as whether they are distinct phenomena [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Postpartum fatigue: clarifying a concept, Milligan [/bib_ref]. In relatively healthy women in the postpartum period, evidence points to depression and fatigue being two different constructs [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. A qualitative study found that women with depressive symptoms reported symptoms such as feelings of emptiness and guilt that were not endorsed by non-depressed but fatigued women [bib_ref] A depressive symptoms responsiveness model for differentiating fatigue from depression in the..., Runquist [/bib_ref]. This is consistent with studies that identified clusters of women with high fatigue but not depressive symptoms [bib_ref] Trajectories of depressive symptoms and fatigue among postpartum women, Kuo [/bib_ref] [bib_ref] Maternal fatigue and depression: identifying vulnerability and relationship to early parenting practices, Wade [/bib_ref]. Two studies examined specific symptom constructs of postpartum depression and fatigue using confirmatory factor analysis (CFA) in community populations within the first year postpartum [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] and one study also at four years postpartum [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] : both studies concluded that a two-factor model of related but separate latent factors of depression and fatigue provided a better fit to the data than a single combined factor at all time-points.
What about women experiencing elevated postpartum fatigue and distress?
The differentiation of depression and fatigue symptoms has not been well examined in a clinical setting. Findings among healthy women may not generalise to those with elevated psychological distress and fatigue symptoms seeking clinical care. Depression and fatigue share many common features that can make them difficult to differentiate in clinical settings [bib_ref] A depressive symptoms responsiveness model for differentiating fatigue from depression in the..., Runquist [/bib_ref] [bib_ref] Distinguishing fatigue and depression in patients with cancer, Jacobsen [/bib_ref]. For example, they may share similar indicators among women seeking clinical help, such as irritability, feeling overwhelmed, and impaired physical and cognitive functioning [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] [bib_ref] A depressive symptoms responsiveness model for differentiating fatigue from depression in the..., Runquist [/bib_ref] [bib_ref] Persevering through postpartum fatigue, Runquist [/bib_ref] [bib_ref] Measurements of sleepiness and fatigue, Shahid [/bib_ref]. Depression and fatigue can also share underlying causes such as sleep disturbance, physiological changes, or situational factors (e.g., unsettled infant behaviours; [bib_ref] Postpartum fatigue: clarifying a concept, Milligan [/bib_ref] [bib_ref] Persevering through postpartum fatigue, Runquist [/bib_ref] [bib_ref] Crying babies, tired mothers: what do we know? A systematic review, Kurth [/bib_ref]. Together, these similarities in presentation and causes present a challenge in differentiating depression and fatigue and can lead to potential over-diagnosis of fatigue as depression [bib_ref] I didn't know it was possible to feel that tired": exploring the..., Giallo [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref].
While there is evidence that fatigue and depression are related but separate constructs in healthy populations, it is possible that as depression and fatigue levels increase, they become less distinct and harder to differentiate [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Distinguishing fatigue and depression in patients with cancer, Jacobsen [/bib_ref]. High fatigue symptoms may reduce self-care behaviours and pleasurable activities, which may contribute to low mood [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref]. Conversely, it is also possible that distinct features of both depression and fatigue may become more apparent as symptom severity increases [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref].
Better understanding of the relationship between depression and fatigue in mothers at risk for both conditions is of critical importance to clinical services for both assessment and treatment. It is currently routine practice in many postpartum settings to screen for depressive disorders, but the assessment of fatigue is not routine [bib_ref] Assessment of postpartum depressive symptoms: the importance of somatic symptoms and irritability, Williamson [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. If symptoms of fatigue and depression largely overlap, existing short screening measures of depressive symptoms may be sufficient, and treatments for postpartum depression may help both sets of symptoms [bib_ref] Psychological treatment of postpartum depression: a meta-analysis, Cuijpers [/bib_ref]. However, if depression and fatigue remain distinct, then separate and more detailed assessment of both constructs could assist with more accurate diagnoses [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] , and targeted interventions for depressive and fatigue symptoms may be warranted [bib_ref] The effectiveness of a self-care intervention for the management of postpartum fatigue, Troy [/bib_ref].
## Current study
Unsettled infant behaviour occurs in~25% of infants, and refers to persistent and inconsolable infant crying, resistance to soothing, short sleep intervals and frequent night awakenings [bib_ref] Understanding and responding to unsettled infant behaviour, Fisher [/bib_ref]. Previous studies among women seeking support for unsettled infant behaviour have shown that many of these women experience elevated depression, anxiety, and fatigue symptoms [bib_ref] Postnatal demoralisation among women admitted to a hospital mother-baby unit: validation of..., Bobevski [/bib_ref] [bib_ref] Nature, severity and correlates of psychological distress in women admitted to a..., Fisher [/bib_ref] [bib_ref] Temperament and behaviour of infants aged 4-12 months on admission to a..., Fisher [/bib_ref] [bib_ref] Fatigue, wellbeing and parenting in mothers of infants and toddlers with sleep..., Giallo [/bib_ref]. Examining the profiles of these symptoms among women presenting at clinical services offering support for unsettled infant behaviour represents a unique opportunity to investigate whether depression and fatigue can be differentiated among women with elevated postpartum fatigue and psychological distress, and thereby address the previous lack of research in the relationship between depression and fatigue in clinical samples.
For this purpose, this study aimed to compare a one-factor model of combined depression and fatigue with a two-factor model of separate but related depression and fatigue. It was hypothesised that a two-factor model of related but separate depression and fatigue latent factors would provide a better fit for the data than a one-factor model, as is the case in community studies. To further demonstrate the predictive utility of the better fitting model, we explored the association(s) between the latent factor(s) and self-reported sleep efficiency given that sleep disturbance is related to both postpartum depression and fatigue [bib_ref] Trajectories of depressive symptoms and fatigue among postpartum women, Kuo [/bib_ref] [bib_ref] Sleep and mood during pregnancy and the postpartum period, Bei [/bib_ref] [bib_ref] The relationship between sleep and mood in first-time and experienced mothers, Calcagni [/bib_ref] [bib_ref] Sleep patterns and fatigue in new mothers and fathers, Gay [/bib_ref] [bib_ref] The relationship between sleep characteristics and fatigue in healthy postpartum women, Rychnovsky [/bib_ref] [bib_ref] Sleep, depression, and fatigue in late postpartum, Thomas [/bib_ref].
# Methods
## Study context and participants
Participants were women with infants aged up to 24 months who had been referred by medical practitioners to attend the Masada Private Hospital Early Parenting Centre (MPHEPC; Melbourne, Australia) for a residential early parenting program that assists with unsettled infant behaviour (for details on the intervention: [bib_ref] Understanding and responding to unsettled infant behaviour, Fisher [/bib_ref] [bib_ref] Nature, severity and correlates of psychological distress in women admitted to a..., Fisher [/bib_ref] [bib_ref] Treatment of maternal mood disorder and infant behaviour disturbance in an Australian..., Fisher [/bib_ref]. All women admitted to the MPHEPC between the 1st June 2015 and 12th October 2015 were invited to participate in the study with no exclusion criteria. Recruitment was carried out via advertisement on the MPHEPC website, a pamphlet in admission documentation, or by researchers on site. Participants completed a survey booklet on the first day of their admission before commencing treatment. The Avenue Hospital Research Ethics Committee (Trial 182) and Monash University Human Research Ethics Committee (CF15/1233) provided ethical approval. Written informed consent was obtained from all participants.
## Procedure
On the day of arrival to the MPHEPC, participants that expressed interest in the research project underwent an informed consent process and provided with a survey booklet that included the measures in this study. The survey booklet was returned to the researchers on site.
## Measures demographics
Maternal and infant demographics were collected through self-report and medical records extraction (see [fig_ref] Table 1: Maternal and infant demographics [/fig_ref].
## Depression
The Depression Anxiety Stress Scales Depression subscale (DASS21-D)is a widely used 7-item measure of depressive symptoms during the last week. The DASS21-D has adequate validity and reliability for postpartum populations [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Anxiety and stress in the postpartum: is there more to postnatal distress..., Miller [/bib_ref]. For this study Cronbach's alpha was .88, Omega was 0.89, and Greatest Lower Bound was 0.92 [bib_ref] The alpha and the omega of scale reliability and validity: why and..., Peters [/bib_ref].
## Fatigue
A revised five-item version of the Fatigue Severity Scale (FSS; [bib_ref] The fatigue severity scale. Application to patients with multiple sclerosis and systemic..., Krupp [/bib_ref] was used to measure the interference of fatigue on functioning. The FSS is a widely used scale of fatigue severity and interference in chronic illness populations. Similar to findings in other chronic illness populations [bib_ref] Psychometric properties of the fatigue severity scale: Rasch analyses of responses in..., Lerdal [/bib_ref] [bib_ref] Psychometric properties of the fatigue severity scale-Rasch analyses of individual responses in..., Lerdal [/bib_ref] [bib_ref] Rasch analysis of the fatigue severity scale in multiple sclerosis, Mills [/bib_ref] [bib_ref] Rasch analysis of the fatigue severity scale in Italian subjects with multiple..., Ottonello [/bib_ref] , the full nine-item FSS had several psychometric issues based on Rasch analysis [bib_ref] Postpartum fatigue: assessing and improving the psychometric properties of the fatigue severity..., Wilson [/bib_ref]. The revised version (FSS-5R) was calculated from Items 4 to 8 of the original FSS with simplified response options (recoded from 1,234,567 to 1,112,345) and had improved psychometric properties [bib_ref] Postpartum fatigue: assessing and improving the psychometric properties of the fatigue severity..., Wilson [/bib_ref]. Scale items are listed in Additional file 1: [fig_ref] Table 1: Maternal and infant demographics [/fig_ref]. For the FSS-5R, Cronbach's alpha was .87, Omega was .88, and Greatest Lower Bound was .89. Scores on the full FSS-9 were also used to calculate the proportion of women reporting fatigue severity above the suggested clinical cut-off (≥ 36) and for comparison with community studies in which the full scale was used.
## Sleep quality
Sleep Efficiency (SE) represents overall sleep quality, and was calculated as the percentage of self-report total sleep time against time spent in bed over the past week. SE ranges from 0% (low) to 100% (high efficiency).
The following well-validated instruments were also used to characterise the overall psychological distress reported by the sample: Depression Anxiety Stress Scale Anxiety (DASS21-A) and Stress (DASS21-S) subscales; Insomnia Severity Index (ISI; [bib_ref] Validation of the insomnia severity index as an outcome measure for insomnia..., Bastien [/bib_ref] ; and the 6 item version of the Irritability Depression Anxiety Scale -Irritability subscale (IDA-I; [bib_ref] Irritability: definition, assessment and associated factors, Snaith [/bib_ref].
# Data analysis
Data analysis was conducted in Mplus Version 7.4. First, one-factor models of depression using the DASS21-D and fatigue using the FSS-5R were assessed separately to confirm the uni-dimensionality of each scale. Then, one-and two-factor models for depression and fatigue were conducted and compared. In the one-factor model, all depression and fatigue items loaded onto a single latent variable representing a single combined construct (see [fig_ref] Figure 1: One-factor model of depression and fatigue [/fig_ref]. In the two-factor model, items from the DASS21-D and the FSS-5R were separately loaded onto their respective latent variables; the depression and fatigue latent variables were allowed to be correlated (see [fig_ref] Figure 2: Two-factor model of depression and fatigue [/fig_ref]. Thus, the two-factor model tests whether depression and fatigue are separate but correlated constructs [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. Confirmatory factor analysis (CFA) analysis was conducted using diagonally weighted least squares (WLSMV) estimation [bib_ref] Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally..., Li [/bib_ref]. The sample size (N = 167) had power of 0.80 to identify an effect size of 0.30and exceeded 10 observations per parameter [bib_ref] Overview of classical test theory and item response theory for the quantitative..., Cappelleri [/bib_ref]. The criteria for adequate model fit were: Chi-Square Test of Model Fit, Root Mean Square Error of Approximation (RMSEA) ≤ 0.05, Comparative Fit Index (CFI) and Tucker-Lewis Index (TLI) > 0.9, and Weighted Root Mean Square Residual (WRMR) < 1.0 [bib_ref] Structural equation modelling: Guidelines for determining model fit, Hooper [/bib_ref]. Comparison of model fit was carried out using the Chi-Square difference test for WLSMV estimation. Discriminant validity of the two-factor model was also assessed by examining the standardised pattern and structure coefficients of the two-factor model of depression and fatigue [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] [bib_ref] The importance of structure coefficients in structural equation modeling confirmatory factor analysis, Thompson [/bib_ref]. Discriminant validity is established if the difference in values of the pattern and structure coefficients is .2 or above [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. Finally, the predictive utility of the better fitting model was assessed by adding SE as the predictor of the latent factor(s). As missing data were low (< 5%), they were handled using pairwise deletion. No model modifications were made.
# Results
During the 19-week recruitment period, 167 of the 380 women admitted to the MPHEPC (44%) completed the study. Maternal and infant demographics and descriptive statistics for the DASS21-D, FSS-5R and SE are reported in [fig_ref] Table 1: Maternal and infant demographics [/fig_ref] and [fig_ref] Table 2: Descriptive statistics [/fig_ref]. Missing data were minimal: 1.1% for the DASS21-D, 0.3% for the FSS-5R, and 4.8% for SE. A correlation matrix of scale items is in Additional file 1: [fig_ref] Table 2: Descriptive statistics [/fig_ref]. Participants reported elevated depressive symptoms, with 50% reporting symptoms at or above the published cut off for mild depressive symptoms (DASS21-D ≥ 5). Fatigue symptoms were also elevated, with 87% of women reporting fatigue severity above the suggested clinical cut-off (≥ 36) for the full FSS-9; scores were higher than those reported in a postpartum community population. Scores on the other measures also point to an overall elevated level of distress in this sample. Forty-eight percent of women reported at least mild anxiety (DASS21-A ≥ 4), 64% reported at least mild stress (DASS21-S ≥ 8), and 46% reported insomnia symptoms in the clinical range (ISI ≥ 15).
# Confirmatory factor analysis
Separate models of the DASS21-D and FSS-5R both showed acceptable fit without modification (see [fig_ref] Table 3: Fit Indices for models [/fig_ref]. For both models, the standardised coefficients all significantly loaded onto the latent factor (all p-values < .001) and all exceeded .78 (see [fig_ref] Table 4 Pattern: Note [/fig_ref] , except for DASS21-D Item 5. Therefore, both scales uni-dimensionally assessed the respective constructs.
The one-factor model with depression and fatigue items loading onto a single construct had a poor fit (see [fig_ref] Table 3: Fit Indices for models [/fig_ref]. All items loaded significantly on the single latent factor (p < .001) and the standardised coefficients ranged from 0.49 to 0.86 (see [fig_ref] Table 4 Pattern: Note [/fig_ref] and [fig_ref] Figure 1: One-factor model of depression and fatigue [/fig_ref]. The two-factor model of depression and fatigue as related but separate latent factors provided an acceptable and improved fit (see [fig_ref] Table 3: Fit Indices for models [/fig_ref]. The standardised coefficients for fatigue items on the fatigue latent factor and depression items on the depression latent factor were all significant (p < .001) (see [fig_ref] Table 4 Pattern: Note [/fig_ref] and [fig_ref] Figure 2: Two-factor model of depression and fatigue [/fig_ref]. Compared to the one-factor model, the two-factor model provided a significantly better fit to the data, Δχ 2 (1) = 67.50, p < .001. The correlation between the fatigue and depression latent factors in the two-factor model was 0.41 (p < .001).
The pattern and structure coefficients of the one and two-factor models are shown in [fig_ref] Table 4 Pattern: Note [/fig_ref]. The differences between the structure and the fixed pattern coefficients ranged from 0.21 to 0.39 for both the depression and fatigue items, suggesting adequate discriminant validity.
In the better fitting two-factor model, SE was added as a simultaneous predictor of both the depression and fatigue latent factors. This model had an acceptable fit to the data without modification (see [fig_ref] Table 3: Fit Indices for models [/fig_ref]. Lower SE was associated with both higher depression (p = .004) and fatigue (p < .001), with no significant difference in the strength of these two paths, Wald χ 2 (1) = 0.131, p = .71.
# Discussion
In this sample of women with elevated psychological distress and fatigue symptoms seeking support for unsettled infant behaviour, depression and fatigue symptoms are best considered as separate constructs that share a moderate correlation. Furthermore, both constructs were simultaneously predicted by a potential common cause (i.e., sleep efficiency), suggesting that the two-factor model may facilitate the understanding of the risk factors for both conditions. This study also supports the DASS21-D and a revised FSS-5R as uni-dimensional measures of depressive and fatigue symptoms in this population. Findings from this study echo results from community postpartum populations where depression and fatigue were also found to be separate constructs [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. However, the correlation between the depression and fatigue latent factors in this study was smaller than the large associations seen in the two studies that applied CFA on non-clinical samples [bib_ref] Assessment of maternal fatigue and depression in the postpartum period: support for..., Giallo [/bib_ref] [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref]. This could be because in this study, depression levels while elevated are not severe based on cut-off scores, while fatigue levels are high based on cut-off scores, thus the difference between the two constructs may be more prominent. Alternatively, the lower correlation in this study could be due to differences in scales: the DASS21-D does not include any fatigue or somatic items, and the FSS-5R assesses fatigue interference rather than specific symptoms. This combination may have led to a weaker correlation between the two factors in this study compared to other combinations of scales. Nevertheless, the correlation between depression and fatigue in this study is comparable to that in other postpartum studies (r = .30 to .45; [bib_ref] I didn't know it was possible to feel that tired": exploring the..., Giallo [/bib_ref] [bib_ref] Perinatal stress, fatigue, depressive symptoms, and immune modulation in late pregnancy and..., Cheng [/bib_ref] [bib_ref] Relationships among infant sleep patterns, maternal fatigue, and development of depressive symptomatology, Dennis [/bib_ref] [bib_ref] Couples' experiences with fatigue during the transition to parenthood, Elek [/bib_ref] [bib_ref] Fatigue in postpartum women, Gardner [/bib_ref] [bib_ref] Does activity matter: an exploratory study among mothers with preterm infants?, Lee [/bib_ref] [bib_ref] Empirical test of an explanatory theory of postpartum fatigue in Korea, Song [/bib_ref].
Our analyses on sleep efficiency serve as an example of many potential uses of the two-factor model in understanding common predictors and mechanisms. In this study, the findings were consistent with the literature linking self-report sleep disturbance with both postpartum fatigue and depressive symptoms [bib_ref] A depressive symptoms responsiveness model for differentiating fatigue from depression in the..., Runquist [/bib_ref] [bib_ref] Trajectories of depressive symptoms and fatigue among postpartum women, Kuo [/bib_ref] [bib_ref] Sleep and mood during pregnancy and the postpartum period, Bei [/bib_ref] [bib_ref] The relationship between sleep and mood in first-time and experienced mothers, Calcagni [/bib_ref] [bib_ref] Sleep patterns and fatigue in new mothers and fathers, Gay [/bib_ref] [bib_ref] The relationship between sleep characteristics and fatigue in healthy postpartum women, Rychnovsky [/bib_ref] [bib_ref] Sleep, depression, and fatigue in late postpartum, Thomas [/bib_ref].
# Limitations and strengths
As participants in this study were predominantly university-educated, born in Australia, and had the necessary resources to access privately funded treatment, this may limit the generalizability of our results. Also, despite overall high distress levels, depressive symptoms reported in our study were not severe. Thus, findings may not generalise to mothers meeting diagnostic criteria for a depressive disorder. A further limitation was that our sample comes from an ongoing clinical service that admits infants of 0-2 years, and infants in this study had an age range spanning 21.5 months. During this period, various psychological, biological and social factors may influence depression and fatigue. It is also possible that our sample could have included women with chronic health difficulties that contribute to their reported symptoms. Finally, given that the service we recruited from only admitted women with their infants, this paper did not examine the how potentially [bib_ref] Psychological distress, alcohol use, fatigue, sleepiness, and sleep quality: an exploratory study..., Wynter [/bib_ref] impact women's experiences and symptoms. Nevertheless, this study represented a unique opportunity to investigate the relationship between depression and fatigue in a clinical postpartum population with elevated psychological distress and fatigue symptoms. Given the high prevalence of infant settling difficulties in the community, these results are likely to be relevant to a high proportion of women who have given birth in the last year or two [bib_ref] Understanding and responding to unsettled infant behaviour, Fisher [/bib_ref]. Other strengths include a large sample size, a relatively high recruitment rate for a help-seeking population (44%), and the use of appropriate statistical modelling. A further strength of this study was that it serves as a demonstration of how a third construct such as sleep efficiency can influence both these constructs.
# Implications and conclusions
Theoretically, our findings add further support for the two-factor model of related but distinct postpartum depression and fatigue and show that depression and fatigue likely remain distinct constructs, even when mothers are experiencing elevated psychological distress and fatigue levels. By showing how sleep efficiency can be independently related to both the depression and fatigue factors, this study demonstrated the potential utility of the two-factor model for understanding other potential physiological, psychological, and situational factors that could underlie both conditions [bib_ref] Postpartum fatigue: clarifying a concept, Milligan [/bib_ref] [bib_ref] The middle-range theory of unpleasant symptoms: an update, Lenz [/bib_ref].
Clinically, our results indicate that among women seeking support for unsettled infant behaviour, and perhaps more broadly, women who present to clinical services with high psychological distress and fatigue in the postpartum period, depression and fatigue symptoms need to be assessed and treated in their own right. Improved assessment and greater awareness that depression and fatigue are related but separate constructs could help prevent the diagnosis of fatigue symptoms as depressive symptoms [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] [bib_ref] A depressive symptoms responsiveness model for differentiating fatigue from depression in the..., Runquist [/bib_ref]. Given that fatigue is one of the DSM-5 diagnostic criteria for Major Depressive Disorder, some overlap between these two constructs is inevitable. However, more detailed assessment of both conditions will assist clinicians to determine whether impaired postpartum functioning is caused by depressive symptoms, fatigue symptoms, or a combination of both.
Women experiencing fatigue but not depression may benefit from targeted interventions for fatigue, rather than potentially more intensive pharmacological treatments or therapy that may be better suited for depression [bib_ref] Differentiating maternal fatigue and depressive symptoms at six months and four years..., Giallo [/bib_ref] [bib_ref] The effectiveness of a self-care intervention for the management of postpartum fatigue, Troy [/bib_ref]. Residential early parenting programs that assist with unsettled infant behaviour have demonstrated efficacy in rapidly reducing fatigue and may be an appropriate treatment in this situation [bib_ref] Temperament and behaviour of infants aged 4-12 months on admission to a..., Fisher [/bib_ref] [bib_ref] Treatment of maternal mood disorder and infant behaviour disturbance in an Australian..., Fisher [/bib_ref].
## Additional file
Additional file 1: [fig_ref] Table 1: Maternal and infant demographics [/fig_ref]. Summary of Items for FSS-5R and DASS21-D. [fig_ref] Table 2: Descriptive statistics [/fig_ref]
[fig] Figure 1: One-factor model of depression and fatigue. Note: Loadings are standardised [/fig]
[fig] Figure 2: Two-factor model of depression and fatigue. Note: Loadings are standardised [/fig]
[fig] .: Means, standard deviations, and Pearson correlations for FSS-5R, DASS21-D and SE (N = 167) (DOCX 41 kb) Abbreviations CFA: Confirmatory Factor Analysis; CFI: Comparative Fit Index (CFI); DASS21: Depression Anxiety Stress Scale; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition; FSS-5R: Fatigue Severity Scale -Revised 5 item version; FSS-9: Fatigue Severity Scale; IDA-I: Irritability Depression Anxiety Scale -Irritability subscale, 6 item version; ISI: Insomnia Severity Index; MPHEPC: Masada Private Hospital Early Parenting Centre; RMSEA: Root Mean Square Error of Approximation; SE: Sleep Efficiency; TLI: Tucker-Lewis Index; WLSMV: Diagonally weighted least squares estimation; WRMR: Weighted Root Mean Square Residual [/fig]
[table] Table 1: Maternal and infant demographics (N = 167)Note. a Among the 58 who reported having received previous mental health treatment, 36 received treatment for depression, 38 for anxiety, and 5 for posttraumatic stress disorder [/table]
[table] Table 2: Descriptive statistics (N = 167) [/table]
[table] Table 3: Fit Indices for models (N = 167) RMSEA Root Mean Square Error of Approximation, SE Sleep Efficiency, TLI Tucker-Lewis Index, WRMR Weighted Root Mean Square Residual [/table]
[table] Table 4 Pattern: Note: a In the two-factor model all correlations were free to be estimated, and factor variances were set to unity for model identification. b Parameters were fixed at 0.00. DASS21-D Depression Anxiety Stress Scale Depression subscale, FSS-5R Fatigue Severity Scale-Revised 5-item version; Values are standardised coefficients elevated mental health symptoms in partners [/table]
|
Pharmacological Activity and Clinical Use of PDRN
PDRN is a proprietary and registered drug that possesses several activities: tissue repairing, anti-ischemic, and anti-inflammatory. These therapeutic properties suggest its use in regenerative medicine and in diabetic foot ulcers. PDRN holds a mixture of deoxyribonucleotides with molecular weights ranging between 50 and 1,500 KDa, it is derived from a controlled purification and sterilization process of Oncorhynchus mykiss (Salmon Trout) or Oncorhynchus keta (Chum Salmon) sperm DNA. The procedure guarantees the absence of active protein and peptides that may cause immune reactions. In vitro and in vivo experiments have suggested that PDRN most relevant mechanism of action is the engagement of adenosine A 2A receptors. Besides engaging the A 2A receptor, PDRN offers nucleosides and nucleotides for the so called "salvage pathway." The binding to adenosine A 2A receptors is a unique property of PDRN and seems to be linked to DNA origin, molecular weight and manufacturing process. In this context, PDRN represents a new advancement in the pharmacotherapy. In fact adenosine and dipyridamole are non-selective activators of adenosine receptors and they may cause unwanted side effects; while regadenoson, the only other A 2A receptor agonist available, has been approved by the FDA as a pharmacological stress agent in myocardial perfusion imaging. Finally, defibrotide, another drug composed by a mixture of oligonucleotides, has different molecular weight, a DNA of different origin and does not share the same wound healing stimulating effects of PDRN. The present review analyses the more relevant experimental and clinical evidences carried out to characterize PDRN therapeutic effects.
# Introduction
In the last decade great attention has been dedicated by pharmacologists to the characterization of the pharmacological properties of substances that are needed and produced by living organisms. DNA derived drugs might have relevant therapeutic effects in a clinical setting. In this context, there are two drugs named PDRN (polydeoxyribonucleotide) and defibrotide that share the same "natural" source: DNA [bib_ref] Polydeoxyribonucleotide (PDRN): a safe approach to induce therapeutic angiogenesis in peripheral artery..., Altavilla [/bib_ref] [bib_ref] Defibrotide: a review of its use in severe hepatic venoocclusive disease following..., Keating [/bib_ref]. However, they differ in DNA origin, molecular weight, and manufacturing procedures; as a consequence, they have different pharmacological properties, mechanism(s) of action and, eventually, clinical effects. While some excellent reviews on defibrotide have been recently published [bib_ref] Defibrotide: properties and clinical use of an old/new drug, Pescador [/bib_ref] [bib_ref] Drug safety evaluation of defibrotide, Richardson [/bib_ref] , there is only one dedicated to PDRN that dates back to 2009 [bib_ref] Polydeoxyribonucleotide (PDRN): a safe approach to induce therapeutic angiogenesis in peripheral artery..., Altavilla [/bib_ref]. However, the last years have testified an increased interest in PDRN with a large number of experimental and clinical studies carried out with this DNA derived drug. Therefore, aim of this paper is to review and up-date the experimental and clinical work carried out on PDRN.
## Pdrn chemistry
PDRN is a proprietary and registered DNA derived drug. It is a mixture of deoxyribonucleotides with molecular weights between 50 and 1,500 KDa and it is derived from Oncorhynchus mykiss (Salmon trout) or Oncorhynchus keta (Chum Salmon) sperm DNA. The most represented molecular weight is 80-200 KDa, with a peak of the Gaussian distribution at 132 KDa. PDRN has a higher molecular weight compared to defibrotide (16.5 ± 2.5 KDa) [bib_ref] Update on the use of defibrotide, Guglielmelli [/bib_ref] [bib_ref] Drug safety evaluation of defibrotide, Richardson [/bib_ref]. Furthermore, defibrotide is derived from the porcine intestinal mucosal DNA [fig_ref] TABLE 1 |: The most important characteristics of PDRN and defibrotide [/fig_ref]. PDRN is extracted and purified at high temperature, a procedure that allows to recover a >95% pure active substance with inactivated proteins and peptides. This latter guarantees the safety of the product and the absolute lack of any immunological side effect. Indeed, the source of raw material (cells vs. organs) is of particular importance: spermatozoa are the most appropriate cells to provide highly purified DNA without risk of impurity such as peptides, proteins and lipids which can remain from the somatic cells.
## Pdrn pharmacokinetics
The pharmacokinetics of PDRN has been evaluated after a single intraperitoneal administration of 8 mg/kg in rat. Measurable levels of PDRN were observed 15 min post-injection, and peak levels 1 h after drug administration, with a bioavailability of 90%. Drug levels then decreased progressively, being PDRN still measurable (0.137 µg/ml) 6 h following injection. The halflife is of 3 h and it is not influenced by dosage. As analyzed below, the drug stimulates the initiation of a cascade of events involving a number of transduction effectors that last much more than its plasma half-life. Therefore, the pharmacodynamics effects of PDRN may be much longer than anticipated by its pharmacokinetic profile. Due to its chemical structure, plasmatic carrier proteins do not bind PDRN, but it is found free in plasma. The distribution of the free drug depends upon tissue blood flow, being higher in those organs with elevated blood supply. The drug is not metabolized by the liver and there is no evidence for a first-passage metabolism. Instead, the drug is mainly degraded by unspecific plasma DNA nucleases, or by nucleases bound to cell membranes leading to oligo and mononucleotides. From a pharmacodynamics point of view, this event is of paramount importance: in fact, PDRN degradation gives rise to the formation of nucleosides and nucleotides that become available for the main activity: the binding to the adenosine A 2A receptor. PDRN fragments are then excreted in the urine (∼65%) and to a lesser extent in the feces.
The pharmacokinetics of PDRN has been also studied in healthy volunteers after intramuscular administration (5.625 mg). The results of this study showed a pharmacokinetic profile overlapping to that observed in experimental animals: more specifically peak levels at ∼1 h; a half-life of ∼3.5 h, with a bioavailability in the range of 80/90%.
## Pharmacological properties of pdrn: experimental in vitro studies supporting the mode of action
Adenosine activates four distinct adenosine receptors indicated as A 1 , A 2A , A 2B , and A 3 . These receptors are widely expressed and implicated in several physiological and pathological functions. The A 2A receptor play a central role in modulating inflammation, oxygen consumption, ischemia, cell growth, and angiogenesis. PDRN was compared to adenosine in primary cultures of human skin fibroblasts [bib_ref] Polydeoxyribonucleotides enhance the proliferation of human skin fibroblasts: involvement of A2 purinergic..., Thellung [/bib_ref] : both PDRN and adenosine induced cell growth. The effects of PDRN were abolished by the concomitant incubation with an adenosine A2 receptor antagonist, 3,7-Dimethyl-1propargylxanthine (DMPX). Indeed, DMPX has greater affinity for A2 A than for A2 B receptor subtype. This leads to hypothesize that PDRN may preferentially act on the adenosine A2 A receptor thus suggesting the involvement of this receptor subtype in PDRN effects. Indeed, it can be speculated that PDRN may represent a pro-drug able to generate active deoxyribonucleotides, nucleosides, and bases exerting their pharmacological effects interacting with the A2 A receptor .
In further experiments, cultured fibroblasts were loaded with radioactive amino acids in the presence of PDRN [bib_ref] Effect of polydeoxyribonucleotides on human fibroblasts in primary culture, Sini [/bib_ref]. Cell growth was accompanied by internalization of PDRN-derived nucleotides to offer purine and pyrimidine rings for the "salvage pathway." In fact, damaged or hypoxic tissue very often cannot undergo to the DNA "de novo" synthesis . Under these conditions, salvage pathways operate to recover bases and nucleosides generated from the breakdown of DNA and RNA. The salvaged bases can be then transformed into nucleotides and reincorporated into DNA. PDRN generates nucleotides and nucleosides that can contribute to DNA formation, thus reactivating normal cell proliferation and growth pattern .
PDRN stimulatory effect on cell growth was also investigated in human cultured osteoblasts [bib_ref] Polydeoxyribonucleotide (PDRN) promotes human osteoblast proliferation: a new proposal for bone tissue..., Guizzardi [/bib_ref]. PDRN (20 mg/ml) promoted cell proliferation with a concomitant increase in alkaline phosphatase and DMPX suppressed this effects. PDRN has been also tested in primary chondrocytes [bib_ref] Protective effects of polydeoxyribonucleotides on cartilage degradation in experimental cultures, Gennero [/bib_ref] , where induced a physiological accumulation of the extracellular matrix with reduced proteoglycan degradation, reducing matrix metalloproteinases 2 and 9. Moreover, PDRN synergizes with glucosamine in reducing extracellular matrix gene expression, thus reducing its degradation [bib_ref] Effects of glucosamine and nucleotide association on fibroblast: extracellular matrix gene expression, Avantaggiato [/bib_ref]. These evidences candidates the DNA derived drug as a potential therapy in regenerative cartilage treatment.
Additionally, PDRN may also protect cells from UVinduced DNA damage. Exposure of human dermal fibroblasts to ultraviolet B radiation causes accumulation of dangerous photoproducts such as cyclobutane pyrimidine dimers (CPDS). PDRN addition to the cell culture immediately after irradiation resulted in p53 protein activation and in enhancing DNA repair likely due to the priming of the salvage pathway [bib_ref] Polydeoxyribonucleotide promotes cyclobutane pyrimidine dimer repair in UVB-exposed dermal fibroblasts, Belletti [/bib_ref]. Another important aspect useful in regenerative medicine could be related to PDRN ability to increase the proliferation of human pre-adipocytes [bib_ref] In vitro polydeoxyribonucleotide effects on human pre-adipocytes, Raposio [/bib_ref]. As a matter of fact, adipose tissue represents a relevant source of adult stem cells, thus PDRN may be used for therapeutic and regenerative purposes.
Collectively these in vitro evidences support the concept that PDRN engages the A 2A and has cell proliferative and regenerative effects.
## Pharmacological properties of pdrn
## Tissue repairing, wound healing and therapeutic angiogenesis
It has been suggested that PDRN accelerates the repair and the growth of bone tissue . Furthermore, PDRN effects were investigated in a model of diabetes-impaired wound healing [bib_ref] Polydeoxyribonucleotide stimulates angiogenesis and wound healing in the genetically diabetic mouse, Galeano [/bib_ref]. Disorders in wound healing are very common in diabetes and they represent a major clinical challenge: in fact, there is still an unmet need for a safe treatment able to counteract this clinical situation. PDRN improved the skin repair process and enhanced woundbreaking strength in diabetic animals. This effect was supported by a marked increase in the expression of Vascular Endothelial Growth Factor (VEGF), a master regulator of angiogenesis that is impaired in diabetes-related wound disorders [bib_ref] Polydeoxyribonucleotide stimulates angiogenesis and wound healing in the genetically diabetic mouse, Galeano [/bib_ref]. Angiogenesis improvement was confirmed by an increase in CD31, transglutaminase-II, and angiopoietin, factors contributing to new vessel formation. The healing-promoting effect was abrogated by DMPX, thus suggesting the involvement of the adenosine A 2A receptor. The positive effect on wound healing and angiogenesis is a characteristic feature of PDRN that is not shared by other DNA-derived drugs that have different DNA origin, molecular weight and manufacturing process. In fact, defibrotide [bib_ref] Defibrotide: a review of its use in severe hepatic venoocclusive disease following..., Keating [/bib_ref] inhibits angiogenesis [fig_ref] TABLE 1 |: The most important characteristics of PDRN and defibrotide [/fig_ref]. This observation led us to speculate that the eventual approval of a DNA containing drug biologically equivalent to the registered PDRN (i.e., a PDRN bioequivalent drug) should be considered with extreme caution. To claim the same therapeutic indication, a new formulation should be more tested as a "bioequivalent drug." More specifically any other DNA containing drug with either smaller or higher molecular weight of DNA polymer should be granted market approval on basis of its own efficacy, safety and clinical data, comparing the new product to the approved PDRN drug.
Another clinical situation characterized by a poor skin repair process and impaired angiogenesis is thermal injury. PDRN effects were investigated in mice with a deep-dermal second degree burn injury [bib_ref] Polydeoxyribonucleotide improves angiogenesis and wound healing in experimental thermal injury, Bitto [/bib_ref] , the treatment enhanced burn wound re-epithelialization and decreased time to final wound closure. PDRN also showed a marked systemic effect: in fact, it reduced the serum levels of the pleiotropic cytokine Tumor Necrosis Factor (TNF-α) and augmented wound VEGF expression and nitric oxide production. The wound healing properties of PDRN might be the consequence of the stimulation of the altered cell-cycle machinery that is deeply impaired in several conditions: in a diabetes setting the drug stimulated the proliferation of the granulation tissue by activating cyclins driven cell-cycle progression and turning off the cell-cycle negative regulators p15 and p27 [bib_ref] Activation of adenosine A 2A receptors restores the altered cell-cycle machinery during..., Altavilla [/bib_ref].
The ability of PDRN to promote therapeutic angiogenesis was also studied in an experimental model of peripheral artery occlusive disease induced by the excision of the femoral artery. PDRN boosted a robust blood flow restoration together with a marked increase in VEGF expression, while DMPX abrogated the beneficial effects of the drug [bib_ref] Polydeoxyribonucleotide (PDRN) restores blood flow in an experimental model of peripheral artery..., Bitto [/bib_ref].
Skin flap technique is commonly used in plastic surgery and esthetic medicine for a faster wound coverage, to reduce the risk of infection and restore organ function. In an experimental ischemic skin flap model PDRN increased blood flow, evaluated by laser Doppler, and again boosted a strong VEGF-driven angiogenesis [bib_ref] Polydeoxyribonucleotide restores blood flow in an experimental model of ischemic skin flaps, Polito [/bib_ref]. This result has been further confirmed in a recent paper [bib_ref] The effects of polydeoxyribonucleotide on the survival of random pattern skin flaps..., Chung [/bib_ref] , suggesting a role for this drug in improving skin flap survival. Overall, all these experimental pre-clinical observations anticipate a marked therapeutic efficacy of PDRN in a clinical setting of impaired wound healing, angiogenesis and disturbed skin repair processes.
## Pdrn anti-inflammatory activity
The adenosine A 2A receptor activation results in a robust antiinflammatory effect, and it represents an interesting target for the molecular design of anti-inflammatory agents. With this scientific background, PDRN was evaluated in collagen-induced arthritis . In this experimental paradigm, PDRN significantly improved the clinical signs of arthritis, reduced the histological damage and blunted the cartilage content and blood levels of several inflammatory cytokine, while increased the expression of the anti-inflammatory cytokine Interleukin-10 (IL-10). All these curative effects were abolished by the concomitant administration of DMPX, further pointing out that the adenosine A2 A receptor is the specific target of PDRN. The immunological and pathological processes occurring in rheumatoid arthritis and another condition named periodontitis are nearly identical. Both conditions are characterized by chronic inflammation in a soft-tissue site adjacent to bone, and periodontitis is one of the most important cause of teeth loss in adults. Experimentally periodontitis can be induced in rodents by ligation of the lower left first molar cervix. In a rat model, PRDN was tested in a gel solution applied for 7 days alone or in combination with an adenosine A2 A receptor antagonist. The drug reduced the histological damage, decreased the tissue levels of several inflammatory cytokines and blunted apoptotic protein expression [bib_ref] Adenosine receptor stimulation by polynucleotides (PDRN) reduces inflammation in experimental periodontitis, Bitto [/bib_ref]. All these effects were abrogated by the A2 A receptor antagonist. The treatment also markedly protected the alveolar bone quality, thus suggesting that PDRN may also promote bone regeneration, as recently confirmed by other studies . Chronic inflammation is also deeply involved in the etiology and development of other conditions as inflammatory bowel disease and it is known that the activation of adenosine A 2A mitigates the inflammatory cascade in colonic epithelial cells. In agreement with this evidence, PDRN was tested in two experimental models of colitis, the drug was given by intraperitoneal injection and it was able to ameliorate tissue repair and to reduce symptomology [bib_ref] Adenosine receptor stimulation by polydeoxyribonucleotide improves tissue repair and symptomology in experimental..., Pallio [/bib_ref].
## Anti-ischemic effects of pdrn
Adenosine contributes to the mechanisms underlying ischaemia/reperfusion injury and the A 2A receptor has been indicated as a therapeutic strategy to modulate the ischemic insult. Testicular twisting and varicocele are peculiar ischemic conditions that create a hypoxic state responsible for testicular damage and long-term complication consisting in disturbed Leydig cell activity and altered spermatogenesis. PDRN has been tested in the experimental models of testicular twisting and varicocele [bib_ref] Activation of adenosine A 2A receptors by polydeoxyribonucleotide increases vascular endothelial growth..., Minutoli [/bib_ref] [bib_ref] Role of inhibitors of apoptosis proteins in testicular function and male fertility:..., Minutoli [/bib_ref] [bib_ref] Polydeoxyribonucleotide administration improves the intra-testicular vascularization in rat experimental varicocele, Arena [/bib_ref]. These studies suggested that PDRN protects against testicular histological damage and markedly improves spermatogenesis, by increasing VEGF expression and angiogenesis, reducing inflammatory cascade and rebalancing the apoptotic machinery. The protective effect might be also ascribed to PDRN ability to limit ischemia reperfusion injury, as observed in the kidney [bib_ref] Protective effect of polydeoxyribonucleotide against renal ischemia-reperfusion injury in mice, Jeong [/bib_ref]. [bib_ref] The effect of PDRN, an adenosine receptor A 2A agonist, on the..., Squadrito [/bib_ref].
In summary, the weight of available evidence from in vivo studies indicates that PDRN possesses several activities: tissue repairing, anti-ischemic, and anti-inflammatory.
## Pharmacological properties of pdrn: clinical studies
## Tissue repairing, wound healing and regenerative effects
In a first pilot study PDRN was tested in the healing of autologous skin graft donor sites. The patients were randomized into two groups: the control group received dressings with gauzes embedded in chloramine solution; the other group received the same treatment plus PDRN (5.625 mg diluted in 3 ml embedded in the gauzes) that improved re-epithelialization and, in addition, the time to complete wound healing [bib_ref] Evaluation of the efficacy of polydeoxyribonucleotides in the healing process of autologous..., Valdatta [/bib_ref]. In a further experiment PDRN eye drops (0.75/ml) were evaluated on corneal epithelial tissue repair in patients after photorefractive keratectomy. The DNA-derived drug markedly stimulated corneal epithelium regeneration, while no significant adverse event was observed [bib_ref] Clinical evaluation of corneal epithelialization after photorefractive keratectomy in a patients treated..., Lazzarotto [/bib_ref].
Diabetic foot ulcers are a leading cause of hospitalization and readily become chronic with poor healing properties [bib_ref] Wound healing and its impairment in the diabetic foot, Falanga [/bib_ref] [bib_ref] The molecular biology of chronic wounds and delayed healing in diabetes, Blakytny [/bib_ref]. Therefore, the opportunity to have a treatment proven to be effective for affected subjects is a breakthrough in this field. Diabetic patient with Wagner grade 1 or 2 ulcers were randomly assigned to receive placebo (n = 106) or PDRN (n = 110) for 8 weeks. The drug was injected daily by intramuscular route (5.625 mg in a 3 ml, vial) for 5 day/week and by perilesional route (5.625 mg, in a 3 ml vial) 2 day/week for 8 weeks. The treated group nearly doubled the rate of complete healing of difficult-to-heal diabetic foot ulcers compared to placebo (Table 2) as early as 8 weeks after start of treatment [bib_ref] The effect of PDRN, an adenosine receptor A 2A agonist, on the..., Squadrito [/bib_ref]. This study is one of the largest trial ever carried out in diabetic patients with poor diabetic skin repair that points out a dramatic efficacy of PDRN in improving hard-to-heal chronic diabetic foot ulcers.
The healing efficacy of the drug was also confirmed in a smaller group of patients suffering for pressure ulcers [bib_ref] Viscoelastic interactions between polydeoxyribonucleotide and ophthalmic excipients, Kim [/bib_ref]. In a further study, a topical application of a gel containing PDRN and hyaluronic acid was compared with a gel containing only hyaluronic acid in patients suffering from venous ulcers of the lower limbs. The endpoint was full skin repair 45 days after the start of treatment. Complete wound healing was obtained in 67% of patients treated with PDRN and hyaluronic acid, while only 22% of patients receiving only hyaluronic acid achieved the therapy target [bib_ref] Trophic effects of polynucleotides and hyaluronic acid in the healing of venous..., De Caridi [/bib_ref].
Degenerative joint disease, or osteoarthritis, was successfully treated in animals with PDRN and in humans there is still a lack of an effective pharmacological therapy aimed at reducing the need of prosthesis and leading to a stabilization of the disease. The topical application of a regenerative gel containing PDRN significantly improved pain and joint mobility with a clear amelioration of the clinical sign and the radiological images (Di [bib_ref] Biological baseline of joint self-repair procedures, Nicola [/bib_ref].
## Other clinical evidences
Lichen sclerosus is an autoimmune inflammatory skin disease that causes a sclerosis process in the male genital tract. The available treatment for these pathological conditions consists in topical local steroids. PDRN subdermal injections (5.625 mg, in a 3 ml vial), in addition to the standard topical treatment, resulted in a marked reduction of most of the clinical signs [bib_ref] Polydeoxyribonucleotide dermal infiltration in male genital lichen sclerosus: adjuvant effects during topical..., Laino [/bib_ref] , thus pointing out that intradermal administration of PDRN, together with the standard steroids treatment, may represent a promising and innovative option for this pathological conditions. This has been recently confirmed in a more recent clinical trial [bib_ref] Genital lichen sclerosus in male patients: a new treatment with polydeoxyribonucleotide, Zucchi [/bib_ref].
Finally, PDRN has been proven effective in the treatment (intradermal administration) of chronic plantar fasciitis [bib_ref] Effectiveness of polydeoxyribonucleotide injection versus normal saline injection for treatment of chronic..., Kim [/bib_ref] and in the therapeutic management (topical application) of female pattern hair loss [bib_ref] The effect of polydeoxyribonucleotide on ischemic rat skin flap survival, Lee [/bib_ref].
## Safety
Acute and chronic toxicity studies in mice and rats were undertaken to evaluate the effects of repeated systemic administration of PDRN. PDRN (8 mg/kg) showed no toxic effect in brain, liver, lungs skeletal muscle and heart and did not cause mortality [bib_ref] Polydeoxyribonucleotide stimulates angiogenesis and wound healing in the genetically diabetic mouse, Galeano [/bib_ref]. In the trial investigating the effects of PDRN on the healing of chronic diabetic foot ulcers for up to 56 days, the safety and tolerability were excellent [bib_ref] The effect of PDRN, an adenosine receptor A 2A agonist, on the..., Squadrito [/bib_ref].
Finally, post-marketing surveillance study carried out throughout 5 years and involving the selling of more than 300,000 PDRN dispensed prescriptions confirmed the excellent safety profile of the drug.
## Discussion and concluding remarks
PDRN pharmacological properties design a global picture of a drug for the management of poor wound healing caused by different pathological conditions. The lack of effects on the immune system is one of the most important determinants of the good safety profile of the drug. In this context, PDRN represents a new advancement in the pharmacotherapy. In fact adenosine and dipyridamole are non-selective activators of adenosine receptors and they may cause unwanted side effects. Indeed, regadenoson is the only other one A 2A receptor agonist available in the market but has been approved by the FDA as a pharmacological stress agent in myocardial perfusion imaging, a well-established noninvasive modality for the diagnosis and prognosis of coronary artery disease (Al [bib_ref] Regadenoson: a new myocardial stress agent, Jaroudi [/bib_ref] [bib_ref] Regadenoson stress for myocardial perfusion imaging, Reyes [/bib_ref]. It displays fewer side effects that adenosine or dipyridamole, but its potential use in other pathological conditions has not yet explored.
Besides engaging the A 2A receptor, PDRN offers nucleosides and nucleotides for the so called "salvage pathway." The tissue repairing and healing effects are unique features of PDRN that are not shared with other DNA derived drug from different origin, molecular weight and manufacturing process. Therefore, caution must be used when extrapolating the pharmacological properties of PDRN to other DNA derived products.
# Author contributions
FS contributed to conception, drafting the article, and final approval; AB contributed substantially to revision for important intellectual content and final approval of the version to be published; NI and GPa made substantial contributions to revising the article; GPi made substantial contributions to revising the article; LM and DA contributed substantially to revision for important intellectual content and final approval of the version to be published. All the authors gave final approval of the version to be published.
# Funding
This paper was supported by Departmental funding assigned to Prof. FS.
[table] TABLE 1 |: The most important characteristics of PDRN and defibrotide.FIGURE 1 | PDRN mode of action: activation of adenosine A2 receptors and the "salvage pathway". [/table]
[table] TABLE 2 |: Outcome achieved, according to treatment group. [/table]
|
Emerging from the ice‐fungal communities are diverse and dynamic in earliest soil developmental stages of a receding glacier
We used amplicon sequencing and isolation of fungi from in-growth mesh bags to identify active fungi in three earliest stages of soil development (SSD) at a glacier forefield (0-3, 9-14, 18-25 years after retreat of glacial ice). Soil organic matter and nutrient concentrations were extremely low, but the fungal diversity was high [220 operational taxonomic units (OTUs)/138 cultivated OTUs]. A clear successional trend was observed along SSDs, and species richness increased with time. Distinct changes in fungal community composition occurred with the advent of vascular plants. Fungal communities of recently deglaciated soil are most distinctive and rather similar to communities typical for cryoconite or ice. This indicates melting water as an important inoculum for native soil. Moreover, distinct seasonal differences were detected in fungal communities. Some fungal taxa, especially of the class Microbotryomycetes, showed a clear preference for winter and early SSD. Our results provide insight into new facets regarding the ecology of fungal taxa, for example, by showing that many fungal taxa might have an alternative, saprobial lifestyle in snow-covered, as supposed for a few biotrophic plant pathogens of class Pucciniomycetes. The isolated fungi include a high proportion of unknown species, which can be formally described and used for experimental approaches.
# Introduction
Since the last Little Ice Age around 1858, glaciers are receding all over the world [bib_ref] Quantifying changes and trends in glacier area and volume in the, Abermann [/bib_ref]. Glacier forefields provide valuable areas for studying the development and colonization of newly exposed soils. Microorganisms are the first life forms to colonize freshly exposed substrates and play a tremendous role in soil formation [bib_ref] Colony-forming analysis of bacterial community succession in deglaciated soils indicates pioneer stress-tolerant..., Sigler [/bib_ref] [bib_ref] Primary succession of soil Crenarchaeota across a receding glacier foreland, Nicol [/bib_ref] [bib_ref] Heterotrophic microbial communities use ancient carbon following glacial retreat, Bardgett [/bib_ref] [bib_ref] Microbial community succession in an unvegetated, recently deglaciated soil, Nemergut [/bib_ref] [bib_ref] The earliest stages of ecosystem succession in high-elevation (5000 metres above sea..., Schmidt [/bib_ref] [bib_ref] Microbial communities and primary succession in high altitude mountain environments, Ciccazzo [/bib_ref]. Biological soil crusts consist of typical pioneer organisms (bacteria including cyanobacteria, fungi, algae and lichens) colonizing the soil surface and subsurface. They perform a key role in stabilizing mobile surfaces, thus protecting soil from erosion and cryoturbation [bib_ref] Biological soil crusts: an eco-adaptive biological conservative mechanism and implications for ecological..., Bu [/bib_ref]. Biological soil crust productivity plays an important role in the carbon cycle during the early stage of succession: In a high Arctic glacier foreland, the estimated biological soil crust net primary production was higher than the net primary production of the other vegetation [bib_ref] Production of biological soil crusts in the early stage of primary succession..., Yoshitake [/bib_ref]. There is a mutual influence between microbes and the environment they colonize. Soil microbial community (SMC) composition and thus primary succession is strongly influenced by soil development in the glacier forefield [bib_ref] Soil fungal community assembly in a primary successional glacier forefront ecosystem as..., Jumpponen [/bib_ref] [bib_ref] Plant succession and rhizosphere microbial communities in a recently deglaciated alpine terrain, Tscherko [/bib_ref] , and especially, active fungal community composition changes over soil developmental stages [bib_ref] Potential sources of microbial colonizers in an initial soil ecosystem after retreat..., Rime [/bib_ref] [bib_ref] Assimilation of microbial and plant carbon by active prokaryotic and fungal populations..., Rime [/bib_ref].
SMCs follow distinct seasonal patterns across different habitats including boreal forest and arctic/alpine habitats [bib_ref] Seasonal dynamics of previously unknown fungal lineages in tundra soils, Schadt [/bib_ref] [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref] [bib_ref] Seasonal dynamics of nutrients and bacterial communities in unvegetated alpine glacier forefields, Lazzaro [/bib_ref] [bib_ref] The seasonal pattern of soil microbial community structure in mesic low arctic..., Buckeridge [/bib_ref] [bib_ref] Seasonal variation in functional properties of microbial communities in beech forest soil, Koranda [/bib_ref] [bib_ref] Seasonal dynamics of fungal communities in a temperate oak forest soil, Voříšková [/bib_ref] [bib_ref] Structures of microbial communities in alpine soils: seasonal and elevational effects, Lazzaro [/bib_ref] [bib_ref] Distinct summer and winter bacterial communities in the active layer of Svalbard..., Schostag [/bib_ref] [bib_ref] Microbial activity in forest soil reflects the changes in ecosystem properties between..., Zifcakova [/bib_ref] and are not at all dormant during winter. A community shift from bacteria towards cold-adapted fungi has been observed during winter [bib_ref] Seasonal dynamics of previously unknown fungal lineages in tundra soils, Schadt [/bib_ref] [bib_ref] Seasonal dynamics of nutrients and bacterial communities in unvegetated alpine glacier forefields, Lazzaro [/bib_ref] [bib_ref] The seasonal pattern of soil microbial community structure in mesic low arctic..., Buckeridge [/bib_ref] , with saprobial Ascomycota and basidiomycete yeasts-dominating fungal communities [bib_ref] Detection of soil fungal communities in an alpine primary successional habitat: does..., Oberkofler [/bib_ref] [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref]. These winter-active decomposers respire considerable amounts of CO 2 [bib_ref] Microbial growth under the snow: implications for nutrient and allelochemical availability in..., Schmidt [/bib_ref] [bib_ref] Winter forest soil respiration controlled by climate and microbial community composition, Monson [/bib_ref] [bib_ref] Microbial diversity in alpine tundra soils correlates with snow cover dynamics, Zinger [/bib_ref] [bib_ref] Sensitivity of soil respiration and microbial communities to altered snowfall, Aanderud [/bib_ref] and build up about 10 times more biomass during winter than during summer [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref] , thus being main players for soil organic matter (SOM) formation and soil development in successional sites.
New fungal lineages discovered from snow-covered habitats [bib_ref] Seasonal dynamics of previously unknown fungal lineages in tundra soils, Schadt [/bib_ref] [bib_ref] Widespread occurrence and phylogenetic placement of a soil clone group adds a..., Porter [/bib_ref] boosted the scientist's interest in SMCs of cold habitats. This resulted in the description of many new fungal taxa, like snow chytrids [bib_ref] Phylogeny and biogeography of an uncultured clade of snow chytrids, Naff [/bib_ref] , psychrophilic yeasts [bib_ref] Mrakiella cryoconiti gen. nov., sp nov., a psychrophilic, anamorphic, basidiomycetous yeast from..., Margesin [/bib_ref] [bib_ref] Relative incidence of Ascomycetous yeasts in Arctic coastal environments, Butinar [/bib_ref] and coldtolerant filamentous fungi [bib_ref] Psychrophilic fungi from the world's roof, Wang [/bib_ref] [bib_ref] Myrmecridium hiemale sp nov from snowcovered alpine soil is the first eurypsychrophile..., Peintner [/bib_ref]. Novel fungal taxa associated with snow or cryoconite were because continuously detected (e.g., [bib_ref] A distinctive fungal community inhabiting cryoconite holes on glaciers in Svalbard, Edwards [/bib_ref] [bib_ref] Contrasting primary successional trajectories of fungi and bacteria in retreating glacier soils, Brown [/bib_ref] , indicating that snow-covered soil also harbours a high diversity of unknown fungal taxa.
In this work, we focus on the earliest 25 years of soil development of a retreating glacier to identify first fungal colonizers and investigate how long it takes to develop a stable community. In this context, we especially address fungi actively growing underneath the snow cover by using a combination of in-growth mesh bags (MBs) [bib_ref] Estimation of the biomass and seasonal growth of external mycelium of ectomycorrhizal..., Wallander [/bib_ref] with both, cultivation as well as amplicon sequencing. This combination of approaches provides reliable information on seasonal variation in active fungal communities at three stages of soil development (SSD) within our 25-year chronosequence. The obtained NGS data and fungal isolates shall provide a solid base for future studies on the diversity of coldadapted fungi.
# Results
## Som content is extremely low and increases within 25 years of deglaciation
Sampling was carried out close to the glacier tongue in three earliest stages deglaciation ranging from 0 to 25 years [fig_ref] Figure 1: A-E Sampling sites at the Rotmoosferner glacier forefield at 2280-2450 m above... [/fig_ref]. Within these first 25 years of soil development, soil pH was neutral (7.5) in all sites due to the relatively high carbonate content, which increased along SSDs (Supporting . SOM was extremely low in all sites, ranging from 0.3% to 0.9% and increased significantly with successional age. Ammonium, nitrate and phosphate also increased. Availability of ammonium and phosphate was significantly correlated with SOM content. Furthermore, ammonium concentrations were significantly correlated to soil water content (Supporting Information .
Diversity of fungi isolated from in-growth MBs is high, with high proportions of fungi with seasonal preferences and of psychrophilic fungi
In total, 138 fungal cOTUs (cultivated strains merged into operational taxonomic units) with 99% sequence similarity were isolated from the MBs. When using a threshold of The presence/absence of fungal cOTUs (97% ITS sequence identity) isolated from in-growth mesh bags, which were buried during the snow-free vegetation period (summer = S) or the snow-covered period (winter = W), respectively. The three early stages of soil development are site 0 (0-3 years), site 1 (7-14 years) and site 2 (18-25 years). Seasonal patterns are especially pronounced during the first 0-3 years of soil development. [Color figure can be viewed at wileyonlinelibrary.com] . Cumulative step function on amplicon data: number of OTUs identified at a specific taxonomic level (total number of OTUs = 220) from the first 25 years of soil development after deglaciation. The proportion of OTUs, which could not be identified on species level, is very high. 97%, several cOTUs collapsed, resulting in 91 cOTUs total; of these, 89 contained a full ITS2 region and were used for comparison with amplicon sequencing data. More than half (54%) of the isolated fungi were Ascomycota, about one-third (34%) Basidiomycota, 7% Mortierellomycota and 5% Mucoromycota. On the genus level, 12% of the fungal isolates could not be identified (Supporting . The most speciesrich genera were Tetracladium (7 cOTUs), Mortierella (6 cOTUs), Mucor (5 cOTUs) and Phenoliferia ss. lato (10 cOTUs). Yeasts made up of 30% of the isolated fungi (single-celled, budding fungi; 41 cOTUs). More than half of the isolated cOTUs (n = 56) showed seasonal preferences, as they were either isolated from winter-MBs (n = 33) or summer-MBs (n = 23), only [fig_ref] Figure 2: Fig [/fig_ref]. Furthermore, the proportion of psychrophilic fungi (no growth at 25 C) was generally high (44%, n = 40 cOTUs). The number of cOTUs obtained was highest in the earliest SSD (50 OTUs in site 0 compared to 26 OTUs in site 1 and 44 OTUs in site 2). Some fungal isolates showed preferences for specific SSDs: 35% (n = 32) were found only in barren ground of site 0, 11% (n = 10) only in site 1 and 29% (n = 26) exclusively in site 2.
Amplicon sequencing confirms that fungal diversity increases with time because deglaciation, but most species remain unidentified Of all forward and reverse reads (n = 937,459), 84.8% passed paired-end joining and initial quality-filtering and 67.7% were retained after extraction of fungal ITS2 and disposal of non-fungal sequences as proposed by ITSx (ver. 1.0.11) [bib_ref] Improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS..., Bengtsson-Palme [/bib_ref]. Sequencing depth per sample spanned from 1821 to 45,280 reads with an average of 15,610 AE 9352 SD. The rarefaction to a minimum of 4020 reads per sample led to the exclusion of soil sample B19 (SSD 2). The rarefied data set consisted of 938 OTUs. Removing low abundance and low prevalence OTUs reduced the total number of OTUs to 220, of which 40.5% could not be assigned a phylum level taxonomy. The inclusion of manual NCBI BLASTn results (included . A,B. Alpha diversity of fungal communities for the summer (S) and winter (W) period at the three earliest stages of soil development of the glacier forefield based on NGS data of in-growth mesh-bag (MB) samples. In addition, a methodological comparison is carried out for winter samples from site 2: MB samples are compared to two types of soil samples: with snow cover (snow) and immediately after snow melt (bare). Asterisks indicate that data are from soil samples. A. Observed species diversity. B. Shannon diversity. Dots are representing single samples. Site 0 = 0-4 years, site 1 = 9-14 years, site 2 = 18-25 years after deglaciation. Significant differences were detected for the observed species richness, but not for Shannon diversity. Bars with the same letter do not differ significantly (ANOVA, Tukey-HSD). [Color figure can be viewed at wileyonlinelibrary.com] in all downstream analyses) reduced the fraction of OTUs unidentified at phylum level to 26.4% . The majority of OTUs was assigned to Ascomycota (39.5%), followed by Basidiomycota (22.3%), Chytridiomycota (5.0%), Mortierellomycota (4.5%) and a low prevalence of other phyla: Rozellomycota, Monoblepharidomycota and Glomeromycota (2.3% in total). On the genus level, 56.4% of all OTUs could not be assigned any taxonomic identification (78.0% without additional NCBI information).
Alpha-diversity (Shannon, observed) showed a general increase of fungal richness from the sites closest to the glacier tongue towards older SSDs . No significant differences in alpha diversity between the two contrasting seasons were found in any of the three SSDs (ANOVA, Tukey-HSD). Soil samples did not differ significantly from MB samples (p = 0.997), and soil samples retrieved about 2 weeks after snowmelt did not show significant differences compared to those obtained from beneath snow cover (p = 0.534).
Fungal community composition is not different between MB and soil samples, and community shifts at snowmelt take longer than 2 weeks No significant differences between the fungal communities were found with the two different sampling strategies used in this study (MBs vs. soil samples; cf. results of ANOSIM and PERMANOVA in Supporting Information . Furthermore, soil samples clustered relatively close to the corresponding MB samples in ordination plots [fig_ref] Figure 5: Non-metric multidimensional scaling [/fig_ref].
The comparison of soil samples retrieved from under snow-cover, to soil samples sampled 2 weeks after snowmelt showed no significant differences. Fungal communities from the same plot clustered close to each other, irrespective of sampling date [fig_ref] Figure 6: Ordination-based [/fig_ref].
## Fungal community composition shifts radically between the earliest ssd
On the community level (beta-diversity), different SSD formed distinct clusters. Differences in community composition are highly significant, as shown by both ANOSIM and PERMANOVA (p < 0.001, Supporting Information , providing proof for a distinct succession of fungal communities. The earliest SSD (site 0) is most distinctive regarding species composition compared to sites 1 and 2 (Figs. 5 and 6). The majority of OTUs did not appear in site 0 at all, or at least, not in a regular pattern. Rather, many OTUs started to first appear in site 1 and stayed present along the successional gradient (e.g., most Tetracladium OTUs, some Helotiales, Dioszegia).
Ascomycota, Basidiomycota and Mortierellomycota were found in all SSDs, with Ascomycota clearly dominating in relative abundance. However, the variability of abundances in plots of the same group (SSD and season), especially for Ascomycota was high, for example, spanning in a range of 22%-98% of reads in site 1 in summer. in site 2 during winter, however, the Ascomycota reads were generally dominating, ranging from 74% to 95% of total reads per plot and within-group variability for the phylum was low. Similarly, Basidiomycota tended to show smaller within-group variability in site 2 during winter, however, concentrating their relative abundance in the lower abundance area (2%-12%). The relative abundance of Basidiomycota decreased in older SSDs during winter. Chytridiomycota were almost absent in site 0 but found in high relative abundance in single plots of sites 1 and 2. Another group of zoosporic fungi, the Monoblepharidomycota, was almost exclusively found in the earliest SSD, with a slightly higher relative abundance in summer (Supporting Information [fig_ref] Figure 1: A-E Sampling sites at the Rotmoosferner glacier forefield at 2280-2450 m above... [/fig_ref].
Seasonal community shifts are typical, and most pronounced in barren, recently deglaciated sites Fungal communities of all SSDs also differed significantly between summer and winter (ANOSIM and PERMANOVA p < 0.05; Supporting Information . Ordination plots of fungal communities showed distinct clusters for the individual successional stages and seasons, apart from individual sample MB-S2 and the outlier samples of plot 23 [fig_ref] Figure 5: Non-metric multidimensional scaling [/fig_ref]. However, the separation by season became less prominent in the latest SSD (site 2), where clusters started to intermix. The odd placement of plot 23 samples can be explained by the location of this plot directly in line of a melting water stream. The deviation of sample MB-S2 could not be explained, but it highlights local heterogeneities as typical for alpine sites.
Some fungal taxa, especially of the class Microbotryomycetes (including Rhodotorula) showed a high abundance during the winter season in both sites 0 and 1 but were rare in the same plots during summer. Differential abundance testing (Welch's t-test) resulted in no significant differences for any OTUs when using FDRcorrection. However, when omitting multiple-testing corrections, the graphic display of significant OTUs (raw pvalue < 0.05) produced a strong visual pattern of contrasting seasonal abundances for several OTUs . In total, 28 OTUs displayed a seasonal preference in at least one of the three SSDs. For OTUs of the class Microbotryomycetes, the pattern is present over two consecutive successional stages (sites 0 and 1), reinforcing the hypothesis that these OTUs prefer one of the two contrasting seasons. This pattern is only indicative and not statistically significant. Especially in habitats with high heterogeneity, multiple-testing corrections might often be too stringent if sampling efficiency not high enough.
Amplicon sequencing detects a higher diversity, but rare fungal isolates are not always detected Amplicon sequencing detected a higher fungal diversity than the cultivation approach (220 OTUs vs. 91 cOTUs). Whole taxonomic groups were not detectable using cultivation techniques (e.g., Chytridio-, Rozello-, Monoblepharido-and Glomeromycota). However, 24 cOTUs did not have a close match in the amplicon sequencing data set (BLAST), which means that they were not detected. Most of these cOTUs were rare: 9 cOTUs were isolated only once and 14 cOTUs were isolated twice (Supporting .
# Discussion
Soil fungal communities closest to the glacier tongue are diverse and unique and rather resembling microbial communities of cryoconite than of soil We used amplicon sequencing and isolation of fungi from in-growth MBs, to identify active fungi in a rather marginally addressed transition phase from cryosphere to early soil development (0-25 years).
As soon as 2 years after melting of the ice sheet, specific community-style patterns of fungal mycobiota were detected on barren ground in the glacier forefield. Ascomycetes (Phialophora, Sarocladium, Mortierella) and psychrophilic basidiomycete yeasts preferentially colonized soil closest to the glacier. The latter were also . Heat map of fungal OTUs detected in MBs showing significant differential abundances (raw p-values, without multiple testing correction) in at least one of the three stages of soil development (site 0: 0-3 years, site 1: 14-19 years, site 2: 18-25 years free of ice). Winter fungal communities are most similar during the first two stages of soil development. The comparison of sequencing data retrieved from ingrowth mesh-bags (MB) compared to soil samples was carried out for winter samples in site 2. Data retrieved from soil samples are very similar to data retrieved from MBs. Black bars indicate the snow-free period, white bars the snow-covered period. [Color figure can be viewed at wileyonlinelibrary.com] reported to be highly abundant in other alpine glacier forefields [bib_ref] Vertical distribution of the soil microbiota along a successional gradient in a..., Rime [/bib_ref] , indicating an important role of these fungi as primary colonizers of mineral soil in the glacier forefield. Of all OTUs detected via amplicon sequencing, 35% were exclusively found in site 0. These early colonizers confirm a clear successional trend in the earliest SSD, because most of them disappeared in later SSDs, especially site 2. They are especially interesting, as they represent the transition from life adapted to ice to life adapted to soil. Soil fungal communities closest to the glacier tongue are unique and rather resembling microbial communities of cryoconite than of soil [bib_ref] A distinctive fungal community inhabiting cryoconite holes on glaciers in Svalbard, Edwards [/bib_ref]. This implies that glacial environments might be important reservoirs of fungal diversity. [bib_ref] A distinctive fungal community inhabiting cryoconite holes on glaciers in Svalbard, Edwards [/bib_ref] also described a potential aeroaquatic-endophyte cycle for cryoconite-associated Ingoldian fungi. We detected a high diversity of the genus Tetracladium, a typical representative of Ingoldian fungi. Cryoconite as inoculum explains very convincingly the high abundance of these endophytes in soil, as soon as the first pioneer plants occur, and further strengthens our hypothesis for dispersal between cryoconite and early soil.
The earliest hyphomycete communities are usually characterized by typical representative species of Helotiales [bib_ref] Vertical distribution of the soil microbiota along a successional gradient in a..., Rime [/bib_ref]. We detected one of them, namely a Phialophora species (OTU 790) almost exclusively in soil close to the glacier tongue (site 0), with high abundance throughout the year. The related ITS2 sequence was identical (100% identity, 100% coverage) to environmental sequences from arctic and alpine regions (Midre Lovenbreen Glacier Norway JN113039, Qinghai-Tibet plateau JX001631). The corresponding UNITE species hypothesis (SH013873.07FU) is typical for cold regions and was reported from Antarctica, the North American Arctic, Svalbard and Tibet. Besides being considered as a saprobe, it was also reported as endophyte of Chorisodontium aciphyllum, a moss frequently found along the Antarctic coast in the Drake Passage region [bib_ref] Diversity and cold adaptation of culturable endophytic fungi from bryophytes in the..., Zhang [/bib_ref]. Morphological characteristics of the related isolates (cOTU97) indicate that this could possibly be a new species of the recently described genus Psychrophila [bib_ref] Psychrophilic fungi from the world's roof, Wang [/bib_ref].
Drastic changes in fungal species composition coincide with the occurrence of first pioneer plants and the buildup of SOM. An increased nutrient availability due to SOM build-up has been shown to drastically influence yeast diversity in alpine glaciers [bib_ref] Influence of abiotic variables on culturable yeast diversity in two distinct Alpine..., Turchetti [/bib_ref]. On the Rotmoos glacier forefront, a 4-year period of ecosystem development (corresponding to our site 0) is too short to enable plant species to colonize bare soil. After about 10 years of exposure, soil is scattered with singlevascular plants (e.g., Saxifraga aizoides), and a high percentage of fungal species now appear for the first time on site, remaining present as soil development proceeds. Many of these fungal taxa are known for their various associations with plants, for example, as endophytes or pathogens (Helotiales, Botrytis and Neonetrica).
## Fungal communities follow a distinct successional pattern
The observed increase in fungal species richness and diversity with successional age appears to be typical for glacier foreland successional sites: Soil fungal community development is reaching a mid-successional diversity maximum at about 40 years of soil development [bib_ref] Soil fungal community development in a high Arctic glacier foreland follows a..., Dong [/bib_ref]. Three recent successional studies carried out in glacier forefields from Tierra del Fuego (Chile) [bib_ref] Microbial succession dynamics along glacier forefield chronosequences in Tierra del Fuego (Chile), Fernandez-Martinez [/bib_ref] , the Altai mountains (Russia) [bib_ref] The last 50years of climate-induced melting of the Maliy Aktru glacier (Altai..., Gatti [/bib_ref] and Tibet (China) [bib_ref] Ecological succession pattern of fungal community in soil along a retreating glacier, Tian [/bib_ref] also included the earliest sites in close vicinity to the glacier tongue. These studies share our observation that earliest fungal communities from barren ground closest to the glacier terminus are clearly segregated from the rest of the chronosequence. These abrupt changes in community composition at early SSD are also prevalent for other pioneer organisms like bacteria and algae.
## Seasonality is especially prominent in the earliest ssds
The fact that fungi are actively growing in soil during winter has been widely neglected for a long time, although early reported by. The rediscovery of fungal communities which are physiologically active under snow-cover [bib_ref] Seasonal dynamics of previously unknown fungal lineages in tundra soils, Schadt [/bib_ref] caused a renewal of interest in the scientific community, leading to the discovery of many unknown fungal species or even lineages. On this basis, an earlier study on fungal seasonal dynamics was carried out at a late successional site (~150 years of soil development) of the Rotmoosferner glacier forefront. This study demonstrated that fungal biomass can increase more than 10-fold underneath the snow cover [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref]. We also observed a change in soil fungal community composition. Seasonal shifts in soil fungal species composition are also known for other well-developed soils [bib_ref] Fungal community on decomposing leaf litter undergoes rapid successional changes, Voříšková [/bib_ref] [bib_ref] Seasonal dynamics of fungal communities in a temperate oak forest soil, Voříšková [/bib_ref].
In the present study, we focused on seasonal dynamics in the earliest SSD and detected clear seasonal shifts in fungal community composition between the short summer and the long winter period. However, seasonality became less prominent in the latest SSD. Several fungal taxa with seasonal preferences were detected: OTUs with higher abundance in winter (especially in early SSDs 0, 1) were predominantly found in the class of Microbotryomycetes. Taxonomic resolution at the genus level was available only for OTUs belong to the genus Rhodotorula (psychrotolerant yeasts). Strains of Microbotryomycetes (cOTUs 62 and 64) were also frequently isolated from winter MBs buried in these sites. They were preliminarily identified at the genus level and are probably representing new psychrophilic yeast species of Cryptococcus, Rhodotorula and Mrakia. Studies on cultivable yeast communities from other Alpine glaciers (Glacier du Geant, France, and Miage Glacier, Italy) [bib_ref] Influence of abiotic variables on culturable yeast diversity in two distinct Alpine..., Turchetti [/bib_ref] confirm that distinct psychrotolerant basidiomycete yeasts are representing a stable core of the sub-glacial yeast communities; these communities are not only typical for glaciers [bib_ref] Yeasts in high Arctic glaciers: the discovery of a new habitat for..., Butinar [/bib_ref] [bib_ref] Yeast and yeast-like diversity in the southernmost glacier of Europe (Calderone Glacier, Branda [/bib_ref] , cryoconite [bib_ref] Characterization of heterotrophic microorganisms in alpine glacier Cryoconite, Margesin [/bib_ref] [bib_ref] Characterization of yeast and filamentous fungi isolated from cryoconite holes of Svalbard, Singh [/bib_ref] , but also for glacial melting water [bib_ref] Biodiversity of cold-adapted yeasts from glacial meltwater rivers in Patagonia, De García [/bib_ref] and arctic coastal environments [bib_ref] Relative incidence of Ascomycetous yeasts in Arctic coastal environments, Butinar [/bib_ref]. This implies that cryoconite and melting water must be considered as a possible inoculum for the native soil close to the glacier tongue.
Ascomycota growing as hyphomycetes were found to dominate winter communities in site 2 (18-25 years of soil development), where they made up two-third of the mycobiota over all plots; especially, various OTUs of the genus Tetracladium and other non-closer defined OTUs within the same order Helotiales were responsible for these high abundance values.
The detection of a biotrophic plant pathogenic fungus belonging to Pucciniomycetes (OTU 208) in the intermediate SSD (site 1) during winter only was surprising. This could suggest a saprobial lifestyle during winter, when the host is not active or alive, and a switch back to biotrophic lifestyle as soon as the host is physiologically active again during the vegetation period. Due to the presence of this OTU in 4 of 5 plots a contamination is highly unlikely.
Furthermore, we also detected Chrysomyxa sp. (OTU 205), another biotrophic rust fungus in low relative abundances (<1%) in all SSDs, but especially in barren ground (site 0: 4 plots in summer, 1 plot in winter). Due to the obvious absence of vascular plants, we assume an alternative lifestyle (e.g., saprobial yeast) or unknown intermediate host (e.g., algae). This indicates that the teliospores (the thickwalled resting spores of rust fungi) produce basidia and basidiospores in snow-covered soil. Basidiospores of rusts bud and persist as yeasts, and they have been detected in soil habitats [bib_ref] Impact of Endochitinase-transformed white spruce on soil fungal biomass and ectendomycorrhizal yymbiosis, Stefani [/bib_ref]. Habitats of host generalist and specialist fungi (both pathogens and mutualists) can exceed those of their host species [bib_ref] Spatial patterns of pathogenic and mutualistic fungi across the elevational range of..., Merges [/bib_ref]. However, the presence of a host plant has indeed a positive effect on the abundance of hostspecialized pathogenic fungi.
However, we also found several OTUs displaying higher relative abundances in summer than winter. Alternaria, Epicoccum and Pithomyces were found to be more active during the vegetation period. In addition, we detected a species of Cladosporium (OTU 786), which was active in both seasons over all SSDs, but whose abundance was generally higher in summer than in winter . Cladosporium species have multiple relationships with plants (endophytes, epiphytes, pathogens) [bib_ref] The genus Cladosporium, Bensch [/bib_ref] , explaining higher abundances during the vegetation period.
Several patterns of seasonal preference were visible over multiple successional stages, but not detected when applying multiple-testing correction (FDR). This indicates that FDR was too stringent in this case and stresses how the visualization of significant results without multipletesting correction might prevent missing relevant biological information.
In-growth MBs allow for a meaningful detection and isolation of winter-active fungi, although having some drawbacks
In-growth MBs are generally used to capture biomass of hyphae-forming fungi, which were actively growing during the incubation period. The biomass colonizing or growing through the sterile substrate (quartz-sand) is then used for isolation or DNA-based identification of fungi. The use of MBs instead of soil samples aims for preventing detection/isolation of inactive (spores, sclerotia) or dead fungal material [bib_ref] Estimation of the biomass and seasonal growth of external mycelium of ectomycorrhizal..., Wallander [/bib_ref]. In addition, fungal hyphae can be visualized and quantified from MB content [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref].
Although the MB approach might be suitable to detect and quantify in-growth of hyphae-forming fungi, we did notice problems occurring with non-hyphal, single-celled organisms. Yeasts or spores can be washed into MBs with a mesh size of 53 μm and revive upon isolation/cultivation procedure. A smaller mesh size should be used to prevent that in the future.
The in-growth MB approach is an interesting alternative to soil sampling, when trying to assess fungal organisms growing during defined periods Based on amplicon sequencing of winter MBs and the surrounding soil, we could not detect any significant differences in fungal community composition (ANOSIM, PERMANOVA). This was surprising, because we expected a higher diversity in soil samples due to co-amplification of inactive fungal structures. We hypothesise that fungal inoculum is either generally introduced into this environment in very small amounts only, or that it is quickly removed from the system by high turnover rates, as introduced fungal propagules might easily get washed out, blown away or consumed by soil invertebrates. This means that fungal spore banks have not been established yet in these SSD. Based on differential abundance test of soil versus MBs samples, only two OTUs were exclusively found in soil and one exclusively in MBs: A Cadophora-/Phialophoralike fungus (OTU875) appeared to be typical for summer MB samples but was absent in most (4/5) winter MBs. Still, this fungus appeared to be frequent in winter soil samples (4/5), but only in those retrieved from under the snow cover. Cadophora-/Phialophora-like fungi are typical dark septate endophytes, which are frequently occurring in the roots of arctic/alpine pioneer plants on a global scale [bib_ref] The potential for mycobiont sharing between shrubs and seedlings to facilitate tree..., Hewitt [/bib_ref]. During summer, (extra-radical) mycelia are foraging through the ground, explaining why they are frequently detected in MBs. During winter, roots harbouring these fungi are degraded and spores are liberated in the soil, where they were detected. They were not growing and therefore not detected as in winter MBs. In this case, MBs might have worked as intended by excluding inactive fungal propagules. Low abundances in combination with high turnover rates or washout could explain why these were not detected in soil shortly after snow melt. Murispora (OTU725) **a plant-associated, dothideomycete is known for its large, muriform ascospores. This OTU was detected in summer MBs from middle and late SSDs and in snow-covered soil samples. Arbuscular mycorrhizal fungi (Glomeromycota, OTU364) were only detected in sites with plant cover (site 2). At this site, they were detected in summer MBs as well as in winter soil samples, however, not in winter MBs. The spores of arbuscular mycorrhizal fungi are typically very large (50-200 μm) and therefore too big to enter MBs. Still, they are present in the surrounding soil as inactive states and could therefore easily be detected in winter soil samples. This indicates that the concept of assessing only actively growing fungi by using in-growth MBs is working for Glomeromycota with a spore size above mesh size.
Summing up, the MB approach can be regarded as an interesting alternative to soil sampling, when trying to assess active fungal organisms during certain seasons. However, yeasts and small fungal spores pose problems to this method. Due to the large mesh-size used (53 μm), most detected OTUs could also have been washed into MBs. We found no differences in the retrieved fungal communities with the two sampling strategies (MBs vs. soil), but we compared only winter MBs with the surrounding soil, and this is a habitat with extremely low SOM content. Results might be different in other environments. Nevertheless, we regard the MB approach as very valuable, especially for the isolation and cultivation of fungi. It might also be appropriate for studying the activity of specific fungal groups (e.g., Glomeromycota).
Species turnover takes more than 2 weeks after snow-melt, showing that there are no short-term effects of snow melt on soil fungal communities
The comparison of fungal communities derived from snow-covered versus 2-weeks snow-free soil samples in SSD 2 did not show any significant differences between these two time points. Thus, it can be assumed that changes in community composition take several weeks. Factors affecting the rate of species turnover are still obscure, but there is obviously a lag-phase between seasons. Nutrient availability or the slow development of trophic interactions after snowmelt could be one important factor. Snowmelt itself might intensify the nutrient scarcity due to washout by melting water. Still, it has to be considered that this comparison was carried out only for the most developed SSD-site 2. In this stage, the seasonal shift of soil fungal communities was not as prominent as in less developed SSDs. Site 2 already has a mosaic-like plant cover with a substantial amount of organic matter, able to hold water and nutrients on site. Such organic layers were only present as isolated colonies in site 1 and were completely lacking in site 0. The investigation of this short-term effect was included in our study due to an unexpected event in the field study: An avalanche had knocked down the snow poles, which made the localization of plots from site 1 under the snow cover impossible. Winter samples could therefore be retrieved immediately after snowmelt, only. To estimate the effect of this late retrieval, we used site 2 as the nearest proxy for a comparison of short-term effects following snowmelt. The results clearly show that 2 weeks of being snow-free does not significantly influence fungal community composition. This was important for validation of the winter samples of site 1.
## Cultivation of winter-active fungi is highly biased but has immense importance and potential
Results of cultivation-based and cultivation-free methods were widely in accordance, at least when considering abundant fungal taxa. However, there is a restriction to fungal groups, which can be isolated with the methods applied, and the cultivation approach is much more tedious and time consuming than the DNAbased approach. Therefore, it is not surprising that fungal diversity studies based on targeted environmental sequencing combined with cultivation are comparatively rare (e.g., [bib_ref] Fungal diversity in oxygen-depleted regions of the Arabian Sea revealed by targeted..., Jebaraj [/bib_ref].
As expected, the proportion of psychrophilic fungi (no growth at 25 C) among isolates (cOTUs) was generally high, considering the soil temperatures range between 0 C and 10 C throughout the year in the glacier forefield. Several cultivation approaches were tested to find the best, but most effective way of cultivation. The results of cultivationbased methods were strongly depending on the applied method of isolation (direct vs. dilution plating). Thus, the combination of both methods might provide a better picture of the fungal diversity. The enormous workload of cultivation and identification of isolate does not allow for extensive replicates, and thus, a statistical evaluation of the results is not possible. However, it enables an estimation of the diversity of the most abundant fungi in a habitat. More importantly, it allows for a thorough study of the obtained isolates. This is especially important considering the high degree of unknown biology of the identified fungal taxa. It is very likely that most of our isolates are saprobial, but representatives for many other functional groups were also isolated: for example, potentially mycorrhizal (Russuales sp.1, Sebacinaceae sp.), plant endophytes (Tetracladium spp.), plant pathogens (Taphrinaceae sp., Heterobasidion sp.) and entomopathogenic fungi (Lecanicillium spp., Hypocreales sp.). Fungal isolates do not only offer the possibility to study, test and describe new fungal taxa. They also offer the possibility to test strains for a variety of applications of psychrophilic fungi, for example, for special coldstable enzymes. The number of cOTUs obtained via isolation was highest in the youngest SSD, which is in contrast to the NGS data, where fungal richness increased along the successional gradient (cf. rarefaction plots). This could either be explained by the lack of 'ecological' replicates for each SSD (only one plot per season and SSD tested). However, we rather prefer a second hypothesis, namely that this bias was caused by the strong selectivity of cultivation methods in contrast to cultivation-free methods. Along the chronosequence, we find more and more fungal groups strictly associated to plants. Many of these plant-associated fungi rely on interactions with their hosts for growth, for example, biotrophic parasites or mycorrhizal fungi, and can therefore not be cultivated. Thus, species diversity of cultivable fungi decreases with increased proportion of plantassociated fungi. When assessing fungal diversity, amplicon NGS is without question the more sensitive method than cultivation-based approaches. However, more than a quarter of the cOTUs (27%) did not have a matching BLAST reference in the amplicon data set (Supporting Information . The number of cOTUs without match could be reduced when using the non-filtered dataset as reference, although. Nevertheless, 17 cOTUs could still not be detected by NGS sequencing. Their absence could be explained by their low abundance (30% of all isolates were retrieved only once), and the stringent quality control measures in the NGS data analysis. Furthermore, there are several technical issues related to NGS generally, which are known to alter the recorded biodiversity. These include biases regarding DNA extraction, PCR primers, variation in rDNA copy numbers per genome, number of nuclei per 'individual', and so forth; the use of abundance of sequence reads as a proxy for abundance of individuals is working fine for the most abundant taxa but is far from perfect [bib_ref] Quantifying microbial communities with 454 pyrosequencing: does read abundance count?, Amend [/bib_ref] [bib_ref] Efficient and Accurate OTU Clustering with GPU-Based Sequence Alignment and Dynamic Dendrogram..., Nguyen [/bib_ref].
# Conclusion
Global warming accelerates glacier retreat and causes a reduction of snow-covered landscapes on a global scale.
Snow cover is an excellent thermal insulator, keeping the ground temperature warmer than the air temperature during cold seasons. This insulation effect enables and promotes the growth of the winter-active soil microbiota. A pronounced seasonal species turnover between snow covered soil and soil with plant cover has already been reported for different habitats (e.g., forests, agricultural soil, tundra soil). However, this is the first study showing that fungal diversity is not only comparatively high in barren ground closest to the glacier tongue, but it is also subject to distinct seasonal species turnover. Environmental studies are usually based on sampling during the vegetation period only. However, by doing so, the generated data reveal only one side of the coin, and a high proportion of the real microbial diversity and function remains overlooked.
Fungal communities of recently deglaciated soil are significantly different from communities of later SSD and are rather similar to fungal communities typical for cryoconite or ice. This indicates different inoculum sources for the earliest ice-related barren ground and for later plant-covered soil. The highly dynamic ice surfaces of a glacier are characterized by continuous melting and freezing. It is therefore very likely that the earliest ice-free barren ground obtains its fungal inoculum from cryoconite via wind-borne propagules or via fungal structures transported by melting water.
## Experimental procedures
## Sampling site description
The study site is located in the forefield of the Rotmoos glacier, in the Central Austrian Alps (Ötztal, 46 50 0 N, 11 03 0 E) at 2280-2450 m above sea level. Since 1858, the Rotmoos glacier has retreated for about 2 km, and due to its low slope and wide profile, the forefield is well suited for studies of primary succession [bib_ref] Pflanzliche Sukzession im Gletschervorfeld. Vegetation und Besiedlungsstrategien, Erschbamer [/bib_ref]. Towards the glacier tongue, all soils are influenced by the basement rocks of the Ötztal-Stubai complex and by the Schneeberg complex (part of the Austrian nappe stack system), which are a remnant of a strongly carbonate-influenced sediment layer [bib_ref] Geologie und Geomorphologie von Obergurgl und Umgebung, Krainer [/bib_ref].
Three areas with distinct SSD (sites 0, 1, 2) were chosen close to the glacier tongue for comparative studies of active soil mycobiota in the earliest stages of soil formation. The three different sites have been free from glacial ice for 0-3 years (site 0), 9-14 years (site 1) and 18-25 years (site 2) respectively. The SSD reach from non-vegetated, barren ground (site 0), over barren ground colonized by individual pioneer plants (site 1), to soil with a partial, mosaic-like vegetation cover (site 2) [fig_ref] Figure 1: A-E Sampling sites at the Rotmoosferner glacier forefield at 2280-2450 m above... [/fig_ref]. In each site, eight representative subplots of 1 m 2 were chosen as random replicates for each SSD. Soils are usually covered with snow from mid-October to late May [bib_ref] Invertebrate succession on an alpine glacier foreland, Kaufmann [/bib_ref]. Soil temperatures range from −0.6 to −0.1 C during the snow-covered period and from 0 to 22 C during the vegetation period [bib_ref] Fungal growth and biomass development is boosted by plants in snow-covered soil, Kuhnert [/bib_ref].
## Sampling strategy
In each plot (8 per SSD), eight in-growth MBs were incubated for a season of prolonged snow-cover (referred to as 'winter MBs') or the vegetation period ('summer MBs') respectively to assess active fungal communities in two contrasting seasons. Active hyphae-forming fungi move through the soil by growth. Occasionally, these hyphae will pass through the randomly distributed MBs, leaving behind biomass on the filling (acid-washed quartz sand). The biomass can then be used either for the isolation of cultivatable fungi or for their detection via DNA-based methods [bib_ref] Estimation of the biomass and seasonal growth of external mycelium of ectomycorrhizal..., Wallander [/bib_ref]. The applied MBs were comparatively small (1 g quartz-sand), facilitating nutrient diffusion, as well as easy passage of fungi. Summer MBs were buried in soil in mid-July (11 th and 16 th ) 2015, incubated for 2 months during the vegetation period and retrieved on September 18, 2015. Winter MBs were buried in soil on September 18, 2015, and incubated during winter. Plots were marked with 2 m long aluminium rods (serving as snow poles) to enable localization underneath the snow cover. Winter MBs were retrieved from soil while still covered by snow in June and July 2016, with the exception of plot W2 (site 0) and all plots of site 1, which were sampled about 2 weeks after snow melt. An avalanche had knocked down the aluminium rods of site 1 and made localization of plots under the snow sheet unfeasible. Upon retrieval, MBs from each plot were placed into separate plastic bags to avoid crosscontamination and stored on ice during transportation and sample processing. Soil samples for physicochemical analyses were retrieved at the beginning of the vegetation period in June 2015. in site 2 (18-25 years ice free), additional soil samples for DNA-based analyses were retrieved from both, snow-covered soil and 2 weeks after snow melt in 2016, to check for differences between MB and soil samples, as well as for short-term changes in fungal communities upon snow melt.
## Soil physicochemical analyses
From each plot, 5-8 small spades of soil (ca. 30 g in total) were taken randomly from 1-5 cm depth. Samples from each plot were mixed, sieved to 2 mm and stored at −20 C. During analytical procedures, soil was stored at 4 C in polyethylene bags. Soil pH was measured potentiometrically in 0.01 M CaCl 2 . Soil dry weight was determined by drying samples at 105 C. The determination of organic matter by the loss-on-ignition method and the semi-quantitatively estimation of carbonate (CaCO 3 ) were carried out as described by. Ammonium (NH 4 ), nitrate (NO 3 ), plant-available and total phosphate (PO 4 ) were measured as described by. Soil physicochemical data were analysed in STATISTICA (ver. 9.1 Statsoft Inc.). Normal distribution of variables was checked via QQ-plots. Depending on distribution, either a parametric ANOVA and Tukey HSD test or non-parametric Kruskal-Wallis tests were used to examine differences of variables between soil development stages (sites 0, 1, 2). Furthermore, spearman correlations were calculated between all soil parameters over all sites.
## Cultivation of soil fungi from in-growth mbs
Fungi were isolated from MBs by direct-plating and dilution plating [bib_ref] Influence of abiotic variables on culturable yeast diversity in two distinct Alpine..., Turchetti [/bib_ref]. For each site and season, a single plot was chosen, of which three MBs were processed. For direct plating, 30 quartz-sand grains were used per MB and medium. For dilution plating, 1 g of quartz-sand particles was suspended 1:10 (w/v) in 0.1% tetrasodium pyrophosphate and shaken overhead for 30 min. Serial dilutions (10 −1 to 10 −3 ) were prepared, and 100 μl were plated on potato dextrose agar (PDA, Carl Roth, Germany) and synthetic nutrient-poor agar (SNA), with three replicates for each medium. To suppress bacterial growth, media were supplemented with antibiotics (100 mg L −1 streptomycin, 50 mg L −1 tetracycline). Plates were incubated at 4 C and 10 C in parallel for >4 weeks and checked for growth regularly. Pure cultures were transferred to PDA without antibiotics and further incubated at both the isolation temperature and at 25 C to allow for a standardized morphological investigation and determination of psychrophilic growth. Morphological identification of isolates was based on general literature for soil fungiand monographs on the respective genera. Moreover, sequencing of rDNA ITS barcoding region was performed via direct colony PCR as described by using the primers ITS1 and ITS4 [bib_ref] Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics, White [/bib_ref]. If direct colony PCR failed, DNA was extracted from pure cultures using the E.Z.N.A. HP Fungal DNA Kit (Omega Biotek) before PCR amplification. Sequencing of PCR products was carried out at MicroSynth AG (Switzerland). Quality-checked sequences were assembled into 'cultivated operational taxonomic units' (cOTUs, term to discern from NGS-generated OTUs) at 99% and 97% sequence similarity using Sequencher . Taxonomy was assigned by closest hits after manual BLAST runs against International Nucleotide Sequence Database** Collaboration (INSDC; http://www.insdc.org/ [bib_ref] The international nucleotide sequence database collaboration, Karsch-Mizrachi [/bib_ref] and UNITE [bib_ref] Towards a unified paradigm for sequence-based identification of fungi, Kõljalg [/bib_ref]. A representative sequence of each 99% OTU was submitted to GenBank (Supporting . Pure cultures are stored in the culture collection at the University of Innsbruck, Institute of Microbiology .
The presence/absence data of isolated fungi were used to check if patterns of fungal activity derived from the cultivation approach are in concordance with amplicon data. Furthermore, the ITS2 region of each 97% OTU was extracted using ITSx and blasted locally against reference sequences of the filtered NGS data set to check if cultivation can detect OTUs that might go unnoticed by the NGS approach.
## Dna extraction and amplicon sequencing
To reduce the effect of small-scale habitat heterogeneity, the content of three MBs per plot were pooled and used for DNA extraction. MBs were opened with a flamesterilized scalpel, and the content was poured into a sterile petri dish. Five ml of quartz sand (roughly content of 3 MBs) were filled into a 5 ml Eppendorf tube, and 3 ml extraction buffer were added. Tubes were sealed with Parafilm ® and put on a vortex mixer for 5 min at 3000 rpm. The addition of an extra bead-beating matrix was omitted as the MB samples consist of sharp grains (silicate), which were expected to shear attached biomass. The lysate was incubated at 65 C for 60 min and regularly mixed by inverting tubes every 10 min. The liquid lysate (~2 ml) was withdrawn with a sterile plastic syringe (Braun Injekt™) and needle (Braun Sterican™ 20 G × 1.5 00 ). Keeping sample amount high and buffer volume low was necessary to obtain a sufficient concentration of DNA. The lysate (1.2 ml) was aliquoted into two 1.5 ml tubes, and extraction was continued with the E.Z.N.A. HP Fungal DNA Kit (Omega Bio-tek). The protocol was modified by including a column equilibration with 3 M NaOH (optional, Product manual November 2015) and the withdrawal of 400 μl of aqueous phase (instead of 300 μl) following chloroform-isoamyl alcohol extraction to increase yield. Volumes of the CXD buffer and ethanol were adjusted accordingly. Tubes were incubated 5 min at 65 C upon addition of elution buffer to increase DNA yield. Extraction success was checked by DNA concentration measurement using Picogreen ® and by test PCR reactions targeting the ITS2 region. Amplification success was checked via gel electrophoresis. Based on the results of test PCRs (criterion: strong product in both summer and winter MB samples), five plots of each SSD (sites 0, 1, 2) were selected for fungal community profiling using amplicon metagenomics.
For soil DNA, 250 mg of sieved soil were used for DNA extraction using the E.Z.N.A. Soil DNA kit (Omega Biotek). Soil samples were included for comparison with MB data. One SSD and one season (site 2, winter) were used for this comparison, leading to 40 NGS-analysed samples in total.
A two-step PCR protocol was used to generate sequencing libraries targeting the rDNA ITS2 region. The first reaction was carried out in a volume of 25 μl, with 5 μl template, 1× KAPA HiFi buffer, 0.2 mg ml −1 bovine serum albumine, 0.3 mM of each dNTP, 0.3 μM of each forward and reverse primer and 1 unit of a high-fidelity polymerase (KAPA HiFi HotStart, Kapa Biosystems). Primers consisted of a locusspecific sequence (ITS3, ITS4 [bib_ref] Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics, White [/bib_ref] and the Illumina adapter). For the forward primer (ITS3), five random bases (N) were included in between the primer regions to increase sequence diversity and cluster template registration by the Illumina system. This increases overall sequencing quality of low diversity libraries runs like any amplicon-based library (Anonymous, 2018). After initial denaturation at 95 C for 4 min, PCR was performed with 25 cycles of 98 C for 20 s, 55 C for 30 s and 72 C for 30 s, with a final elongation of 5 min at 72 C. PCR was performed in triplicates for each sample to reduce PCR-related biases and increase sensitivity . Triplicates were pooled and cleaned up using the GeneElute™ PCR Clean-Up Kit (Sigma-Aldrich). Purified PCR products were sent to Microsynth AG (Switzerland) for second step PCR. The number of cycles in the second step PCR was raised from 15 to 20 to increase yield. Libraries were pooled in an equimolar manner, and paired-end sequencing (2 × 250 bp) was conducted on an Illumina MiSeq platform. Sequencing results were provided as de-multiplexed fastq files.
## Bioinformatics and statistical analysis of amplicon data
Raw sequence reads (n = 1,874,918, forward and reverse) were processed with the pipeline PIPITS (ver. 1.5.0, Gweon et al.) on a standard desktop PC with 8 GB RAM running Ubuntu 16.04. The fungal ITS2 region was extracted using ITSx to increase resolution in downstream sequence clustering and remove reads of non-fungal origin [bib_ref] Improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS..., Bengtsson-Palme [/bib_ref]. Short (<100 bp) and unique sequences were removed before OTU finding, as these are likely to have emerged as sequencing errors [bib_ref] UPARSE: highly accurate OTU sequences from microbial amplicon reads, Edgar [/bib_ref]. Sequences were clustered to OTUs at 97% sequence identity. Chimera detection was run with UCHIME [bib_ref] UCHIME improves sensitivity and speed of chimera detection, Edgar [/bib_ref] based on the UNITE UCHIME reference dataset (01.01.2016). Taxonomy was assigned using the RDP classifier and the latest trained UNITE fungal ITS reference dataset (UNITE 7.2; 28.06.2017) as implemented in PIPITS. Sequence data were deposited in the European Nucleotide Archive (ENA) under the project accession PRJEB30879. The biom file produced by PIPITS was combined with sample metadata using biom and analysed in R (3.3.2) (Supporting . Non-fungal lineages overseen by ITSx (n = 20, mostly Chromista) were removed from the dataset. Rarefaction plots were drawn to check whether sequencing depth sufficiently covered fungal diversity (Supporting Information [fig_ref] Figure 2: Fig [/fig_ref]. Based on these results, all samples were rarefied (subsampled with replacement) to 4020 reads per sample (size of second smallest sample) to mitigate the effect of uneven sample sizes on downstream analyses, while retaining as much sequence diversity as possible. Furthermore, only OTUs with an abundance of at least 3 reads in a minimum of 3 samples (note: total number of replicates per SSD and season is 5) were retained to remove low abundance OTUs that might represent contaminations. For the filtered dataset, manual BLASTn searches were run against the International Nucleotide Sequence Database Collaboration (INSDC) with exclusion of uncultured and environmental sample sequences. Taxonomy was assigned up to genus level, when a reliable match was obtained within 97% sequence identity and > 90% coverage. Hits were considered reliable when multiple references displayed the same or equivalent results. The results provided by the UNITE and NCBI databases were checked for concordance in the assigned taxonomy. If NCBI provided a reliable higher taxonomic resolution, the taxonomy was adjusted in the data. Fungal richness and diversity was investigated based on observed OTUs and the Shannon-diversity index. Differences between diversity indices were statistically evaluated using ANOVA and post hoc tests (Tukey-HSD). Data requirements for parametric tests have been confirmed beforehand (Shapiro-Wilk test). Dissimilarities between samples were displayed by an ordination-based heat map [bib_ref] NeatMap-non-clustering heat map alternatives in R, Rajaram [/bib_ref] and by non-metric multidimensional scaling (NMDS) plots, based on a Bray-Curtis dissimilarity matrix as calculated from the rarefied, filtered and square root transformed OTU abundance data. Venn diagrams were used to show shared species between SSDs for periods of snow-cover (W) and the vegetation period (S) respectively (Supporting Information . Differences among a-priori defined groups were tested by analysis of similarity (ANOSIM) and permutational nonparametric MANOVA (PERMANOVA). Homogeneity of dispersion among groups was confirmed [bib_ref] Distance-based redundancy analysis: testing multispecies responses in multifactorial ecological experiments, Legendre [/bib_ref] using the function betadisper in the R package vegan. This is necessary as significant differences detected by PERMANOVA may arise due to differences in dispersions between compared groups [bib_ref] Distance-based tests for homogeneity of multivariate dispersions, Anderson [/bib_ref]. To test for the main working hypotheses the following subsets were analysed individually using ANOSIM and PER-MANOVA: (i) >MB samples< to investigate the effect of season and SSD on fungal community composition. (ii) >Soil samples< to compare fungal community composition in soil samples with or without snow-cover (short-term effects). (iii) >Winter samples from site 2 taken from beneath snow-cover< to compare fungal communities retrieved from MBs versus soil samples (methodological comparison). For the MB samples, unequal variances ttests ('Welch's t-test') were conducted for each SSD separately to identify OTUs with a significant difference in their mean abundance between seasons. A correction for false discovery rate (FDR; [bib_ref] Controlling the false discovery rate: a practical and powerful approach to multiple..., Benjamini [/bib_ref] was included in multiple testing. OTUs, which displayed a seasonal pattern in at least one of the three SSD where visualized in a heat map (even if the FDR-adjusted pvalues were >0.05, if raw p-values were <0.05).
[fig] Figure 1: A-E Sampling sites at the Rotmoosferner glacier forefield at 2280-2450 m above sea level. A. Rotmoosferner glacier forefield, the arrow is indicating the sampling area close to the glacier tongue. B. Representative sampling plot of site 0 (0-3 years ice free). C. Sampling plot of site 1 (7-14 years ice free). D. Overview of the sampling area, black frames indicate the sampling sites with different stages of soil development ranging from 0 to 25 years (site 0, site 1 and site 2). E. Sampling plot of site 2 (18-25 years ice free). [Color figure can be viewed at wileyonlinelibrary.com] [/fig]
[fig] Figure 2: Fig. 2. The presence/absence of fungal cOTUs (97% ITS sequence identity) isolated from in-growth mesh bags, which were buried during the snow-free vegetation period (summer = S) or the snow-covered period (winter = W), respectively. The three early stages of soil development are site 0 (0-3 years), site 1 (7-14 years) and site 2 (18-25 years). Seasonal patterns are especially pronounced during the first 0-3 years of soil development. [Color figure can be viewed at wileyonlinelibrary.com] [/fig]
[fig] Figure 5: Non-metric multidimensional scaling (NMDS) plot of fungal communities detected in mesh bags (MB) and soil samples. NMDS is based on a Bray-Curtis dissimilarity matrix of the rarefied, filtered and squareroot transformed OTU abundance data (NGS data). Stress = 0.157. Colours indicate site 0 (lilac), site 1 (green), site 2 (blue). Soil samples were investigated in site 2 during winter only and are represented by pale blue colours. Circles represent summer samples, full triangles winter samples under snow-cover and nested triangles winter samples retrieved 2 weeks after now-melt. Fungal communities occurring during the first 3 years of soil development most are distinct and seasonally different. [Color figure can be viewed at wileyonlinelibrary.com] [/fig]
[fig] Figure 6: Ordination-based (NMDS) heat map of 100 most abundant OTUs detected by amplicon sequencing. Data show a clear separation of samples between stages of soil development (site 0: 0-3 years, site 1: 14-19 years, site 2: 18-25 years free of ice). The comparison of sequencing data retrieved from in-growth mesh-bag (MB) samples compared to soil samples was carried out for winter samples in site 2. Both winter soil samples (with snow cover and 2 weeks after snow-melt) have a highly similar pattern compared to the MB sample from the same site. Black bars indicate the snow-free period, white bars the snow-covered period. [Color figure can be viewed at wileyonlinelibrary.com] [/fig]
|
Leukoencephalopathy in Mitochondrial Neurogastrointestinal Encephalomyopathy-Like Syndrome with Polymerase-Gamma Mutations
# Introduction
Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE) syndrome is caused by mutations in the thymidine phosphorylase (TYMP) gene that lead TYMP deficiency and elevated plasma thymidine levels. Clinical manifestations include severe gastrointestinal dysmotility, cachexia, ptosis and ophthalmoparesis, peripheral neuropathy, and leukoencephalopathy. Patients with a clinical phenotype indistinguishable from MNGIE have been reported in the absence of TYMP gene mutations; their condition has been called MNGIE-like syndrome. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalopathy due to mutations in RRM2B, Shaibani [/bib_ref] [bib_ref] Altered cerebral glucose metabolism in a family with clinical features resembling mitochondrial..., Lehnhardt [/bib_ref] A handful of cases of MNGIE-like syndrome have been reported in association with polymerase-gamma (POLG) gene. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] van Goethem et al. first described two sisters with clinical features mimicking MNGIE linked to mutations in POLG. Tang et al. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] and Prasun et al. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] reported four patients presenting MNGIE with mutations in POLG. All cases shared similar brain magnetic resonance imaging (MRI) findings indicating the absence of leukoencephalopathy. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] Here, we report, for the first time, a patient with MNGIE-like syndrome involving leukoencephalopathy and associated with novel POLG mutations.
## Case report
The Chinese male had a history of good health until 42 years old when he developed mild gastrointestinal dysmotility leading to diarrhea and episodes of abdominal pain. At age 48, he was hospitalized for ptosis, diplopia, and weakness in the extremities. At the same time, gastrointestinal manifestations developed significantly, such that he began to suffer chronic diarrhea and persistent abdominal distention, leading to loss of 22 kg.
Colonoscopy revealed colonic diverticulum. Abdominal computed tomography showed gastroduodenal expansion. Peripheral nerve conduction velocity findings were consistent with demyelinating and axonal sensory motor neuropathy. Protein in the cerebrospinal fluid was elevated (119 mg/dL; normal: 15-45 mg/dL) while no leukocytes were detected. The patient was initially diagnosed with chronic inflammatory demyelinating polyneuropathy and given corticosteroids and intravenous immunoglobulin. These treatments had no effect.
At age 49, the patient was hospitalized again for progressive limb weakness. His appearance was cachectic, body mass index was 13.8 kg/m 2 , and he showed muscle atrophy. Neurological examination showed ophthalmoplegia, ptosis, muscle weakness, and absence of deep tendon reflexes. In light of his disease course, we suspected that he suffered from MNGIE or MNGIE-like syndrome.
Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE) syndrome, caused by mutations in the thymidine phosphorylase gene, manifests as a multisystemic disorder characterized by severe gastrointestinal dysmotility, cachexia, ptosis and ophthalmoparesis, peripheral neuropathy, and leukoencephalopathy. These clinical manifestations, with the exception of leukoencephalopathy, are mimicked by MNGIE-like syndrome, linked to polymerase-gamma (POLG) gene. Here, we report a 49-year-old Chinese man with MNGIE-like syndrome involved leukoencephalopathy and was associated with novel POLG mutations. This case expands the clinical spectrum of MNGIE-like syndrome. Subsequently, we arranged the muscle biopsy, and the result revealed ragged-red fibers and cytochrome oxidase (COX)-negative fibers. Brain MRI showed bilateral periventricular white matter hyperintensities in fluid-attenuated inversion recovery and T2-weighted images [ [fig_ref] reFerences 1: Hirano M, Nishigaki Y, Martí R [/fig_ref] ]. The medical history of the patient did not suggest alternative causes of leukoencephalopathy, such as intoxication, infections, or hypertension. Next-generation sequencing of mitochondrial and nuclear genomes from skeletal muscle tissues identified two novel heterozygous variants in POLG: c.3643 + 1G > A (splicing), near exon 23 and c.2396C > A (p. S799Y), in exon 14. These mutations are located in the polymerase domain and were verified by direct DNA sequencing.
Neither variant was found in any of the databases consulted, including dbSNP, HapMap, and 1000 Genomes, or among 500 healthy Chinese samples. Protein function prediction using Polyphen-2, SIFT, and MutationTaster suggested that the missense variant is likely to be damaging.
# Discussion
Here, we report a Chinese man with clinical features of MNGIE. Histological examination showed ragged-red fibers and COX-negative fibers. Moreover, the clinical manifestations could not be attributed to other phenotypes of POLG-related disorders, such as Alpers-Huttenlocher syndrome, myoclonic epilepsy myopathy sensory ataxia, ataxia neuropathy spectrum, or chronic progressive external ophthalmoplegia. [bib_ref] Clinical and molecular features of POLG-related mitochondrial disease, Stumpf [/bib_ref] The patients carried two novel variants in POLG which were not detected in 500 normal control DNA samples. We considered the two variants are likely pathogenic. Other POLG mutations have been shown to cause MNGIE-like syndrome through autosomal recessive or dominant transmission. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] The patient has no family history. We do not know whether the two POLG variants in our patients occurred in trans since no DNA from their relatives was available. Further work is needed to establish a clear genotype-phenotype correlation and to confirm the pathogenicity of these two novel POLG variants.
The leukoencephalopathy in our patient appeared as patchy white matter lesions in contrast to the confluent leukoencephalopathy usually observed in MNGIE. [bib_ref] The role of brain MRI in mitochondrial neurogastrointestinal encephalomyopathy, Scarpelli [/bib_ref] In recent years, white matter lesions have been increasingly recognized in mitochondrial disorders. [bib_ref] Mitochondrial syndromes with leukoencephalopathies, Wong [/bib_ref] Other mutations in the POLG gene have similarly been linked to leukoencephalopathy, such as in Alpers syndrome. [bib_ref] Mitochondrial syndromes with leukoencephalopathies, Wong [/bib_ref] [bib_ref] Novel POLG splice site mutation and optic atrophy, Milone [/bib_ref] Encephalopathy is closely related to mitochondrial dysfunction resulting in energy deficiency. [bib_ref] Mitochondrial syndromes with leukoencephalopathies, Wong [/bib_ref] To our knowledge, this is the first case of MNGIE-like syndrome with leukoencephalopathy involving POLG variants. Our findings expand the clinical spectrum of MNGIE-like syndrome linked to POLG variants and challenge the previous view that the presence or absence of leukoencephalopathy differentiates MNGIE from MNGIE-like syndrome associated with POLG variants. [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype: An expanded clinical spectrum of POLG1 mutations, Tang [/bib_ref] [bib_ref] Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE)-like phenotype in a patient with a novel heterozygous..., Prasun [/bib_ref] Diagnosis of MNGIE or MNGIE-like is challenging. Appropriate attention to gastrointestinal manifestations and brain MRI may help identify these syndromes in the early stage.
## What is known?
Polymerase-gamma gene mutations have been linked to mitochondrial neurogastrointestinal encephalomyopathy-like (MNGIE-like) syndrome, which shares the manifestations of classical MNGIE syndrome except leukoencephalopathy.
## What is new?
Two novel polymerase-gamma mutations have been linked to mitochondrial neurogastrointestinal encephalomyopathy-like syndrome that involves leukoencephalopathy.
# Acknowledgments
We wish to acknowledge the patient and his family for their participation in this study and their efforts to help us.
## Financial support and sponsorship
Nil.
## Conflicts of interest
There are no conflicts of interest.
[fig] reFerences 1: Hirano M, Nishigaki Y, Martí R. Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): A disease of two genomes. Neurologist 2004;10:8-17. 2. Van Goethem G, Schwartz M, Löfgren A, Dermaut B, Van Broeckhoven C, Vissing J, et al. Novel POLG mutations in progressive external ophthalmoplegia mimicking mitochondrial neurogastrointestinal encephalomyopathy. Eur J Hum Genet 2003;11:547-9. [/fig]
[fig] Figure 1: Leukoencephalopathy in a man with mitochondrial neurogastrointestinal encephalomyopathy-like associated with polymerase-gamma gene mutations and fluid-attenuated inversion recovery (a and c) and T2-weighted (b and d) axial brain magnetic resonance images show bilateral periventricular white matter hyperintensity [/fig]
|
Seeding of hepatocellular carcinoma into the stomach wall following endoscopic ultrasound and fine-needle aspiration biopsy
Delayed gastrointestinal metastasis is a rare complication of hepatocellular carcinoma (HCC). We present the case of a patient who presented with melaena and microcytic anaemia 6 years after receiving an orthotopic liver transplant for hepatitis B-induced HCC. Oesophagogastroduodenoscopy revealed a fungating gastric mass at the lesser curve and histology from biopsies confirmed metastatic recurrence of HCC in the stomach. The route of metastasis is likely due to iatrogenic seeding of tumour cells during pre-transplant endoscopic ultrasound (EUS) and fine needle aspiration (FNA) biopsy. Subsequent positron emission tomography and magnetic resonance imaging failed to reveal further metastatic disease and the patient was managed with a total gastrectomy. This is the first reported description in the literature of needle-track metastasis in the stomach due to liver EUS-FNA for HCC.
## Case review
Gastric metastasis from hepatocellular carcinoma (HCC) is a rare complication, the incidence varies from 0.08 to 2% [bib_ref] Gastrointestinal tract involvement in hepatocellular carcinoma: clinical, radiological and endoscopic studies, Chen [/bib_ref] [bib_ref] Gastrointestinal metastasis in hepatocellular carcinoma: radiological and endoscopic studies of 11 cases, Lin [/bib_ref] [bib_ref] Gastric metastasis of hepatocellular carcinoma via a possible existing retrograde hematogenous pathway, Hu [/bib_ref]. A biopsy is considered in patients with HCC when there are diagnostic challenges as risk of tumour seeding is reported to be 2.7% [bib_ref] Needle track seeding following biopsy of liver lesions in the diagnosis of..., Silva [/bib_ref]. Endoscopic ultrasound (EUS) guided trans-gastric biopsy of the lesion has been used where accessible to establish diagnosis. Tumour seeding through this has not yet been reported.
A 43-year-old male of oriental origin, with background chronic hepatitis B infection related cirrhosis developed HCC detected during routine surveillance tests (Raised AFP-69 and on Ultrasound). He had undetectable viral load with lamivudine monotherapy. Magnetic resonance imaging (MRI) showed a 3.3×3 cm 2 lesion in the caudate lobe of liver but not entirely consistent with HCC. Hence, he had a EUS guided transfundal fine needle aspirationFNA of the lesion which confirmed HCC.
Staging MRI and CT scans confirmed the disease as T2N0M0. As he was within the Milan criteria, he was listed for orthotopic liver transplantation and received bridging trans-arterial chemo-embolization (TACE). Orthotopic liver transplantation from a non-heart beating donor was carried out~5 months after diagnosis. The explant histological analysis showed that the HCC was poorly differentiated with a final histology of T2N0M0 with no nodular or micro-vascular metastatic invasion.
His immune-suppression during the immediate posttransplant period was azathioprine, prednisolone and sirolimus (trough levels of [bib_ref] Extrahepatic metastases of hepatocellular carcinoma, Katyal [/bib_ref]. The choice of sirolimus over a calcineurin inhibitor was made at that time due to the risk of metastasis given the worrying features of the explant histology. He received adefovir, lamivudine and hepatitis B immunoglobulin for chronic hepatitis B prophylaxis. Post-transplant follow up included 6-monthly contrast-enhanced CT imaging, routine blood tests and alfa-fetoprotein levels for surveillance of HCC recurrence. After 3 years of CT surveillance, the lack of any findings and the general excellent clinical condition of the patient the decision was made to stop CT surveillance.
Seven years in to post-transplant follow up, he was investigated for iron deficiency anaemia, with a gastroscopy and colonoscopy. Gastroscopy [fig_ref] Figure 1: Index gastroscopy showing large ulcer in lesser curvature [/fig_ref] performed at local Hospital showed a 25 mm ulcerated mass in the lesser curvature of the stomach, just below the cardia. Biopsies obtained from the mass lesion were consistent with HCC. Further gastroscopies are performed [fig_ref] Figure 2: Repeat gastroscopy in 6 weeks-showing ulcer healing [/fig_ref] at 6 weeks following the index gastroscopy showed a polypoid growth in the fundus with healed ulceration over this. Further gastroscopy [fig_ref] Figure 3: Endoscopic images of the recurrence of HCC in the gastric fundus projecting... [/fig_ref] in transplant centre as a perioperative investigation showed large ulcerated mass in the fundus.
## Histology
The biopsyshowed gastric type mucosa and squamous lined mucosa consistent with the gastro-oesophageal junction sample site. There were tumour fragments with extensive areas of necrosis. Large polygonal tumour cells were seen and arranged in trabeculae with frequent mitoses.
Immunohistochemistry was performed to characterize these cells. The tumour cells stained positive for CAM 5.2, Inhibin, Hepatocyte Paraffin 1 (focal) with a high Ki67 proliferative index (>50%). The tumour cells did not stain for CK7, Calretinin, Synaptophysin, Chromogranin, S100, CD117 and DOG1.
He underwent staging for his recurrence by positron emission tomography (PET) (PET image shown in [fig_ref] Figure 5: PET showing the recurrence of HCC at the gastric fundus projecting into... [/fig_ref] and MRI (MRI image shown in [fig_ref] Figure 6: MRI showing the recurrence of HCC at the gastric fundus projecting into... [/fig_ref] which has confirmed the diseases the localized to stomach.
Based on these features and the clinical and radiological findings the patient was diagnosed with metastatic HCC that was localized to stomach and underwent total gastrectomy.
The patient had an uncomplicated total gastrectomy and no other residual disease was found intra-operatively. Histology of the resected specimen showed variable areas of moderate-topoorly differentiated HCC. Overall, 15 retrieved local lymph nodes showed no evidence of tumour metastasis. The patient made an uneventful recovery from surgery and was discharged on Day 10. However, follow up imaging at 2 months showed multiple liver lesions, including a cluster of lesions in the caudate lobe (where his native liver tumour was). Three further small lesions in segments I, II and VI. This was confirmed as biopsy proven metastatic HCC of his graft. Unfortunately the only options now are palliative.
# Discussion
The likely mechanism of dissemination of this patient's HCC was by needle-track seeding from EUS-FNA that was performed for initially diagnosis of primary HCC after 7 years of initial diagnosis. HCC most often metastasizes to lungs, bone, abdominal lymph nodes, adrenal glands, peritoneum or brain. In a study of 403 patients with HCC not a single case metastasized to the stomach making it unlikely that a mechanism other than seeding led to metastasis in this patient [bib_ref] Extrahepatic metastases of hepatocellular carcinoma, Katyal [/bib_ref]. Moreover, the pre-transplantation FNA and the gastric recurrence were both located at the lesser curvature of the stomach and the explant specimen did not reveal direct extension of the primary HCC lesion to the lesser curvature of the stomach. Hypothesizing that needle-track seedling as most likely mode of spread. Levy et al. has provided some basic science support for the idea of needle track implantation. They demonstrated the presence of malignant cells within gastrointestinal tract luminal fluid following EUS-FNA in 3 out of 26 patients with pancreatic cancers. Importantly, malignant cells were not found prior to EUS-FNA in these patients or in those following EUS-FNA for nonmalignant indications.
To our knowledge this is the first case of needle-track implantation of HCC into the stomach wall following EUS-FNA sampling of the liver. One argument in favour of using EUS- FNA as opposed to percutaneous liver FNA is the purported greater risk of needle-track metastasis with percutaneous approaches [bib_ref] Endoscopic ultrasound-guided liver biopsy, Parekh [/bib_ref]. A systematic review of 1340 patients evaluated the risk of needle track seeding following percutaneous FNA of HCC as 2.7% overall or 0.9% per year [bib_ref] Needle track seeding following biopsy of liver lesions in the diagnosis of..., Silva [/bib_ref]. As experience with liver EUS-FNA increases, studies with longer follow up periods are needed to establish the risk of needle track seeding for EUS-FNA. Our patient has been established on Sirolimus, (mTORmammalian target-of-rapamycin inhibitor) which has shown to have a protective effect from de novo cancers and HCC recurrence [bib_ref] Sirolimus-based immunosuppression is associated with increased survival after liver transplantation for hepatocellular..., Toso [/bib_ref]. However, Sirolimus is not shown to improve longterm recurrence free survival beyond 5 years [bib_ref] Sirolimus use in liver transplant recipients with hepatocellular carcinoma: a randomized, multicenter,..., Geissler [/bib_ref]. He underwent loco-regional therapies for his native HCC prior to transplantation which are proven to control tumour burden. His HBV DNA remains undetectable throughout the post-transplant period implying that there is no reactivation of hepatitis B. Despite of this he has developed delayed metastasis. As the lesion correlates well with the biopsy site, the recurrence might be secondary to tumour seeding.
There are no central guidelines suggesting any HCC surveillance post-transplant; however, these patients are followed up closely with 6 monthly contrast CT scans for a maximum 5 years depending on the local practice. In this particular case metastasis was identified outside the follow up period.
Treatment algorithm for recurrence of HCC has been proposed in 2013 by Geneva group that summarizes for extrahepatic metastasis mTOR inhibitors should be considered along with reduction of their immunosuppression. Resection should be considered for isolated metastasis and for cancers which are not resectable Sorafenib remains the choice of treatment [bib_ref] Integrating sorafenib into an algorithm for the management of post-transplant hepatocellular carcinoma..., Toso [/bib_ref]. In our case as his ECOG performance state is zero and we have identified isolated metastasis resection is considered as the initial option of management.
# Conclusion
There is a possibility of needle-tract metastasis of HCC following liver EUS-FNA and delayed presentation of gastric metastasis. This report also highlights the need for meticulous surveillance using serial tumour markers and regular haematology and biochemistry bloods post-liver transplantation for HCC.
[fig] Figure 1: Index gastroscopy showing large ulcer in lesser curvature. [/fig]
[fig] Figure 2: Repeat gastroscopy in 6 weeks-showing ulcer healing. [/fig]
[fig] Figure 3: Endoscopic images of the recurrence of HCC in the gastric fundus projecting into the cardia. [/fig]
[fig] Figure 4: (A and B) Haematoxylin and eosin staining of polygonal tumour cells showing pleomorphic, hyperchromatic nuclei arranged in a trabecular pattern. (C, D and E) Tumour cells staining strongly positive for BAF47/INI1 (nuclear), CAM5.2 (cytoplasmic) and HerPar1 (cytoplasmic). (F) Tumour cells shows negative staining and benign gastric mucosa shows strong cytoplasmic staining (internal positive control). [/fig]
[fig] Figure 5: PET showing the recurrence of HCC at the gastric fundus projecting into the cardia. [/fig]
[fig] Figure 6: MRI showing the recurrence of HCC at the gastric fundus projecting into the cardia. [/fig]
|
Walking-speed estimation using a single inertial measurement unit for the older adults
BackgroundAlthough walking speed is associated with important clinical outcomes and designated as the sixth vital sign of the elderly, few walking-speed estimation algorithms using an inertial measurement unit (IMU) have been derived and tested in the older adults, especially in the elderly with slow speed. We aimed to develop a walking-speed estimation algorithm for older adults based on an IMU.MethodsWe used data from 659 of 785 elderly enrolled from the cohort study. We measured gait using an IMU attached on the lower back while participants walked around a 28 m long round walkway thrice at comfortable paces. Best-fit linear regression models were developed using selected demographic, anthropometric, and IMU features to estimate the walking speed. The accuracy of the algorithm was verified using mean absolute error (MAE) and root mean square error (RMSE) in an independent validation set. Additionally, we verified concurrent validity with GAITRite using intraclass correlation coefficients (ICCs).ResultsThe proposed algorithm incorporates the age, sex, foot length, vertical displacement, cadence, and step-time variability obtained from an IMU sensor. It exhibited high estimation accuracy for the walking speed of the elderly and remarkable concurrent validity compared to the GAITRite (MAE = 4.70%, RMSE = 6.81 cm/s, concurrent validity (ICC (3,1)) = 0.937). Moreover, it achieved high estimation accuracy even for slow walking by applying a slowspeed-specific regression model sequentially after estimation by a general regression model. The accuracy was higher than those obtained with models based on the human gait model with or without calibration to fit the population.ConclusionsThe developed inertial-sensor-based walking-speed estimation algorithm can accurately estimate the walking speed of older adults.Walking-speed estimation using a single inertial measurement unit for the older adults PLOS ONE | https://doi.org/10.
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
Human gait is the bipedal, biphasic, forward propulsion of the center of gravity. Because gait is achieved through complex cognitive-motor interactions,it can become abnormal in movement disorders and in cognitive disorders such as Alzheimer's disease (AD).Furthermore, gait impairment frequently precedes cognitive impairments in cognitive disorders.Among the various gait parameters, walking speed has been studied the most. Moreover, it has been observed to be predictive of a range of outcomes including response to rehabilitation, functional dependence, frailty, mobility disability, cognitive decline, falls, institutionalization, hospitalization, and mortality in both clinical and research settings.Therefore, the White Paper published in 2009 designated walking speed as the sixth vital sign.In the majority of previous clinical research, walking speed was manually measured using a stop watch over a short walking distance.However, these manual measurements were subject to human error and raters' subjectivity. Furthermore, the measuring procedures such as walking distance , start (static versus dynamic), path (straight versus turn), and speed (self-selected versus maximal) varied considerably.Laboratory-based motion capture systems and instrumented walkways provided more accurate and instantaneous walking speeds than the manual measurements did. However, these systems are expensive, require trained personnel, and have spatial constraints. These make it infeasible to evaluate walking over sufficient distance and duration and to evaluate natural walking in real-life conditions.
Recently, inertial sensors began to be employed for analyzing gait features including walking speed. Because they are light, wearable, and inexpensive, they can measure gait for longer durations in unsupervised and real-life conditions. However, because inertial sensors do not directly measure walking speed, previous research estimated walking speed using inertial sensors based on one or more of three approaches: kinematic human gait modeling, direct integration, and regression modeling.Apart from the direct integration methods, which tend to drift over time or have constraints of sensor-attachment locations to use zero velocity updates, or human gait model methods, which are less accurate without calibration and exact measurement, regression models have emerged as a preferred method of walking speed estimation based on a set of inertial features.In the regression-based approaches, walking speed is estimated based on input features from the inherent pattern of the inertial measurement unit (IMU) signals and a set of other personal characteristic parameters such as age, gender, height and weight.Because predefined kinematic human gait models are not necessarily required in these approaches, signals from various body parts for which the widely used kinematic human gait model is not established can potentially be used as input to a regression model.This implies that accurate estimation and flexibility of application could be attained by incorporating appropriate datadriven terms representing the gait characteristics of the population. Additionally, the location of the sensor is highly flexible as opposed to that in direct integration (e.g., can be attached to the waist or wrist for long-time monitoring without discomfort), and there is lower disadvantage of drifting over time. However, simultaneously, the performance of the models is highly dependent on the completeness of the training data and the set of parameters in the regression models. To the best of our knowledge, a limited number of studies based on the regression model have been published.
In three of these, the sensor was attached to the lower body; in one, the head; and in another, the wrist.Although the studies have revealed highly potential results (mean absolute error = 4.5-5.4%, or concordance correlation = 0.93) for walking speed estimation in the specified dataset, the generalizability would be restricted because the models were derived with training dataset which contained 8-30 subjects (mean age 27-37 years). It is challenging to prevent overfitting with this number. In addition, there is no study that included older adults, although older adults have different gait characteristics from younger adults.Because the slowing of gait is a key feature of older adults and the errors in estimating gait speed increase as the gait speed decreases, the algorithm for estimating walking speed in older adults should be designed to reduce errors at low walking speeds.
In this study, we developed a regression-based algorithm for estimating walking speed using IMU-signal data from a large dataset of older adults. In addition, we comprehensively evaluated demographic, anthropometric, and clinical parameters to determine their validity for incorporation in the algorithm. Moreover, we developed the algorithm that can improve the estimation accuracy in the elderly population by additionally applying a model trained specifically in the group with slow walking speed. Finally, the study describes the concurrent validity of the algorithm in comparison with the walking speed measured by an instrumented walkway gold standard, and compares the estimation accuracy of the novel algorithm with those of other algorithms based on the human gait model.
# Materials and methods
## Participants
We enrolled 785 participants from two cohort studies: 495 from the Korean longitudinal study on cognitive aging and dementia (KLOSCAD)and 290 from the Korean frailty and aging cohort study (KFACS).The KLOSCAD and KFACS are population-based prospective multicenter cohort studies conducted in Korea. The KLOSCAD was launched in 2009. Under this study, 6,818 elderly Koreans aged 60 years and over are being followed every two years. The KFACS was launched in 2016. Under this study, 3,000 elderly Koreans aged 70-84 years are being followed every 2 years. Among the 785 participants, 659 (289 men and 370 women) were included in the present analysis after excluding the participants who were pre-frail or frail according to the Korean version of fatigue resistance ambulation illness low weight scale (K-FRAIL), or had obtained 20 points or below in the Tinetti performance oriented mobility assessment (POMA).All the participants had provided written informed consent themselves or via their legal guardians. This study had been approved by the Institutional Review Board of the Seoul National University Bundang Hospital.
## Measurement of gait
We evaluated parkinsonian symptoms and gait disturbances using the UPDRS (range 0-108, with higher scores indicating more severe parkinsonian motor symptoms)and Tinetti POMA. The Tinetti instrument consists of three scales: a Gait Scale, a Balance Scale, and an overall Gait and Balance score. The maximum score is 28. A higher score represent a higher performance. We also measured the gait of each participant by using a FITMETER 1 (FitLife Inc., Suwon, Korea,) laced over the center-of-body-mass (CoM) and a GAITRite™ (CIR Systems Inc., Havertown, PA) simultaneously. The FITMETER is an IMU and incorporates a digital tri-axial accelerometer (BMA255, BOSCH, Germany) and gyroscope (BMX055, BOSCH, Germany). The sensor is a hexahedron (35 × 35 × 13 mm) with smooth edges and weighing 14 g. It can measure tri-axial acceleration up to ±8g (with resolution 0.004g) and tri-axial angular velocity up to ±1,000˚/s (with resolution 0.03˚/s). Measurements were made at a sample rate of 250 Hz. The GAITRite is a portable gait analysis walkway system that measures temporal and spatial gait parameters via an electronic walkway connected to the USB port of a computer. Its walkway size is 520 (L) × 90 (W) × 0.6 cm (H), with an active sensing area of 427 (L) × 61 cm (W). It contains 16,128 sensors placed with a spatial accuracy of 1.27 cm. Measurements were made at a sample rate of 100 Hz.
In the present study, we fixed a FITMETER at the level of the third-fourth lumbar vertebrae of each participant by using a Hypafix. Then, we asked each participant to walk back and forth three times on a 14 m flat straight walkway at a comfortable self-selected pace, and to start turning after passing the 14 m line. We placed the GAITRite electronic mat in the middle of the walkway to measure steady state walking. We excluded the 2 m long walks prior to the turns to eliminate the influence of turning. Similarly, the 2 m walk after each start was eliminated.
## Signal preprocessing
Because an IMU attached to each participant had its own local coordinates, a coordinate transform from the local Cartesian coordinate to the global Cartesian coordinate was performed by applying a complementary filter to the raw acceleration and angular velocity data.The anterior-posterior (AP) direction, medio-lateral (ML) direction, and vertical (V) direction were defined as the x-, y-, and z-axes, respectively, of the acceleration data in this study. We applied a low pass filter to the acceleration data in the global Cartesian coordinate to enhance the discrimination rate of step and direction. Two consequential moving average filters (Hanning filters) were applied. The window sizes of the filters were 0.32 s and 0.08 s, respectively, according to the subjects' average cadence, which is 0.5-0.6 s.We analyzed the data of the central 10 m of the 14 m walking distance to measure steady state walking. Each step was identified from the vertical acceleration data by using the peak detection method. The time gap between consecutive peaks was set to at least 0.24 s.The left and right steps were discriminated using the sign of the yaw angle at the peaks in
## Feature extraction and selection
For each subject, the following demographic, anthropometric, and IMU features were calculated using data from six 10-m-long straight walks. Age and gender are the two demographic features, whereas height, weight, leg length, upper body length, waist circumference, and two feet-lengths of the subjects are the seven anthropomorphic features. The UPDRS and the POMA scores are clinical features reflecting performance of gait and balance. The IMU features include cadence, coefficient of variance of step time (CV step time), magnitude of the 3D acceleration and vertical acceleration, and vertical displacement of CoM within a step cycle. For each step detected in the IMU signal, the magnitude of acceleration and the vertical displacement of CoM were calculated. To do so, we updated the vertical initial velocity at the start of each step, assuming that the vertical heights of CoM and end of a stride were the same in steady state walking, and computed the vertical displacement by integrating the vertical acceleration value during a single step cycle. We used the mean of those values as the features for each subject. The linear relationship between each candidate's features and walking speed measured with the gold standard was then evaluated using univariate linear regression analysis. We selected the features that could be included in the regression model for walking speed estimation, by considering statistically significant correlation with walking speed and collinearity. We calculated the variance inflation factor (VIF) to evaluate multicollinearity and excluded features with a VIF of 2.5 or higher among the selected features to avoid multi-collinearity according to Allison's criteria.To do so, the variable with the highest VIF score was dropped until all remaining variables had a score of <2.5. The above process of selecting the suitable features for estimating walking speed was carried out one time each in a whole dataset and a subset group with low gait speeds. This was because features for estimating low speed accurately may be different from those for estimating the whole range of gait speed. Low walking speed was defined as a speed less than 100 cm/s, which is equal to the threshold in a previous study.Age, gender, cadence, vertical displacement CoM, and foot lengths were the final five features selected using multivariate regression analysis with stepwise selection, from the whole dataset. The features exhibited significant correlation with gait speed and satisfied the condition of collinearity. For the group of subjects exhibiting low walking speeds, age, gender, vertical displacement of CoM, CV step time, and foot length were selected for the regression model. The measurement or computation methods of the selected features in the models are provided in the.
## Model development and validation
We developed two regression models to be applied sequentially: the general model and the low speed-specific model. To develop the models, we randomly divided the entire dataset into the derivation dataset (70%) and validation set (30%). The general model used five features selected from the whole dataset. The model was fitted using data from the derivation dataset and tested on the validation data. The model derivation and validation processes were carried out in MATLAB (version R2016b) and SPSS v..0 (IBM corp., New York, NY). The "regress" function (multiple linear regression using least squares) in MATLAB was applied to obtain the vector of regression coefficients for the linear combination of the features attributed to walking speed. To investigate whether the inclusion of demographics such as gender and age in the estimation model improves accuracy, the models excluding gender, age, or both and the Step discrimination points are indicated by crosses on the filtered signal. Whether the step is left or right is determined by the sign of the yaw angle at the step discrimination point.
https://doi.org/10.1371/journal.pone.0227075.g001 models including these were compared. The final criterion for inclusion in the general model was the minimization of the Akaike information criterion (AIC). The AIC is a likelihood-based measure in which lower values indicate better fit and which extracts a penalty for increasing the number of variables. Thus, the variables selected for inclusion should provide the best fit as well as a parsimonious prediction model. Once the general model was developed, the low speed-specific model was fitted using the low-speed group data (data from individuals whose speeds estimated using the general model were less than 100 cm/s) in the derivation dataset. As in the case of the development of the general model, we compared the AIC values of the model including the five features and the model that excluded gender and selected better-fit model. The final algorithm we propose estimates the walking speed by applying the general model and the low speed-specific model consecutively. First, the general model is applied to estimate the walking speed. If the estimated speed is below 100 cm/s, we applied the low speed-specific model instead of general model to the subject to get more accurate estimation result.As measures of the accuracy of the walking speed estimation algorithm, the mean absolute error (MAE) and the root mean square error (RMSE) on the validation set were determined. We demonstrated that the algorithm, a combination of the general and low speed-specific models as described in Fig 2 (referred to as M 1 hereafter.), exhibits higher performance than the algorithm applying the one-step general model, by comparing the estimation accuracy of the entire validation set with that of its subset with low speed.
## Comparison with other models
The novel regression-based algorithm recommended in this study is denoted as M 1 . M 2 -M 5 represent four comparative algorithms based on the human gait model. Among the four algorithms, M 2 and M 3 are based purely on the human gait model. M 2 , proposed by Zijlstra, is based on the inverted pendulum gait model and combines step length with step frequency. M 3 is its modification obtained by adding the foot length to the term 2 ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi 2lh À h 2 p in the steplength estimation equation:respectively, obtained by multiplying the calibration coefficient C 1 to the step length term and then adding the intercept constant C 0 to the product. Because we fitted these models to the linear regression equation using the human gait model term, we refer to the models as fitted human gait models in this paper.
[formula] M 4 ¼ C 0 þ C 1 � Cadence ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi 2lh À h 2 p M 5 ¼ C 0 þ C 1 � Cadence � ðFoot length þ 2 [/formula]
ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi 2lh À h 2 p Þ
# Statistical analysis
We determined the demographic and anthropometric parameters as means and standard deviations (for continuous variables) or percentages (for the discrete variable). We performed We demonstrated the concurrent validity of each model based on the intraclass correlation coefficients (ICCs) calculated using a two-way mixed model (absolute agreement type) by comparing the walking speeds estimated from each model and those obtained from GAITRite. Additionally, we measured the accuracy of the walking speed estimation models with the MAE and RMSE. The Statistical analysis was performed using SPSS v.19.0.presents the subjects' characteristics in the derivation and validation sets. The subjects have an average age of 73.9 ± 4.7 years (mean ± standard deviation), average weight of 61.9 ± 9.0 kg, average leg length of 84.5 ± 5.6 cm, average feet length of 23.4 ± 1.4 cm, average UPDRS score of 0.8 ± 2.2, average POMA of 27.7 ± 0.8, average walking speed of 113.7 ± 17.9 cm/s, and average cadence of 115.4 ± 9.1 cm/s. The subjects in the derivation and validation sets exhibit comparable clinical characteristics and gait parameters.
# Results
## Model derivation and development
The general model with the smallest AIC value includes five features: age, gender, cadence, vertical displacement CoM, and foot length. The β coefficients, standard errors, standardized beta, t values, and p values of the best-fitting general model are presented in. Age displayed a negative correlation with walking speed. Meanwhile, the female gender, vertical displacement, cadence, and foot length correlated positively with walking speed. The vertical displacement of CoM exhibited the highest standardized beta value (B = 0.582, p < .001), closely followed by cadence (B = 0.581, p < .001).also presents the β coefficients, standard errors, and P values of the features in the low speed-specific model. In the derivation data set, the AIC value for the general model (AIC = 3136.0) was smaller than those of the simplified general models (age-excluded model, AIC = 3157.9; gender-excluded model, AIC = 3143.1; age-and-gender-excluded model, AIC = 3167.0), indicating the increase in predictive capability with the incorporation of the age or gender term. (S1 The low speed-specific model with the smallest AIC value incorporates five features: age, cadence, CV step time, vertical displacement, and foot length. Unlike the general model, the gender term was replaced by the CV step time term. The walking speed exhibited a negative correlation with age and gait variability, and a positive correlation with vertical displacement, cadence, and foot length. The vertical displacement of CoM displayed the highest standardized beta value (B = 0.713, p < .001), followed by cadence (B = 0.628, p < .001) and CV step time (B = -0.153, p = .028). In the derivation data set, the AIC of the low speed model (= 3138.6) was smaller than those of the simplified model (age-excluded model, AIC = 3242.2), indicating the increase in the predictive capability with the incorporation of the age term. (S1 The best-fitting general and low speed-specific models are as follows:
Generalpresents the accuracy of M 1 and the general model in terms of walking speed estimation in the 197 older adults (90 men, 107 women) assigned to the validation data set. Although the general model estimated the walking speed with high accuracy, the accuracy varied depending on the walking speed to be estimated. The accuracy for the entire validation set was 4.96% for the MAE and 6.93 cm/s for the RMSE. However, for groups with walking speeds below 100 cm/s (21.3% of the validation dataset), the estimation error was significantly increased by 58% (from 4.96% to 7.84%) in terms of the MAE and 16% (from 6.93 cm/s to 8.01 cm/s) in terms of the RMSE. Moreover, the estimation error increased gradually as the speed decreased. The accuracy of slow speed model for the slow speed group was MAE = 6.07%, RMSE = 7.24 ??/?, and concurrent validity (ICC (3,1)) = 0.916. Meanwhile, the newly proposed M 1 algorithm estimates the walking speed more accurately than the general model does, particularly in the low-speed group. The accuracy for the low-speed group in terms of MAE is 6.69%. This is approximately 1.2% lower than that of the general model. For the entire validation set, M 1 also provides improvement over the general model. However, the improvement is less apparent than that observed in the low-speed group in the validation set.show the regression analysis and Bland-Altman plot, respectively, for the predicted walking speed based on M 1 . The solid line inis the best fit line (y = 0.851x + 17.1). The dotted line is the ideal line (y = x), which represents a perfect correlation between walking speed from GAITRite and walking speed predicted by M 1 . Although the best fit line deviates marginally from the ideal line, the analysis reveals a very strong linear correlation between the predicted and reference walking speeds (Pearson's r = 0.942, p < .001). The Bland-Altman plot reveals that the error is largely within the 95% limit of agreement even in the range of speeds less than 100 cm/s.presents summary statistics regarding the performance of M 1 and its four comparative models (M 2 -M 5 ) in terms of walking-speed estimation. For each pre-specified summary statistic (MAE, RMSE, or ICC), M 1 provided improvement in accuracy over M 2 and M 3 , the prediction algorithms based on the human gait model (MAE of 4.90% vs 20.9% and 19.5%, respectively; RMSE of 6.81 cm/s vs 26.0 cm/s and 22.1 cm/s, respectively). It also exhibited remarkable concurrent validity with GAITRite (ICC of 0.937 with 95% confidence interval [0.918 0.952]). The algorithms purely based on the human gait model exhibited concurrent validities with the GAITRite at the low-to-moderate level (ICC of 0.446 for M 2 and 0.585 for M 3 ). The mean absolute error was over 20%, and the RMSE exceeded one standard deviation of walking speed. The fitted human gait models have considerably higher accuracy. The MAE and RMSE of M 4 were 6.25% and 8.63 cm/s, respectively, approximately one-third of those of M 2 . Moreover, the concurrent validity was remarkable. The accuracy of the M 5 , which incorporates the foot length term, was higher than that of M 4 and lower than that of M 1 . M 1 exhibited the highest accuracy and concurrent validity among the five models.
## Model testing and comparison with other models
# Discussion
In this study of 659 Korean older adults, we developed and validated an algorithm for walking speed estimation based on the regression method. The algorithm consists of variables such as age, gender, and foot length, which are already known to the individual and readily available, as well as vertical displacement of CoM, cadence, and CV step time, which were obtained from an IMU sensor. It exhibited high accuracy in estimating the walking speed of the older adults and a remarkable concurrent validity compared to the gold standard. Although the estimation error, in general, increased gradually as the speed decreased, the algorithm we propose achieved high estimation accuracy even at the low walking speed by applying a low speed-specific regression model, sequentially after estimation by a general regression model. The newly developed algorithm exhibited a substantially higher accuracy when compared with models based on the human gait model without calibration to fit the population. This high accuracy was verified in a completely exclusive validation dataset of a sufficiently large size.
We demonstrated that the regression-based algorithms proposed in this study exhibit substantially higher estimation accuracy than pure-model-based algorithms. This is consistent with the previous study in that model-based methods require subject-or population-specific model calibration, which is a disadvantage of the method.The fitted human gait models that completed the population-specific calibration exhibited substantially higher accuracy than the pure-model-based algorithms. However, their accuracy was less than that of the regression-based algorithm M 1 . In addition to the substantially higher accuracy, the novel algorithm circumvents the use of the leg length term included in the gait-model-based algorithms. This is another advantage because if anthropometric measurement is required for each individual, it will be inconvenient in that each individual needs to visit an institution or consult trained personnel for reliable leg length measurement, prior to the application of the model to the individuals. The accuracy measure values obtained from the linear regression model proposed in this study are MAE = 4.70% and RMSE = 6.81 ??/?. Moreover, the concurrent validity is ICC (3,1) = 0.937 in the older adult population. These are comparable to the results of previous regression-based methods (MAE 4.5-5.4% or concordance correlation of 0.93) in the young adult population.Zihajehzadeh et al. developed and validated the Gaussian process regression (GPR) model for walking speed estimation with accuracy of MAE = 4.50% and RMSE = 5.30 ??/? in the young healthy adult population (average age = 27 ± 4 years). Recently, Soltani et al. developed an algorithm that estimates step length to less than 5% of MAE using machine learning (regression-based) in 30 healthy adults. It is noteworthy that even a simple linear regression model comprising age, gender, foot length, and three gait features from an IMU proposed in this study exhibits a high accuracy of estimating walking speed in older adults. Both the low-speed-specific model and the general model incorporated the age term in the best-fitting speed estimation equations. Although the general model includes the age term as well as the gender, the age term plays a more important role in speed estimating because the AIC reduction is substantially larger when the age item is incorporated. The negative beta value of the age term indicates that as age increases, the speed estimated from the IMU features and anthropometrics should be adjusted downward. The observations indicate that to estimate gait speed more accurately, the age term is required in addition to the terms used in the inverted pendulum model, such as vertical displacement, cadence, and foot length. This may imply that the human gait is not accurately modeled by the inverted pendulum model, with ageing.
It is also an important observation that notwithstanding the remarkable overall performance of the general model in the older adults, the estimation accuracy of the model is considerably lower at walking speeds less than 100 cm/s. This is consistent with a previous study regarding regression-model-based walking speed estimation using a wrist-worn inertial sensor.The accuracy in the group with walking speeds of 50-100 cm/s was RMSE = 7.5 cm/s and MAE = 8.9% in the previous study, and RMSE = 8.01 cm/s and MAE = 7.84% in the present study. The lower performance of the regression model at low walking speeds should be considered particularly while estimating walking speed in the elderly by using a regression model with the IMU sensor features. The walking speed of 100 cm/s, used as the threshold value of low walking speed in this study, was widely used as the threshold for frailty or sarcopenia in previous studies.In a previous study,.8% belonged to the low gait speed (less than 100 cm/s) group.Meanwhile, in our study, 19.9%, or one in five, belonged to the low gait speed group in the elderly population. In older adults, various pathological conditions such as stoke, Parkinsonism, excessive white matter hyperintensity, cardiovascular disease, and spinal stenosis, as well as frailty and sarcopenia are more prevalent than in the younger population. These conditions cause a slowing of gait.Moreover, such pathological conditions may cause variation in gait patterns during walking, hindering the estimation of the walking speed using a generalized regression-based model. Our observations demonstrate that a low-speed-specific model can reduce the MAE from 7.84% to 6.69% in the low walking speed groups. In addition, the MAE difference between the general model and the low-speedspecific model tended to increase with lowering of speed, within the low walking speed group. This may cause the difference between the two models to increase in a population with a high proportion of elderly with low walking speeds. Therefore, we recommend that an algorithm implementing a specific model that reflects the characteristics of groups with low walking speeds be used for older adults. We consider this to be even more vital in the hospitalized or institutionalized elderly population than in the community-dwelling elderly population, the target population of the present study.
Our algorithm exhibited slightly lower but comparable accuracy to the previous speed estimation algorithms using waist or pelvis worn single IMU.An algorithm using the kinematic human gait model showed accuracy of 3.36% for MAE and 3% for MAE in studies using direct integration. The previously proposed algorithms were validated in healthy young adults, and the difference in study population with our study may explain the small difference in accuracy. Even for the older adults, however, there were a few studies on the direct integration-based walking speed estimation algorithmand the accuracies reported in these studies were 1.4%, and 3%, which were higher than ours was. The studies used a pair of IMUs attached to anklesand feet, respectively. Since the farther from the contact point the IMU is placed, the more difficult the gait events identification is due to the disadvantage of zero velocity updates, the difference in accuracy between previous studies and ours could be explained in part by differences in sensor attachment sites as well as in estimation method. Considering that gait measurement using the IMU is based on the advantage of monitoring for longer durations in unsupervised and real-life conditions at lower cost, it should be taken into account that the waist-placement causes less constraint in body movement and minimal discomfortand using a single sensor rather than a pair of sensors demands lower costs.
A strength of this study is the development of the algorithm from a sufficient number of older adults and verification of the accuracy of the algorithm by applying it on an independent validation set. Another strength of this study is that a low-speed model is independently developed for speed estimation and that the estimation accuracy is improved by applying it as the next step after the application of the general model. However, there are certain limitations. We used an independent validation dataset to verify whether the proposed algorithm is over-fitted to the specific dataset. Nonetheless, it is necessary to apply it to other populations to verify its generalizability. We used the features selected based on the whole dataset in general model. It may cause biased evaluation for fitted model trained by derivation dataset. Nevertheless, the bias might be small because the whole dataset and derivation dataset were homogeneous groups with similar clinical and gait characteristics. In addition, an algorithm for estimating the walking speed of the elderly with speeds of less than 100 cm/s based on the IMU sensor needs to be further studied and improved. In the elderly with walking speeds less than 100 cm/ s, additional terms may be required in the fine-tuned regression model in order to reflect divergent gait patterns. In the present study, interaction terms between features were not selected as the features because of the multicollinearity with the raw feature terms in the feature selection stage, but including the interaction terms in the regression model could potentially improve the accuracy of the model. Feature selection methods that are more contemporary such as Least Angle Regression (LARS) or Elastic Net Regression, or models other than linear regression may be required to improve performance of the algorithm. Considering the flexibility to develop a better-fit model by incorporating an appropriate term that reflects the gait characteristics without constraints on the sensor-attachment site, the regression-based model can be effective for measuring the gait in the elderly. Future studies are warranted to develop regression-based models with improved performance, especially in older adults with slow walking speeds.
Supporting information S1 AIC, and Delta AIC for the different regression models.
|
Tuberculosis of the Testis: A Case Report
A 45-year-old man visited our clinic with a painless swelling of the left scrotum and an ulcer as chief complaints. A hard and indurated mass was palpable with ulcerating foci that were proximal and distal, measuring 3 × 2 cm and 2 × 1 cm respectively and about 2 cm apart. Laboratory data were normal except for an elevated erythrocyte sedimentation rate (ESR), and white blood cell (WBC) differential showed neutropenia and lymphocytosis. A diagnosis of left testicular tumor was made and the patient had a left orchidectomy with fistulectomy. Histopathology results showed a stratified squamous epithelium with tuberculous granuloma and necrotic caseation. Patient is currently on anti-tubercular medication. The rarity of this condition makes these findings important to report.
# Introduction
Most people think of tuberculosis as affecting only the lungs, but in reality it affects almost every part of the body. Pulmonary tuberculosis (TB) is most common type, accounting for about 70% of cases. Sometimes, pulmonary TB will spread, though this usually happens only in immune-suppressed patients and young children.
Dissemination of TB to the testis may result in secondary infection of epididymis. In many of these cases, there is associated tuberculous prostatitis and seminal vesiculitis, and it is believed that epididymitis usually represents a secondary spread from these other involvements of the genital tract.The result is great thickening and the formation of a caseous mass. Caseous masses are very dense, but after a certain amount of time they generally breakdown, and, having burst externally, result in tedious fistulae. When mycobacterium tuberculosis (MTB) droplet nuclei are inhaled, they go to the alveoli where they are taken up by phagocytosis into air-space macrophages. Here the bacilli are processed into phagosomes that fail to acidify. The bacilli evade intracellular killing and can therefore survive and multiply for long periods of time. These infected macrophage carry viable bacilli in the lymphatics to regional lymph nodes or in the blood stream to any part of the body. Disseminated tuberculosis spreads haematogenously when blood vessels erode into blood vessels. The disseminated tubercles appear like millet seed in various organs (for example, liver, spleen, kidneys, brain meninges, bone marrow and other tissues). Reactivation of TB in later years can occur as a result of reactivation of viable bacilli (quiescent or dormant tubercle), which have persisted within nodes. On histology, areas of granuloma and caseous necrosis are seen.
## Case presentation
A 45-year-old car spare parts dealer presented to the Abnira medical centre with a five-year history of left scrotal swelling and a two-week history of left scrotal ulcer. The swelling, which had being increasing in size, regressed some months before presentation and became hardened with ulcerated foci some weeks after. There was a history of contact with persons with pulmonary TB in the past, for over a year.
On examination, he was not in obvious painful distress, not pale, anicteric, afebrile, not dehydrated, and had no obvious weight loss. His pulse rate was 80/minute and regular with full volume, and his blood pressure was 120/80 mmHg. Heart sounds were 1st and 2nd only, his respiratory rate was 14 cycles per minute and both lung fields were clear. The abdomen was full, moved with respiration with no palpably enlarged abdominal mass, and kidneys were not ballotable.
The right scrotum was normal whereas the left was swollen with two foci of discharging ulcer, with a proximal focus of 3 cm × 2 cm and a distal focus of 2 cm × 1 cm. The foci were approximately 2 cm apart. On palpation, the area was found to be non-tender, hard, able to go above the mass, not attached to the overlying skin, and not transilluminable.
The laboratory data showed an elevated erythrocyte sedimentation rate (ESR) of 40 mmfall/hr. Full blood count (FBC) showed a hemoglobin (Hb) of 12.3 g/dL, a white blood cell count (WBC) of 3,250 cells/mm,and platelets of 259000 with a WBC differential showing 28 neutrophils, 69 lymphocytes, 1 monocyte, 2 eosinophils, and 0 basophils. Blood film showed reactive lymphocytes +++, toxic granulation ++, macrocytes ++, microcyte +, neutropenia, lymphocytosis, hypochromasia, and platelets to be adequate and normal. Swab microscopy culture and sensitivity (M/C/S) yielded a growth of the staphylococcus species after a 24-hour incubation at 37 °C, sensitive to gentamicin, levofloxacin and sparfloxacin. Chest x-ray, spinal x-ray, urea, electrolyte, creatinine and liver function tests were found to be normal. Retroviral screening and venereal Disease Research Laboratory (VDRL) tests were non-reactive. An assessment of left testicular tumor was made and the patient was prepared for left orchidectomy (after obtaining an informed consent). At surgery, a tract was found connecting the proximal and distal ulcer foci. Fistulectomy and Orchidectomy.
The histological findings of the excised left testis were: macroscopic description (A): SCO soft-to-firm testicular tissue measuring 7 × 5 × 3 cm and weighing 57 g, with a cut section showing a mass of 4 cm diameter and 3 k. (B): SCO a skin tissue measuring 3 × 3 × 2 cm and weighing 8 g, with a cut section showing a grayish surface.
Microscopy: Testicular tissue showed stratified squamous within the stroma, several granuloma and extensive caseous necrosis. The histological diagnosis was tuberculosis.
Anti-tuberculosis chemotherapy was subsequently initiated. Three months after surgery, recurrence has not been found.
# Discussion
Tuberculosis (formerly known as consumption, pthisis, or 'wasting') is an infectious disease caused by the bacillus mycobacterium tuberculosis. M. tuberculosis infects about one third of the world population and kills about three million patients each year and so is the single most important infectious cause of death on Earth.There has being a dramatic increase in TB in the world and especially in Africa due to the emergence of Acquired Immune Deficiency Syndrome (AIDS).The primary phase of M.TB infection begins with inhalation of the mycobacterium and ends with a T cell-mediated immune response that induces hypersensitivity to the organisms and controls 95% of infections. In secondary and disseminated TB, some individuals become re-infected with mycobacterium or reactivate dormant disease, or they progress directly from the primary mycobacterium lesion into disseminated disease. Following M.TB infection, 5%-10% of people will develop the active disease while more than 90% of M.TB infections do not result in disease in an individual's lifetime.Primary disease occurs soon after the primary infection in the context of a naïve immune system, and the bacilli multiply and spread. Primary disease commonly occurs in the lungs and then disseminates to multiple sites, resulting in TB meningitis and miliary TB. Post-primary disease occurs long after primary infection in the context of a sensitized immune system, either as reactivation of a latent infection, or as a result of re-infection with a new strain. This may be because the strain of mycobacterium is particularly virulent or the host is particularly susceptible.
This patient had no history of chronic cough or weight loss; however, he had a history of drenching night sweats for over a year, with a positive history of contact with a person with pulmonary tuberculosis for over a year. He neither smoked nor worked in a place that would predispose him to such a disease.
In a reported case from Japan by Sensaki and colleagues in 2001, [bib_ref] A case of very rare tuberculosis of the testis, Senzaki [/bib_ref] it was shown that the presentation and findings of tuberculosis of the testis were similar to those reported here, other than the fact that we were unable to do a computerized tomography scan (CT-scan), beta-human chorionic gonodotropin (hcG) or magnetic resonance imaging (MRI).
In the absence of a histologic finding, the diagnosis of TB testis may be impossible. Patients with testicular masses should be fully investigated preoperatively and post-operatively in order to increase the diagnostic index of this condition.
# Conclusion
Although it is a very rare disease, the clinician should consider tuberculosis of the testis as a possible differential of a scrotal mass. This will increase the possibility of early diagnosis, as well as proper and early management.
# Author contributions
# Funding
Author(s) disclose no funding sources.
## Competing interests
Author(s) disclose no potential conflicts of interest.
# Disclosures and ethics
As a requirement of publication author(s) have provided to the publisher signed confirmation of compliance with legal and ethical obligations including but not limited to the following: authorship and contributorship, conflicts of interest, privacy and confidentiality and (where applicable) protection of human and animal research subjects. The authors have read and confirmed their agreement with the ICMJE authorship and conflict of interest criteria. The authors have also confirmed that this article is unique and not under consideration or published in any other publication, and that they have permission from rights holders to reproduce any copyrighted material. Any disclosures are made in this section. The external blind peer reviewers report no conflicts of interest.
[fig] Figure 1: Section of the testis showing granulomatous and necrotic foci. notes: Arrow downwards shows granulomatous focus; arrow left and horizontal shows area of necrosis. [/fig]
[table] Table 1: The Walgrens timetable of TB: it was published by Walgren in 1984, it provides useful guide to the time sequence for development of various manifestations of the disease. [/table]
|
Efficacy and safety of dulaglutide 3.0 and 4.5 mg in patients aged younger than 65 and 65 years or older: Post hoc analysis of the AWARD‐11 trial
Aim: To evaluate the efficacy and safety of dulaglutide 3.0 and 4.5 mg versus 1.5 mg when used as an add-on to metformin in subgroups defined by age (<65 and ≥65 years).Materials and Methods:Of 1842 patients included in this post hoc analysis, 438 were aged 65 years or older and 1404 were younger than 65 years. The intentto-treat (ITT) population, while on treatment without rescue medication, was used for all efficacy analyses; the ITT population without rescue medication was used for hypoglycaemia analyses; all other safety analyses used the ITT population.Results: Patients aged 65 years or older and those younger than 65 years had a mean age of 69.5 and 53.2 years, respectively. In each age subgroup, the reduction from baseline in HbA1c and body weight (BW), and the proportion of patients achieving a composite endpoint of HbA1c of less than 7% (<53 mmol/mol) with no weight gain and no documented symptomatic or severe hypoglycaemia, were larger for dulaglutide 3.0 and 4.5 mg compared with dulaglutide 1.5 mg, but the treatment-by-age interactions were not significant. The safety profile for the additional dulaglutide doses was consistent with that of dulaglutide 1.5 mg and was similar between the age subgroups.Conclusion: Dulaglutide doses of 3.0 or 4.5 mg provided clinically relevant, doserelated improvements in HbA1c and BW with no significant treatment-by-age interactions, and with a similar safety profile across age subgroups.K E Y W O R D Sbody weight, dulaglutide, elderly, GLP-1 RAs, glycaemic control, type 2 diabetes
# | introduction
Type 2 diabetes (T2D), which accounts for more than 90% of patients with diabetes, 1 is a health condition increasing in prevalence among the ageing population. [bib_ref] Recent trends in the prevalence of type 2 diabetes and the association..., Caspard [/bib_ref] Approximately 40% of the adult population with diabetes in the United States is older than 65 years,and the number of older adults living with this condition is expected to increase rapidly in the coming decades.Older patients with T2D have an increased risk of many co-morbidities, including cognitive dysfunction, cardiovascular disease (CVD), frailty, chronic kidney disease, retinopathy and peripheral neuropathy, affecting manual dexterity and physical ability. [bib_ref] Effects of glucagon-like peptide 1 receptor agonists on comorbidities in older patients..., Onoviran [/bib_ref] These may predispose this patient population to an increased risk of recurring hypoglycaemia and hypoglycaemiaassociated morbidity, particularly in patients treated with insulin.
Co-morbidities may affect accurate insulin dosing and insulin clearance, and hypoglycaemic events increase the risk of falls and/or additional cardiovascular-related morbidity. Therefore, the use of insulin therapy, particularly in elderly patients with T2D diabetes, is limited by the risk of hypoglycaemia. [bib_ref] Management of comorbidities in older persons with type 2 diabetes, Morley [/bib_ref] [bib_ref] Similar efficacy and safety of once-weekly dulaglutide in patients with type 2..., Boustani [/bib_ref] Guideline changes by the American Diabetes Association (ADA) now recommend personalizing treatment for older patients with T2D
based on their functional status and co-morbidities. Less stringent glycaemic treatment goals are recommended for frail elderly patients to keep the risk of hypoglycaemia low while also taking co-morbid illness and/or limited life expectancy into consideration.Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and sodium-glucose co-transporter-2 inhibitors (SGLT-2is) are recommended as first-line therapy in combination with metformin, irrespective of the HbA1c level, in patients who either are at high risk or have pre-existing atherosclerotic cardiovascular disease (ASCVD), heart failure or chronic kidney disease. Additionally, GLP-1 RAs and SGLT-2is are the preferred second treatment option after metformin for patients who would benefit from weight loss, while GLP-1 RAs, SGLT-2is, dipeptidyl peptidase-4 inhibitors and thiazolidinedione are recommended after metformin for patients with increased hypoglycaemic risk. [bib_ref] A consensus report by the American Diabetes Association (ADA) and the European..., Buse [/bib_ref] These guidelines also pertain to older adults, particularly those with pre-existing CVD and increased risk of hypoglycaemia; however, as with any other pharmacotherapy, they need to be used cautiously according to a patient's history, profile and individualized needs.
Prior Assessment of Weekly AdministRation of LY2189265 in Diabetes (AWARD) studies have shown that dulaglutide at 0.75 and 1.5 mg once-weekly is effective for glycaemic control and well tolerated in elderly patients with T2D. [bib_ref] Similar efficacy and safety of once-weekly dulaglutide in patients with type 2..., Boustani [/bib_ref] [bib_ref] Efficacy and safety of once-weekly dulaglutide in elderly Chinese patients with type..., Kuang [/bib_ref] In the AWARD-11 study, dulaglutide 3.0 and 4.5 mg once-weekly provided clinically relevant, dose-related improvements in glycaemic control and body weight (BW) in patients with T2D inadequately controlled with metformin monotherapy. [bib_ref] Efficacy and safety of Dulaglutide 3.0 mg and 4.5 mg versus Dulaglutide..., Frias [/bib_ref] The safety profile was comparable with the 1.5 mg dose through 52 weeks and consistent with prior dulaglutide studies in the AWARD trial programme. The AWARD-11 study enrolled nearly one quarter of patients aged 65 years or older. We conducted a post hoc analysis to examine the efficacy and safety of these additional dulaglutide doses in this elderly population.
## | materials and methods
## | study design
The study design of the AWARD-11 trial was previously described in detail. [bib_ref] Efficacy and safety of Dulaglutide 3.0 mg and 4.5 mg versus Dulaglutide..., Frias [/bib_ref] Briefly, this randomized, phase 3, double-blind, multicentre, parallel-arm study (ClinicalTrials.gov identifier: NCT03495102) included a 2-week lead-in period, followed by a 52-week treatment period (with primary efficacy endpoint at 36 weeks), and a 4-week safety follow-up period. Patients initiated treatment with dulaglutide 0.75 mg for 4 weeks, followed by stepwise dose escalation every 4 weeks to the randomized dose of 1.5, 3.0 or 4.5 mg .
## | key eligibility criteria
The key eligibility criteria of the AWARD-11 trial were age 18 years or older, HbA1c of 7.5% or higher (≥58 mmol/mol) and 11.0% or less (≤97 mmol/mol), body mass index (BMI) of 25 kg/m 2 or higher, and patients were taking commercially available metformin.
## | efficacy measures and safety assessments
For this post hoc exploratory analysis, the primary efficacy measure was the change in HbA1c from baseline to 36 weeks in subgroups defined by age (<65 and ≥65 years). Secondary efficacy measures (all assessed at 36 weeks) were the change from baseline in BW and the proportion of patients achieving an HbA1c of less than 7.0% (<53 mmol/mol). Patients performed fasting plasma glucose measurements once-daily, four-point self-monitored blood glucose (SMBG) measurements once-weekly, and six-point SMBG during the week preceding clinic visits. Safety assessments at 52 weeks included incidence of treatment-emergent adverse events (TEAEs), discontinuation of study drug because of adverse events (AEs), adjudicated and confirmed cardiovascular (CV) and pancreatic AEs, and occurrence of hypoglycaemic episodes. Hypoglycaemic episodes were collected on a dedicated case report form, including cases where SMBG was 70 mg/dL or less (≤3.9 mmol/L), regardless of whether symptoms were experienced. As defined by the ADA, events were categorized as documented symptomatic hypoglycaemia any time patients felt they were experiencing symptoms and/or signs associated with hypoglycaemia and had a plasma glucose level of 70 mg/dL or less (≤3.9 mmol/L). Severe hypoglycaemia was defined as an episode requiring the assistance of another person to actively administer carbohydrate, glucagon or other resuscitative actions. Total hypoglycaemic events were defined as an episode with a plasma glucose level below the defined threshold, regardless of symptoms, an episode of symptomatic hypoglycaemia where the plasma glucose level was not measured, and all severe hypoglycaemia episodes.Although 65 years is the most common age cut point used for subgroup analyses reported in clinical trials, [bib_ref] Efficacy and safety of esketamine nasal spray plus an oral antidepressant in..., Ochs-Ross [/bib_ref] [bib_ref] Efficacy and safety of dulaglutide in older patients: a post hoc analysis..., Riddle [/bib_ref] [bib_ref] Semaglutide as a therapeutic option for elderly patients with type 2 diabetes:..., Warren [/bib_ref] [bib_ref] Efficacy and safety of empagliflozin in older patients in the EMPA-REG OUTCOME..., Monteiro [/bib_ref] [bib_ref] Cardiovascular safety and lower severe hypoglycaemia of insulin degludec versus insulin glargine..., Pratley [/bib_ref] [bib_ref] Prediction of 6-year incidence risk of chronic kidney disease in the elderly..., Zhou [/bib_ref] some studies define 70 years as the cut point for elderly patients. [bib_ref] The risk of diabetes on clinical outcomes in patients with coronavirus disease..., Chung [/bib_ref] Thus, we also compared the effects of dulaglutide dose escalation on change in HbA1c and BW from baseline and the proportion of patients achieving an HbA1c of less than 7.0% (<53 mmol/mol) at 36 weeks using a cut point of 70 years. Safety assessments were also carried out in these subgroups of patients aged less than 70 and 70 years or older.
# | statistical analysis
The safety and efficacy analyses were performed using the intent-totreat (ITT) population, defined as all patients randomized who received at least one dose of study drug. Efficacy analyses excluded measurements collected after discontinuation of study drug or initiation of another antihyperglycaemic medication ('on-treatment without rescue analysis'). The analysis for hypoglycaemia was performed using the ITT population excluding observations after rescue medication. All tests of treatment effects were conducted at a two-sided alpha level of .05, unless otherwise specified. All tests of interactions between treatments and factors of interest were conducted at a two-sided alpha level of .10. were younger than 65 years of age.
# | results
## | patient baseline characteristics
At baseline, in the subgroup aged 65 years or older, the average age was 69.5 years, with a mean HbA1c of 8.4% (68.3 mmol/mol), mean BW of 90.0 kg and mean BMI of 32.9 kg/m 2 . The mean duration of disease was 9.9 years, and females comprised 48.4% of patients. In the subgroup younger than 65 years of age, the average age was 53.2 years, with a mean HbA1c of 8.7% (71.6 mmol/mol), mean BW of 97.5 kg and mean BMI of 34.7 kg/m 2 . The mean duration of disease was 6.9 years, and females comprised 48.9% of patients. Baseline data for fasting serum glucose, systolic blood pressure, diastolic blood pressure, mean baseline estimated glomerular filtration rate (eGFR), eGFR categories and patient CV risk factors are also presented in . Within each age subgroup, these baseline characteristics were comparable among treatment groups .
As expected, patients aged 65 years or older had a longer duration of T2D and a higher prevalence of renal impairment, hypertension, dyslipidaemia, history of CVD and atrial fibrillation.
## | efficacy: glycaemic control and bw
In patients younger than 65 years, the least-squares mean (LSM) change in HbA1c from baseline to week 36 was À1.74% and À1.94% with dulaglutide 3.0 and 4.5 mg, respectively, compared with À1.59% for dulaglutide 1.5 mg . In the subgroup aged 65 years or older, similar results were observed, as the LSM change in HbA1c from baseline to week 36 was À1.58% and À1.65% with dulaglutide 3.0 and 4.5 mg, respectively, compared with À1.33% for dulaglutide 1.5 mg . The results for both age subgroups were in line with those seen in the overall study population , with no statistically significant treatment-by-age subgroup interaction for HbA1c reduction (interaction P = .591).
In patients younger than 65 years, the LSM change in BW from baseline to week 36 was À3.9 and À4.8 kg with dulaglutide 3.0 and 4.5 mg, respectively, compared with À3.1 kg for dulaglutide 1.5 mg . Comparing these results with patients in the 65 years or older subgroup, a similar trend was observed. The LSM change in BW from baseline was À4.3 and À4.5 kg with dulaglutide 3.0 and 4.5 mg, respectively, compared with À3.1 kg for dulaglutide 1.5 mg at 36 weeks . The results for both age subgroups were in line with those seen in the overall study population , with no statistically significant treatment-by-age subgroup interaction for change in BW (interaction P = .510).
In both age subgroups, the proportion of patients achieving the HbA1c target of less than 7% with no weight gain or documented symptomatic or severe hypoglycaemia at the primary endpoint of 36 weeks was significantly higher (P ≤ .034) for the additional dulaglutide doses (3.0 and 4.5 mg) in comparison with dulaglutide 1.5 mg . The proportions of patients achieving this composite target in both age subgroups were in line with those seen in the overall study population , with no statistically significant treatment-by-age subgroup interaction (interaction P = .506).
Similar results were obtained for patients younger than 70 and those aged 70 years or older, with no statistically significant treatment-by-age subgroup interaction for HbA1c reduction from baseline (interaction P = .937), weight change from baseline (interaction P = .376), or the proportion of patients reaching the HbA1c target of less than 7% with no weight gain or documented symptomatic or severe hypoglycaemia (interaction P = .717)
## | safety
The most frequent TEAE experienced in both age subgroups was nausea, which ranged from 12.1% to 18.7%, followed by diarrhoea (range, 7.5% to 12.7%), vomiting (range, 6.4% to 10.7%), dyspepsia (range, 2.3% to 8.7%) and nasopharyngitis (range, 4.2% to 8.3%). There was no statistically significant treatment-by-age subgroup interaction for any of these TEAEs (interaction P = .383). The incidence of TEAEs related to a composite of supraventricular arrythmias, conduction disorders and adjudicated CV events was low and was similar across dose groups, with no significant treatment-byage subgroup interaction (interaction P = .963).
The incidence of all AEs reported as serious (serious adverse events [SAEs]) was low across dose groups in each age subgroup ( .904
Atrial fibrillation Incidence of documented symptomatic hypoglycaemia (<70 mg/ dL) was not different between treatment groups, and the treatmentby-age subgroup interaction was not significant (interaction P = .293).
Similarly, no difference was observed between treatment groups with a non-significant treatment-by-age subgroup interaction (interaction P = .422) for total hypoglycaemia incidence ( Similar safety results were obtained for patients younger than 70 and those aged 70 years or older with non-statistically significant treatment-by-age subgroup interactions for safety assessments, except for nausea, which was higher in the 70 years or older subgroup (interaction P = .054) . This was probably driven by the small number of patients in the 70 years or older subgroup in combination with the higher incidence in the lowest dulaglutide dose group (1.5 mg) relative to the higher dose groups (3.0 and 4.5 mg). No statistically significant treatment-by-age subgroup interactions were observed for discontinuation of study drug because of AEs, SAEs or incidences of hypoglycaemia .
# | discussion
The AWARD-11 trial showed that dulaglutide 3.0 or 4.5 mg versus The tolerability profile of dulaglutide across all three dose groups was comparable between age groups. As expected, the most common TEAEs reported across all doses in AWARD-11 were nausea, vomiting and diarrhoea. In the overall study population, the incidence of nausea was similar across dose groups, whereas diarrhoea and vomiting were more frequently reported in the 3.0 and 4.5 mg groups. [bib_ref] Efficacy and safety of Dulaglutide 3.0 mg and 4.5 mg versus Dulaglutide..., Frias [/bib_ref] In the current subgroup analysis, there was no dose relationship between reports of common gastrointestinal (GI) events in the older age subgroup compared with the younger age subgroup, and the overall incidences of these events among older patients was generally similar to those reported in the younger age group. Cardiac complications (arrythmias, conduction disorders and CV events) were also similar across doses and age subgroups. Discontinuations of study drug as a result of any AE or specifically because of nausea, vomiting or diarrhoea were not significantly different among older versus younger patients, further supporting the conclusion that overall and GI tolerability of dulaglutide across the dose range studied in AWARD-11 are similar between age groups, consistent with prior studies with lower doses of dulaglutide. [bib_ref] Similar efficacy and safety of once-weekly dulaglutide in patients with type 2..., Boustani [/bib_ref] [bib_ref] Efficacy and safety of once-weekly dulaglutide in elderly Chinese patients with type..., Kuang [/bib_ref] The dulaglutide safety profile related to SAEs was also comparable between age groups across all doses. The incidence of SAEs was not dose-related in either age subgroup.
This analysis has certain limitations that may influence the interpretation and generalizability of the results. With no evidence of treatment-by-age interaction, the most appropriate estimate of the effect size for any of the endpoints is that observed in the overall population of the trial; however, it is reassuring that further exploratory subgroup analysis by age remains consistent with the overall results for the
## Peer review
The peer review history for this article is available at https://publons. com/publon/10.1111/dom.14469.
# Data availability statement
Lilly provides access to all individual participant data collected during the trial, after anonymization, with the exception of pharmacokinetic or genetic data. Data are available to request 6 months after the indication studied has been approved in the US and EU and after primary publication acceptance No expiration date of data requests is currently set once they are made available. Access is provided after a proposal has been approved by an independent review committee identified for this purpose and after receipt of a signed data sharing agreement. Data and documents, including the study protocol, statistical analysis plan, clinical study report, blank or annotated case report forms, will be provided in a secure data sharing environment for up to 2 years per proposal. For details on submitting a request, see the instructions provided at vivli.org.
## Orcid
[fig] Figure 2A -: C). [/fig]
[fig] 1. 5 F: mg once-weekly provides dose-related improvements in glycaemic control and BW reduction at 36 weeks that are sustained through 52 weeks. Similarly, this exploratory post hoc analysis showed that treatment with once-weekly dulaglutide 3.0 and 4.5 mg resulted in clinically relevant reductions in HbA1c and BW compared with dulaglutide 1.5 mg in the younger than 65 and 65 years or older age subgroups, and also the younger than 70 and 70 years or older age subgroups. The safety profile for the additional dulaglutide doses was . Data are presented as LSM ± SE unless otherwise indicated. Abbreviations: BL, baseline; LSM, least-squares mean; n, sample size; SE, standard error consistent with that of dulaglutide 1.5 mg and similar between age subgroups. These results were in line with the findings of previous published studies that showed similar glycaemic and BW effects and safety profiles irrespective of baseline age when treated with dulaglutide doses of 0.75 and 1.5 mg. 9,12 These additional dulaglutide doses may benefit elderly patients who are not achieving glycaemic targets on a lower dulaglutide dose, as they can remain on their current therapy with familiar administration and tolerability experience. The growing demographic of elderly adults with T2D, spanning those who are healthier and robust and others who are more frail, will require individualized glycaemic targets based on each patient's health status. Medication classes with a low risk of hypoglycaemia such as GLP-1 RAs have been recommended by the ADA and the European Association for the Study of Diabetes as the preferred treatment over insulin in adults with T2D with established and high-risk ASCVD, obesity, and in those who are at an increased risk of hypoglycaemia. The current guidelines for treating T2D recommend an HbA1c of less than 7% (<53 mmol/mol) for adults who are considered comparatively healthy. For comparatively healthy older people with an extended life expectancy, a glycaemic target similar to that for younger patients may be most appropriate. Less stringent glycaemic treatment goals are recommended for more frail elderly patients to keep the risk of hypoglycaemia low while also taking co-morbid illness and/or limited life expectancy into consideration. 4,10,14,23 The availability of four dulaglutide doses (0.75, 1.5, 3.0 and 4.5 mg) provides additional options to individualize a patient's treatment to achieve the respective glycaemic target tailored to the patient's health status and life expectancy.The proportion of patients achieving the composite endpoint target of an HbA1c of less than 7% (<53 mmol/mol) with no weight gain and no documented or severe hypoglycaemia at 36 weeks was also similar, irrespective of baseline age. Sustaining glycaemic control targets over time is associated with a reduced risk of microvascular complications (retinopathy, nephropathy, neuropathy) and ASCVD in patients with diabetes[24][25][26] ; achieving such control with a low risk of hypoglycaemia is particularly important in older adults because they generally have a higher risk of hypoglycaemia, a longer duration of T2D and a higher risk of microvascular complications and ASCVD. [/fig]
[table] Table 2: The treatment-by-age subgroup interaction for SAEs was T A B L E 1 Baseline characteristics and demographics of the <65 and ≥65 year subgroups [/table]
|
Pelvic Actynomyces Infection: Report of Two Cases Occurred in the Hospital of San José
The actynomyces infection is a rare cause of chronic pelvic inflammation, which can be manifested in multiple ways. It is caused by the actynomyces bacteria, usually by the israelii type, which can be a part of the normal flora of the genital tract in patients who use intrauterine device (IUD). There is a discussion about the importance of considering this infection disease as part of the differential diagnosis in patients using the IUD, with atypical manifestations and bizarre presentation of infections of the genital tract, severe pelvic adherent syndromes, tubo-ovarian complexes (abscesses) barely symptomatic, and in the case of intraoperatory suspicion of pelvic carcinomatosis among others.
The patient is a 38-year old female. She was admitted to the service for a scheduled extraction of IUD by hysteroscopy. The patient was asymptomatic. She has had the IUD for 9 years. During the hysteroscopy, it was evidenced that the IUD was perforating the uterine wall at the level of the fundus. There was visualization of intestinal loops through the defect. The laparoscopic procedure was converted into a laparotomy due to the finding of sealed pelvis. The following are the findings from the exploratory laparotomy: uterine perforation, right adnexal mass in plastron, left tuboovarian complex perforated comprising the bowel wall at the level of the sigmoid colon, and secondary peritonitis. Hysterectomy and bilateral salpingo-oophorectomy were done. A Hartmann hemicolectomy plus drainage of the peritonitis and lavage of the peritoneal cavity were also done. The postoperative progress of the patient was adequate with absence of inflammatory response signs. The patient was discharged home. The histopathological analysis of the specimen showed infection by actynomyces comprising the uterus adnexal tissues and with severe inflammatory process of the uterine wall. Based on the results of the histopathology, the patient was readmitted to the service for intravenous antibiotic therapy. She was given crystalline penicillin 1 000 000 IU per hour by continuous infusion for 15 days. Then, she was discharged home on amoxicillin 2 grams BID, PO for six months. The patient has been followed up in two occasions one month apart in the outpatient clinic. The physical examination was within normal limits and laboratory workup showed no findings of infection or sepsis.
## Case 2
The patient is a 23-year old female. She was admitted to the emergency room with a main complain of three-month abdominal-pelvic pain associated with the feeling of abdominal mass in some occasions, subjective fever (not assessed by thermometer) in multiple occasions, asthenia, adynamia, and weight loss. As a relevant past medical history, the patient has been diagnosed with bilateral adnexal masses, which have been followed up with tumor marker studies that are negative until this date. The past surgical history is positive for appendectomy and two c-sections. The contraception method used by the patient for the last three years was the IUD. When the patient was admitted to the service she presented a severe abdominal pain. Physical examination revealed a right adnexal mass of approximated 6 centimeters in diameter that was bounded to the body of the uterus and was not painful at manipulation. There was also a left adnexal mass of approximated 10 centimeters of diameter which was painful at palpation. Based on the findings of suffering left adnexal mass, the patient was taken to an exploratory laparotomy due to acute abdomen diagnosis. Intraoperative findings were the following: severe adhesion syndrome with sealed pelvis, and 3 centimeter diameter cystic mass in the omentum with signs of ischemia. During Infectious Diseases in Obstetrics and Gynecology the surgical resection there was an incidental dissection of the intestinal serosa near the tubo-ovarian complex. Bilateral salpingo-oophorectomy was done with complete resection of the plastron and suture of the serosa was also done. Three days later the patient was readmitted by acute abdomen, she was taken back to the operation room for a new laparotomy. The findings during the laparotomy were as follows: generalized peritonitis secondary to a perforation at the level of the ileum and rectum, which required management with a colostomy. The histopathological study was compatible with bilateral tubo-ovarian abscess with infection by actynomyces. The patient received antibiotic therapy with crystalline penicillin 24 000 000 IU per day intravenously. Once the histopathology diagnosis was confirmed the patient continued therapy with ampicillin for another 6 months.
# Discussion
The actynomyces infection is a chronic infection. It is caused by the actynomyces organism which is a gram positive anaerobe. The infection is usually localized around the cervicofacial area, but it could also be present with manifestations at the lungs, abdominal, or urogenital levels. The infection that presents at the pelvic level corresponds to a 10 to 15% of all reported cases. There is a higher incidence of the infection in males than females, except in the abdominal and pelvic localization where the higher incidence is in females. It has been reported as an incidence of 1 to 16% of all of the pelvic infections [bib_ref] Tuboovaric actinomycosis. A report of a case, Salazar [/bib_ref]. Actynomyces is a rare cause of chronic infection; it causes local infiltration and induration of the tissues. Its role as a pathologic entity in the genital female tract is discussed because its manifestations have a wide spectrum that includes: asymptomatic infection, chronic pelvic pain, adnexal mass, tubo-ovarian complex, tumor-like carcinoma, and pelvic adhesion syndrome. The preoperative diagnosis is hard to achieve due to the finding of the microorganism by culture or immunofluorescence studies of vaginal secretion without being pathologic or any association to abdominal pain or sepsis. Although the chance of making a precise diagnosis is very attractive because it could change the therapeutic options, these facts we must consider: it is a nonmalignant process; it is a microorganism very sensible to B-lactams antibiotics; the radical surgery in some cases can be complicated with dissection of the intestinal and urinary tracts [bib_ref] Pelvic actinomycosis. Urologic perspective, Marella [/bib_ref]. Once the diagnosis has been confirmed, the patients should be on long term therapy with B-lactams antibiotics.
|
Contemporary Management Strategies for Chronic Type B Aortic Dissections: A Systematic Review
BackgroundCurrently, the optimal management strategy for chronic type B aortic dissections (CBAD) is unknown. Therefore, we systematically reviewed the literature to compare results of open surgical repair (OSR), standard thoracic endovascular aortic repair (TEVAR) or branched and fenestrated TEVAR (BEVAR/FEVAR) for CBAD.MethodsEMBASE and MEDLINE databases were searched for eligible studies between January 2000 and October 2015. Studies describing outcomes of OSR, TEVAR, B/FEVAR, or all, for CBAD patients initially treated with medical therapy, were included. Primary endpoints were early mortality, and one-year and five-year survival. Secondary endpoints included occurrence of complications. Furthermore, a Time until Treatment Equipoise (TUTE) graph was constructed.ResultsThirty-five articles were selected for systematic review. A total of 1081 OSR patients, 1397 TEVAR patients and 61 B/FEVAR patients were identified. Early mortality ranged from 5.6% to 21.0% for OSR, 0.0% to 13.7% for TEVAR, and 0.0% to 9.7% for B/FEVAR. For OSR, one-year and five-year survival ranged 72.0%-92.0% and 53.0%-86.7%, respectively. For TEVAR, one-year survival was 82.9%-100.0% and five-year survival 70.0%-88.9%. For B/FEVAR only one-year survival was available, ranging between 76.4% and 100.0%. Most common postoperative complications included stroke (OSR 0.0%-13.3%, TEVAR 0.0%-11.8%), spinal cord ischemia (OSR 0.0%-16.4%, TEVAR 0.0%-12.5%, B/FEVAR 0.0%-PLOS ONE | 12.9%) and acute renal failure (OSR 0.0%-33.3%, TEVAR 0.0%-34.4%, B/FEVAR 0.0%-3.2%). Most common long-term complications after OSR included aneurysm formation (5.8%-20.0%) and new type A dissection (1.7-2.2%). Early complications after TEVAR included retrograde dissection (0.0%-7.1%), malperfusion (1.3%-9.4%), cardiac complications (0.0%-5.9%) and rupture (0.5%-5.0%). Most common long-term complications after TEVAR were rupture (0.5%-7.1%), endoleaks (0.0%-15.8%) and cardiac complications (5.9%-7.1%). No short-term aortic rupture or malperfusion was observed after B/FEVAR. Long-term complications included malperfusion (6.5%) and endoleaks (0.0%-66.7%). Reintervention rates after OSR, TEVAR and B/FEVAR were 5.8%-29.0%, 4.3%-47.4% and 0.0%-53.3%, respectively. TUTE for OSR was 2.7 years, for TEVAR 9.9 months and for B/ FEVAR 10.3 months.ConclusionWe found a limited early survival benefit of standard TEVAR over OSR for CBAD. Complication rates after TEVAR are higher, but complications after OSR are usually more serious. Initial experiences with B/FEVAR show its feasibility, but long-term results are needed to compare it to OSR and standard TEVAR. We conclude that optimal treatment of CBAD remains debatable and merits a patient specific decision. TUTE seems a feasible and useful tool to better understand management outcomes of CBAD.
# Introduction
Currently, the optimal management of chronic type B aortic dissections (CBAD) remains undetermined as there have been no randomized controlled trials comparing open surgical repair (OSR) and thoracic endovascular aortic repair (TEVAR) [bib_ref] ESC Guidelines on the diagnosis and treatment of aortic diseases: Document covering..., Erbel [/bib_ref]. Furthermore, branched and fenestrated TEVAR (B/FEVAR) are emerging as new techniques to treat more complicated cases with an endovascular approach [bib_ref] Chronic dissectionindications for treatment with branched and fenestrated stent-grafts, Sobocinski [/bib_ref]. The initial treatment objective for uncomplicated acute type B dissections is clinical stabilization of the patient through optimal medical therapy (OMT) to prevent propagation of the dissection, malperfusion, rapid aortic dilatation and/or rupture. However, secondary interventions after initial OMT are common, with intervention rates ranging between 9.0% and 40.6% [bib_ref] Effects of the patent false lumen on the long-term outcome of type..., Akutsu [/bib_ref] [bib_ref] Long-term outcome of aortic dissection with patent false lumen: predictive role of..., Evangelista [/bib_ref] [bib_ref] Early and late outcome of operated and non-operated acute dissection of the..., Gysi [/bib_ref] [bib_ref] Usefulness of fibrinogen/fibrin degradation product to predict poor one-year outcome of medically..., Kitada [/bib_ref] [bib_ref] Predictors of surgical indications for acute type B aortic dissection based on..., Kunishige [/bib_ref] [bib_ref] Degree of fusiform dilatation of the proximal descending aorta in type B..., Marui [/bib_ref] [bib_ref] Long-term follow-up of acute type B aortic dissection: ulcer-like projections in thrombosed..., Miyahara [/bib_ref] [bib_ref] Open repair of chronic distal aortic dissection in the endovascular era: Implications..., Pujara [/bib_ref] [bib_ref] Growth rate of aortic diameter in patients with type B aortic dissection..., Sueyoshi [/bib_ref] [bib_ref] Growth rate of affected aorta in patients with type B partially closed..., Sueyoshi [/bib_ref] [bib_ref] Tight heart rate control reduces secondary adverse events in patients with type..., Kodama [/bib_ref]. Most common indications for secondary interventions for CBAD include aneurysm formation, rapid aneurysmal sac enlargement, extension of dissection and malperfusion [bib_ref] Effects of the patent false lumen on the long-term outcome of type..., Akutsu [/bib_ref] [bib_ref] Long-term outcome of aortic dissection with patent false lumen: predictive role of..., Evangelista [/bib_ref] [bib_ref] Early and late outcome of operated and non-operated acute dissection of the..., Gysi [/bib_ref] [bib_ref] Usefulness of fibrinogen/fibrin degradation product to predict poor one-year outcome of medically..., Kitada [/bib_ref] [bib_ref] Predictors of surgical indications for acute type B aortic dissection based on..., Kunishige [/bib_ref] [bib_ref] Degree of fusiform dilatation of the proximal descending aorta in type B..., Marui [/bib_ref] [bib_ref] Long-term follow-up of acute type B aortic dissection: ulcer-like projections in thrombosed..., Miyahara [/bib_ref] [bib_ref] Open repair of chronic distal aortic dissection in the endovascular era: Implications..., Pujara [/bib_ref] [bib_ref] Growth rate of aortic diameter in patients with type B aortic dissection..., Sueyoshi [/bib_ref] [bib_ref] Growth rate of affected aorta in patients with type B partially closed..., Sueyoshi [/bib_ref] [bib_ref] A prospective study of medically treated acute type B aortic dissection, Winnerkvist [/bib_ref].
Both endovascular therapies and OSR show up-and downsides; endovascular management is less invasive, however successful treatment during the chronic phase may be challenging due to thickening of the intimal flap. Standard TEVAR for CBAD patients has shown acceptable mid-term outcomes, however complete aortic remodeling was seen in only 36% of cases [bib_ref] Endovascular stent-graft placement of aneurysms involving the descending aorta originating from chronic..., Czerny [/bib_ref] , mostly precluded due to abdominal extension of the dissection. Such extended involvement determines a thoracoabdominal aortic issue that may require a more extensive repair. In such a setting, branched and fenestrated procedures may offer an endovascular solution. However, anatomical limitations like narrow lumens and technical difficulties, such as the orientation of the branches and fenestrations, and the presence of the lamella inside the lumen, make the procedure challenging. Nevertheless, in general, any type of endovascular management could be of value in chronic patients, reducing operative risks of OSR. An open approach is more invasive with higher operative risks [bib_ref] Surgical management of descending thoracic aortic disease: open and endovascular approaches: a..., Coady [/bib_ref] , but unlike endovascular management, it is rarely affected by anatomical constraints. Currently, OSR is the most commonly adopted strategy, in particular in younger patients and those affected by connective tissue disorders, while endovascular treatment has been adopted for specific clinical scenarios such as limited extent of the dissection, intramural hematoma evolving with penetrating aortic ulcer, and older patients. Our aim was to systematically review the literature and compare outcomes of CBAD patients managed with OSR, TEVAR and B/FEVAR, who were initially treated with OMT alone.
# Materials and methods
## Search strategy
The EMBASE and MEDLINE databases were searched for eligible studies from January 1 st , 2000 up to October 1 st , 2015. The following search terms were used: 'follow-up', 'chronic', 'post-dissection' 'type B', 'aortic/aorta dissection' and 'outcome', or synonyms (S1 Appendix).
## Article selection
The Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines were used for analysis of the studies in this systematic review [bib_ref] Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of..., Stroup [/bib_ref]. Included studies were critically assessed for study design, heterogeneity, possible bias, and other limitations. Two reviewers (AK and HB) performed eligibility for this systematic review independently. Disagreement between reviewers was resolved during a consensus meeting. Inclusion criteria were: (1) English language; (2) case series describing outcomes of OSR, TEVAR, B/FEVAR or multiple, for CBAD; and (3) follow-up of at least one year. Exclusion criteria were: (1) studies before 2000 to ensure contemporary practice in all included studies; (2) patients initially treated with invasive therapy; (3) case reports; (4) studies describing mixed populations without separate outcomes listed; and (5) articles from the same institution or author were studied critically to ensure no overlapping patient populations were included in the final analysis.
## Extracted data and endpoints
Data extracted included demographics, patient history, intervention details, and follow-up outcomes. The primary endpoints were early mortality, and one-and five-year survival. Secondary endpoints included the occurrence of complications. Early outcomes were defined as either inhospital or 30-day outcome. Long-term outcomes were defined as occurring during follow-up. Rapid aortic enlargement was defined as 0.5 cm increase in diameter per year.
## Tute
The concept of "time until treatment equipoise" (TUTE) has been described in an attempt to better and easier advise patients of relative risks of different management modalities [bib_ref] Time until treatment equipoise: a new concept in surgical decision making, Noorani [/bib_ref]. It is defined as the duration of time that elapses after an intervention, before the risk of the intervention is canceled out and reversed by the cumulative risk of conservative management. In other words, it is the point in time during follow-up after which an intervention is most beneficial, because the mortality risk of the intervention is lower than the mortality risk of continuing conservative management. TUTE may guide decision making for asymptomatic patients on prognostic grounds. The equipoise is set at the point where the areas between the survival curves of no intervention and intervention are equal. This point is chosen instead of the crossing of the lines, because the intervention itself also carries operative mortality risk, which needs to be accounted for. To estimate TUTE for OSR, TEVAR and B/FEVAR, we performed a TUTE analysis as described by the creators of the concept [bib_ref] Time until treatment equipoise: a new concept in surgical decision making, Noorani [/bib_ref]. In this analysis, the gender, age, mortality rate for the intervention and the expected added mortality rate per year due to the natural history of the condition, are entered. We used the mean age for each of the interventions and the 30-day mortality rate for each intervention. The expected mortality without intervention was adopted from recent available literature [bib_ref] Management of complicated and uncomplicated acute type B dissection. A systematic review..., Moulakakis [/bib_ref]. Based on these risk factors, survival curves are calculated, and the point in time where the area between the two curves before and after the crossing of the lines is equal (intervention vs. no intervention) is given. This is the point in time after which an intervention improves survival compared to only medical management, e.g. TUTE.
# Statistical analysis
We discussed the end-points and our rationale for this study with our institutional statistical center (Julius Support Center, UMC Utrecht, the Netherlands). After initial investigation of available studies, it was concluded that a meta-analysis was not feasible and not advisable. This decision was made because of the large heterogeneity among the available literature, since all studies used different in-and exclusion criteria, diverse definitions, and reported different follow-up times. Furthermore, in many studies the original data were not present.
Data are presented as mean ± SD or as percentage. Percentages per variable were calculated by dividing the observed total by the combined total of patients from the studies reporting the characteristic. Values of <0.05 were considered significant. Data analysis and graphing were performed using Microsoft Excel (Microsoft Inc.) and SPSS (SPSS 22 Inc., Chicago, Ill, USA).
# Results
## Included studies
A total of 35 articles were selected for systematic review [fig_ref] Fig 1: Flowchart of article selection [/fig_ref]. The initial search of EMBASE and MEDLINE provided 702 articles. After removal of duplicates, 579 articles remained. Of these, 404 articles could be excluded based on the content of the abstract. Seventy-one full-text articles were assessed more closely, after which another 40 articles were excluded. Thirty-one articles were deemed eligible for this systematic review. Cross-referencing of the remaining articles yielded four articles, leaving a final number of 35 articles. No qualitative analysis, e.g. meta-analysis, was performed since the heterogeneity of included studies was large, and therefore a quantitative analysis was most suitable.
## Patient characteristics
The OSR cohort comprised of 1081 patients, with a mean age of 58.2 ± 3.8 years [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref]. The reported follow-up for these studies ranged between 34 months and 102 months. Overall, there were 816 males (74.2%) [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] [bib_ref] Chronic aortic dissection not a risk factor for neurologic deficit in thoracoabdominal..., Safi [/bib_ref]. Demographic details and medical history of these patients are listed in S1 [fig_ref] Table: Demographics and OSR details [/fig_ref] The TEVAR cohort consisted of 1397 patients with a mean age of 59.4 ± 4.2 years and 76.0% was male (n = 1051) [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref]. Follow-up ranged between 12 and 90 months. All TEVAR patient characteristics are shown in S2 [fig_ref] Table: Demographics and OSR details [/fig_ref] Sixty-one B/FEVAR patients were included, mean age 65.7 (± 8.0 years) [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Follow-up ranged between 17 months and 20.4 months. Overall, 83.6% of patients were males. All other details are listed in S3 [fig_ref] Table: Demographics and OSR details [/fig_ref] Interventional details Several studies described the timing of OSR; elective interventions were performed between 53.2% and 95.7% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] , while urgent and emergent interventions were performed between 10.5% and 12.5% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] , and 3.1% to 7.7% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] , respectively. The exact interval between incident dissection and OSR was available in three studies, ranging between a mean of 32.4-61.0 months [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] and a median of 31.0 (thoraco-abdominal extent) to 43.0 months (limited to descending aorta) [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref]. Mean time to intervention was 35.5 months. For 37.4% of the patients the intervention was limited to the descending aorta , while for the remaining patients the intervention was thoraco-abdominal . The operative details are reported in S1 [fig_ref] Table: Demographics and OSR details [/fig_ref] For TEVAR, elective interventions ranged between 62.5% and 100.0% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Midterm results of stent-graft repair of acute and chronic aortic dissection with..., Kato [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref] , urgent operations between 18.8% and 97.2% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] , and emergency procedures were performed between 2.0% and 18.8% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] [bib_ref] Aortic remodeling after endovascular repair with stainless steel-based stent graft in acute..., Yang [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref]. Reported time intervals from incident dissection to TEVAR ranged between a median of 3 weeks up to 36.0 months [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Midterm results of thoracic endovascular aortic repair in patients with aneurysms involving..., Czerny [/bib_ref] [bib_ref] The results of stent graft versus medication therapy for chronic type B..., Jia [/bib_ref] [bib_ref] Technical and clinical success after endovascular therapy for chronic type B aortic..., Oberhuber [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] and a mean of 3 weeks and 53.8 months [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Comparison of long-term clinical outcome between patients with chronic versus acute type..., Chen [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Midterm results of stent-graft repair of acute and chronic aortic dissection with..., Kato [/bib_ref] [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] [bib_ref] Early and midterm results of thoracic endovascular aortic repair of chronic type..., Xu [/bib_ref]. Mean time to intervention was 24.4 months. Indications for TEVAR were aortic aneurysm (74.5% to 100.0%) [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Endovascular repair of Stanford type B aortic dissection: early and mid-term outcomes..., Guangqi [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Transluminal stent-graft placements for the treatments of acute onset and chronic aortic..., Shimono [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] , failure of OMT (12.3%) [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] , rupture (2.7% to 10.0%) [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Endovascular repair of Stanford type B aortic dissection: early and mid-term outcomes..., Guangqi [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] , rapid aortic enlargement (11.8% to 100.0%) [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] , recurrent/refractory pain (4.3% to 57.7%) [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] , malperfusion (2.5% to 18.8%) [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] , patent false lumen (FL) (64.2%) [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref] and other indications (6.6% to 23.5%) [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref]. Double indications could be present in a single patient. Complete details are listed in S2 Table. All B/FEVAR procedures were elective and the only reported indication for these procedures was aneurysmal degeneration [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Reported time from incident dissection to the intervention was noted in one study and was 31.0 months [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Technical success was achieved in 93.5% in one case series [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref] and 100.0% in another study [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref]. Complete details are listed in S3 [fig_ref] Table: Demographics and OSR details [/fig_ref] Survival Short-term mortality after OSR ranged between 5.6% and 21.0% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref]. One-, five-and tenyear survival was between 72.0% and 92.0% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] , 53.0% and 86.7% , and between 32.0% and 60.0% [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] , respectively. One study reported a 15-year survival of 49.0% [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [fig_ref] Table 1: Complications and survival OSR [/fig_ref].
For TEVAR, early mortality was reported to be between 0.0% and 13.7% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref]. One-year survival was between 82.9% and 100.0% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Endovascular repair of Stanford type B aortic dissection: early and mid-term outcomes..., Guangqi [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Efficacy of thoracic endovascular stent repair for chronic type B aortic dissection..., Scali [/bib_ref]. Five-year survival was between 64.0% and 88.9% [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Efficacy of thoracic endovascular stent repair for chronic type B aortic dissection..., Scali [/bib_ref] [bib_ref] Early and midterm results of thoracic endovascular aortic repair of chronic type..., Xu [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref]. Ten-year survival was reported in one study, with a survival rate of 63.0% [bib_ref] Key success factors for thoracic endovascular aortic repair for non-acute Stanford type..., Kitamura [/bib_ref]. One-year dissection-related survival was 97.1% in one study [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] and five-year dissection-related survival ranged between 92.1% and 98.3% [bib_ref] Intermediate to long-term outcomes of endoluminal stent-graft repair in patients with chronic..., Kim [/bib_ref] [bib_ref] Outcomes of endovascular management for complicated chronic type B aortic dissection: effect..., Lee [/bib_ref] [bib_ref] Midterm results with thoracic endovascular aortic repair for chronic type B aortic..., Parsa [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref]. Early mortality after B/FEVAR ranged between 0.0% and 9.7% [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. One-year survival was between 76.4% and 100.0% [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Five-year survival was not available in these studies. Two-and three-year survival was noted in one study, being 85.0%-100.0% and 75.0%-85.0%, respectively [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref]. All details can be found in [fig_ref] Table 3: Complications and survival B/FEVAR [/fig_ref].
Three studies reported outcomes for both OSR and TEVAR. These studies might be of most predictive value and most informative, since they compared both interventions using a similar population of patients. No differences were reported for one-year survival between OSR and TEVAR [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] , as well as for five-year survival (p-value 0.23) [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref].
## Complications
For OSR, the most common early complications were stroke (0.0% and 13.3%) [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Outcomes of open repair for chronic descending thoracic aortic dissection, Estrera [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] , spinal cord ischemia 0.0%-16.4%) [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Effect of chronic dissection on early and late outcomes after descending thoracic..., Conrad [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref] [bib_ref] Chronic aortic dissection not a risk factor for neurologic deficit in thoracoabdominal..., Safi [/bib_ref] and acute renal failure (0.0% -33.3%). Long-term complications after OSR included aneurysm formation (5.8%-20.0%) [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] and new type A dissection (1.7-2.2%) [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref]. Complete results are shown in [fig_ref] Table 1: Complications and survival OSR [/fig_ref].
Most common early complications after TEVAR included stroke (0.0%-11.8%) were common during follow-up. Other late or long-term complications included rupture (0.5%-7.1%) , malperfusion (0.0%-3.6%) [bib_ref] Comparison of long-term clinical outcome between patients with chronic versus acute type..., Chen [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Midterm results of stent-graft repair of acute and chronic aortic dissection with..., Kato [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref] and cardiac complications (5.9%-7.1%) [bib_ref] Midterm results of thoracic endovascular aortic repair in patients with aneurysms involving..., Czerny [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref]. All results are listed in.
Early complications after B/FEVAR included spinal cord ischemia (0.0%-12.9%) , acute renal failure (0.0%-3.2%) [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref] , and cardiac complications (3.2%) [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Late complications included malperfusion (0.0%-6.5%) and endoleaks (0.0%-66.7%) [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. All results are listed in [fig_ref] Table 3: Complications and survival B/FEVAR [/fig_ref].
## Reinterventions
Reoperations after OSR were common, ranging between 5.8% and 29.0% [fig_ref] Table 1: Complications and survival OSR [/fig_ref] [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref]. Most reinterventions after OSR were managed with another open repair; only in a small number of cases an endovascular procedure was performed [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Mid-term outcome with surgery for type B aortic dissections: a single center..., Goksel [/bib_ref] [bib_ref] Open thoracoabdominal aortic repair for chronic type B dissection, Kouchoukos [/bib_ref] [bib_ref] Optimal treatment for patients with chronic Stanford type B aortic dissection: endovascularly,..., Nozdrzykowski [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Operative Results and Clinical Features of Chronic Stanford Type B Aortic Dissection:..., Fujikawa [/bib_ref]. Reasons for reintervention included retrograde dissection [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] , bleeding [bib_ref] Long-term outcomes in thoracoabdominal aortic aneurysm repair for chronic type B dissection, Bashir [/bib_ref] , aneurysm formation [bib_ref] Outcomes of open surgical repair for chronic type B aortic dissections, Conway [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] and renal failure [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref]. One study described mortality after the reintervention (30.0%). Indications for these procedures were graft infection, thoraco-abdominal aneurysm and aneurysmal growth of distal dissection. Procedures performed were TEVAR (n = 2) and hybrid (n = 1) [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref].
For TEVAR, reinterventions were reported between 4.3% and 47.4% [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref] [bib_ref] Endovascular repair of complicated chronic distal aortic dissections: intermediate outcomes and complications, Kang [/bib_ref] [bib_ref] Aortic morphology following endovascular repair of acute and chronic type B aortic..., Sayer [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref] , distal FL perfusion [bib_ref] Current management and outcome of chronic type B aortic dissection: results with..., Andersen [/bib_ref] [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] [bib_ref] Technical and clinical success after endovascular therapy for chronic type B aortic..., Oberhuber [/bib_ref] [bib_ref] Endograft exclusion of acute and chronic descending thoracic aortic dissections, Song [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref] , rupture [bib_ref] Midterm results of thoracic endovascular aortic repair in patients with aneurysms involving..., Czerny [/bib_ref] [bib_ref] The results of stent graft versus medication therapy for chronic type B..., Jia [/bib_ref] [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref] , and malperfusion syndromes [bib_ref] Aortic pathology determines midterm outcome after endovascular repair of the thoracic aorta:..., Patterson [/bib_ref]. Only a few studies described outcomes after secondary intervention: Andacheh et al. reported no mortality [bib_ref] Patient outcomes and thoracic aortic volume and morphologic changes following thoracic endovascular..., Andacheh [/bib_ref] , while van Bogerijen reported two deaths due to type III endoleaks (40.0%) [bib_ref] Propensity adjusted analysis of open and endovascular thoracic aortic repair for chronic..., Van Bogerijen [/bib_ref]. Jia et al. reported a mortality of 66.6% after secondary intervention; reasons for mortality were multi-organ failure (n = 1), type A dissection (n = 1) and unknown (n = 4) [bib_ref] The results of stent graft versus medication therapy for chronic type B..., Jia [/bib_ref]. Lastly, Nathan et al. reported one death after open surgical reintervention (11.1%) [bib_ref] Outcomes after stent graft therapy for dissection-related aneurysmal degeneration in the descending..., Nathan [/bib_ref].
Reinterventions after B/FEVAR were between 0.0% and 53.3% [fig_ref] Table 3: Complications and survival B/FEVAR [/fig_ref] [bib_ref] Fenestrated and branched endovascular aortic repair for chronic type B aortic dissection..., Kitagawa [/bib_ref] [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref]. Reported reinterventions were all endovascular, always for treating endoleaks. Only one study described outcomes after secondary intervention, with no observed mortality [bib_ref] Outcomes of fenestrated/ branched endografting in post-dissection thoracoabdominal aortic aneurysms, Oikonomou [/bib_ref].
## Time until treatment equipoise (tute)
The following parameters were entered in the TUTE analysis for each intervention: OSR (Male, 58 years, intervention mortality 9.9%, no intervention mortality 7.5%), TEVAR (male, 59 years, mortality intervention 3.1%, no intervention mortality 7.5%), and B/FEVAR (male, 65 years, mortality intervention 3.2%, no intervention mortality 7.5%). This resulted in TUTE for OSR of 2.7 years, for regular TEVAR this was 9.9 months and for B/FEVAR 10.3 months [fig_ref] Fig 2: Time until Treatment Equipoise [/fig_ref]. This suggests TEVAR is the treatment that will have the earliest beneficial impact, compared to OSR and B/FEVAR. This is because TEVAR has lower operative risks compared to OSR. The available results of B/FEVAR are limited in current literature, making comparison vulnerable to bias. However, B/FEVAR seems to become more beneficial than just medical management after a similar timeframe as standard TEVAR, about 9-10 months after the incident dissection.
# Discussion
In this systematic review, short and long-term mortality rates of TEVAR for CBAD seemed to be favorable compared to those managed with OSR. To actually compare the results of B/ FEVAR to the other two intervention strategies is challenging because of the small number of studies available. Most complications were observed after TEVAR, mainly related to the stentgraft, such as endoleak, rupture, and malperfusion. Although more reinterventions were required after TEVAR, the type of procedure was usually less invasive. A large percentage (>60.0%) of the reinterventions was represented by another endovascular procedure, an embolization or a ballooning of the stent-graft. Complications following OSR were usually more severe and the subsequent reintervention was frequently another open procedure. Complication rates after B/FEVAR were generally low, usually represented by endoleaks, and reintervention rates were high, always endovascular. Our results show that it is difficult to distinguish which intervention is most suitable for CBAD. Currently, no randomized controlled clinical trials exist, mainly due to the rarity of the disease, to provide definitive evidence on optimal management strategy for CBAD. Therefore, comparison of observational data of these management modalities is important.
Actually, management of complicated acute and subacute type B dissection is usually performed by TEVAR while OSR is reserved for those patients affected by connective tissue disorders or with unsuitable anatomy for endovascular approach [bib_ref] ESC Guidelines on the diagnosis and treatment of aortic diseases: Document covering..., Erbel [/bib_ref]. For CBAD patients this choice is more challenging, because other factors play a role in decision-making. In the chronic phase the TL is often small due to chronic compression of the lumen and scarring and thickening of the intimal flap occurs. Therefore, TL expansion and aortic remodeling is more challenging to accomplish when compared to treatment in the acute phase [bib_ref] ESC Guidelines on the diagnosis and treatment of aortic diseases: Document covering..., Erbel [/bib_ref]. Moreover, the frequent involvement of the abdominal aorta in type B dissection explains why TEVAR for chronic dissections, although liberally used, is associated with high reintervention rates.
OSR is often used as the treatment of choice for CBAD, especially for extensive dissections involving visceral arteries and for patients that are deemed unsuitable for TEVAR. Besides several technical problems, as mentioned above, short landing zones or strong angulation in the arch could inhibit the use of endovascular techniques. B/FEVAR allow for treatment of more challenging cases by endovascular means. This novel technique has been reported in highly selected cases in limited expert centers because of challenging issues related to narrow lumens, correct orientation of branches and fenestrations, and diminished sealing capacity in such setting. Because of limited reporting on B/FEVAR, it is difficult to compare it to standard TEVAR and OSR, and additional and long-term results are highly anticipated.
TUTE has been recently introduced to educate patients, but also to determine appropriate timing of an intervention [bib_ref] Time until treatment equipoise: a new concept in surgical decision making, Noorani [/bib_ref]. Our analysis showed that the TUTE for regular TEVAR was 9.9 months, 10.3 months for B/FEVAR, and 2.7 years for OSR. Such results are in agreement with the increasing CBAD standard TEVAR management. The reason lies in the lower operative risk compared to OSR, associated with a relevant percentage of positive outcomes, despite higher rates of reintervention.
This systematic review has several limitations; first, we did not perform qualitative analyses. After careful consideration with our affiliated statistical center (Julius Support Center, UMC Utrecht, The Netherlands), it was considered to be not feasible and advisable to perform a meta-analysis. The heterogeneity among the data was too large, since all studies used different in-and exclusion criteria, diverse definitions, and reported different follow-up times. Furthermore, in many studies the original data were not present. Another limitation is that the rate of elective or urgent/emergent interventions differed fundamentally between studies, and a large number of studies did not report any procedural details. This might have caused differences in occurrence of complications and mortality.
# Conclusion
In conclusion, this systematic review suggests a limited early survival benefit of standard TEVAR over OSR for CBAD. The complication rates after TEVAR are higher, but the complications after OSR are usually more serious. Initial experiences with B/FEVAR show that this is a safe and feasible approach but long-term results are needed to compare it to OSR and standard TEVAR. Nevertheless, further development of dedicated branched and fenestrated stent-grafts for CBAD is needed to continuously improve their performance and broaden its indications. Until then, optimal treatment of CBAD remains debatable and merits a patient specific decision based on anatomy, life expectancy, general patient condition, and available expertise. TUTE seems a feasible and useful tool to better understand management outcomes of CBAD.
Supporting Information S1 Appendix. Search strategy MEDLINE.
(DOCX) S2 Appendix. PRISMA Checklist.
(DOC) S1
[fig] Fig 1: Flowchart of article selection. doi:10.1371/journal.pone.0154930.g001 [/fig]
[fig] Fig 2: Time until Treatment Equipoise. Results of TUTE analysis for OSR (top), TEVAR (middle) and B/ FEVAR (bottom) for CBAD. doi:10.1371/journal.pone.0154930.g002 [/fig]
[table] Table 1: Complications and survival OSR. [/table]
[table] Table 2: Complications and survival TEVAR. [/table]
[table] Table 3: Complications and survival B/FEVAR. ARF = acute renal failure; B/FEVAR = branched and fenestrated thoracic endovascular aortic repair; FU = follow-up; MI = myocardial infarction; SCI = spinal cord ischemia; ST = short term a Extensive dissection cohort (Type II/III) b Focal dissection cohort (without visceral involvement) doi:10.1371/journal.pone.0154930.t003 [/table]
[table] Table: Demographics and OSR details. (DOCX) S2 Table. Demographics and TEVAR details. (DOCX) S3 Table. Demographics and B/FEVAR details. (DOCX) [/table]
|
High Performance Solution Processed Organic Field Effect Transistors with Novel Diketopyrrolopyrrole-Containing Small Molecules
The donor-acceptor (D-A)-type diketopyrrolopyrrole (DPP)-based small molecules (LGC-D117 andLGC-D118) were synthesized and used as the active layer of solution-processable organic field-effect transistors (OFETs). Both LGC-D117 and LGC-D118 contain silaindacenodithiophene as electrondonor units with DPP as an electron-accepting linker, and octylrhodanine as the electron-accepting end group. The molecules were functionalized with different side chains to study their effects on OFET characteristics. LGC-D117 has a simple branched alkyl side chain, whereas LGC-D118 features a bulky siloxane-terminated hybrid alkyl chain. The siloxane side chains of LGC-D118 account for its better crystallinity, leading to significantly high field-effect mobility (max 3.04 cm 2 V −1 s −1 ). In particular, LGC-D118 is well soluble and sustains the high mobility in the environmentally friendly 2-methyltetrahydrofuran solvent with low temperature annealing at 100 °C due to the bulky siloxaneterminated alkyl side chain.Conjugated molecules have been actively researched due to their great potential for realizing ultra-lightweight flexible devices on plastic substrates 1-5 . In particular, side chain engineering of these molecules enables device fabrication using cost-effective graphic-art printing processes 6-12 . Among the emerging organic devices, organic field-effect transistors (OFETs) are considered a core component of various analog and digital integrated circuits. The realization of high-performance OFETs requires the development of novel organic semiconductors (OSCs) with excellent π-orbital planarity and small intermolecular distance to facilitate inter-and intramolecular charge carrier transport in the transistor channel 13-18 . Conjugated polymers composed of alternating electron donor (D) and acceptor (A) moieties (D-A) show impressively high field-effect mobilities (μ FET ) 19-23 . Strong attractive interactions between the D units in one molecule and the A units in the neighboring one(s) provide a large π-orbital transfer integral for efficient charge carrier hopping 24 . Recently D-A conjugated polymers have shown rapid progress of their charge carrier mobility, showing values above 10 cm 2 V −1 s −1 for both holes and electrons due to the active utilization of various large planar building blocks, such as isoindigo, naphthalenediimide, perylenediimide, and diketopyrrolopyrrole (DPP) 8, 19, 25-29 . Despite the impressive success of D-A polymers, the development of D-A small molecules for the active layer of solution-processable OFETs lags behind. Solution-processable small-molecule OSCs have several advantages over polymer semiconductors, exhibiting less batch-to-batch variation (i.e., better reproducibility), easy purification, and functionalization potential. Thus, the active development of solution-processable D-A small molecules is essential 20 . In particular, DPP-based D-A small molecules have been widely investigated because of their promising performance in organic electronic applications 30, 31 . Zhou et al.32proposed a new D-A small-molecule OSC that features a tetrathienoacene donor core with DPP as the end acceptor and exhibits hole mobility in the range of 0.02-0.09 cm 2 V −1 s −1 . Kim et al. 33,34also used DPP as an acceptor and dithiensilole as the core donor, Published: xx xx xxxx OPEN www.nature.com/scientificreports/ 2 Scientific RepoRts | 7: 164 |
achieving ambipolar transport characteristics with hole mobility of 0.011 cm 2 V −1 s −1 and electron mobility of 0.015 cm 2 V −1 s −1 . However, OFETs utilizing DPP-based small molecules show a much lower performance than those utilizing DPP-based polymers.
In addition to the main backbone design, side chain engineering is also important for OFET performance, determining solubility, molecular packing, polarity, and film forming properties. In particular, molecular packing is greatly affected by alkyl chain length and branching point position [bib_ref] Influence of alkyl chain branching positions on the hole mobilities of polymer..., Lei [/bib_ref]. Nevertheless, a limited number of side chain designs have been reported, featuring mostly linear and branched hydrocarbon alkyl chains of different lengths [bib_ref] Record High Hole Mobility in Polymer Semiconductors via Side-Chain Engineering, Kang [/bib_ref] [bib_ref] Critical Role of Alkyl Chain Branching of Organic Semiconductors in Enabling Solution-Processed..., Zhang [/bib_ref] [bib_ref] Effect of side chain length on the charge transport, morphology, and photovoltaic..., Duan [/bib_ref] [bib_ref] Alkoxy-Functionalized Thienyl-Vinylene Polymers for Field-Effect Transistors and All-Polymer Solar Cells, Huang [/bib_ref]. For example, Minari et al. [bib_ref] Alkyl chain length dependent mobility of organic field-effect transistors based on thienyl-furan..., Minari [/bib_ref] studied the Alkyl chain length dependence of OFET performance. Compared to branched alkyl chains, linear alkyl chains generally promote closer molecular packing due to interchain interdigitation [bib_ref] Liquid-crystalline semiconducting polymers with high charge-carrier mobility, Mcculloch [/bib_ref] [bib_ref] A New Poly(thienylenevinylene) Derivative with High Mobility and Oxidative Stability for Organic..., Lim [/bib_ref]. However, DPP based semiconductors utilizing linear alkyl chains frequently show poor or no solubility in common solvents. Recently, Mei et al. [bib_ref] Siloxane-terminated solubilizing side chains: Bringing conjugated polymer backbones closer and boosting hole..., Mei [/bib_ref] demonstrated OFETs with improved performance employing bulky siloxane-terminated linear alkyl chains in the polymer, the enhancement can be ascribed to combination of smaller π-stacking distances and larger crystalline coherence length. Lee et al. also reported highly improved ambipolar OFET performance for DPP based polymer with bulky siloxane-terminated linear alkyl chains [bib_ref] Boosting the Ambipolar Performance of Solution-Processable Polymer Semiconductors via Hybrid Side-Chain Engineering, Lee [/bib_ref].
In this study, we report and compare two new D-A type DPP-based small molecules, LGC-D117 and LGC-D118 [fig_ref] Figure 1: Molecular structures of LGC-D117 and LGC-D118 [/fig_ref] for the active layer of solution-processable OFETs.
LGC-D117 and LGC-D118 feature silaindacenodithiophene as the electron-donating core, DPP as the electron-accepting linker, and octylrhodanine as the electron-accepting end group. Both LGC-D117 and LGC-D118 have the same backbone, but the side chains attached to the DPP linker are different. Because of the change from a normal alkyl chain to a bulky siloxane-terminated hybrid alkyl chain [bib_ref] Siloxane-terminated solubilizing side chains: Bringing conjugated polymer backbones closer and boosting hole..., Mei [/bib_ref] [bib_ref] Boosting the Ambipolar Performance of Solution-Processable Polymer Semiconductors via Hybrid Side-Chain Engineering, Lee [/bib_ref] , LGC-D118 exhibited improved crystallinity even in case of the as-spun film, leading to a significantly enhanced μ FET after annealing at 100 °C (up to 3.04 cm 2 V −1 s −1 ). To the best of our knowledge, this is the first report on siloxane-terminated hybrid bulky side chain in small molecular OSCs for OFETs. Interestingly, this side chain makes the semiconductor film highly crystalline, accounting for its rather high mobility (max 3.04 cm 2 V −1 s −1 ) with low temperature annealing at 100 °C, which is the highest field-effect hole mobility reported in literature for DPP-based solution-processable small-molecule OSCs. Additionally, the bulky side chain makes LGC-D118 readily soluble in various eco-friendly solvents such as 2-methyltetrahydrofuran (M-THF). The LGC-D118-based OFETs fabricated using the M-THF solution showed impressive (max 2.60 cm 2 V −1 s −1 ) after low temperature annealing at 100 °C. These results indicate that the newly synthesized small molecule OSC with siloxane-terminated hybrid bulky side chain, LGC-D118, can be a promising active material for high-performance flexible OFETs with low temperature processes for the commercialization.
# Results
LGC-D117 and LGC-D118 were obtained via palladium-catalyzed coupling and Knoevenagel condensation, and were characterized by 1 H NMR and mass spectroscopy, with synthetic routes and detailed procedures provided in the Supporting Information. The new OSCs exhibit good solubility in common organic solvents such as chloroform, chlorobenzene, and 1,2-dichlorobenzene, with LGC-D118 displaying better solubility than LGC-D117 due to the siloxane-terminated hybrid bulky side chain. Differential scanning calorimetry (DSC) analysis of LGC-D117 and LGC-D118 revealed sharp endotherms at 218 and 248 °C, while exotherms were detected at 168 and 231 °C, respectively, as shown in [fig_ref] Figure 1: Molecular structures of LGC-D117 and LGC-D118 [/fig_ref] and summarized in . Due to stiff siloxane side chain,
LGC-D118 has higher melting and crystallized temperature compared with LGC-D117.
The UV-vis spectra of LGC-D117 and LGC-D118 as chlorobenzene solutions and thin films are shown in [fig_ref] Figure 2: UV-vis absorption spectra of [/fig_ref] , with the corresponding absorption properties summarized in . The absorption spectra of all compounds were similar for both solutions and films. The optical bandgaps of LGC-D117, and LGC-D118 were estimated as 1.41 and 1.40 eV, respectively. Both LGC-D117 and LGC-D118 exhibited similar absorption peaks around 710 nm in solution, whereas the absorption peak of the LGC-D118 film was more red-shifted than that of the LGC-D117 film, indicating stronger π-π stacking in the LGC-D118 film. After annealing at 110 °C, however, the absorption peak of the LGC-D117 film was slightly red-shifted, while that of LGC-D118 was marginally blue-shifted. The electrochemical properties of all compounds were investigated by cyclic voltammetry, as shown in [fig_ref] Figure 2: UV-vis absorption spectra of [/fig_ref] and summarized in . The lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) energy levels were calculated using the onset of the reduction/oxidation peaks. The LUMO energy levels of LGC-D117 and LGC-D118 were estimated as −3.60 and −3.66 eV, respectively, based on the onset reduction potential referenced to ferrocenium/ferrocene (−4.8 eV). The respective HOMO energy levels were estimated as −5.19, and −5.18 eV, while the electrochemical bandgaps were calculated as 1.59 and 1.52 eV, respectively.
The charge carrier mobility (μ FET ) in OFETs shows a strong dependence on the transfer integrals, which are sensitive to the π-π stacking distance and molecular packing conformation. To investigate and compare the OFET properties of LGC-D117 and LGC-D118, top gate/bottom contact (TG/BC) transistors were fabricated. Each OSCs was dissolved in chlorobenzene (CB) at a concentration of 3 mg/mL, and the obtained solutions were spin-coated onto glass substrates with pre-patterned electrodes, which were used with and without annealing. Subsequently, CYTOP TM was spin-coated onto the OSCs as a dielectric layer and annealed at 90 °C. All devices were characterized under N 2 atmosphere in a glovebox, showing good linear and saturation regimes. [fig_ref] Figure 3: Typical transfer and output curves of organic field-effect transistors based on LGC-D117... [/fig_ref] shows typical transfer and output curves of OFETs based on LGC-D117 and LGC-D118, with their performance summarized in (transfer curves of OFETs after annealing are shown in Figs S3 and S4). The transistor parameters, including μ FET , threshold voltage (V th ), and the on/off ratio were calculated in the saturation regime using the standard OFET formula. For LGC-D117, the spin-coated film exhibited a lowest mobility of 0.41 cm 2 V −1 s −1 on average after annealing at 100 °C. Devices based on LGC-D117 annealed at higher temperatures show slightly LGC higher performance [fig_ref] Figure 3: Typical transfer and output curves of organic field-effect transistors based on LGC-D117... [/fig_ref] and . In contrast, spin-coated LGC-D118 film exhibited average mobility of 1.76 cm 2 V −1 s −1 . For LGC-D118, having a solubilizing siloxane-terminated linear alkyl chain, the OFETs exhibited average mobilities of 1.76, 1.42, and 1.04 cm 2 V −1 s −1 after annealing at 100, 120, and 140 °C, respectively, and highest mobility of 3.04 cm 2 V −1 s −1 at 100 °C. Thus, the mobility increased with lower temperature annealing. Environmentally friendly manufacturing processes are valuable for the commercialization of solution-processed OFETs, attracting much interest to the synthesis of high-performance semiconducting materials soluble in environmentally friendly solvents [bib_ref] Hydrofluoroethers as Orthogonal Solvents for the Chemical Processing of Organic Electronic Materials, Zakhidov [/bib_ref] [bib_ref] High-Performance Polythiophene Thin-Film Transistors Processed with Environmentally Benign Solvent, Liu [/bib_ref]. Herein, we used the eco-friendly 2-methyltetrahydrofuran (M-THF) as M-THF solvent [bib_ref] 2-Methyltetrahydrofuran (2-MeTHF): A Biomass-Derived Solvent with Broad Application in Organic Chemistry, Pace [/bib_ref] [bib_ref] 2-Methyltetrahydrofuran as a suitable green solvent for phthalimide functionalization promoted by supported..., Pace [/bib_ref] for device fabrication.
LGC-D118 was readily soluble in M-THF, and OFET fabricated using the corresponding M-THF solutions showed stable p-type operation . In particular, LGC-D118 OFETs exhibited an average mobility of 1.53 cm 2 V −1 s −1 and highest mobility of 2.60 cm 2 V −1 s −1 after annealing at 100 °C as shown in .
# Discussion
The electronic wavefunction overlap determining the charge transfer integral is a sensitive function of precise molecular packing [bib_ref] Highly sensitive flexible pressure sensors with microstructured rubber dielectric layers, Mannsfeld [/bib_ref]. Therefore, we investigated the crystallinity of OSCs, which is important for understanding the high μ FET of LGC-D118 OFETs, as the high mobility of LGC-D118 likely due to the higher crystallinity of . Summary of devices incorporating different materials and annealing conditions. *Error in average mobility = standard deviation (σ) of values. Averaging was performed for 4-8 devices. **Contact resistance at channel width of 1000 µm obtained using the Y-function method 52 .
their films. Polarization microscopy was utilized to investigate the crystalline morphologies of the fabricated thin films. shows polarized images of different as-spun films (images of annealed films are shown in Atomic force microscopy (AFM) imaging of OSC films was further performed to confirm their microstructure. [fig_ref] Figure 4: Atomic force microscopy imaging of the as-spun organic films for [/fig_ref] show the topography of LGC-D117 and LGC-D118 films, with a textured structure formed for all materials. In each domain, the microstructures are aligned parallel to one direction. We found that the surface of the LGC-D117 film featured small domains, with slight morphology changes observed after annealing at high temperatures. Compared to LGC-D117, AFM imaging of LGC-D118 showed a different morphology, the domain size being very large even for the as-spun film. However, the LGC-D118 film showed a larger grain boundary, and the gap between the disordered grains increased after thermal annealing, creating trapping sites for limiting μ FET by charge transport at high annealing temperature (140 °C, [fig_ref] Figure 4: Atomic force microscopy imaging of the as-spun organic films for [/fig_ref]. This microstructure data can explain the mobility trend observed for different annealing temperatures.
Two-dimensional (2D) grazing incidence X-ray diffraction (GIXRD) was used to analyze molecular packing in thin films (Figs 5 and S10), with the results summarized in . GIXRD data of all as-spun and annealed (140 °C) films showed strong edge-on orientation of molecules, with three visible orders of lamellar stacking peaks in the out-of-plane direction. The largest out-of-plane peak intensity was recorded for the as-spun LGC-D118 film . Materials with different side chains exhibited different out-of-plane d-spacings. Compared to LGC-D117, which have alkyl side chains in their DPP units, LGC-D118 exhibited the longest d-spacing of 18.04 Å due to its siloxane-terminated solubilizing groups. The reason might be due to difference of intermolecular stacking distance and the side-chain ordering among those molecules [bib_ref] Some interesting things about polysiloxanes, Mark [/bib_ref]. The siloxane side chains in LGC-D118 also account for its high crystallinity (even for the as-spun film), being involved in strong intermolecular interactions [bib_ref] Use of intermolecular hydrogen bonding for the induction of liquid crystallinity in..., Kumar [/bib_ref]. The as-spun LGC-D118 film showed high crystallinity, which slightly decreased after annealing at 140 °C. The results of GIXRD analysis were strongly correlated with those of polarizing optical microscopy, AFM and OFET tests.
LGC-D118 showed the highest mobility for as-spun samples, while OTFTs after annealing with increased temperature showed a lower performance. This crystallinity and mobility trend is rationalized by the properties of the employed side chains.
In conclusion, we report new D-A type DPP-based small molecules for high-performance OFETs. The highest μ FET of 3.04 cm 2 V −1 s −1 and excellent solubility were achieved for LGC-D118 OFETs due to the strong interaction between introduced siloxane side chains. Additionally, LGC-D118 was readily soluble in an environmentally friendly solvent [fig_ref] Figure 1: Molecular structures of LGC-D117 and LGC-D118 [/fig_ref] , sustaining its high mobility. GIXRD, AFM, and polarization microscopy results indicated that this remarkably high mobility could be attributed to the higher crystallinity and shorter intermolecular distance achieved by mild annealing. Our results demonstrate that small molecule with siloxane side chain can be used to prepare OSCs for organic electronics, showing high crystallinity, sufficient solubility, and low temperature processing, and possibly further boosting device performance.
# Methods
Characterization and Measurements. Materials were characterized by 1 H NMR spectroscopy (Agilent DD1, 500 MHz). Differential scanning calorimetry (DSC) was performed using a TA Instrument Q20 differential scanning calorimeter at heating rate of 10 °C/min in a nitrogen atmosphere. UV-vis spectra were obtained with a Mecasys Optizen Pop spectrophotometer. Cyclic voltammetry (CV) experiments were performed with an AutoLab analyzer. All CV measurements were carried out in 0.1 M tetrabutylammoniumtetrafluoroborate (Bu4NBF4) in acetonitrile with platinum as the counter electrode, indium tin oxide (ITO) coated with a thin film as the working electrode, and Ag/Ag+ electrode as the reference electrode, at a scan rate of 100 mV/s. To estimate the polymer energy levels from the vacuum energy level, we used the ferrocene/ferrocenium (Fc/Fc+) redox couple as a calibration reference. The half-wave potential (E1/2) for oxidation of the Fc/Fc+ redox couple was assumed to be 4.8 eV, below the vacuum level.
Film Characterizations. Optical images were obtained using an OLYMPUS polarized microscope. 2D GIXRD measurements of the thin film microstructure were performed at the 9A beamlines of the Pohang Accelerator Laboratory (PAL). 11.07 KeV photons with grazing angle 0.13° were directed onto the sample to produce 2D scattering patterns. The surface morphology measurements were performed using an atomic force microscopy (AFM) (NX10, Park systems).
Device fabrication and Characterization. OFET devices were fabricated in a TG/BC structure. Glass substrates were employed, and a lithographed electrode (Au/Ni = 13 nm/3 nm) was used as the source and drain electrodes. The glass substrates were sequentially cleaned with acetone, DI water, and isopropanol, and oven dried at 110 °C for 1 h. After drying, the substrates were treated with UV/Ozone for 30 min and then moved into a N 2 filled glovebox. The pristine organic semiconductor layer was spin coated onto glass substrate from the solution (3 mg/ml in CB) at 2000 rpm and annealed at different temperatures for 1 h. CYTOP was spin coated on to the organic semiconductor as the dielectric layer (1:1 diluted) at 1500 rpm and annealed at 90 °C for 1 h. Al (50 nm) was used for the gate electrode and was thermally evaporated under vacuum (~10 −6 Torr). Electrical characterization was measured under nitrogen using a Keithley semiconductor parametric analyzer (Keithley 4200-SCS). Hole mobility (μ) was determined using I ds = (WC i /2 L) × μ × (V g − V th ) 2 in the saturation regime, where C i is the capacitance measured from OFET structure device [fig_ref] Figure 1: Molecular structures of LGC-D117 and LGC-D118 [/fig_ref] , I ds is the drain-source current, V g is gate voltage, and V th is the threshold voltage.
[fig] Figure 1: Molecular structures of LGC-D117 and LGC-D118. Scientific RepoRts | 7: 164 | DOI:10.1038/s41598-017-00277-7 [/fig]
[fig] Figure 2: UV-vis absorption spectra of (a) LGC-D117, (b) LGC-D118 solutions and films. [/fig]
[fig] Figure 3: Typical transfer and output curves of organic field-effect transistors based on LGC-D117 and LGC-D118 OSCs after annealing at 100 °C. (V sd = −40 V). [/fig]
[fig] Figure 4: Atomic force microscopy imaging of the as-spun organic films for (a) LGC-D117, (b) LGC-D118 and of films annealed at 140 °C for (c) LGC-D117, (d) LGC-D118. [/fig]
[fig] Figure 5: Grazing incidence X-ray diffraction patterns of the as-spun and thermally annealed (140 °C) LGC-D117 [(a) and (c)], LGC-D118 [(b) and (d)]. [/fig]
|
Secret Sarcoma: A Cardiac Mass Disguised as Influenza
This case presentation discusses an extremely rare diagnosis presenting with common symptoms, attributed to influenza, which were not investigated further when clear cardiac symptoms developed. The patient initially presented with symptoms consistent with influenza, but when orthopnea and dyspnea on exertion developed, these cardiac symptoms continued to be attributed to a postviral syndrome and were not further evaluated. Premature closure bias contributed to a delay in diagnosing a rare cardiac condition. The diagnostic momentum, or the continuing of a diagnostic label, occurred across multiple providers and settings. This case demonstrates the risk of premature closure and diagnostic momentum and reminds clinicians to reframe the differential diagnosis as more information on history or physical exam becomes available.
# Introduction
A cardiac sarcoma is a rare malignant cardiac tumor that can present with nonspecific symptoms (similar to most cancers) in combination with clinical signs of heart failure due to the mass effect of the tumor. Timely diagnosis is crucial to prevent further obstruction as the tumor grows. If the tumor is recognized in the earlier stages, complications including heart failure, arrhythmias possibly leading to sudden cardiac death, and ischemia secondary to obstruction of coronary arteries can be avoided. Additionally, early detection can allow treatment before the cancer can metastasize.
## Case presentation
A 50-year-old Asian male with a past medical history of supraventricular tachycardia and obstructive sleep apnea on CPAP at night presented with one month of intermittent flu-like symptoms, orthopnea, and dyspnea on exertion. At the onset of these symptoms, he presented to a walk-in clinic and was diagnosed with influenza. He was treated symptomatically and noted improvement, but one week later he had a recurrence of symptoms while playing volleyball. From that time on, he noticed dyspnea on exertion, continued malaise, fevers, and diffuse joint pains so he presented multiple times to outpatient providers. He received doxycycline without improvement, and follow-up testing showed a mild leukocytosis, negative EBV, and an unremarkable chest X-ray. He was diagnosed with lingering postviral symptoms from influenza. He ultimately presented as a walk-in patient to the cardiology clinic when he started having chest tightness, palpitations, and his dyspnea progressed to occurring at rest, relieved only with a tripod position.
EKG on presentation [fig_ref] Figure 1: EKG on presentation showing right axis deviation and ST-T wave changes [/fig_ref] showed right axis deviation and abnormal ST-T wave segments in V1 through V3 which was new compared with a prior EKG. Due to the concern for pulmonary embolism, a CT angiogram of the chest was obtained which displayed moderate bilateral pleural effusions, a mass in the right ventricle, and a mass in the left atrium extending through the mitral valve invading into the left ventricle [fig_ref] Figure 2: CT angiogram displayed moderate bilateral pleural effusions, a mass in the right... [/fig_ref]. Echocardiogram exhibited normal LVEF but some mitral valve occlusion due to the mass. Cardiac MRI was obtained [fig_ref] Figure 3: MRI confirmed masses in the right and left ventricle [/fig_ref] and confirmed the masses. The patient required debulking of the left atrial tumor, and pathology revealed an undifferentiated, highgrade pleomorphic sarcoma. Due to tumor infiltration into the left pulmonary veins, as well as focal areas of uptake in the small bowel at a site of intussusception, he was started on pembrolizumab chemotherapy with concurrent radiation therapy to the heart and small bowel.
# Discussion
A cardiac sarcoma is a primary malignant cardiac tumor. These are exceedingly rare, with the incidence of primary neoplasms of the heart found at autopsy being 0.017%-0.056% [bib_ref] A 30-year analysis of cardiac neoplasms at autopsy, Butany [/bib_ref] [bib_ref] Tumors of the heart. A 20-year experience with a review of 12,485..., Lam [/bib_ref]. Of these rare primary cardiac tumors, up to 75% are benign and are usually myxomas, lipomas, or rhabdomyomas. Only 25% of primary cardiac tumors are malignant and, of these, 75% are sarcomas [bib_ref] Primary cardiac tumors, Leja [/bib_ref]. Tumors presenting in the heart are more frequently secondary tumors than primary. The incidence of secondary tumors is 1.23% with the three most common malignant neoplasms to the heart being from lung cancer, esophageal carcinoma, and lymphoma. Risks of a cardiac sarcoma include malignant spread and physiologic effects on the heart including compression of the coronary arteries or pericardial space, arrhythmias, and outflow obstruction. Sudden death is associated with primary neoplasms approximately 0.0025% of the time. Although 86% of these were classified as benign, their locations within the heart lead to conductive and hemodynamic abnormalities resulting in sudden death [bib_ref] Primary cardiac tumors causing sudden death: a review of the literature, Cina [/bib_ref]. In previous studies, most patients with a cardiac sarcoma were dead within one year of their diagnoses [bib_ref] Primary tumors of the heart and pericardium, Mcallister [/bib_ref] and had a median survival time of 6 months [bib_ref] Primary sarcomas of the heart, Burke [/bib_ref]. In a newer study, patients who underwent resection with curative intent and survived surgery (surgical mortality was 7.4%), the median survival was 23.5 months [bib_ref] Outcomes after surgical resection of cardiac sarcoma in the multimodality treatment era, Bakaeen [/bib_ref]. Nonsurgical options for treatment of cardiac sarcomas can be difficult due to radiation effects on the surrounding heart and the high levels of radiation (6000-6500 cGy) that is typically needed. Also, chemotherapy effective against sarcomas, such as Adriamycin, is typically avoided due to its cardiotoxic side effect profile [bib_ref] Primary cardiac sarcoma: a novel treatment approach, Movsas [/bib_ref].
The epidemiology indicates the rarity of this case, but the importance of recognizing this presentation in an outpatient setting before it progresses should not be lost. Beyond adding to the differential diagnoses of a patient presenting with progressive dyspnea on exertion, this case reinforces the importance of reframing the diagnosis when seeing a patient multiple times within a short timeframe. This patient was seen four times within a month by a combination of outpatient providers. Premature diagnostic closure bias creates difficulties for busy practitioners to reassess a case and develop a new differential when the previous diagnosis no longer applies. Our patient had a viral-like illness at the onset, but persistent symptoms of dyspnea on exertion and orthopnea should prompt a cardiac workup. An EKG showing right axis deviation and ST-T wave changes may have alerted practitioners that further workup is needed, and an echocardiogram or CT angiogram would have discovered the rare underlying diagnosis and allowed for expedited diagnosis and treatment.
[fig] Figure 1: EKG on presentation showing right axis deviation and ST-T wave changes. [/fig]
[fig] Figure 2: CT angiogram displayed moderate bilateral pleural effusions, a mass in the right ventricle, and a mass invading into the left atrium through the mitral valve into the left ventricle. [/fig]
[fig] Figure 3: MRI confirmed masses in the right and left ventricle. Shown above is an image of the mass located in the right ventricle. [/fig]
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.